Value Creation in Multinational Enterprise, Volume 7 (International Finance Review) (International Finance Review)

VALUE CREATION IN MULTINATIONAL ENTERPRISE INTERNATIONAL FINANCE REVIEW Series Editor: J. Jay Choi Volume 1: Asian F...

Author: J. Jay Choi | Reid W. Click

28 downloads 1133 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

VALUE CREATION IN MULTINATIONAL ENTERPRISE

INTERNATIONAL FINANCE REVIEW Series Editor: J. Jay Choi Volume 1:

Asian Financial Crisis: Financial, Structural and International Dimensions Edited by: J. J. Choi

Volume 2:

European Monetary Union and Capital Markets Edited by: J. J. Choi & J. Wrase

Volume 3: Global Risk Management: Financial, Operational & Insurance Strategies Edited by: J. J. Choi & M. Powers Volume 4:

The Japanese Finance: Corporate Finance and Capital Markets in Changing Japan Edited by: J. J. Choi & T. Hiraki

Volume 5:

Latin American Financial Markets: Developments in Financial Innovations Edited by: H. Arbelaéz & R. W. Click

Volume 6:

Emerging European Financial Markets: Independence and Integration Post-Enlargement Edited by: J. A. Batten & C. Kearney

INTERNATIONAL FINANCE REVIEW VOLUME 7

VALUE CREATION IN MULTINATIONAL ENTERPRISE EDITED BY

J. JAY CHOI Temple University

REID W. CLICK George Washington University

Amsterdam – Boston – Heidelberg – London – New York – Oxford Paris – San Diego – San Francisco – Singapore – Sydney – Tokyo JAI Press is an imprint of Elsevier

JAI Press is an imprint of Elsevier The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2007 Copyright r 2007 Elsevier Ltd. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN-13: 978-0-7623-1392-1 ISBN-10: 0-7623-1392-7 ISSN: 1569-3767 For information on all JAI Press publications visit our website at books.elsevier.com Printed and bound in The Netherlands 07 08 09 10 11 10 9 8 7 6 5 4 3 2 1

CONTENTS ABOUT THE SERIES

ix

LIST OF CONTRIBUTORS

xi

PART I: AN OVERVIEW INTRODUCTION TO VALUE CREATION IN MULTINATIONAL ENTERPRISE J. Jay Choi and Reid W. Click

3

PART II: MULTINATIONALS AND FOREIGN DIRECT INVESTMENT STRATEGIC AND FINANCIAL DETERMINANTS OF FOREIGN DIRECT INVESTMENTS Jongmoo Jay Choi and Eric C. Tsai

19

MODELING THE EVOLUTIONARY SEQUENCE OF INTERNATIONAL JOINT VENTURES Elmar Lukas

61

NEW TRENDS AND PERFORMANCES OF KOREAN OUTWARD FDI AFTER THE FINANCIAL CRISIS Seong-Bong Lee

75

v

vi

CONTENTS

PART III: STRATEGIES AND FIRM PERFORMANCE THE VALUE CREATION PERSPECTIVE OF INTERNATIONAL STRATEGIC MANAGEMENT Reid W. Click

99

OWNERSHIP STRUCTURE, DIVERSIFICATION STRATEGY, AND PERFORMANCE: IMPLICATIONS FOR ASIAN EMERGING MARKET MULTINATIONAL ENTERPRISES Juichuan Chang

125

SUCCESSFUL ADAPTATION STRATEGIES OF MULTINATIONAL ENTERPRISES IN CENTRAL AND EASTERN EUROPE Roxana Wright

149

THE RELATIONSHIP BETWEEN ORGANIZATIONAL STRUCTURES AND PERFORMANCE: THE CASE OF THE FORTUNE 500 Nicole Avdelidou-Fischer

169

PART IV: MERGERS AND ACQUISITIONS DIFFICULTIES IN VALUE CREATION: TELECOM NEW ZEALAND’S ACQUISITION OF AAPT LTD Alireza Tourani-Rad and Zoltan Toth

209

ANALYSIS OF GLOBAL COMPETITORS’ REACTION TO MEGA MERGER ANNOUNCEMENTS BY AN MNC: THE CASE OF THE CITICORP–TRAVELERS MERGER Isaac Otchere and Suhadi Mustopo

229

Contents

vii

A COMBINED CASCADE MODEL TO EXPLAIN INDUSTRIAL CONSOLIDATION: THEORY AND AN APPLICATION TO STEEL Huaichuan Rui

255

PART V: FINANCE AND GOVERNANCE MEASURING AND MANAGING THE FOREIGN EXCHANGE EXPOSURE OF CHINESE COMPANIES Patrick J. Schena UK MEASURES OF FIRM-LIVED EQUITY DURATION Richard A. Lewin, Marc J. Sardy and Stephen E. Satchell

285

307

MULTINATIONALS AND EXCHANGE RATE PASS-THROUGH Alexandra Lai and Oana Secrieru

339

CORPORATE GOVERNANCE IN RUSSIA: A CASE STUDY OF TIMELINESS OF FINANCIAL REPORTING IN THE TELECOM INDUSTRY Robert W. McGee

365

PART VI: THE FINANCIAL SERVICES SECTOR MERGERS AND CONSOLIDATION OF FINANCIAL SERVICE FIRMS: GLOBAL TRENDS AND STRATEGIES FOR VALUE CREATION Edward C. Boyer and Jongmoo Jay Choi

393

viii

CONTENTS

CROSS-BORDER INVESTMENT IN THE LATIN AMERICAN BANKING SECTOR Jesu´s Arteaga-Ortiz, Harvey Arbelaéz and Wendy M. Jeffus

419

PART VII: THE INFLUENCE OF THE STATE GOVERNANCE AND POLITICAL RISK Arvind K. Jain

441

STRATEGIC TRADE POLICY FOR SMALL STATES: A POLITICAL ECONOMY PERSPECTIVE Jaleel Ahmad

461

STRATEGIC IPO UNDERPRICING: THE ROLE OF CHINESE STATE OWNERSHIP Yong Wang and Xiaotian (Tina) Zhang

475

MULTINATIONAL VERSUS STATE POWER IN AN ERA OF GLOBALIZATION: THE CASE OF MICROSOFT IN CHINA, 1987–2004 Jean-Marc F. Blanchard

497

INTERNATIONALIZATION AND VALUE CREATION IN THE GLOBAL TEXTILES AND APPAREL INDUSTRY: A COMPARATIVE ANALYSIS OF LITHUANIA AND MOLDOVA Sanford L. Moskowitz

535

STRATEGY DURING AN INDUSTRY CRISIS: THE POST-QUOTA EXPERIENCE OF TURKISH APPAREL MANUFACTURERS Liesl Riddle

565

ABOUT THE SERIES International Finance Review is an annual book series in the international finance area (IFR, broadly defined). The IFR will publish theoretical, empirical, institutional or policy-oriented articles on multinational business finance and strategies, global capital markets and investments, global risk management, global corporate finance and institutions, currency markets and international financial economics, emerging market finance, or related regional or country-specific issues. In general, each volume will have a particular theme. Those interested in contributing an article or editing a volume should contact the Series Editor, J. Jay Choi at E-mail: [email protected]

EDITORIAL ADVISORY BOARD M. Adler Columbia University NY, USA

V. Errunza McGill University, Montreal, Quebec, Canada

W. Bailey Cornell University, Ithaca, NY, USA

R. Grosse Thunderbird Business School, Glendale, AZ, USA

I. Cooper London Business School, UK

Y. Hamao University of Southern California, Los Angeles, CA, USA

J. Doukas Old Dominion University/European Financial Management, Norfolk, VA, USA

C. R. Harvey Duke University, Durham, NC, USA

G. Dufey University of Michigan, Ann Arbor, MI, USA

R. Hawkins Georgia Institute of Technology, Atlanta, GA, USA ix

x

ABOUT THE SERIES

J. E. Hodder University of Wisconsin, Madison, WI, USA

R. Marston University of Pennsylvania, Philadelphia, PA, USA

M. Levi University of British Columbia, Vancouver, B C, Canada

R. Roll University of California at Los Angeles, CA, USA

D. Logue Dartmouth College, Hanover, NH, USA

A. Saunders New York University, NY, USA

J. Lothian Fordham University, NY, USA

R. Sweeney Georgetown University, Washington, DC, USA

LIST OF CONTRIBUTORS Jaleel Ahmad

Concordia University, Montreal, Canada

Harvey Arbelaéz

Monterey Institute of International Studies, Monterey, CA, USA; Groupe ESC Lille, Lille, France

Jesu´s Arteaga-Ortiz

Universidad de Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain

Nicole Avdelidou-Fischer

Queen Mary, University of London, London, UK

Jean-Marc F. Blanchard

San Francisco State University, San Francisco, CA, USA

Edward C. Boyer

Temple University, Philadelphia, PA, USA

Juichuan Chang

Ling Tung University, Taichung City, Taiwan

J. Jay Choi


Reid W. Click

George Washington University, Washington, DC, USA

Arvind K. Jain

Concordia University, Montreal, Canada

Wendy M. Jeffus

Southern New Hampshire University, Manchester, NH, USA

Alexandra Lai

Bank of Canada, Ottawa, Canada

Seong-Bong Lee

Korea Institute for International Economic Policy, Seoul, Korea

Richard A. Lewin

University of Cambridge, Cambridge, UK xi

xii

LIST OF CONTRIBUTORS

Elmar Lukas

University of Paderborn, Paderborn, Germany

Robert W. McGee

Barry University, Miami Shores, FL, USA

Sanford L. Moskowitz

Saint John’s University, Collegeville, MN, USA

Suhadi Mustopo

Deutsche Bank AG, Jakarta, Indonesia

Isaac Otchere

Carleton University, Ottawa, Canada

Liesl Riddle

George Washington University, Washington, DC, USA

Huaichuan Rui

Brunel University, Uxbridge, UK

Marc J. Sardy

Rollins College, Winter Park, FL, USA

Stephen E. Satchell

University of Cambridge, Cambridge, UK

Patrick J. Schena

The Fletcher School, Tufts University, Medford, MA, USA

Oana Secrieru

Bank of Canada, Ottawa, ON, Canada

Zoltan Toth

Auckland University of Technology, Auckland, New Zealand

Alireza Tourani-Rad

Auckland University of Technology, Auckland, New Zealand

Eric C. Tsai

State University of New York at Oswego, Oswego, NY, USA

Yong Wang


Roxana Wright

Keene State College, Keene, NH, USA

Xiaotian (Tina) Zhang


PART I: AN OVERVIEW

This page intentionally left blank

INTRODUCTION TO VALUE CREATION IN MULTINATIONAL ENTERPRISE J. Jay Choi and Reid W. Click OVERVIEW In a fundamental sense, creation of value is the purpose of a firm. Values – measured by profits, cash flows, stock prices, or some strategic objectives – are the ultimate reasons why a firm exists. The ongoing and ever-expanding discussion of globalization, whether based on trade flows or financial flows, draws attention to the value of multinational enterprise. Existing empirical work on the impact of multinational firms, however, is inconclusive. Some observers point to the valuation discount with international operations due to the costs of agency and control and the difficulty of coordinating complex organizations and cultures. Others emphasize the value of a multinational network and the operational efficiency of a multinational enterprise (MNE). Thus, issues related to value creation are important and lively areas of business and finance. In fact, value creation is now at the frontier between the functional areas of finance and strategy. In international finance, the topic also interacts with economics in the areas of strategic trade policy and exchange rate behavior, as well as business strategy, as it relates to the management of an MNE. The role of government policy is also part of the debate, because the importance of public policy and the behavior of Value Creation in Multinational Enterprise International Finance Review, Volume 7, 3–15 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1569-3767/doi:10.1016/S1569-3767(06)07001-4

3

4

J. JAY CHOI AND REID W. CLICK

policymakers are elevated when finance and business become international, as evidenced by consistent attention to political risk. The papers in this volume of International Finance Review provide a reflection on the role of international finance – and its relationship to strategy, economics, political science, and public policy – in examining value creation in multinational enterprise. These are 22 original papers, not published elsewhere, submitted specifically for this volume based on its theme. The papers present a breadth of methodologies, including theoretical, empirical, conceptual, and case study approaches. Several papers offer combinations of these different categories. Among the empirical papers, there are many kinds of data sets analyzed, ranging from macroeconomic data to firm-level financial data to survey data. In addition, the data sets are rigorously analyzed in many different ways. This volume also takes a broad perspective on multinational enterprise, which not only allows discussion of traditional areas in the study of MNEs as corporations but also includes topics related to multinational enterprise as an undertaking and not just as corporation. For example, there is attention to small- and medium-sized companies as well as larger MNEs. There are also papers that consider exporting enterprises and the environment of multinational enterprise. In this spirit, the volume covers multinational enterprise from a variety of perspectives, including views from private corporations and government policymakers, and the authors of the papers include both academics and practitioners. Altogether, the papers offer insights into value creation through a variety of lenses. One prominent feature of the volume is that the papers collectively cover nearly all areas of the world. As some acknowledgment of its current importance in the world economy, there are four papers on China, ranging from financial studies to industry studies. There are specific country inquiries for Korea, Russia, the U.K., Spain, Turkey, Lithuania and Moldova, and Australia and New Zealand. There are also regional studies of Asia, Latin America, and Central and Eastern Europe. In addition, four papers consider general foreign operations of U.S. enterprises. Another feature of the volume is that many different industries are examined. There are naturally several papers on financial services, but there are also papers on telecommunications firms, the steel industry, and textiles and apparel. The papers are divided into seven parts, including, as Part I, this overview. Part II begins our examination of value creation in multinational enterprise with broad attention to MNEs and foreign direct investment (FDI). In Part III, there is greater focus on strategies and firm performance. Parts IV–VI offer deeper analysis in three specific, yet prominent, areas of

Introduction to Value Creation in Multinational Enterprise

5

value creation and MNEs: mergers and acquisitions, finance and governance, and the financial services sector. Part VII returns to a broader picture, examining the influence of the state in value creation in multinational enterprise, as a way to wrap up the volume.

MULTINATIONALS AND FOREIGN DIRECT INVESTMENT Part II examines multinationals and FDI from both empirical and theoretical perspectives. Jay Choi and Eric Tsai open the discussion with ‘‘Strategic and Financial Determinants of Foreign Direct Investments.’’ They point out that conventional theories regard FDIs as strategic moves based on operational or industrial organization considerations. Choi and Tsai, however, demonstrate that financial factors are also important in corporate FDI decisions. The paper offers a comprehensive assessment through an integrated model with both strategic and financial factors, and empirically investigates the factors using firm-level data from the FDIs of U.S. corporations. The financial factors concern internal capital market strength and corporate governance and include exchange rate changes, internal and external financing cost, risk diversification, and agency costs. The results of the study demonstrate that there is variability in the significance of financial variables depending on industries and destinations. The integrated model with both strategic and financial factors is shown to be superior to either component model in explaining FDIs, and financial factors are shown to be no less important in their ability to explain the prevailing FDI phenomena than strategic or operational variables. Specific attention to international joint ventures is provided in Elmar Lukas’s ‘‘Modeling the Evolutionary Sequence of International Joint Ventures.’’ The paper is a theoretical contribution to the literature on FDI under uncertainty and underscores the importance of modeling the evolutionary sequence pattern of foreign market entry. The model in the paper helps refine the application of real options in the international joint venture context by providing a closed-form solution in continuous time to value their overall strategic flexibility. Moreover, the analysis provides a novel perspective on existing empirical results and generates a number of new testable predictions. The final paper in this section is an examination of FDI from a different part of the world. Seong-Bong Lee examines the empirical patterns of

6


Korean FDI in ‘‘New Trends and Performances of Korean Outward FDI after the Financial Crisis’’ using macroeconomic data and firm-level data aggregated to the macro level. Korean outward FDI increased rapidly during the 1990s, and the Korean chaebols were the main players in this increase. However, Lee points out that outward FDI decreased significantly during the three to four years after the financial crisis and then reversed in 2002, increasing once again. Lee documents that a feature of the current rise in outward FDI is that not only are large enterprises engaging in FDI, but small and medium-sized companies are too. Therefore, Korean outward FDI now shows a mixed investment pattern in which two different types of investment coexist: one is performed by big companies to secure large local markets and the other by small- and medium-sized companies attempting to ensure low labor costs. This new pattern reflects structural changes in the Korean economy as well as the growing trend of reciprocal influences. Lee not only analyzes these new outward FDI patterns, but also examines the performance of outward FDI after the financial crisis in Korea. Lee also offers a concluding discussion of the anticipated impact of these changes in outward FDI on the Korean economy.

STRATEGIES AND FIRM PERFORMANCE With a broad picture generally established, Part III moves on to consider strategies and firm performance. The four papers in this section address the issue of value creation in MNEs through empirical investigations. Two papers utilize data on the foreign operations of U.S. MNEs and the other two utilize data from the Asia/Pacific and Central/Eastern European regions. A general background to strategy and firm performance is provided in the first paper of this section, ‘‘The Value Creation Perspective of International Strategic Management.’’ In this paper, Reid W. Click uses the principle of value creation to analyze the strategic management of MNEs. International strategic management is first defined as the process through which value is created by managers operating across a national border. The domain of international strategic management is thus determined by activities that distinguish international management from domestic management in the process of value creation. Click then uses this perspective on value creation to answer three questions pertaining to international strategic management. First, how important is international strategic management? Simple statistics presented in the paper demonstrate that the international component of value creation is important in the U.S. economy. Second, what is the


7

domain of international strategic management? The paper presents a framework in which international strategic management is the aggregation of value created through international production, marketing, and financial activities, and reveals that the domain of international management is vast. Third, does international strategic management make the whole MNE worth more than the sum of its parts? Empirical evidence uncovered using stock market returns suggests that the answer is yes, at least for U.S. multinationals in the early 1990s. The second paper in this section builds upon the first paper by addressing ‘‘Ownership Structure, Diversification Strategy and Performance: Implications for Asian Emerging Market Multinational Enterprises.’’ Juichuan Chang, the author, presents additional discussion of the issues and turns our attention to a different part of the world. Chang provides an integrated framework that conceptualizes the antecedents of international expansion by emerging market businesses in relation to firm performance. The data set uses 115 MNEs from four Asian countries: Hong Kong, Taiwan, Korea, and Singapore. The methodology develops multiple-item measures of multiple dimensions to clarify ownership structure and categorizes three diversification strategies. Chang then tests how ownership structure and diversification strategy affect the financial performance of emerging market MNEs using return on assets and return on sales. The results indicate that the relationship between ownership structure and firm performance is nonlinear (in fact, an S shape). The results also indicate that excessive international diversification, product diversification, or geographic scope of the expansion process negatively moderate the impact of Asian MNEs’ performance. The third paper in this section moves the discussion to a different part of the world. Roxana Wright examines the ‘‘Successful Adaptation Strategies of Multinational Enterprises in Central and Eastern Europe.’’ The paper explores strategies of adaptation to the environment as employed by multinational corporations and empirically investigates determinants of success using a sample of 100 MNE subsidiaries operating in 19 Central and Eastern European countries. Organizations are treated as adaptive systems that have to match the complexity of their environments. The research recognizes the complex nature of the market institutions emerging from transition, which emphasizes the need for new managerial frameworks. Adaptive approaches such as vertical integration and/or value chain development, leveraging autonomy and integration, local knowledge acquisition, and embedding in the social and political environment are explored in their relationship to success in the region.

8


The last paper in this section on strategy and performance moves the discussion back to the U.S. in an empirical investigation of the Fortune 500 companies. ‘‘The Relationship Between Organizational Structures and Performance: The Case of the Fortune 500,’’ by Nicole Avdelidou-Fischer, focuses on Fortune 500 companies because they have generally been among the most profitable and admired in the world. After a discussion of whether companies should organize regionally, nationally, or globally, the important assumption is made that each structural type utilizes resources differently in generating profit. Performance is conceptualized as return on capital employed and return per employee. Statistical tests reveal that structural types are positively related to financial performance, calculated as return on capital employed, with multidivisional-structured companies outperforming functional-structured ones. Tests also suggest that structural types are not related to human resource performance, calculated as return per employee.

MERGERS AND ACQUISITIONS Part IV considers a specific aspect of the investigation of value creation and MNEs: mergers and acquisitions. Three papers in this section demonstrate that we can learn a lot about value creation in MNEs by studying large, perhaps extreme, events. The papers, in fact, consider mergers and acquisitions in three very different industries: telecommunications, financial services, and steel. Alireza Tourani-Rad and Zoltan Toth take a case study approach and examine a single acquisition that proved to be generally unsuccessful. Their paper, ‘‘Difficulties in Value Creation: Telecom New Zealand’s Acquisition of AAPT Ltd.,’’ begins with an overview of the Australian and New Zealand telecommunications markets and Telecom New Zealand’s (TCNZ) NZ$2 billion acquisition of AAPT Ltd., Australia’s third largest telecommunications firm, in 2000. The paper documents that a few years later, after writing off approximately NZ$1 billion, TCNZ is considering a sell-off at a considerable loss. Tourani-Rad and Toth discuss the strategic reasons behind the acquisition and explain how smaller telecom companies are struggling to compete with the incumbent telecom in Australia. They further conduct an event study to assess the impact of the acquisition on both TCNZ’s and AAPT’s share prices and look at some of the post-acquisition issues. In a second case study, ‘‘Analysis of Global Competitors’ Reaction to Mega Merger Announcements by an MNC: The Case of the Citicorp


9

Travelers Merger,’’ Isaac Otchere and Suhadi Mutopo find that global competitors, especially banks in Europe and the U.S., reacted positively to the Citicorp and Travelers merger announcements in 1998. Uncertainties created by investigations into the merger proposal had a significant impact on competitors’ stock price. The announcement that the merger had been consummated also elicited a significantly positive reaction from the rivals following the resolution of uncertainties emanating from the regulatory challenges. The positive reaction by competitors suggests that the merger was a wealth-creating event for the large firms in the financial services industry. The expected benefits outweighed any competitive effects resulting from the merger. The competitors’ reaction was, however, not homogenous. Cross-sectional analysis in the paper shows that the abnormal returns earned by the competitors were higher the larger the competitor. In addition, the abnormal returns were greater for the U.S. rivals. Otchere and Mutopo conclude that the fact that global competitors reacted positively to the Citicorp–Travelers mega merger announcement is consistent with their assertion that the merger had ramifications that go beyond regulatory concerns in the U.S. Huaichuan Rui’s paper, ‘‘A Combined Cascade Model to Explain Industrial Consolidation: Theory and an Application to Steel,’’ offers a rigorous theoretical examination of mergers and acquisitions along with a test of the theory using data from the global steel industry and, particularly, the steel industry in China. Rui points out that expansion through mergers and acquisitions continues to be a viable international strategy utilized by industrial firms. A striking feature of this is that global giant firms lead the wave and generate an unimaginable impact on relatively small and weak firms across sectors and even nations. Rui observes that there seems to be a ‘‘cascade effect’’ between the industrial consolidations in these areas. A combined cascade model developed in the paper explains that the power imbalance caused by the degree of consolidation of the players within a firm’s value system determines the movement and direction of the cascade effect. Rui concludes that with the existence of such an effect, mergers and acquisitions will be a mutually interdependent, dynamic, reversible, and endless process among industries.

FINANCE AND GOVERNANCE Following on the heels of our detailed discussions of mergers and acquisitions, Part V offers detailed discussion of finance and governance in

10


MNEs. There is one theoretical paper, and the others offer empirical investigations of China, the U.K., and Russia. Patrick J. Schena examines the traditional topic of foreign exchange exposure for a new set of companies in ‘‘Measuring and Managing the Foreign Exchange Exposure of Chinese Companies.’’ Specifically, the paper explores the sensitivity of Chinese stock returns to changes in trade-weighted indices of the Chinese currency (the RMB) and the currencies of China’s trading partners from 1999 to 2003. The paper analyzes an exposure elasticity crosssectionally using accounting variables to proxy for size and costs of financial distress. Schena finds that internationally oriented Chinese companies have experienced exchange exposure particularly against the Japanese yen. He also finds that, against a trade-weighted index, there is no empirical evidence that Chinese firms are engaged in hedging activities. However, when exposures are measured in yen terms, Schena finds that Chinese firms, particularly exporters, engage in active currency hedging. A second type of exposure, interest rate exposure, is examined in ‘‘U.K. Measures of Firm-Lived Equity Duration’’ by Richard Lewin, Marc Sardy, and Stephen Satchell. Investors often have much of their portfolios invested in equities that are exposed to interest rate risk. Hedging underlying equity exposures is not easy; in contrast, fixed-income investors have duration to immunize bond portfolios from small fluctuations in interest rates. U.S. equity duration estimates from a dividend discount model result in long durations – often in excess of 50 years. Based on U.K. data, Lewin, Sardy, and Satchell develop an alternative approach to generate equity duration as a by-product of asset pricing. Their analysis suggests that the equity premium puzzle may comprise an important element in reconciling this approach to equity duration with traditional dividend discount model alternatives. Alexandra Lai and Oana Secrieru study another traditional topic in international finance that is closely related to foreign exchange exposure: exchange rate pass-through. Their paper, ‘‘Multinationals and Exchange Rate Pass-Through,’’ offers a rigorous theoretical approach to the topic. The paper examines the impact an MNE has on the degree of exchange rate pass-through when it engages in Cournot competition with domestic and foreign rivals. The MNE can locate its production for the foreign market domestically – resulting in intra-firm trade – or in the foreign country – resulting in international production. In addition to incomplete exchange rate pass-through, Lai and Secrieru show that an MNE increases the sensitivity of domestic market prices and reduces the sensitivity of foreign market prices to exchange rate movements. In addition, Lai and Secrieru


11

show that intra-firm trade prices are more sensitive to exchange rate movements than their international production counterparts and react in the opposite direction. The final paper in this section considers the role of financial information within the context of corporate governance in a transitional country. In ‘‘Corporate Governance in Russia: A Case Study of Timeliness of Financial Reporting in the Telecom Industry,’’ Robert McGee address timeliness of financial reporting as an important element of transparency. Specifically, McGee looks at the telecommunications industry in Russia and computes the number of days it takes companies to receive an audit opinion, then compares the time lag to the number of days it takes non-Russian companies in the telecommunications industry to receive an audit opinion. McGee finds that Russian companies take longer to report financial results than do nonRussian companies. In addition, larger Russian companies take less time to report their financial condition than do small Russian firms, but the difference is not significant. Companies using Russian Accounting Standards take significantly less time to report financial results than companies using either International Financial Reporting Standards (IFRS) or U.S. Generally Accepted Accounting Principles (GAAP). In addition, companies using IFRS take significantly longer to report financial results than companies using U.S. GAAP. McGee also finds that the dominant auditor in the Russian telecommunications industry did not complete audits in significantly less time than did non-dominant auditors. Finally, in discussion, McGee concludes that although Russian companies take far less time to issue financial statements now than they did a few years ago, it is premature to definitively conclude that the improvement is significant due to the limited data set.

THE FINANCIAL SERVICES SECTOR The financial services sector receives particular attention in Part VI. The two papers in this section offer a detailed discussion of many issues already introduced in this volume, including mergers and acquisitions, and macrolevel data on capital flows. Edward Boyer and Jay Choi take up the issue of rapid consolidation globally in the financial services industry in ‘‘Mergers and Consolidation of Financial Service Firms: Global Trends and Strategies for Value Creation.’’ Consolidation has proceeded not only in the same market but also across different market segments and across national boundaries. Boyer and Choi

12


begin by outlining the general trend of the mergers and acquisitions and consolidation of the financial service industry in both the U.S. and the global economy. They next identify and analyze the reasons that contribute to the consolidation of the financial service industry, and examine some cases of successful and unsuccessful financial service mergers and acquisitions. Finally, they draw some strategic implications. The second paper builds on the first paper in this part by looking at the same phenomenon in a specific area of the world. Wendy Jeffus, Jesu´s Arteaga-Ortiz, and Harvey Arbelaéz look at ‘‘Cross-Border Investment in the Latin American Banking Sector.’’ They point out that cross-border investment in the banking sector has been a large part of overall merger and acquisition activity in Latin America. Spain and the U.S. have been the largest investors, participating in almost 70% of the total transaction value. Specifically, the total value of Spanish investment within that region is over 50 billion dollars. After an introduction and an explanation of the importance of FDI and the implications for cross-border investment in banking, the paper focuses on the largest investor in the Latin American banking sector and attempts to find an explanation for the increasing participation of Spanish banks. Jeffus, Arteaga-Ortiz, and Arbelaéz discuss a potential new reality: Latin America could be the geographic location where major contenders in the banking industry worldwide will be engaged in battles for global dominance. The data for this analysis consist of worldwide investment in Latin America from 1985 to 2002.

THE INFLUENCE OF THE STATE The final section of the volume, Part VII, more closely addresses the influence of the state, although the range of discussion is quite broad. The topic has come up in several other papers. For example, the paper by Otchere and Mutopo in Part IV demonstrated the influence of the U.S. government in the Citicorp–Travelers merger through regulation. The papers in this section offer more detail and depth on the influence of the state and consider countries other than the U.S. The first paper, Arvind Jain’s ‘‘Governance and Political Risk,’’ looks at corruption. Jain sees political risk as arising from renegotiation of implicit or explicit contracts under which foreign investors enter a host country. Governments will legitimately enter into renegotiation to increase the share of rents earned by society. Corrupt political leaders, however, will use their powers to extract rents from foreign investors for personal gains rather than


13

for the good of society. Jain, therefore, recommends that political risk assessment should assess the intentions of government as well as the strengths of political and social institutions that keep leaders under control. In addition, he points out that firms should also understand that their own actions may contribute to creating political risk. In the second paper, Jaleel Ahmad addresses ‘‘Strategic Trade Policy for Small States: A Political Economy Perspective.’’ Ahmad points out that there is a literature on strategic trade policy dealing with industries and sectors characterized by international rivalry for market shares and the struggle to capture ‘‘rents’’ over and above normal factor rewards. Ahmad builds on this literature by exploring the validity and implications of strategic trade policy for small states and small firms that are not major players in international markets. The smallness of the firms may, in fact, be an advantage rather than a hindrance. The implications of smallness for strategic behavior are examined in a simple game-theoretic framework. The insights become sharper when extended to intra-industry trade in differentiated products. Ahmad concludes that the desirable policy interventions for small countries and firms are quite different from those for large firms. The third paper considers a traditional topic in finance, the underpricing of initial public offerings (IPOs), and demonstrates that in certain circumstances the influence of the state is very important. Yong Wang and Xiaotian (Tina) Zhang extend the traditional finance topic to account for the role of the state in ‘‘Strategic IPO Underpricing: The Role of Chinese State Ownership.’’ Wang and Zhang point out that the stock markets in emerging economies are attractive to international investors but their unique characteristics need to be examined. In particular, Chinese stock markets experienced much more significant IPO underpricing than most other stock markets in the world. The paper offers a two-period wealth maximum model to explain the strategic IPO underpricing by state owners. Given the fact that the entire IPO procedure (including the IPO price) is regulated and controlled by state owners, Wang and Zhang argue that state owners strategically underprice the IPO because they care less about the IPO proceeds and more about the wealth gain after IPO. Empirically, Wang and Zhang find a positive relationship between IPO underpricing and state ownership in the Chinese stock market, and this is consistent with the wealth maximization hypothesis of IPO pricing. Jean-Marc Blanchard considers another approach to the issue of the influence of the state in ‘‘Multinational versus State Power in an Era of Globalization: The Case of Microsoft in China, 1987–2004’’ the fourth paper in this section. Blanchard examines relative power by comparing and

14


contrasting two schools of thought: a ‘‘multinationals in command’’ perspective and a ‘‘states in command’’ perspective. The paper offers an alternative analytical framework, a modified bargaining power model highlighting three factors as playing a decisive role in shaping the power relationship between states and multinationals: the balance of needs, allies/ adversaries, and the institutional environment. To illustrate the model and demonstrate its explanatory value, Blanchard examines the experience of Microsoft in China. In the fifth paper on the influence of the state, ‘‘Internationalization and Value Creation in the Global Textiles and Apparel Industry: A Comparative Analysis of Lithuania and Moldova,’’ Sanford Moskowitz considers the political affiliations of the state as a determinant of the success of local industries. The study examines the internationalization process within the textiles and apparel industry and shows how the evolution of an industry toward greater internationalization is intricately linked to its ability to move up its specific value chain. The analysis compares and contrasts the ability of this industry within a Western European country that is a member of the European Union (Lithuania) and a non-accession Eastern European country (Moldova) to move up the textiles and apparel value chain and so achieve higher levels of internationalization. In examining and relating the relevant factors, the analysis provides insights into – and suggests important modifications to – such important concepts and themes as the stage theory of internationalization, the role of ‘‘inward–outward’’ linkages in the value creation process, the mechanism of small- and medium-size firm internationalization, and the part played by the European Union in the internationalization (and thus globalization) process. Liesl Riddle builds on the discussion of the textile and apparel industry in the sixth and final paper in this part, ‘‘Strategy During an Industry Crisis: The Post-Quota Experience of Turkish Apparel Manufacturers.’’ Riddle points out that discussions about the impact of quota elimination in the apparel industry have focused on countries that obviously benefit from their elimination (such as China) and those that clearly are harmed by their demise (such as the sub-Saharan African countries). However, the global production landscape of the apparel trading system also includes numerous countries for which the post-quota environment is uncertain – and vital, but little attention has been paid to the specific impact of the quota elimination on these countries. At the dawn of quota elimination, uncertainty loomed particularly large for Turkey, the world’s fourth largest apparel exporting nation. Riddle’s investigation draws on secondary data and a survey of Turkish clothing exporters to chronicle Turkey’s quota elimination


15

experience during 2005 and 2006. The results indicate that most Turkish apparel exporters expected their market share in the European Union and the U.S. to remain the same or improve in 2005. The reality was far worse: by 2006 the industry had incurred large losses and production had declined sharply. The case provides fashion managers and scholars with insight into the experience of a country for which the post-quota elimination future was uncertain but for whom the apparel industry was vital in 2005.

CONCLUSION Taken together, the 23 original papers in this volume cover a rich array of topics and capture the excitement of studying value creation in MNEs. They provide lessons that are applicable to a variety of participants, ranging from managers in firms to public policymakers. The wide spectrum of fields represented and methodologies implemented suggests that researchers are looking at value creation in MNEs from many different perspectives. Thanks to the generosity of the contributing authors, this volume of International Finance Review captures the essence of current thought on value creation in MNEs, and hopefully will stimulate further investigation of the topic in the future.


PART II: MULTINATIONALS AND FOREIGN DIRECT INVESTMENT


STRATEGIC AND FINANCIAL DETERMINANTS OF FOREIGN DIRECT INVESTMENTS Jongmoo Jay Choi and Eric C. Tsai ABSTRACT Conventional foreign direct investment (FDI) theories regard FDIs as strategic moves based on operational or industrial organization considerations. We demonstrate that financial factors are also important in corporate FDI decisions. The financial factors concern internal capital market strength and corporate governance and include exchange rate changes, internal and external financing cost, risk diversification, and agency costs. There is variability in the significance of financial variables depending on industries and destinations. The integrated model with both strategic and financial factors is superior to either component model in explaining FDIs. However, financial factors are no less important in explaining the prevailing FDI phenomena than strategic or operational variables.

1. INTRODUCTION The mainstream theory of foreign direct investments (FDIs) grew out of international trade and industrial organization traditions. International Value Creation in Multinational Enterprise International Finance Review, Volume 7, 19–60 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1569-3767/doi:10.1016/S1569-3767(06)07002-6

19

20

JONGMOO JAY CHOI AND ERIC C. TSAI

trade theory indicates that international differences in locational endowments lead to international trade. Industrial organization theory suggests firm-specific ownership factors as the primary reasons why multinational corporations (MNCs) can overcome entry costs in making investments abroad. These endowment-focused theories (of either location or firm ownership) can explain FDI flows from developed to less-developed regions of the world. To explain the vast majority of FDI flows between developed countries (DCs), internalization theory was developed.1 It suggests that, in some areas, MNCs as institutions may be more operationally efficient as arbiters of resource allocation than markets. To achieve efficient global resource allocation, i.e., foreign production, an MNC can create its own internal markets to reduce transaction costs and to prevent valuation loss associated with external transactions, especially for information-intensive products in which MNCs specialize. Dunning (1977) suggests an eclectic theory that combines international trade, industrial organization, and internalization theories in an integrated general strategic framework. The advantages for MNCs are due to three strategic sources: ownership, location, and internalization (OLI). The integrated strategic theory is a milestone in the FDI theory, but it leaves some issues unanswered. First of all, the OLI framework does not explain why the most common entry mode of FDIs is international mergers and acquisitions (M&As) rather than greenfield investments. Second, internalization is theoretically beneficial to MNCs regardless of where investment occurs. Internalization therefore does not seem to justify the phenomenon that most FDIs occur between DCs. Third, significant bilateral short-term FDI swings are frequently observed, and it is not clear whether these swings correspond to changes in those strategic factors. Finally, as Itaki (1991) notes, Dunning’s OLI framework does not include any financial factors at all. This is in contrast to normative decision making in practice that usually includes both strategic and financial factors in international investment decisions. Harris and Ravenscraft (1991) suggest three hypotheses for cross-border investments: (a) imperfections in product and factor markets, (b) biases in government and regulatory policies, and (c) imperfections in capital markets. Their empirical work, however, only examined the first hypothesis – which is actually the core of the conventional strategic FDI theories. In this paper, we focus on examining the third hypothesis with the intention of relating it to the first. We do that by introducing the financial variables, as well as the structural ones, as determinants of FDIs. Our financial model depicts FDIs as the consequence of the ‘‘interaction’’ between a strong ‘‘internal’’ capital market and ‘‘internal’’ corporate

Strategic and Financial Determinants of Foreign Direct Investments

21

governance. We show that both financial and strategic variables are significant in explaining corporate international investment behavior, and that these results are fairly robust across industries and for different specifications. We suggest then that the neglect of financial variables in the FDI model may constitute a misspecification of the model. However, compared to the strategic model, the financial model exhibits greater variability depending on the industry sector and investment destinations. By incorporating financial variables, we show that Dunning’s ‘‘eclectic’’ theory would be even more eclectic. Beyond this implication for the academic FDI literature, this paper should demonstrate the usefulness of our integrated model for corporate practitioners who may be examining strategic and financial reasons for investing abroad or for policymakers who may be interested in providing incentives to attract foreign investments. The structure of this paper is as follows. Section 2 discusses the related literature. Section 3 presents the estimation framework and Section 4 describes data used in the empirical work. Empirical results are then presented in Section 5. Section 6 includes a summary conclusion.

2. REVIEW OF RELATED WORK 2.1. Strategic Perspectives on FDIs Traditional international trade theory that emphasizes location endowments as determinants of FDIs is insufficient as a theory of FDI because it cannot explain why international investment takes place in addition to international trade. The modern strategic theory of FDI, developed by Hymer (1976), Kindleberger (1969), and Caves (1971), starts with the notion that excess profit projects are created by ‘‘structural imperfections’’ of some kind in goods or factor markets. FDIs take place when MNCs take on those projects as they try to utilize their firm-specific ownership endowments, such as proprietary technology and others, to generate greater returns (despite their disadvantage of facing significant entry costs) than those of indigenous local firms. MNCs can produce greater profit by switching from an international exchange business (export) to a foreign production business (FDI). Buckley and Casson (1976), Dunning (1977), and Casson (1979) point out that the internalization of intermediate goods markets gives an advantage to a firm that can reduce or eliminate the transaction costs caused by ‘‘natural market imperfections.’’ Furthermore, the uncertainty in the valuation of intangible investments or information-intensive products in the open market

22


implies that internalization is a way of appropriating the potential value of research and development within the firm. Thus, the ability of an MNC to generate rents due to its ownership endowments is enhanced by the benefits of an internal market created within a firm, and when these benefits are important, FDIs make more sense rather than exporting or licensing. Dunning (1977, 1980, 1988) consolidated different strands of mainstream FDI theories in an integrated eclectic framework that combines OLI factors conducive to FDIs. 2.2. Financial Perspectives on FDIs The market imperfection paradigm assumed in the mainstream strategic theory of FDI by Hymer (1976), Kindleberger (1969), and Caves (1971) does not exclude imperfections in financial markets. In addition, the internalization theory emphasizes the distinction between external and internal transactions, which could include financial transactions. Therefore, a financial perspective on FDIs is entirely consistent with the assumptions behind the mainstream strategic FDI theories in the industrial organization framework. From a financial standpoint, variables of major interest include valuation, cost, and risk. For instance, why is a given foreign project valued more for an MNC than for an indigenous firm? Otherwise no FDI will take place. Table 1 indicates that about three-quarters of all direct investments abroad by U.S. firms between 1992 and 1998 are in the form of M&As rather than greenfield investments. The choice between M&As and greenfields would depend, among other things, on how the firm is valued in the market place as opposed to the value of the firm’s assets. If both financial and real markets are efficient and in perfect equilibrium, then there is no disparity in valuation and there is little incentive to utilize M&As versus greenfields or, for that matter, little difference between investing at home and abroad. Aliber (1970) was the first to suggest that FDIs can take place because of financial factors in an environment of partially segmented international capital markets. MNCs generally have superior access to financing in the stable currency zone in international capital markets while the indigenous firm can only raise funds in the relatively weak currency unit. The stability of currency means a lower currency risk premium, and hence a lower cost of capital, for MNCs than for local indigenous firms. As such, Aliber (1970) suggested the possibility that a multinational firm can obtain a higher valuation than an indigenous firm from the same project due to the advantage of lower cost of capital.

Outward and Inward Foreign Direct Investment (FDI) Flows and Stocks of the United States. 1986–1991 Annual Average

1992

1993

1994

1995

1996

1997

1998

1992–1998 Annual Average

Panel A: Outward and Inward FDI Flows of the U.S. Outward flows 26.0 U.S. as % of world 14.4 Inward flows 49.0 U.S. as % of world 30.8

39.0 19.4 18.9 10.8

74.8 31.1 43.5 20.0

73.3 25.8 45.1 18.6

92.1 26.1 58.8 17.8

74.8 19.7 76.5 21.3

110.0 23.2 109.3 23.5

132.8 20.5 193.4 30.0

85.3 23.1 77.9 22.4

1980

1985

1990

1995

1997

1998

Panel B: Outward and Inward FDI Stocks of the U.S. Outflow stock 220.2 251.0 U.S. as % of world 42.9 36.6 Inward stock 83.0 184.6 U.S. as % of world 16.4 23.6

435.2 25.4 394.9 22.3

696.1 24.5 535.6 19.2

860.7 25.1 681.7 19.8

993.6 24.1 875.0 21.4

Note: FDI figures are in billions of U.S. dollars. Data Source: Various issues of World Investment Report, United Nations Conference on Trade and Development.


Table 1.

23

24


The differential access to financing is further attributed to information asymmetry by Froot and Stein (1991). With the presence of information imperfection in otherwise globally integrated capital markets, using an auction example in which a U.S. target firm was sold to the highest bidder, they show that a depreciation of the host country currency systematically reduces the relative wealth of host country agents, facilitating acquisitions of local firms or assets by foreign firms. In their view, information imperfections in capital markets cause external financing to be more expensive because of monitoring costs. No firm, therefore, is able to borrow, without incurring additional cost, the full amount of the project or the value of the acquisition target. Since complete external financing is costly, the relative corporate wealth between foreign and domestic bidders as a source of internal financing becomes relevant. A depreciation of the host country currency systematically raises the relative wealth of foreign firms, who in turn can outbid indigenous local firms. Nonetheless, given additional operational costs associated with FDIs, their explanations based on financing do not extend to why FDIs are sustainable in the long run. Given partially segmented international capital markets and/or the presence of agency costs, the two-country corporate valuation model of Choi (1989) shows that corporate international investment decisions depend on two major factors: (a) the real foreign exchange risk, which affects the cost of capital, and (b) the gains from diversification of operational cash flows, which are necessary at the corporate level given agency costs or market imperfections. The models by Black (1978) and Stulz (1983) emphasize the change in relative distribution of wealth globally as a factor inducing foreign investments. In the empirical literature, exchange rates received the most attention but the results on them are somewhat mixed. A significant linkage is shown between exchange rates and U.S. FDIs in the empirical work of Froot and Stein (1991), Grosse and Trevino (1996), Caves (1989), Ray (1989), and Swenson (1994). Klein and Rosengren (1994) also find real exchange rate and relative wealth to be significant determinants of direct investment inflows in the U.S., but the effects of relative wage rates are insignificant. Choi and Jeon (2006) find a significant cointegrating relationship between real exchange rates and direct investment flows of G-7 countries. Their multivariate cointegration analysis also establishes a linkage between direct investment flows and several financial variables including the cost of capital and real wealth in addition to real exchange rates. Healy and Palepu (1993), however, find the coefficients of both the contemporaneous and lagged changes in exchange rates to be insignificant on


25

FDIs. Stevens (1998) estimates the Froot and Stein (1991) model but finds only weak support for the relationship between FDIs and foreign exchange rates. Dewenter (1995), using the survey data of foreign acquisitions of U.S. targets, finds that, after controlling for relative corporate wealth, the relationship between foreign exchange rate level and foreign investment relative to domestic investment (acquisition activity) is statistically insignificant. Blonigen (1997) suggests that the mixed support for the relationship between FDIs and exchange rates may be related to firm-specific assets. If involved in international acquisitions, firm-specific assets can generate returns in currencies other than that used for purchase, and thus subject acquisitions to exchange risk. He reports that Japanese acquisitions in the U.S. are correlated with real dollar depreciation in industries that tend to have more firm-specific assets, such as manufacturing industries. Cushman (1985) maintains that the reduced cost of capital is a consequence of riskadjusted expected real currency appreciation of the source country, which in turn will stimulate FDIs. However, the net effect also depends on an increase in output prices versus input costs caused by exchange rate changes. That is, the foreign exchange effect is generally ambiguous depending on their effects through output and input markets. Cushman’s pooled estimation results show that FDIs are negatively related to an expected real appreciation of the source country currency and positively related to exchange rate volatility. This implies that the use of aggregate data may obscure the differential effects of foreign exchange rates on FDIs. The literature on financial determinants of FDIs as discussed thus far has centered on the issues of differential valuation or cost. What has rarely been studied is the aspect that FDIs could also be triggered by agency behaviors. Recent studies on the diversification discount, however, offer some parallel comparisons to this issue. Although they make no distinction between the types of diversification (industrial versus global) in examining why managers diversify their firms, Aggarwal and Samwick (2003) argue that agents enjoy risk reduction and private benefits from diversification. It is the private benefits such as improved career prospects, prestige, power, perquisites, compensation, and managerial entrenchment that in many cases drive managers to diversify. If this is the case, global diversification theoretically could offer managers a greater degree of risk reduction and private benefits. Click and Harrison (2000) find a negative correlation between multinationality and management-owned equity and conclude that private gains are the reason for globalization and the global diversification discount. All these are in line with our proposition that agency issues are one of the driving forces for FDIs.

26


In this paper, we intend to establish that financial factors are as important as strategic ones in explaining FDIs. In fact, our results support FDIs as being driven by the interaction between internal capital market strength and internal corporate governance. We achieve this by examining firm-level data, which have not been used in existing literature on financial determinants of FDIs. In contrast to existing work that focus on strategic factors only or those that focus on a single financial variable, we suggest an integrated financial model of FDIs that include agency issues as well as valuation, cost, exchange rates, and risk diversification.

3. ESTIMATING FRAMEWORK The above discussion of existing literature suggests two different views regarding the determinants of FDIs. One hypothesis is strategic in that FDIs take place as a result of a firm’s strategic move to take advantage of market imperfections and differential costs in product and factor markets. This view is represented by the OLI model, which indicates that corporate FDIs increase with favorable location factors and strong firm-specific ownership endowments, and out of the efficiency of internal markets within an MNC setting. Even though the main focus in this paper is on the financial determinants of FDIs, it is imperative to first verify whether OLI theory maintains its validity in explaining FDIs in the 1990s. H1 (Strategic Hypothesis). Corporate FDIs are determined by strategic considerations pertinent to location factors, ownership endowments, and efficiency of internalization. The other is a financial hypothesis, which depends on imperfections and asymmetries in capital markets. Given the imperfect integration of international financial markets, MNCs may have an advantage because of their superior access to lower cost of financing or in stable currency units. To the extent that there exist deviations from purchasing power parity, uncertain changes in exchange rates would have a real impact on international investments. In addition, given information asymmetry that favors internal over external financing, exchange rates may influence the relative wealth of the firm, which may influence the propensity to acquire international firms. In sum, the existing literature on the financial determinants of FDIs primarily pursues this line of inquiry, namely the interaction of foreign exchange rates, internal financing, and corporate wealth on FDIs. Theoretically, although its scope is a bit narrow, it is plausible.


27

In light of the mainstream view of FDI as a strategic one, some additional remarks and justifications on the financial perspectives may be in order. Our financial model depicts that the combination of internal capital market strength and internal corporate governance factors motivates FDIs. First, to overcome their additional costs, the occurrence of FDIs requires internal capital market strength. A firm could actively take advantage of a relative corporate wealth increase due to foreign exchange rate movements. Of course, it also could be the result of corporate free cash flow building up and the lack of domestic investment opportunity. Regardless of the reasons, however, a stronger internal capital market does provide MNCs an edge over the indigenous local firms in the host country. Given imperfect or informationally asymmetric capital markets, a depreciation of the host country currency effectively raises relative wealth of the foreign MNCs. Given the costly external financing due to monitoring costs, an increased wealth implies an advantage of using cheaper internal source of capital. This financing advantage puts MNCs in a better position than indigenous local firms during bidding competitions. It is a plausible explanation for the dominance of M&As as a mode of entry for FDIs. In comparison, the financing advantage for greenfield investments may be relatively less compelling outside the bidding scenario. Greenfield investments would have a greater operational risk than M&As as well as the uncertainty in valuation. Nevertheless, with or without the influence from foreign exchange, imperfectly integrated international capital markets imply a differential access to financing, which gives an advantage to one group of firms (i.e., MNCs) over the other (i.e., local indigenous firms). Second, internal capital market strength, however, does not necessarily lead to FDIs. Depending on managerial objectives and agency behaviors, it is up to the management to exploit the internal capital market strength for globalization, so FDI is also a consequence of internal corporate governance. One can argue that MNCs’ financial advantages from internal capital market strength are short term in nature. That may well be, although there is no evidence that financial effects are necessarily short-lived. It is well documented that, in the presence of agency costs and other financial market imperfections, diversification can provide a risk-reduction benefit to the firm. International diversification allows overall corporate risk to be lowered through reduction in financial or business risk. If so, it may well be the compelling reason for the firm to initiate and sustain FDIs over the long run despite the added cost and risk associated with operating abroad. Even though risk reduction is a legitimate reason for cross-border investments, it is not at all farfetched that FDIs are conducted for the sake of

28


managers’ pursuit of private benefits at the expense of shareholders and lead to unintended global diversification. Earlier studies have suggested various agency problems pertinent to investments. For instance, investment can deviate from its optimal level due to agency problems, such as the underinvestment and substitution problems suggested by Myers (1977) and Jensen and Meckling (1976). Regardless of whether any given level of international investment is optimal, agency problems can still influence certain investment decisions. And it is not obvious that the agency problems are of short-term nature because they depend on long-standing and internationally dissimilar practices of corporate governance and industrial organization. At the higher level, managers’ pursuit of personal benefits, such as gains in prestige, power, human capital, perquisites and compensation, and entrenchment, can be the predominant force driving cross-border investments. This, of course, directs us to believe that agency behavior could lead to FDI. Although FDI driven by agency behavior has not yet been explicitly proposed and tested, recent studies on the diversification discount are consistent with this suggestion. Again, the existing studies on financial determinants of FDIs extend their research primarily within the scope of factors related to internal capital markets. It is, of course, critical, but not necessarily a sufficient condition. Our financial model describes both the necessary and sufficient conditions for FDI occurrence and provides a more comprehensive framework for explaining FDI behaviors from finance perspectives. In the present estimation, the financial factors related to internal capital market strength include foreign exchange rate movement, corporate free cash flow, and cost of capital (relevant if internal funding is insufficient). The other financial factors related to internal corporate governance are international diversification, financial and business risk, and agency cost. In short, FDI outflows may be affected positively by currency depreciation of the host country, internal financing capacity of the firm, and benefits from international diversification, and negatively by agency cost and the cost of external financing. H2 (Financial Hypothesis). Corporate FDIs are determined by financial factors such as foreign exchange rate movements, financing methods, corporate risks, international diversification, and agency costs. We do not think that either hypothesis, standing alone, is sufficient to fully explain actual direct investment decisions by the firm. We think that an exclusive emphasis on either hypothesis may be indicative of the field of specialization of a researcher more than a complete depiction of reality. After all, it is undisputable that no major corporate decisions are being


29

made regarding international investments without considering both strategic and financial dimensions of the project. More importantly, we consider both theories to be conceptually consistent. The mainstream strategic models are based on imperfections in goods and factor markets, which create an advantage for one class of firms as opposed to another. Similarly, if imperfections and information asymmetry exist in international capital markets, one class of firms may have a financial advantage over another. Recourse to market imperfections of some kind is the same either way. In fact, Oxelheim, Stonehill, and Randoy (2001) classify financial factors within the OLI framework. Thus, we state an integrated hypothesis that includes both strategic and financial factors. H3 (Integrated Hypothesis). Corporate FDIs, in general, are determined by both strategic and financial considerations jointly. The relative importance of these factors, however, may vary depending on industry sectors or the context in which investments are made. To summarize, we can characterize FDI in the following general functional form: FDI ¼ f ðStrategicfactors; Financialfactors; ControlvariablesÞ

(1)

where Strategic factors ¼ ½location factors; ownership endowments, internalization variables, Financial factors ¼ ½internal capital market factors, internal corporate governance factors or more specifically, Financial factors ¼ ½exchange rate; internal and external financial profile,

risk and diversification; agency cost.

Eq. (1) is estimated for a subset as well as for the entire sample. The specific list of variables used in the empirical work is discussed in the next section. Whether a strategic model is sufficient to explain FDIs or whether financial variables are needed in addition is an empirical issue. We use the firm-level panel data to investigate different hypotheses in various ways. Panel data analysis could reveal more information over time and crosssectionally. However, a pooled regression of all panel data across all firms and all time periods horizontally, FDI it ¼ a þ bDit þ it (where D is a vector

30


of financial determinants, i indicates a firm, and t time), shows various degrees of autocorrelation. Hence, the results of pooled estimations are not reported in the interest of space. The identification of correct functional form, however, might resolve this problem. Based on the panel data analysis methodologies developed, we also tried alternative functional forms, which include fixed effect and random effect models, FDI it ¼ ai þ bDit þ it : These estimations assume a common slope but can take different assumptions on the distributions of the intercepts. However, heteroskedasticity is found to be present along with autocorrelation in some cases. To reduce these problems, we assume the stable relationship between the means of each explanatory variable for that is, we estimate FDI i ¼ a þ bDi þ i ; P individual firms,P where FDI i ¼ ðFDIit =TÞ; Di ¼ ðDit =TÞ; and T is the number of estimation periods.

4. DATA SOURCES AND DESCRIPTION OF VARIABLES Perhaps due to data convenience, the existing empirical work on the financial determinants of FDIs is typically based on aggregate national or industry data, such as common macroeconomic measures for the cost of capital or exchange rates (e.g., U.S. dollar per SDR for exchange rate and three-month T-Bill rate for cost of capital). To test our proposed financial hypothesis, firm-specific data are required, because the aggregate data can mask important differences that may exist at the level of firms. For instance, the effect of foreign exchange volatility on relative corporate wealth and cost of capital cannot be precisely measured, given the diversity of corporate international investment activities and dissimilarity of firm characteristics. This may help explain why the existing studies have mixed results. In our analysis, we use disaggregated firm-level data in corporate FDI, and explanatory variables such as firm-specific cost of capital as well as the exchange rate computed as a weighted average that reflects the extent and diversity of international investments by individual firms. We retrieve most of the data for sample firms from the Standard and Poor’s CompuStat Research Insight dataset. The criterion for inclusion in the sample is U.S. incorporations showing an ownership of foreign assets between 1992 and 1998.2 Table 2 shows descriptive characteristics for these firms. A total of 1,838 firms are selected, of which 1,131 (62%) are in manufacturing industries, and 1,551 (84%) and 1,014 (55%) had direct investments in DCs and less developed countries (LDCs), respectively. The


Table 2.

Descriptive Characteristics for the Sample of U.S. Corporations.

Total assets Foreign assets Foreign asset ratio (%) Total sales Foreign sales Foreign operating profits EBIT Earnings volatility (%) Beta Debt ratio (%) Free cash flow Cash reinvestment ratio (%) Agency cost (%) WACC (%)

31

Full Sample (N ¼ 1,838)

Manufacturing Industries (N ¼ 1,131)

Other Industries (N ¼ 621)

3307.00 (16360.00) 936.64 (5372.00) 24.43 (20.68) 2086.00 (8123.00) 659.60 (3704.00) 65.15 (287.00) 257.02 (1062.00) 9.45 (35.45) 0.64 (0.12) 21.55 (31.86) 130.09 (873.00) 4.03 (1.98) 38.79 (22.55) 11.32 (16.44)

2447.00 (13959.00) 810.72 (4548.00) 24.79 (19.27) 2114.00 (8768.00) 831.41 (4596.00) 79.99 (337.00) 228.77 (946.00) 9.25 (39.39) 0.57 (0.17) 20.49 (31.34) 142.26 (1026.00) 3.38 (1.48) 40.95 (20.21) 11.65 (18.89)

2232.00 (15741.00) 514.48 (5225.00) 24.21 (23.29) 1845.00 (6925.00) 333.17 (1039.00) 31.95 (153.00) 187.40 (1021.00) 9.61 (23.13) 1.14 (0.69) 21.64 (33.03) 79.57 (441.00) 5.14 (3.05) 36.85 (26.01) 10.99 (23.00)

Note: This descriptive statistics table shows means of each sample groups of U.S. corporations (standard deviations are in parentheses). Total assets, foreign assets, total sales, foreign sales, earnings before interest and taxes (EBIT), and free cash flow are in millions of U.S. dollars. Free cash flow comprises income before extraordinary items, depreciation, amortization, deferred taxes, equity in earnings of unconsolidated subsidiaries, extraordinary items and discontinued operations, and minority interest. Standard & Poor’s CompuStat Research Insight Database is the data source for all raw data that is the basis of computing other ratios. Foreign asset ratio is foreign assets divided by total assets. Earnings volatility is the standard deviation of EBIT scaled by the mean of total assets. Betas are computed by CompuStat. Debt ratio is long-term debts over total assets. Agency cost is measured by the ratio of operating expense to total sales. The weighted average cost of capital (WACC) is a weighted cost of equity (as measured by capital asset pricing model) and tax-adjusted cost of debt. Manufacturing industries comprise firms with SIC between 2,000 and 3,999. Other industries exclude these manufacturing firms as well as financial firms with SIC between 6,000 and 6,999.

32


computation of several explanatory variables requires certain macroeconomic data, such as exchange rates, risk-free rate, and gross domestic products. They are obtained from the International Monetary Fund’s International Financial Statistics dataset. Finally, geographic distance is obtained from World Almanac and World Atlas. Table 3 summarizes the definitions of all variables, their specific data sources and expected signs, along with a correlation matrix for financial variables. Beyond that, a brief discussion and some justification for the variables selected and their anticipated signs are in order. 4.1. Dependent Variable To analyze financial determinants of FDI, firm-level data representing FDI activities are essential. Another relevant issue is the use of FDI flows or FDI stocks. The distinction is crucial but the choice of FDI stocks is obvious: FDI flows reflect the sum invested in affiliates by foreign companies during a specific time period, which affiliates may spend to accumulate fixed and current assets, to repay past borrowings, or for some other objectives. However, the FDI stock represents the total value of assets attributable to the foreign investor. Thus, the FDI stock is an indicator for the value of assets engaged in international production. (Stephan & Pfaffmann, 2001).

Conventional FDI data collected by government agencies in each country (e.g., U.S. Department of Commerce) or international institutions (e.g., IMF, OECD, or UNCTAD) are available for sectors and regions with some degree of disaggregation, but not to the firm level. Fortunately, the availability of corporate foreign assets in the CompuStat dataset provides a direct measure of FDI stock. In this study, the dependent variable, FDIi, is foreign assets of firm i in natural logarithm form, which is consistent with existing FDI studies such as in Miller and Reuer (1998), where foreign assets are used as an explanatory variable in examining the effect of FDI on exchange exposure of the firm, and FDI in natural logarithm form serves its intended purposes in Buckley, Clegg, Forsans, and Reilly (2001). Independent variables include strategic and financial variables as well as control variables. 4.2. Strategic Variables The strategic framework for FDI tested in this paper follows the mainstream OLI model. As such, the selection of these variables is in accordance with empirical OLI studies. An obvious location factor is production cost (PC).

Definitions and Expected Signs of Variables and Correlation Matrix for Financial Variables.

Variable Name

Description

Expected Sign

Panel A: Summary of Variable Definitions and Expected Signs Dependent variable FDIi (foreign direct Foreign assets of firm i associated with foreign operations identified as ‘‘Identifiable investment) Assets’’ in the Geographic Segment of CompuStat, in natural logarithm Strategic variables PCi (production cost) TCi (transportation cost)

XARi (fixed asset ratio) RDRi (R&D ratio) FSRi (foreign sales ratio) IARi (intangible asset ratio) SERi (sales expense ratio) Financial variables WFXi,t (weighted foreign exchange composite)

Cost of goods sold divided by total sales of firm i Weighted average of geographic distance in miles (in natural logarithm): X FI j TC i ¼ ln Dj ; TF I i j

+ +

where FIj is foreign investment (foreign assets) in country j; TFIi,t is total foreign direct investment (total foreign assets) of firm i in a particular year; Dj is the distance in miles between U.S. and country j where the firm operates in the specific year Fixed assets scaled by total assets of firm i Research and development expense as a percent of total sales of firm i Foreign sales divided by total sales of firm i Intangible assets divided by total assets of firm i

+ + + +

Sales expense divided by total expense of firm i

+

Weighted multilateral foreign exchange rate change for firm i in year t: X FI j;t ej;t WF X i;t ¼ ln ; TF I t ej;t1 j

33

where FIj is foreign assets in country j; TFIt is total foreign direct investment stock of firm i in year t; ej,t (home currency per foreign currency) is the annual bilateral foreign exchange rate between U.S. and country j in which the firm operates in year t


Table 3.

34

Table 3. (Continued ) Variable Name IFi (internal financing)

ACEi (agency cost expense ratio) ACAi (agency cost asset utilization ratio) Control variables Si (firm size) CONi (concentration ratio) USW (U.S. wealth) HCWi (host country wealth)

Free cash flow of firm i including income before extraordinary items, depreciation, amortization, deferred taxes, equity in earnings of unconsolidated subsidiaries, extraordinary items and discontinued operations, and minority interest Weighted average cost of equity and tax-adjusted cost of debt of firm i The number of geographic segments in which firm i operates Long-term debt ratio representing firm i’s capital structure The standard deviation of firm i’s earnings before interests and taxes (EBIT) scaled by the mean of its total assets to avoid size bias Operating expense over total sales of firm i

Expected Sign +

+ +/ /+ +

Total sales divided by total assets of firm i

Total sales of firm i in natural logarithm Total sales of four largest firms in the industry of firm i divided by total sales of the industry The gross domestic products (GDP) of the U.S. in natural logarithm Weighted average GDP (in natural logarithm) of all host countries in which the firm operates: X FI j HCW i ¼ ln GDPj ; TF I i j

+

where FIj is the foreign assets in country j. TFIt is the level of total foreign direct investment of firm i in the particular year


EFi (external financing) IDi (international diversification) FRi (financial risk) BRi (business risk)

Description

IFi

EFi

IDi

FRi

BRi

ACEi

ACAi

Panel B: Correlation Matrix for Financial and Control Variables Financial Variables IFi 0.06 1.00 (o.01) EFi 0.02 0.01 1.00 (0.09) (0.42) IDi 0.06 0.20 0.01 1.00 (o.01) (o.01) (0.21) FRi 0.03 0.21 0.04 0.06 1.00 (0.02) (o.01) (o.01) (o.01) BRi o.01 0.25 0.01 0.01 0.04 (0.91) (o.01) (0.90) (0.42) (o.01) ACEi o.01 0.02 0.01 0.02 0.03 (0.99) (0.13) (0.90) (0.06) (o.01) ACAi 0.02 0.19 0.04 0.04 0.11 (0.09) (o.01) (o.01) (0.01) (o.01)

0.03 (o.01) 0.01 (0.43)

0.05 (o.01)

1.00

Control Variables Si 0.09 (o.01) CONi 0.03 (0.02)

0.09 (o.01) 0.01 (0.47)

0.13 (o.01) 0.02 (0.05)

0.10 (o.01) 0.04 (o.01)

0.75 (o.01) 0.06 (o.01)

0.02 (0.02) 0.01 (0.88)

0.31 (o.01) 0.06 (o.01)

0.27 (o.01) 0.03 (o.01)

Si

CONi

1.00 1.00

1.00 0.05 (o.01)

1.00

USW


WFXi

35

USW HCWi

36

Table 3. (Continued ) WFXi

IFi

EFi

IDi

FRi

BRi

ACEi

ACAi

Si

CONi

USW

0.15 (o.01) 0.30 (o.01)

0.12 (o.01) 0.04 (o.01)

0.01 (0.70) 0.01 (0.29)

0.11 (o.01) 0.03 (0.02)

0.01 (0.36) 0.06 (o.01)

0.03 (o.01) 0.01 (0.84)

0.01 (0.23) 0.04 (o.01)

0.03 (o.01) 0.05 (o.01)

0.07 (o.01) 0.05 (o.01)

0.26 (o.01) 0.10 (o.01)

1.00 0.11 (o.01)


Note: Panel B shows Pearson correlation coefficients between variables. In parentheses is the p value under H0: r ¼ 0, and o.01 indicates the p value is less than 0.01. Data sources: FDI is available from the CompuStat Geographic Segment database. CompuStat defines Identifiable Assets in a geographic segment or country as the tangible and intangible assets that are used by, or directly associated with, each geographic segment or country. The seven geographic segments as classified by CompuStat are Africa, Asia, Europe, Pacific, South America, North America, and Other Foreign Operations. CompuStat only identifies 10 countries with heaviest U.S. foreign investments, South Africa, Japan, Philippines, Great Britain, France, Germany, Australia, Brazil, Canada, and Mexico. To compute TC, geographic distance in miles is obtained from World Almanac and World Atlas. However, if some of a firm’s foreign operation is identified only in a geographic segment or only as foreign, for that portion of investment, the average distance in miles to that region or the total average miles is used. ID is available from the CompuStat Geographic Segment database. CompuStat reports the actual number of geographic segments disclosed by the firms. To compute WFX, bilateral exchange rates are obtained from International Financial Statistics (IFS) dataset. However, if some of a firm’s foreign operation is identified only in a geographic segment or only as foreign, for that portion of investment, the average exchange rate change in that segment or the rate of USD per SDR is used, respectively. To compute USW and HCW, GDPs are obtained from IFS dataset by IMF. In calculating HCW, however, if some of a firm’s foreign operation is identified only in a geographic segment or only as foreign, for that portion of investment, the average GDP of the countries in that region or the average world GDP is used. To calculate all other strategic, financial, and control variables, the accounting data are retrieved from CompuStat dataset.


37

Some empirical work in location analysis (e.g., Culem, 1988; Kravis & Lipsey, 1982; Maki & Meredith, 1986) uses wage rates or unit labor cost. However, several researchers, such as Huang (1997), argue that the labor cost may not be an accurate measure of location cost since the proportion of labor cost in the total production cost has been decreasing. Another location factor, transportation cost (TC), is frequently measured in empirical work (e.g., Grosse & Trevino, 1996) by geographic distance. As for ownership factors, the fixed asset ratio (XAR) is in line with the notion that corporations with higher operating leverage tend to have a greater ownership advantage in expanding their foreign production due to an oligopolistic edge such as economies of scale. Quite common in related work (e.g., Dunning, 1980; Pugel, Kragas, & Kimura, 1996), research and development expense ratio (RDR) measures a firm’s ownership advantage through its investment in knowledge production and technology know-how, which give the firm a competitive edge in the global market. Internalization variables comprise foreign sales ratio (FSR), measuring the degree of foreign involvement; intangible asset ratio (IAR), measuring the level of information-based assets; and sales expense ratio (SER), measuring operating efficiency. Internalization gains through FDIs derive from a higher level of informationbased assets, a greater degree of foreign involvement, and improving operational efficiency. IAR gauges the need for internalization, because internalization reduces loss of efficiency and uncertainty in the valuation of intangible or information-based assets given the market imperfection for such assets in the external markets.3 FSR represents a corporation’s degree of foreign involvement in many empirical studies (e.g., Click & Harrison, 2000; Duru & Reeb, 2001).4 Overall, conventional FDI theories, OLI in particular, would suggest positive effects on FDI and thus positive expected signs for these strategic variables. More favorable location, ownership, and internalization conditions motivate firms to step up their cross-border direct investments. 4.3. Financial Variables ‘‘Internal’’ capital market factors include foreign exchange rate, free cash flow, and cost of capital. In particular, the weighted foreign exchange composite (WFX) takes into account different FDI compositions of firms, and a negative sign for its coefficient is anticipated. Host country currency depreciation encourages FDI due to stronger internal capital markets for MNCs. Financing method variables include those that capture the profile of internal financing (IF) and external financing (EF). Consistent with existing

38


literature in corporate finance, free cash flow represents corporate wealth or internal financing capacity. We expect a positive sign for it, because the buildup of free cash flow (stronger internal capital market) offers firms an edge in the international bidding scenario. We estimate cost of capital for each firm and a negative sign is anticipated since, if internal funding is insufficient, a higher cost of external financing deters foreign investments.5 ‘‘Internal’’ corporate governance factors are related to managerial objectives or agency behaviors that lead to cross-border expansion and diversification. They include international diversification (ID), an overall measure for global diversification; financial risk (FR) and business risk (BR), measures of intended diversification for risk reduction; and expense ratio (ACE) and asset utilization ratio (ACA), measures for agency costs. International diversification (ID) measures a firm’s intention to spread its operations globally and thereby achieve risk-reduction benefits of a more diverse operating cash flow. It should be noted that even when diversification is not a primary objective, any international investment—whether intended or not—has a diversification dimension. We measure the extent of international diversification by the number of geographic segments in which a firm has FDIs (foreign assets). Other than country count, it is most frequently used in geographic diversification studies (e.g., Bodnar, Tang, & Weintrop, 1999; Duru & Reeb, 2001). Overall, there should exist a positive relationship between international diversification and FDI. More specifically, the intent of global diversification is to reduce risk in either business or financial dimension. As is common in the finance literature, we use longterm debt ratio, representing the firm’s capital structure, as a measure of financial risk. Business risk is measured by earnings volatility, the standard deviation of a firm’s earnings before interests and taxes but scaled by the mean of its total assets to avoid size bias. Given a certain level of total risk a firm can or will assume, it is unclear whether firms are more interested in diversifying financial or business risk. However, it appears that the goal of reducing a particular type of risk through FDI commands a positive sign for its coefficient. Furthermore, since they work as substitutes, the other one would have a negative coefficient. Regardless of its significance, the type of risk with a negative sign may be simply irrelevant in a firm’s diversification decision because the negativity is a consequence of the firm assuming a certain level of total risk. As for agency cost, a commonly used proxy for agency cost in the finance literature (e.g., Ang, Cole, & Lin, 2000) is the operating expense ratio (ACE), measuring how well the management controls operating costs including expenditure on excessive perquisites and


39

other direct agency costs.6 Another agency cost measure is the asset utilization ratio (ACA) measuring how well firms deploy and utilize their assets. Generally, low asset utilization ratio, or high expense ratio, indicates the presence of a more severe agency problem. If private benefits encourage managers to diversify their firms globally, ACE is expected to be positive and ACA negative. In the interest of parsimony, however, ACE is our primary measure of agency cost in estimation. These financial variables are selected in accordance with existing empirical literature as well as econometric considerations. However, the effect of agency behavior on FDI has not been tested at all in existing literature. Furthermore, a comprehensive analysis of financial determinants of FDIs based on firm-level data has not yet been done. It is imperative to verify that the collection of financial variables in our estimating framework is appropriate in a certain statistical sense. Panel B of Table 3 shows the correlation matrix for financial variables. The null hypothesis of no correlation cannot be rejected at any significance level for most of the pairs. For those few rejected, the magnitude of correlation is only about 0.02 or less. 4.4. Control Variables We use control variables at different levels: firm size (S) at the firm level, concentration ratio (CON) at the industry level, and the U.S. (USW) and host country wealth (HCW) at the country level. These variables take into account a firm’s resources, competition in the industry, and wealth of home and host countries. Given different patterns of global asset deployment among firms, the computation of certain variables, including WFX, TC, and HCW, is necessary to reflect each firm’s distinctive foreign investment activities. Their weighting schemes are identical as detailed in Table 3.

5. EMPIRICAL RESULTS 5.1. Strategic Factors Table 4 shows the estimation results of the strategic model. Most of the strategic variables in the OLI model are statistically significant and have the expected signs. For the full sample, except for sales expense ratio, all variables are significant at the 1% level and have the expected signs.7 Thus, overall the strategic motives for FDI, as indicated by the OLI model, appear

40

Table 4.

Strategic Determinants of Foreign Direct Investments FDIi ¼ a þ b1 PCi þ b2 TCi þ b3 XARi þ b4 RDRi þ b5 FSRi þ b6 IARi þ b7 SERi þ b8 Si þ b9 CONi þ b10 USW þ b11 HCWi þ i


C (intercept) Strategic variables Location variables PC (production cost) TC (transportation cost) Ownership variables XAR (fixed asset ratio) RDR (R&D ratio)

5.63 (2.39)

Manufacturing Industries (N ¼ 1,131) 2.20 (0.88)

Other Industries (N ¼ 621) 17.67 (2.94)

DCs (N ¼ 1,551) 10.02 (3.70)

LDCs (N ¼ 1,014)

2.63 (0.55)

0.18 (9.43) 0.34 (5.74)

0.17 (8.67) 0.27 (4.17)

0.53 (2.90) 0.55 (4.07)

0.21 (9.85) 0.00004 (2.57)

0.01 (4.20) 0.0001 (2.53)

0.73 (4.36) 0.07 (3.92)

0.71 (3.80) 0.15 (2.91)

0.66 (1.74) 0.04 (1.55)

1.03 (7.04)

1.27 (5.22)


The dependent variable, foreign direct investment (FDI), is corporate foreign asset in natural logarithm. Corporate foreign asset is defined as identifiable asset (including both tangible and intangible assets) used by or directly associated with foreign operations outside the U.S.

IAR (intangible asset ratio) SER (sales expense ratio) Control variables S (firm size) CON (four-firm concentration ratio) USW (U.S. GDP) HCW (weighted host country GDPs) Adjusted R2 LM heteroskedasticity test (H0: homoskedasticity)

2.82 (27.76) 0.96 (5.33) 0.29 (1.66)

3.06 (26.59) 1.11 (5.82) 0.21 (1.09)

2.21 (10.07) 0.14 (0.30) 0.23 (0.56)

2.52 (21.19) 0.96 (4.99)

1.88 (9.46) 0.16 (0.44)

0.98 (81.19) 0.002

0.99 (71.24) 0.005

0.94 (35.24) 0.008

0.99 (76.32) 0.004

0.93 (44.77) 0.001

(0.98) 0.04 (0.17) 0.04 (1.71)

(1.72) 0.28 (1.04) 0.05 (1.74)

(1.95) 1.14 (1.71) 0.008 (0.13)

(2.18) 0.70 (2.31) 0.11 (2.99)

(0.35) 0.32 (0.06) 0.17 (3.96)

0.91 4.94a

0.92 5.20a

0.87 0.41c

0.83 2.52c

0.73 3.82b

Note: LM heteroskedasticity test follows w2 distribution with one degree of freedom. Manufacturing industries comprise firms with SIC between 2,000 and 3,999. Other industries exclude these manufacturing firms as well as financial firms with SIC between 6,000 and 6,999. DCs and LDCs refer to foreign direct investments in developed and less developed countries, respectively. The classification of developed and developing countries is according to United Nations Conference on Trade and Development. Significant at 1% level. Significant at 5% level. Significant at 10% level. The t-statistics are in parentheses. a The null hypothesis of homoskedasticity is rejected at the 5% level, but not rejected at the 1% level. b The null hypothesis of homoskedasticity is rejected at the l0% level, but not rejected at the 5% level. c The null hypothesis of homoskedasticity is not rejected at any of the conventional 10%, 5%, or 1% level.


Internalization variables FSR (foreign sales ratio)

41

42


to be strong during the sample period of 1992–1998. Similarly, all OLI variables are significant for manufacturing industries at the 1% level except for sales expense ratio, with additional significance shown for three control variables (firm size, concentration ratio, and host country GDP). In contrast, other industries (non-manufacturing/non-financing) have their FDI most directly related to location variables. Two out of three internalization variables (intangible asset ratio and sales expense ratio) and one ownership variable (R&D ratio) are insignificant for the other industry category. In addition, the other ownership variable is only marginally significant. This makes sense because compared to manufacturing firms, non-manufacturing firms do not produce products that are heavily embedded with firm-specific ownership assets. They possess less ownership advantage and thus have a lesser need for internalization. It appears that location is the major concern in their FDI decisions. Table 4 further shows the behavioral pattern of FDI by U.S. firms depending on destinations, DCs versus LDCs. There does not exist distinctive patterns between FDI in DCs and in LDCs.8 Again, conventional FDI theories, internalization included, fail to explain the phenomenon that most FDIs occur between DCs. 5.2. Financial Factors Table 5 on the financial model of FDI indicates the general validity of the model. The results for the full sample show all financial variables having the expected signs at a high significance level except for foreign exchange, which is significant for the manufacturing industries and investments in DCs. However, compared to the strategic model, there is a greater variation in the results depending on industry sector and host country. The overall results, nevertheless, support our proposition that FDI is the consequence of the interaction between a stronger internal capital market and internal corporate governance. In sum, home currency appreciation, accumulation of corporate free cash flow, managerial objectives for risk reduction, and agency behavior, collectively, are the driving forces for FDI. Models with various combinations of variables, as shown in Table 6, are also estimated on the full sample and two conclusions may be drawn. First, except for foreign exchange, the results including signs and significance for the variables vary little across different equations. Clearly, as indicated in the previous correlation matrix, these variables do not interact with each other much and our results are relatively robust. Second, the foreign exchange variable has a much greater variability in its significance across the equations but always maintains the correct sign. It seems that the effect of

Financial Determinants of Foreign Direct Investments FDIi ¼ a þ b1 WFXi þ b2 IFi þ b3 EFi þ b4 IDi þ b5 FRi þ b6 BRi þ b7 ACEi þ b8 ACAi þ b9 Si þ b10 CONi þ b11 USW þ b12 HCWi þ i

The dependent variable, foreign direct investment (FDI), is corporate foreign asset in natural logarithm form. Corporate foreign asset is defined as identifiable asset used by or directly associated with foreign operations outside the U.S. Full Sample (N ¼ 1,838)

C (intercept)

0.82 (0.28)


DCs (N ¼ 1,551)

LDCs (N ¼ 1,014)

2.53 (0.39)

7.47 (2.18)

1.91 (0.35)

0.47 (1.61) 0.13 (5.89) 0.17 (2.40)

0.52 (1.92) 0.07 (2.83) 0.14 (2.04)

0.93 (0.96) 0.19 (3.86) 0.05 (0.22)

2.02 (3.75) 0.12 (4.62) 0.02 (0.14)

0.09 (0.28) 0.13 (3.32) 0.30 (2.60)

0.43 (10.39)

0.42 (9.41)

0.50 (5.80)

0.18 (3.61)

0.32 (4.25)

43

Financial variables ‘‘Internal’’ capital market WFX (FDI-weighted foreign exchange composite) IF (internal financing: corporate free cash flow) EF (external financing: cost of capital) ‘‘Internal’’ corporate governance ID (international diversification)

3.27 (1.16)



Table 5.

44

Table 5. (Continued )

FR (financial risk: capital structure) BR (business risk: earnings volatility) ACE (agency cost: expense ratio)

CON (four-firm concentration ratio) USW (U.S. GDP) HCW (weighted host country GDP) Adjusted R2 LM heteroskedasticity test (H0: homoskedasticity)



DCs (N ¼ 1,551)

LDCs (N ¼ 1,014)

0.55 (4.81) 0.92 (2.14) 0.07 (4.67)

0.35 (2.99) 0.66 (1.47) 0.09 (6.37)

0.53 (2.02) 0.63 (0.71) 0.29 (1.28)

0.69 (5.10) 1.03 (2.12) 0.10 (5.73)

0.24 (1.12) 1.65 (1.93) 0.19 (0.84)

0.85 (33.28) 0.0004 (0.16) 0.08 (0.27) 0.08 (2.28)

0.94 (34.31) 0.01 (3.76) 0.37 (1.13) 0.06 (1.44)

0.68 (12.14) 0.0004 (0.09) 0.07 (0.09) 0.08 (1.00)

0.85 (27.64) 0.003 (0.94) 0.37 (0.97) 0.32 (6.04)

0.96 (20.49) 0.006 (1.29) 0.22 (0.36) 0.26 (5.19)

0.79 5.32a

0.84 5.32a

0.68 4.45a

0.73 2.30b

0.69 5.29a

Note: LM heteroskedasticity test follows w2 distribution with one degree of freedom. Manufacturing industries comprise firms with SIC between 2,000 and 3,999. Other industries exclude these manufacturing firms as well as financial firms with SIC between 6,000 and 6,999. DCs and LDCs refer to foreign direct investments in developed and less developed countries, respectively. The classification of developed and developing countries is according to United Nations Conference on Trade and Development. Significant at 1% level. Significant at 5% level. Significant at 10% level. The t-statistics are in parentheses. a The null hypothesis of homoskedasticity is rejected at the 5% level, but not rejected at the 1% level. b The null hypothesis of homoskedasticity is not rejected at any of the 10%, 5%, or 1% conventional level.


Control variables S (firm size)


Further Testing on Financial Determinants of Foreign Direct Investments FDIi ¼ a þ b1 WFXi þ b2 IFi þ b3 EFi þ b4 IDi þ b5 FRi þ b6 BRi þ b7 ACEi þ b8 ACAi þ b9 Si þ b10 CONi þ b11 USW þ b12 HCWi þ i

The dependent variable, foreign direct investment (FDI), is corporate foreign asset in natural logarithm form. Corporate foreign asset is defined as identifiable asset used by or directly associated with foreign operations outside the U.S. Full Sample (N ¼ 1,838) (1) C (intercept)

0.43

3.27 (1.16)

(1.83)

0.47 (1.61) 0.13 (5.89) 0.17 (2.40)

1.30 (3.45) 0.62 (28.94) 0.24 (2.43)

0.43 (10.39)

0.73 (13.09)

(3)

(4)

(5)

(6)

(7)

(9.13)

2.83 (0.96)

6.11 (2.21)

0.03 (1.67) 0.03 (0.51)

0.69 (2.52) 0.13 (5.89) 0.16 (2.20)

0.46 (1.59) 0.13 (5.92) 0.17 (2.40)

0.58 (1.90) 0.12 (5.38) 0.20 (2.61)

0.66 (2.32) 0.13 (6.09) 0.06 (0.93)

0.39 (10.83)

0.43 (10.27)

0.43 (10.40)

0.77 (0.31)

2.99 (17.37)

2.50

0.43 (10.88)

45

Financial variables ‘‘Internal’’ capital market WFX (FDI-weighted foreign exchange composite) IF (internal financing: corporate free cash flow) EF (external financing: cost of capital) ‘‘Internal’’ corporate governance ID (international diversification)

(2)


Table 6.

46

Table 6. (Continued ) Full Sample (N ¼ 1,838) (1) FR (financial risk: capital structure) BR (business risk: earnings volatility) ACE (agency cost: expense ratio) ACA (agency cost: asset utilization)

CON (four-firm concentration ratio) USW (U.S. GDP) HCW (weighted host country GDP) Adjusted R2 LM heteroskedasticity test (H0: homoskedasticity)

1.08

(4.81) 0.92 (2.14) 0.07 (4.67)

(6.95) 3.91 (6.91) 0.24 (3.79)

0.85 (33.28) 0.0004 (0.16) 0.08 (0.27) 0.08 (2.28) 0.79 5.32a

0.59 0.51b

(3)

(4)

0.22

0.56

0.55

0.50

(4.95) 1.00 (2.34) 0.07 (4.69)

(4.85) 0.92 (2.15) 0.07 (4.67)

(4.16) 0.77 (1.72) 0.09 (5.40)

0.96 (43.17) 0.001 (0.53) 0.11 (0.42) 0.09 (3.26)

0.85 (33.36)

0.85 (33.39)

0.08 (2.29)

0.91 (34.99) 0.0003 (0.11) 0.16 (0.49) 0.06 (1.61)

0.87 (36.09) 0.001 (0.59) 0.38 (1.25) 0.09 (2.72)

0.83 6.56a

0.79 6.31a

0.79 5.36a

0.77 1.28b

0.79 4.24a

(2.34) 0.06 (4.40) 0.69 (17.68)

(5)

Note: LM heteroskedasticity test follows w2 distribution with one degree of freedom.

Significant at 1% level. Significant at 5% level. Significant at 10% level. The t-statistics are in parentheses. a

The null hypothesis of homoskedasticity is rejected at the 5% level, but not rejected at the 1% level. The null hypothesis of homoskedasticity is not rejected at any of the 10%, 5%, or 1% conventional level.

b

(6)

(7) 0.49 (4.80) 0.07 (4.34)


Control variables S (firm size)

(2)

0.55


47

foreign exchange on FDI is consistent but its importance is lessened with the presence of some factors. To have a more comprehensive discussion on the proposition in this paper, we shall discuss the results of each financial factor individually both on the full sample and across different industry sectors and destinations. 5.2.1. Foreign Exchange The impact of the foreign exchange effect on FDI is via home currency appreciation, creating a stronger internal capital market for MNCs. In full sample estimations, the FDI-weighted foreign exchange composite variable has statistically significant coefficients with negative signs in most cases, meaning that a depreciation in the foreign currency or an appreciation in the U.S. dollar results in an increase in outward FDI by U.S. firms. This is also true for firms in the manufacturing industry, but the effect is insignificant for the ‘‘other’’ industries. These results are consistent with the view of Blonigen (1997), who argues that the exchange rate effect depends on firm-specific assets, which are more pronounced in manufacturing firms than in nonmanufacturing firms. In the context of acquisition, firm-specific assets can generate returns in currencies other than that used for purchase, and thus subject the acquiring firm to exchange risk. Hence, the exchange rate effect on FDI is more significant for manufacturing than for non-manufacturing firms. It is also interesting that there are significant differences in exchange rate impacts depending on destinations. The results for FDI in DCs are the same as those for manufacturing industry, i.e., outward FDI by U.S. firms is positively impacted by a strong dollar. However, the exchange rate is not a significant factor for FDI in LDCs. This is attributable to the fact that many currencies of the LDCs are effectively fixed or managed by the government. That is, there is little variability in observed data and hence little observed exchange rate impact. However, it does not mean that there is no risk or no risk management problem in developing countries, only that the risk is postponed and resolved ultimately by an eruption of crises or crashes as happened in the Asian financial crisis of 1997. In addition, DCs have more firms with advanced technologies or information-based assets. In the case of international acquisitions, firm-specific assets can be relevant to the effect of foreign exchange. Another point worth mentioning for the full sample estimations in Table 6 is that some factors seem to overshadow foreign exchange. It appears that, without controlling for firm size, foreign exchange is a very important factor (Eq. (2)). However, after controlling for firm size and host

48


country GDP and with the presence of international diversification and business risk, the exchange rate is not a factor for FDI (Eqs. (1) and (5)). One possible explanation for its lesser importance for large firms is that they tend to have more resource and expertise in managing foreign exchange risk and may need less help from home currency appreciation to improve their internal capital markets. Furthermore, when large firms consider various factors simultaneously (e.g., international diversification, business risk, and host country wealth), foreign exchange seems to take a back seat in the FDI decision. These two reasons, foreign exchange being overshadowed by some other factors and the link between firm-specific assets and foreign exchange, may explain why the results of foreign exchange in the existing FDI studies are mixed. 5.2.2. Internal and External Financing Internal financing capacity shows the strength of a corporation’s internal capital market, but external financing becomes relevant when internal funding is insufficient. Internal financing or corporate wealth is usually measured in empirical work by corporate free cash flow. With an increase in free cash flow, a firm is likely to conduct more FDI, especially if it runs out of domestic investment opportunities. If a firm requires external financing, typically measured by its cost of capital, the higher rate undoubtedly hinders its ability to make cross-border investments. Both these financing effects, positive for internal and negative for external financing, are confirmed for the full sample, but internal financing tends to be more significant than external financing. These results are consistent with the view of Froot and Stein (1991) that internal financing is more important than external financing in international acquisitions. The internal financing variable is significantly positive both across all industry sectors and regardless of destinations. The effects of external financing are more selective. The coefficients are significant for manufacturing industries but insignificant for ‘‘other’’ industries. Utility firms (which are included in the other category) would raise equity capital to support initial investments, and given relatively stable income, their need for external financing might be relatively small on an ongoing basis. Manufacturing firms, in contrast, emphasize growth and the requirement of a higher degree of operating leverage could mean less free cash flow during the expansion stage. As a result, external financing is relatively more important.9 In addition, the internal financing variable is significant for FDI in both DCs and LDCs. However, the coefficient of external financing is significant


49

only in LDCs. This appears to be related to country risk. Given the complexity of FDI in LDCs, more funding, management expertise, experience in international business, and resource base for risk management may be required.10 It is entirely likely that firms with FDIs in LDCs must also rely on external financing. Another possibility, although further investigation is necessary, is the corporate concern for country risk or mere government regulation. As a means of minimizing their country risk exposure, multinational firms may prefer partial local financing for their investments in LDCs. In addition, some host country governments may limit the percentage of ownership by foreign investors. 5.2.3. International Diversification and Risk Reduction International diversification has both intended and unintended dimensions. The intended global diversification, in general, is regarded as the consequence of risk reduction in that operations are spread out over various countries or regions. The international diversification variable is highly significant and has the expected positive sign across the board – for the full sample, for all industry sectors, and for both DCs and LDCs. Most international investments, as it seems, are motivated by the desire to diversify globally for financial or business risk-reduction purposes. Of course, it is also possible that, for some investments, such a result is unintended. That is, even if diversification does not figure prominently in terms of motivation, these international investments, nonetheless, achieve a degree of geographical diversification. The underlying goal, however, could have a serious effect if managers are set to pursue their own personal gains at the expense of shareholders, which is to be discussed in the next section. In Table 6, financial risk and business risk variables are both significant and have exactly opposite signs as expected in all the equations in the full sample estimation. This is due to the fact that, to an extent, both risks work as substitutes. For a given amount of total risk a firm is willing or able to take, a firm with lower business risk can afford to increase its financial risk, or vice versa. However, since financial risk has a positive coefficient with greater significance, it appears that the target for risk reduction, if intended, is financial risk. Furthermore, for both manufacturing and other industries, financial risk is significant but business risk is not. Another plausible interpretation of this result is that risk diversification may apply to financial risk rather than business risk. This interpretation is consistent with the finding of Burgman (1996), who shows that international diversification does not necessarily lower earnings volatility – a measure of business risk – for MNCs.

50


Furthermore, financial risk is relevant only for FDIs in DCs but not for those in LDCs. Apparently, firms consider DCs the better destination for their financial risk diversification. This is not surprising because most international acquisitions are financed by cash or equity swaps. DCs have well-developed capital markets that facilitate fair equity valuation. The notion that diversifying financial risk through investments in DCs would lower MNCs’ debt ratio over time is consistent with results of Lee and Kwok (1988) and Fatemi (1988), who suggest that MNCs have a lower debt ratio than domestic firms. 5.2.4. Agency Cost Agency problem could lead to unintended global diversification as well as potential firm value destruction. Of the two measures for agency cost, the expense ratio is the primary proxy in the interest of parsimony. However, in the full sample estimation, positive expense ratio and negative asset utilization ratio coefficients point in the same direction, a positive link between agency costs and FDI – more severe agency problems leading to higher FDI. Given the strong significant and consistent results on the full sample, it is not at all difficult to conclude that FDI is also driven by agency behavior – the pursuit of managers’ private benefits. This is obviously consistent with the findings in recent diversification discount studies. For instance, Click and Harrison (2000) establish the inverse relationship between multinationality and managerial equity ownership, and Duru and Reeb (2001) find managers diversify their firms in part due to their private benefits. It is also in line with the finding of Lee and Kwok (1988) that MNCs tend to have higher agency cost.11 The grouping results in Table 5 indicate that, after controlling for firm size, the agency problem exists only for manufacturing industries and for FDI in DCs. One plausible explanation is that the agency problem is more severe for corporations with more firm-specific or information-based assets. Given the relative abundance of these kinds of assets in manufacturing firms, in particular, such firms face more complex potential agency conflicts. These assets also make corporations less transparent and create more difficulty for corporate governance and monitoring. This explanation is in line with the finding in Harris and Ravenscraft (1991) that cross-border takeovers are more likely to take place in R&D intensive industries (more likely manufacturing). The result on DCs can be similarly understood. DCs have more advanced manufacturing firms (with greater firm-specific and information-based assets) than LDCs. Firms with such assets are preferable targets for managers of acquiring firms who pursue their own personal gains.


51

Overall, the financial factors appear to be important in FDI decisions in addition to strategic factors. Given imperfections and information asymmetry in international capital markets, corporate FDIs are influenced by such financial variables related to internal capital markets and internal corporate governance. However, relative to strategic variables, there appears to be greater variability in the level of significance across industry sectors and destinations of investments for financial variables. In addition, all these financial variables are significant in the full sample estimations, but financial considerations on FDI appear to be more critical for manufacturing than for other industries and more essential for FDI in developed than in LDCs.12 5.3. An Integrated Model To put the relative importance of strategic and financial factors into perspective, we now combine these two sets of variables, in addition to control variables, in an integrated empirical model. Estimation results of an integrated model generally preserve the previous results obtained in each of the two models, especially for the full sample and FDI in DCs and LDCs.13 On the full sample, Table 7 shows that financial factors are as important as strategic factors for FDI by U.S. firms, with virtually all the variables significant. The table shows the same result – an equal importance of financial and strategic factors – for manufacturing. The performance of the integrated model declines a bit for other industries, but here strategic variables appear to be much more important than financial variables. The integrated model performs better for U.S. FDIs in DCs than for those in LDCs. However, there is no clear evidence in terms of the relative importance of strategic and financial variables in either region. We also calculated Schwartz’s Bayesian criterion for the three representative models: strategic, financial, and integrated.14 The Bayesian criterion should be as small as possible because the addition of an independent variable would raise this score if there is no incremental explanatory power. By this criterion, the strategic model performs better than the financial model for other industries. For the manufacturing industries, both models perform just about equally. The integrated model works best for the full sample. We did not pursue this line of inquiry in depth in this paper because, technically, it is contingent on the selection of a representative strategic or financial model. We are content, in this paper, with establishing the basic point that financial variables are necessary in addition to strategic variables in understanding FDI decisions fully, and leave the exact determination of their relative contribution for future research.

52

Table 7.


FDIi ¼ a þ b1 WFXi þ b2 IFi þ b3 EFi þ b4 IDi þ b5 FRi þ b6 BRi þ b7 ACEi þ b8 PCi þ b9 TGCDi þ b10 XARi þ b11 RDRi þ b12 FSRi þ b13 IARi þ b14 SERi þ b15 ACAUi þ b16 Si þ b17 CONi þ b18 USW þ b19 HCWi þ i


C (intercept) Financial variables WFX (weighted foreign exchange) internal capital market variable IF (internal financing) internal capital market variable EF (external financing) – internal capital market variable ID (international diversification) – internal corp. governance var. FR (financial risk) – internal corporate governance variable

4.32 (1.88) 0.19 (0.63) 0.07 (3.85) 0.24 (3.99) 0.14 (3.90) 0.24 (2.49)

Manufacturing Industries (N ¼ 1,131) 0.83 (0.35) 0.09 (0.31) 0.07 (3.52) 0.33 (5.68) 0.16 (4.15) 0.24 (2.48)


DCs (N ¼ 1,551)

LDCs (N ¼ 1,014)

13.16 (2.40)

1.71 (0.61)

4.37 (0.82)

0.96 (1.16) 0.08 (1.82) 0.02 (0.14) 0.08 (1.07) 0.25 (1.06)

0.23 (0.48) 0.05 (2.22) 0.12 (1.37) 0.12 (2.75) 0.24 (2.02)

0.10 (0.28) 0.04 (0.86) 0.45 (3.50) 0.60 (7.72) 0.08 (0.40)


The dependent variable, FDI, is corporate foreign asset in natural logarithm. Corporate foreign asset is identifiable asset, tangible or intangible, used by or directly associated with foreign operations occurring outside the U.S.

0.80 (2.35) 0.12 (4.16)

0.62 (1.75) 0.03 (0.95)

0.91 (1.23) 0.29 (1.06)

1.65 (4.13) 0.23 (6.40)

1.78 (1.92) 0.02 (0.04)

Strategic variables PC (production cost) – location variable TC (transportation cost) – location variable XAR (fixed asset ratio) – ownership variable FSR (foreign sales ratio) – internalization variable IAR (intangible asset ratio) – internalization variable SER (sales expense ratio) – internalization variable

0.27 (2.74) 0.20 (3.71) 1.59 (10.28) 2.90 (24.90) 0.83 (4.24) 0.22 (1.34)

0.06 (0.57) 0.19 (3.36) 0.97 (5.44) 2.96 (23.29) 0.97 (4.66) 0.38 (2.00)

1.17 (2.94) 0.31 (2.54) 2.02 (5.90) 2.89 (12.22) 0.47 (1.10) 0.15 (0.43)

0.57 (4.98) 0.12 (1.85) 1.35 (6.99) 2.91 (20.70) 1.33 (5.85) 0.45 (2.19)

1.34 (2.61) 0.32 (2.21) 2.14 (5.64) 2.59 (9.87) 0.14 (0.30) 0.48 (1.20)

0.89 (39.93) 0.002 (1.00) 0.02 (0.06) 0.02 (0.71)

0.91 (38.44) 0.01 (2.72) 0.53 (2.07) 0.07 (2.07)

0.89 (16.97) 0.003 (0.92) 0.82 (1.37) 0.09 (1.34)

0.90 (33.57) 0.003 (1.06) 0.12 (0.39) 0.01 (0.21)

0.99 (20.72) 0.002 (0.36) 0.28 (0.48) 0.21 (4.00)

0.88

0.92

0.84

0.77

Control variables S (firm size) CON (four-firm concentration ratio) USW (U.S. GDP) HCW (weighted host country GDP) Adjusted R2

0.82


BR (business risk) – internal corporate governance variable ACE (agency cost: expense ratio) – internal corp. governance var.

53

54

Table 7. (Continued )

LM heteroskedasticity test (H0: homoskedasticity)




DCs (N ¼ 1,551)

LDCs (N ¼ 1,014)

5.10a

1.12c

3.82b

0.41c

6.40a


Note: LM heteroskedasticity test follows w2 distribution with one degree of freedom. Other industries exclude manufacturing and financial firms. DCs and LDCs refer to FDIs in developed and less developed countries, respectively. Significant at 1% level. Significant at 5% level. Significant at 10% level. The t-statistics are in parentheses. a The null hypothesis of homoskedasticity is rejected at the 5% level, but not rejected at the 1% level. b The null hypothesis of homoskedasticity is rejected at the l0% level, but not rejected at the 5% or 1% level. c The null hypothesis of homoskedasticity is not rejected at any of the conventional 10%, 5%, or 1% level.


55

6. CONCLUDING REMARKS The mainstream FDI literature emphasizes strategic factors. This almost exclusive emphasis on strategic factors does not seem to be justified in practice. We introduce financial factors in addition to strategic factors as determinants of FDI. The estimation based on 1,838 U.S. firms for the period from1992 to 1998 indicates that the traditional OLI strategic factors are in fact significant in explaining FDI behavior in full sample estimations. The results do not differ much across destinations, although evidence suggests that strategic considerations have been more important for manufacturing than other industries. At the same time, evidence strongly indicates the importance of financial factors as well. The relevance of financial variables in FDI decisions is based on the assumption of imperfections in financial markets. Hence, the financial model of FDI shares the same assumption – market imperfections of some kind – with the mainstream FDI literature in the industrial organization framework. Estimations show that almost all financial variables identified – foreign exchange rate, profile of internal and external financing, risk and diversification variables, and agency costs – are significant for the full sample, for firms in manufacturing industries, and for investments in DCs. In general, these results support the basic notion underlying our financial model that FDI is the consequence of interaction between a stronger internal capital market and internal corporate governance. International diversification and internal financing variables, in particular, are important in all estimations regardless of destinations or industries. FDI in other industries is dependent only on the firm’s financial risk in addition to these two variables. The fact that FDI made by manufacturing industries (but not other industries) is significantly affected by foreign exchange appears to be related to the presence of firm-specific assets. Beyond the two common financial determinants – international diversification and internal financing – an additional list of financial variables is significant for FDI in DCs: foreign exchange, financial risk, and agency cost. For LDCs, only external financing is relevant in addition to internal financing and diversification. This reflects a different degree of development of capital markets and country risk in developed and developing countries. If one has to choose, there is a sense in which, overall, the strategic factors are the slightly more important considerations in FDI decisions. However, we also show that financial factors are important in FDI decisions in addition to strategic factors. Empirically, our financial model works quite well

56


by itself or in conjunction with the OLI factors. Given imperfections and information asymmetry in international capital markets, generally, FDIs are influenced by such financial variables as foreign exchange, internal and external financing, risk and diversification, and agency problems of the concerned parties. However, relative to strategic variables, there appears to be greater variability in the level of significance of financial variables across industry sectors and across investment destinations. Most importantly, however, our financial model is better than the mainstream OLI framework in one sense – in that it is able to explain the prevailing FDI phenomena when OLI factors fail. First of all, why is the most popular FDI entry mode consistently and predominantly M&As rather than greenfield investments? It could be explained by the strength of the internal capital market and the consequence of internal corporate governance. Host country currency depreciation or corporate free cash flow buildup leads to a stronger internal capital market. It favors foreign acquisitions because the edge from currency movement could be short-lived and there is no reason not to use up the excess capacity from free cash flow quick, thus eliminating the inefficiency. Typically, greenfield investment is a longer-term project to plan, deploy, and complete. Agency behavior also favors foreign acquisitions simply because it is easier and quicker for managers to gauge and materialize their potential gains from acquiring an established firm than by building one from scratch. Second, why most FDIs occur between DCs can similarly be explained from financial perspectives. A stronger internal capital market facilitated by host country currency devaluation is more likely to happen in DCs than in LDCs because the foreign exchange rate oftentimes is fixed or managed by LDC governments. International acquisitions driven by agency behavior also favor DCs, where there resides a larger set of ideal targets for maximizing managers’ personal benefits. Another relevant point is that the popularity of M&A as the primary entry mode implies that DCs are the ideal destinations, as M&A is more likely to occur in a country having well-developed capital markets and legal systems. In addition, our evidence also shows that DCs are the destinations for firms to diversify their financial risk through foreign investments. Finally, financial factors are logical candidates to explain the often-observed short-term bilateral FDI swings between two countries. After all, financial variables such as foreign exchange, free cash flow, and cost of capital could vary significantly in a short period of time. Ever changing managerial objectives regarding global diversification and agency behavior in pursuing private benefits also assure that FDI is not likely to be a constant toward a particular country for an extended period of time.


57

NOTES 1. For instance, in 1998, 91.6% of all FDI outflows were from developed countries and 71.5% of all FDI inflows were to developed countries. Source: World Investment Report, 1999, United Nations Conference on Trade and Development (UNCTAD). 2. Foreign assets are defined as identifiable assets, which include both tangible and intangible assets, used by or directly associated with operations abroad by U.S. firms. 3. Intangible assets in CompuStat dataset include a variety of patents, rights and agreements, licenses, goodwill, designs and drawings, franchises, quotas, subscription and client lists, and trademark. 4. One legitimate concern here is the potential problem of simultaneity. Granted, foreign sales are generated through both domestic and foreign productions. However, internalization focuses on the need for MNCs to switch from export (domestic production) to FDI (foreign production) to improve efficiency. If the sole purpose of FDI is to boost foreign sales, total sales go up too. FSR does not necessarily change significantly, especially if the ratio is relatively large. Therefore, simultaneity may be of little concern because the variable is measured by ratio rather than level. 5. The cost of equity is estimated by the capital asset pricing model, where the market risk beta is estimated for each firm based on the monthly data for the previous 60-month period, which is reported by CompuStat. The weights are based on the market value of equity and the book value of debt. 6. In comparison, the agency cost in Myers (1977) is advertising and R&D expenses divided by total sales. We did not use this because it blurs the distinction with internalization or ownership advantage. R&D expenses are often used as a measure of internalization or ownership factor. 7. Although the intercept is significant, it is negative, which is simply an indication that the starting point of FDI is relatively small given the dependent variable in natural logarithm form. The same goes for other equations with significant and negative intercepts. 8. Choi, Kim, and Chandran (1995) provide a summary of literature regarding ‘‘conventional’’ and ‘‘non-conventional’’ FDIs by multinationals from developed and developing countries, respectively. 9. As Table 2 shows, manufacturing industry has a lower cash reinvestment ratio – roughly 60% of that for ‘‘other’’ industry. 10. Not surprisingly, the significance of external financing on FDIs made by manufacturing firms and in LDCs is consistent with firm size effect (S). The size effects are greater for both groups than their counterparts. 11. To entertain classic agency theories, the agency problem related to FDIs is likely to be a substitution problem due to foreign exchange risk and country risk involved in international investments. If a foreign project is riskier than a comparable domestic project, the agency problem is more of the substitution problem as illustrated in Jensen and Meckling (1977) rather than underinvestment problem as posited in Myers (1978). The substitution problem describes the tendency of the management to substitute a riskier for a safer project for the interest of certain parties (in this case, their own).

58


12. Incidentally, the firm size has a positive impact on FDI for the full sample and across all industry sectors and destinations. However, there is evidence that the absolute magnitudes of the size coefficients are somewhat larger for manufacturing than for other industries, indicating the scale of resources necessary for FDI. The size effects are greater for LDCs than for DCs, suggesting a notion that operations in LDCs generally require a larger resource base for risk management and experience in international business. 13. A couple of different variables does lose their significance for manufacturing and other industries and weighted foreign exchange also becomes insignificant for manufacturing firms and FDIs in DCs. However, this is not considered severe, given that all other results are consistent with previous estimations and that foreign exchange is overshadowed by firm size and few other variables. 14. Schwartz’s Bayesian criterion is believed to be superior to Akaike’s information criterion because it is asymptotically consistent for large sample (see Wei, 1990).

REFERENCES Aggarwal, R. K., & Samwick, A. A. (2003). Why do managers diversify their firms? Agency reconsidered. Journal of Finance, 58, 1613–1649. Aliber, R. A. (1970). A theory of foreign direct investment. In: C. Kindleberger (Ed.), The international corporation. Cambridge: MIT Press. Ang, J. S., Cole, R. A., & Lin, J. W. (2000). Agency costs and ownership structure. Journal of Finance, 55, 81–106. Black, F. (1978). The ins and outs of foreign investment. Financial Analysts Journal, 34, 1–7. Blonigen, B. A. (1997). Firm-specific assets and the link between exchange rates and foreign direct investment. American Economic Review, 87, 447–465. Bodnar, G. M., Tang, C., & Weintrop, J. (1999). Both sides of corporate diversification: The value impacts of geographic and industrial diversification. Working Paper. Buckley, P. J., & Casson, M. C. (1976). The future of multinational enterprise. London: Macmillan. Buckley, P. J., Clegg, J., Forsans, N., & Reilly, K. T. (2001). Increasing the size of the ‘‘country’’: Regional economic integration and foreign direct investment in a globalised world economy. Management International Review, 41, 251–274. Burgman, T. A. (1996). An empirical examination of multinational corporate capital structure. Journal of International Business Studies, 27, 553–570. Casson, M. (1979). Alternatives to the multinational enterprise. London: Macmillan. Caves, R. E. (1971). International corporations: The industrial economics of foreign investment. Economica, 38, 1–27. Caves, R. E. (1989). Exchange-rate movements and foreign direct investment in the United States. In: D. B. Audretsch & M. P. Claudon (Eds), The internationalization of U.S. markets. New York: New York University Press. Choi, J. J. (1989). Diversification, exchange risk, and corporate international investment. Journal of International Business Studies, 20, 145–155. Choi, J. J., & Jeon, B. N. (2006). Financial factors in foreign direct investments: A dynamic analysis of international data. Research in International Business and Finance, (In Press, Corrected Proof).


59

Choi, J. J., Kim, K., & Chandran, R. (1995). Foreign direct investments by firms from developing and developed countries: Stylized facts and theoretical interpretations. In: Y. Kim & K. Oh (Eds), The US–Korea economic partnership. London: Avebury. Click, R. W., & Harrison, P. (2000). Does multinationality matter? Evidence of value destruction in U.S. multinational corporations. Working Paper. Culem, C. G. (1988). The locational determinants of direct investments among industrialized countries. European Economic Review, 32, 885–904. Cushman, D. O. (1985). Real exchange rate risk, expectations, and the level of direct investment. Review of Economics and Statistics, 67, 297–308. Dewenter, K. (1995). Do exchange rates drive foreign direct investment? Journal of Business, 68, 405–433. Dunning, J. H. (1977). Trade, location of economic activity and the MNE: A search for an eclectic approach. In: B. Ohlin (Ed.), The international allocation of economic activity. London: Macmillan. Dunning, J. H. (1980). Toward an eclectic theory of international production: Some empirical tests. Journal of International Business Studies, 11, 9–31. Dunning, J. H. (1988). The eclectic paradigm of international production: A restatement and some possible extensions. Journal of International Business Studies, 19, 1–31. Duru, A., & Reeb, D. M. (2001). Evidence on the relation between geographic diversification and firm value. Working Paper. Fatemi, A. M. (1988). The effect of international diversification on corporate financing policy. Journal of Business Research, 16, 17–30. Froot, K. A., & Stein, J. C. (1991). Exchange rates and foreign direct investment. Quarterly Journal of Economics, 106, 1191–1217. Grosse, R., & Trevino, L. J. (1996). Foreign direct investment in the United States: An analysis by country of origin. Journal of International Business Studies, 27, 139–155. Harris, R. S., & Ravenscraft, D. (1991). The role of acquisitions in foreign direct investment: Evidence from the U.S. stock market. Journal of Finance,, 46, 825–844. Healy, P. M., & Palepu, K. G. (1993). International corporate equity associations: Who, where, and why? In: K. A. Froot (Ed.), Foreign direct investment. Chicago, IL: University of Chicago Press. Huang, G. (1997). Determinants of United States–Japanese foreign direct investment: A comparison across countries and industries. New York: Garland. Hymer, S. H. (1976). The international operations of national firms: A study of direct foreign investment. Cambridge, MA: MIT Press. Itaki, M. (1991). A critical assessment of the eclectic theory of the multinational enterprise. Journal of International Business Studies, 22, 445–460. Jensen, M. C., & Meckling, W. H. (1976). Theory of firm managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Kindleberger, C. P. (1969). American business abroad; six lectures on direct investment. New Haven, CT: Yale University Press. Klein, M. W., & Rosegren, E. S. (1994). The real exchange rate and foreign direct investment in the United States: Relative wealth vs. relative wage effects. Journal of International Economics,, 36, 373–389. Kravis, I. B., & Lipsey, R. E. (1982). The location of overseas production and production for export by US multinational firms. Journal of International Economics, 12, 201–223.

60


Lee, K., & Kwok, C. (1988). Multinational corporations vs. domestic corporations’ international environmental factors and determinants of capital structure. Journal of International Business Studies,, 19, 195–217. Maki, D. R., & Meredith, L. N. (1986). Production cost differentials and foreign direct investment: A test of two models. Applied Economics, 18, 1127–1134. Miller, K. D., & Reuer, J. J. (1998). Firm strategy and economic exposure to foreign exchange rate movements. Journal of International Business Studies, 29, 493–514. Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175. Oxelheim, L., Stonehill, A., & Randoy, T. (2001). On the treatment of finance specific factors within the OLI paradigm. International Business Review, 10, 381–398. Pugel, T. A., Kragas, E. S., & Kimura, Y. (1996). Further evidence on Japanese direct investment in U.S. manufacturing. Review of Economics and Statistics, 78, 208–213. Ray, E. J. (1989). The determinants of foreign direct investment in the United States, 1979–85. In: R. C. Feenstra (Ed.), Trade policies for international competitiveness. Chicago, IL: University of Chicago Press. Stephan, M., & Pfaffmann, E. (2001). Detecting the pitfalls of data on foreign direct investment: Scope and limit s of FDI data. Management International Review, 41, 189–218. Stevens, G. V. G. (1998). Exchange rates and foreign direct investment: A note. Journal of Policy Modeling, 20, 393–401. Stulz, R. (1983). On the determinants of net foreign investment. Journal of Finance, 38, 459–468. Swenson, D. L. (1994). Impact of U.S. tax reform on foreign direct investment in the United States. Journal of Public Economics, 54, 243–266. Wei, W. W. S. (1990). Time series analysis: Univariate and multivariate methods. Redwood City, CA: Addison-Wesley.

MODELING THE EVOLUTIONARY SEQUENCE OF INTERNATIONAL JOINT VENTURES Elmar Lukas ABSTRACT This paper contributes to the literature on foreign direct investment under uncertainty and underscores the importance of modeling the evolutionary sequence pattern of foreign market entry. The model presented helps to refine the application of real options in the international joint venture context by providing a closed-form solution in continuous time to value their overall strategic flexibility. Moreover, the analysis provides a novel perspective on existing empirical results and generates a number of new testable predictions.

1. INTRODUCTION As a result of an increasingly volatile economic environment, one of the main changes in the nature of multinational enterprises’ (MNEs) operations is related to the institutional choice of achieving access to foreign markets. Unlike identifying situations where a MNE can achieve majority control,

Value Creation in Multinational Enterprise International Finance Review, Volume 7, 61–73 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1569-3767/doi:10.1016/S1569-3767(06)07003-8

61

62

ELMAR LUKAS

globalization forces have affected the need for firms to cooperate. Due to increased competition and the rapid pace of technological development, inter-company collaboration is now crucial to the pursuit of competitive advantage. As a consequence, the number of international joint ventures (IJVs), both horizontal (between competing companies) and vertical (between companies within a given value chain), has risen dramatically during the past two decades, especially when firms operate in highly uncertain industries (e.g., Broll & Marjit, 2005; Dunning & Narula, 2004; Todeva & Knoke, 2005). The increase is, however, not only driven by achieving access to markets but by seeking foreign tangible or intangible assets as well. Hence, sourcing of complementary assets internationally favors the choice of IJVs and acts as an ever-growing driving force for international collaboration in general (e.g., Hagedoorn, 2002). A joint venture (JV) is a legal organization formed by two or more parties to undertake jointly economic activity for mutual benefit. One can broadly distinguish between two basic organizational modes: equity JV and nonequity JV. The former is created when each partner has an equity share in the new venture. Non-equity JVs, in contrast, are agreements to cooperate in some way, but they do not involve the creation of new firms.1 A JV becomes international if at least one partner has its headquarter outside the venture’s country of operation or if the venture has a significant level of operation in more than one country (Geringer & Hebert, 1991). The search for determinants that drive the use and success of IJVs has a long tradition in the business and economics literature.2 The research conducted was primarily empirically in nature (e.g., Blodgett, 1992; Gatignon & Anderson, 1988; Hennart, Kim, & Zeng, 1998; Kogut & Chang, 1996; Leung, 1997). While being extensively analyzed empirically, less effort, however, has been made in scrutinizing the properties of IJVs through rigorous theoretical modeling (e.g., Buckley & Casson, 1996; Marjit, Broll, & Mallick, 1995). Further, all attempts do not account for the fact that a firm’s commitment to invest into a new market is associated with sunk costs that cannot be recovered once the project is initiated. Moreover, foreign direct investment (FDI) decisions are to a large portion investment decisions under uncertainty and are only the first commitment of subsequent expansion. Hence, with respect to the initial switching decision, i.e., whether to abandon export or not, one has additionally to consider the impetus of subsequent expansion (Gilroy & Lukas, 2006). Another criticism stems results from the fact that these models only consider the unidirectional case. Thus, they lack explanation of divestment or strategic reorientation (e.g., Buckley & Tse, 1996; Rivoli & Salorio, 1996).

Modeling the Evolutionary Sequence of International Joint Ventures

63

In the last decade, researchers have highlighted the importance of a more dynamic perspective in FDI theory. Consequently, real options theory has gained significant attention in the international business field because it not only captures dynamic aspects of decision theory but also represents an advance in conceptualizing and measuring the value of strategy under uncertainty (e.g., Chi & Seth, 2002). In brief, real options theory suggests viewing real investments as options buying the firm the rights such as to make investments later, the right to defer or alter scale or to initiate subsequent investments.3 Besides others, Buckley and Casson (1998) have drawn attention to this by arguing that the existing models do value FDI decisions only with respect to their immediate effects rather than in terms of possible new investment opportunities. In other words, in most cases initial foreign production serves as a platform for expansion abroad, indicating that the initial investments carry a high option value due to possible new investment opportunities, e.g., regional technology platforms (e.g., Chang & Rosenzweig, 2001; Howells & Wood, 1993; Kogut & Chang, 1996). It is clear that this fact is most obvious for IJVs and it is Kogut (1991) who puts this thought further. Possible project interdependencies within the IJV allow for strategic flexibility, calling for an interpretation of IJVs as platform investments. Thus, although unprofitable from a stand-alone perspective, the value of a JV can be much higher due to the flexibility to acquire later stakes of the venture in the future. Consequently, the termination of an IJV does not indicate its failure but the exploitation of its flexibility. Based on a two-stage binomial model, Chi and McGuire (1996) model a situation in which the MNE has the option to acquire or sell out the partner’s stake in an equity JV.4 They depict that both options create economic value for the partners, especially if the partners foresee different valuation expectations of the venture ex post and in the presence of transactions costs. By treating two sources of uncertainty explicitly in their model, the authors were also able to show how the options serve to diminish the risk of misappropriation and thus alleviate the difficulty of JV contracting under information asymmetry. Besides the theoretical analysis, the authors also present a number of testable hypotheses related to their work. In a more advanced model setting, Pennings and Sleuwaegen (2004) design an option model where both the timing of market entry and the entry mode are determined simultaneously. The switch from export, whether to a wholly owned subsidiary, a JV, or to licensing, is dependent on uncertainty of payoffs, cost structure, competitive stance of incumbents, tax differences, and the degree of cooperation between the JV partners. In particular, the timing of a JV is related to transfer prices, amount of equity share, market

64

ELMAR LUKAS

structure, and to the degree of governmental regulation. However, possible subsequent actions were neglected. While the option idea has become a building block for empirical research on IJVs lately (e.g., Folta & Miller, 2002; Reuer & Leiblein, 2000; Reuer & Tong, 2005), in-depth research is still lacking in the international business literature with respect to the consequent modeling of the evolutionary sequence pattern of an IJV. Thus, the goal of this paper is to model a JV induced market entry under uncertainty in a continuous time setting. The rest of the paper is structured as follows. In Section 2, we will present the model: a two-phase market entry sequence. The main results are presented in Section 3, which also provides a synopsis of major comparative-static results. Finally, Section 4 summarizes the main findings and provides suggestions for further research.

2. THE MODEL This paper focuses on a representative equity-based JV between two private firms. Both firms combine complementary resources comprising firmspecific knowledge, which is either related to technology or marketing expertise or both. For the sake of simplicity, it is assumed that only a subset of overall knowledge is shared, however, in an amount that secures the agreed objective, such as R&D collaboration. Moreover, it is assumed that only one firm is a foreigner to the new market, namely the MNE, which has chosen a local partner in the host country. The choice of which entry strategy an enterprise chooses has no influence on the profit rates of other enterprises in the foreign market. Moreover, the value of the chosen FDI mode v(t) is ex ante unknown and follows a geometric Brownian motion. Assuming a perfect capital market, the existence of a unique martingale measure Q can be used to modify the stochastic differential equation, which results in dv=v ¼ ðr dÞ dt þ s dZ Q

(1) 2

where rd is the growth rate of the project value, s designates the variance of dv/v, r is the risk-free interest rate, d represents the opportunity cost of waiting, and dZQ indicates a Wiener process with non-zero drift. Let e refer to the initial equity stake the MNE investor has invested in and ¯ be the country-specific upper boundary for owning equity stakes held by foreign investors in a given country.5 Consequently, this initial strategy generates two exclusive strategic options. However, it is worthwhile to


65

consider a time span in which the partners become acquainted and can check if joint work is possible for the sake of the venture. We account for this by assuming that the MNE has a certain time period [0,t1] in which it can decide how to continue with its market entry strategy. At the end of this period of length T, the MNE can decide whether it prefers to continue collaboration with the host partner by receiving the right to convert the IJV into a cross-border merger and acquisition (M&A), i.e., by acquiring the remaining shares ð¯ Þ at a later date. Counterbalancing this, the MNE might prefer the right to dissolve the IJV over the growth option by selling its own interest e to the local partner at a later date.6 Let us denote the optimal threshold separating both strategies with c. The option value F, the optimal trigger points vU, vL (representing the actual timing of the subsequent investment/divestment), and c may be solved recursively. From Dixit and Pindyck (1994) as well as Merton (1973) the results for a perpetual investment option and a perpetual divestment option, respectively, are commonly known. Thus, they are just summarized briefly. Under the assumption of a perpetual time to maturity and aforementioned boundary conditions the solution for a perpetual call option results in ( Avb1 vovU CðvÞ ¼ (2) ð¯ Þv I v vU with I as the corresponding cost for acquiring the rest of the equity stake ð¯ Þ; ð1b1 Þ ð¯ Þ 1 b1 I A¼ and b1 ð¯ Þ b1 1 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 ðr dÞ ðr dÞ 1 2 2r þ þ 2 b1 ¼ 2 s2 s2 2 s as constants.7 From this, the optimal trigger value vU for the M&A strategy can be deduced, which results in vU ¼

1 b1 I ð¯ Þ b1 1

(3)

On the other hand, if the MNE decides to dissolve the IJV it will receive a perpetual put option. To be more precise, on exercising the second stage the MNE gives up an existing project with value ev and receives its abandonment value k (see, e.g., Chi, 2000). The value of this strategic option is thus

66

ELMAR LUKAS

given by

( PðvÞ ¼

Bvb2 k v

v4vL v vL

(4)

whereby s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 b2 k 1b2 1 ðr dÞ ðr dÞ 1 2 2r and b2 ¼ þ 2 B¼ b2 ðb2 1Þ 2 s2 s2 2 s are again constants. The corresponding optimal threshold value vL for initiating a divestment strategy results in vL ¼

b2 k b2 1

(5)

Consequently, the value of the chooser option is determined by F ¼ erT E Q ½maxfPðvÞ; CðvÞg

(6)

Q

with E [y] as the expectations operator under the martingale measure Q. This results in solving the following integral: 2 v ZL Zc rT 4 F ¼e ðk vÞ dPðvÞ þ Bvb2 dPðvÞ vL

1

Zvu þ

Avb1 dPðvÞ þ

c

Z1

3

_ 7 ðð Þv IÞ dPðvÞ5

ð7Þ

vu

where dP(v) denotes the implied probability measure. To derive a closedform solution for the complex chooser option one has to determine the aforementioned optimal threshold c. Thus, c is determined by the intersection of P(c) and C(c). From Acb1 ¼ Bcb2 ; we get !1=g 1b b1 vL 2 c¼ (8) 1 ð¯ Þ b2 v1b U with g ¼ b1 b2 : Solving Eq. (7) results in F ¼ kerT Nðd 3 Þ vedT Nðd 4 Þ þ Bvb2 Nðd 7 Þ Bvb2 Nðd 8 Þ þ Avb1 Nðd 5 Þ Avb1 Nðd 6 Þ þ ð¯ ÞvedT Nðd 1 Þ IerT Nðd 2 Þ

ð9Þ


67

with v as the value of the overall IJV at time 0, N(y) as the cumulative normal distribution and v 1 v 1 ln ln þ r d þ s2 T þ r d s2 T vU 2 vU 2 pffiffiffiffi pffiffiffiffi d1 ¼ ; d2 ¼ s T s T ln d3 ¼

v 1 L r d s2 T 2 v pffiffiffiffi ; s T

b1 ln d5 ¼

v U

v

1 r þ b21 s2 T 2 pffiffiffiffi ; sb1 T

c 1 2 2 b2 ln r þ b2 s T v 2 pffiffiffiffi d7 ¼ ; sb2 T

v 1 L ln r d þ s2 T 2 v pffiffiffiffi d4 ¼ s T b1 ln d6 ¼

c 1 r þ b21 s2 T v 2 pffiffiffiffi sb1 T

v 1 2 2 L b2 ln r þ b2 s T 2 v pffiffiffiffi d8 ¼ sb2 T

Given the derived results, we can state that the MNE will initiate its foreign expansion strategy via the establishment of an IJV if the expanded net present value, i.e., the net present value rule adjusted by the value of subsequent flexibility exceeds the associated costs of entry. Thus, the MNE’s timing decision results in v þ F ðvÞ K 0

(10)

where K comprises the costs for setting up such a market entry strategy and ev designates the initial value of the IJV.

3. RESULTS This section presents a summarization of the results and the comparativestatic analysis. If not noted specifically, we will assume the following values: r ¼ 0.055, s ¼ 0.4, d ¼ 0.01, and k ¼ 0.8. The value of the IJV’s flexibility F is composed of the option value to dissolve the IJV (i.e., the first to fourth terms) and the growth option value, which reflects the value of the subsequent cross-border M&A strategy (i.e., the remaining terms). As the result indicates, F increases with uncertainty, abandonment value, and time to

68

ELMAR LUKAS

0.85 0.80 0.75

F(v)

0.70 0.65 0.60 0.55 0.0 0.1 0.2 0.3 0.4 T 0.5 0.6 0.7 0.8 0.9 1.0

Fig. 1.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

v(t)

Value of Combined Acquisition/Divestment Option F With Respect to v(t) and Time of Joint Collaboration T.

maturity T ¼ t1. In addition, the value of the IJV’s flexibility decreases with high initial equity shares, i.e., the lower the initial equity share the more valuable is the option, and higher acquisition costs for the remaining shares. Fig. 1 summarizes some of the results graphically. The comparative-static results for the trigger values vL, and vU are wellknown from the standard literature.8 The threshold value vU becomes larger and so does the propensity to wait when turning the IJV into a merger; the higher the costs of acquiring the remaining shares I are, the smaller b1 is. Moreover, an increase in involved aggregate investment uncertainty leads to an increase in vU. In addition, the trigger value is also dependent on the size of equity share e. If the MNE already holds a majority in the IJV, 1=ð¯ Þ becomes significantly large, thus indicating an increased propensity to wait before acquiring the remaining shares (i.e., higher threshold value). Due to the concavity of vU(s), the effect of e and I is more significant when the aggregate uncertainty of the overall project is higher. The opposite can be observed for the trigger value of the divestment stage. Low uncertainties correlate with a high threshold value. Due to the dependence of b2 on s, vL decreases as uncertainty increases, indicating that the propensity to wait also increases. This effect is further amplified, the lower the initial equity share e


69

or the higher the recovery value k is. However, due to the convexity of vL(s), the effect of k and e is more significant the lower the aggregate uncertainty of the overall project is. The chooser option is a path-dependent derivative. Hence, implications about the kind of termination the MNE chooses at time t1 can only be made in conjunction with the threshold c. As noted earlier, at t1 the MNE chooses that strategy, giving maximum return option value according to maxfCðvÞ; PðvÞg: Consequently, if vt1 is greater than c, the MNE will stick to its current strategy and further collaborate until the above-mentioned threshold vU is reached, turning it into a merger. If vt1 is lower than c, the MNE will further collaborate while at the same time preferring to dissolve the IJV. From the results derived, c shows now two different trends with respect to its dependence on project uncertainty. If the MNE holds a majority in the IJV, the threshold increases the higher the aggregate uncertainty is. Consequently, with increased uncertainty there is a perceived trend toward sell out because the MNE demands a higher project value for compensating the associated risks accompanied with a merger strategy. For minority IJVs, however, c is inversely dependent on project uncertainty. Thus, the chance for a subsequent merger is even greater the higher the project uncertainty becomes. Furthermore, given the fact that c is (not) reached, the propensity to initiate the investment (divestment) is even faster the lower the uncertainty s is (i.e., because only small upward (downward) movements of v(t) are needed to hit the corresponding threshold value). Both trends are dampened by a decrease in recovery value k. Fig. 2 depicts graphically the dependence of c on uncertainty and equity stake. In most cases, host governments are concerned about the actual amount invested in their country. While waiting for new information generally improves the overall risk exposure of the MNE in such a setting, i.e., given sunk costs and uncertainty, value for the host country is only generated if the firm exercises its real options right. Treating the initial decision of whether to start the IJV or not as an option right, the firm will invest as soon as Eq. (10) is satisfied. Because F(v) is always positive, the generated flexibility can offset the transaction cost comprised by K, which is, e.g., the case for highly uncertain projects (see, e.g., Sanchez, 1993).

4. SUMMARY The expansion of MNEs is a path-dependent process that is reflected in the fact that the observed internationalization processes of MNEs are not only a

70

ELMAR LUKAS

2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.0

1.0 0.9 0.8 0.1

0.7 0.6

0.2 0.3

0.5

0.4

0.4

0.5 0.6 0.7

Fig. 2.

0.3

Critical Threshold Value c(r,d,s,e,I,k) for the IJV Termination Strategy With Respect to Uncertainty s and Equity Stake ð¯ ¼ 1Þ:

unidirectional paths. Therefore, strategic reorientation, divestment, or withdrawal must also be considered as serious strategies. The results show the new complementary insight that the evolutionary sequence of market entry via IJVs is not only driven by the growth option, as commonly modeled in the literature, but also by the flexibility to dissolve the JV. While it has been commonly agreed that IJVs are a transitional form of foreign market expansion, less emphasis has been placed on what triggers the choice of termination form. Consequently, the model provides a solution that allows revealing which kind of termination is chosen by the MNE given the initial equity stake in the venture and uncertainty. Moreover, implications for governmental policies to attract FDI can be deduced from the model. The study presented provides new opportunities for further empirical research under an option framework or to extend the model by introducing, e.g., tax differentials.

NOTES 1. See Glaister, Husan, and Buckley (1998). However, in non-equity JVs sharing or exchange of equity may occur between partners (see ibid, p. 170).


71

2. See, e.g., Gilroy (1993). 3. A detailed introduction to real options is given by Dixit and Pindyck (1994) and Trigeorgis (1998). 4. Although not explicitly stated, this model can be interpreted as a chooser option, albeit only two discrete states are considered. 5. Thus, the venture will be an equity joint venture JV if 0:05oo¯ (see, e.g., Gomes-Casseres, 1987). 6. The last step may be justified because a subsequent innovation renders an existing partner’s technology obsolete or due to misappropriation risk. Consequently, the venture is abandoned for the sake of a new venture or for withdrawal from the foreign market. 7. It is assumed that the acquisition price is fixed right from the start. For a justification of this assumption refer to, e.g., Chi and McGuire (1996). 8. See, e.g., Dixit and Pindyck (1994).

ACKNOWLEDGMENTS The author would like to thank Udo Broll, B. Michael Gilroy, and participants of 4th Korea and the World Economy Conference in Seattle for helpful comments and discussions. Any remaining errors are the sole responsibility of the author.

REFERENCES Blodgett, L. L. (1992). Factors in the instability of international joint ventures: An event history analysis. Strategic Management Journal, 13(6), 475–481. Broll, U., & Marjit, S. (2005). Foreign investment and the role of joint ventures. South African Journal of Economics, 73(3), 474–481. Buckley, A., & Tse, K. (1996). Real operating options and foreign direct investment: A synthetic approach. European Management Journal, 14(3), 304–314. Buckley, P. J., & Casson, M. C. (1996). An economic model of international joint venture strategy. Journal of International Business Studies, 27, 849–876. Buckley, P. J., & Casson, M. C. (1998). Models of the multinational enterprise. Journal of International Business Studies, 29(1), 21–44. Chang, S. J., & Rosenzweig, P. (2001). The choice of entry mode in sequential foreign direct investment. Strategic Management Journal, 22(8), 747–776. Chi, T. (2000). Option to acquire or divest a joint venture. Strategic Management Journal, 21(6), 665–687. Chi, T., & McGuire, D. J. (1996). Collaborative ventures and value of learning: Integrating the transaction cost and strategic option perspectives on the choice of market entry modes. Journal of International Business Studies, 27(2), 285–307.

72

ELMAR LUKAS

Chi, T., & Seth, A. (2002). Joint ventures through a real options lens. In: F. J. Contractor & P. Lorange (Eds), Cooperative strategies and alliances: What we know 15 years later (pp. 71–87). Amsterdam: Elsevier. Dixit, A. K., & Pindyck, R. S. (1994). Investment under uncertainty. Princeton, NJ: Princeton University Press. Dunning, J. H., & Narula, R. (2004). Multinationals and industrial competitiveness: A new agenda. Cheltenham, UK: Edward Elgar. Folta, T. B., & Miller, K. D. (2002). Real options in equity partnerships. Strategic Management Journal, 23(1), 77–88. Gatignon, H., & Anderson, E. (1988). The multinational corporation’s degree of control over foreign subsidiaries: An empirical test of a transaction cost explanation. Journal of Law, Economics and Organization, 4, 305–336. Geringer, M., & Hebert, L. (1991). Measuring performance in international joint ventures. Journal of International Business Studies, 22(2), 249–263. Gilroy, B. M. (1993). Networking in multinational enterprises: The importance of strategic alliances. Columbia: University of South Carolina Press. Gilroy, B. M., & Lukas, E. (2006). The choice between greenfield investment and cross-border acquisition: A real option approach. Quarterly Review of Economics and Finance, 46(3), 447–465. Glaister, K. W., Husan, R., & Buckley, P. J. (1998). UK international joint ventures with the triad: Evidence for the 1990s. British Journal of Management, 9(3), 169–180. Gomes-Casseres, B. (1987). Joint venture instability: Is it a problem? Columbia Journal of World Business, 22(2), 97–102. Hagedoorn, J. (2002). Inter-firm R&D partnerships: An overview of major trends and patterns since the 1960. Research Policy, 31(4), 477–492. Hennart, J.-F., Kim, D.-J., & Zeng, M. (1998). The impact of joint ventures on the longevity of Japanese stakes in the U.S. manufacturing affiliates. Organization Science, 9(3), 382–395. Howells, J., & Wood, M. (1993). The globalisation of production and technology. London: Belhaven Press. Kogut, B. (1991). Joint ventures and the option to expand and acquire. Management Science, 37(1), 19–33. Kogut, B., & Chang, S. J. (1996). Platform investments and volatile exchange rates: Direct investment in the U.S. by Japanese electronic companies. Review of Economics and Statistics, 78(2), 221–231. Leung, W. F. (1997). The duration of international joint ventures and foreign wholly-owned subsidiaries. Applied Economics, 29(10), 1255–1269. Marjit, S., Broll, U., & Mallick, I. (1995). A theory of overseas joint ventures. Economics Letters, 47(3–4), 367–370. Merton, R. C. (1973). Theory of rational option pricing. Bell Journal of Economics and Management Science, 4, 141–183. Pennings, E., & Sleuwaegen, L. (2004). The choice and timing of foreign market entry under uncertainty. Economic Modelling, 21(6), 1101–1115. Reuer, J. J., & Leiblein, M. J. (2000). Downside risk implications of multinationality and international joint ventures. Academy of Management Journal, 43(2), 203–214. Reuer, J. J., & Tong, T. (2005). Real options in international joint ventures. Journal of Management, 31(3), 403–423.


73

Rivoli, P., & Salorio, E. (1996). Foreign direct investment and investment under uncertainty. Journal of International Business Studies, 27, 335–357. Sanchez, R. (1993). Strategic flexibility, firm organization, and managerial work in dynamic markets: A strategic options perspective. Advances in Strategic Management, 9, 251–291. Todeva, E., & Knoke, D. (2005). Strategic alliances and models of collaboration. Management Decision, 43(1), 123–148. Trigeorgis, L. (1998). Real options: Managerial flexibility and strategy in resource allocation. Cambridge: MIT Press.

APPENDIX A While the first two and the last two integrals of Eq. (7) are similar to the Black–Scholes integral and are solved for accordingly, one has to apply Itoˆ’s lemma to dvb and vb, respectively, to solve the remaining integrals. Thus, neglecting dividends this time we get dvb ¼ rvb dt þ sbvb dZ Q

(A.1)

b TþsbZQ T

(A.2)

vbT ¼ vb erT1=2s

2 2

The last two terms of the exponential function can be substituted into a stochastic process XV Nð1=2s2 b2 T; s2 b2 TÞ: Exemplarily, the solution is drafted for only one integral. The resulting integral e

rT

ZvU

Avb1 dPðvÞ

(A.3)

c

can be transformed by substituting Eq. (A.2) into ¼e

rT

Zb

Avb1 erTþs f Nð1=2s2 b2 T; s2 b2 TÞ ðsÞ ds

(A.4)

a

with a ¼ b1 lnðc=vÞ rT and b ¼ b1 lnðvU =vÞ rT as lower and upper boundaries. Utilizing the symmetry features of the normal distribution, i.e., Rb Rb Ra a f ðxÞ dx ¼ 1 f ðxÞ dx 1 f ðxÞ dx; the integral can easily be solved.


NEW TRENDS AND PERFORMANCES OF KOREAN OUTWARD FDI AFTER THE FINANCIAL CRISIS Seong-Bong Lee ABSTRACT This paper analyzes new outward foreign direct investment (OFDI) patterns, OFDI performance after the financial crisis in Korea, and the anticipated impact of these changes in OFDI on the Korean economy. This paper examines current trends in Korean OFDI activities from various viewpoints, including the geographical distribution of investments, industry, size, and investors’ equity share. This paper also verifies the relationship between OFDI and the Korean economy through an in-depth analysis of motives, performance, and prospects of Korean OFDI. Finally, two emerging issues regarding Korean OFDI are discussed.

1. INTRODUCTION Korean outward foreign direct investment (OFDI) increased rapidly during the 1990s. The Korean chaebols were main players in this OFDI increase (Bae & Hwang, 1997; Dent & Randerson, 1997; Park, 2000). However, it Value Creation in Multinational Enterprise International Finance Review, Volume 7, 75–96 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1569-3767/doi:10.1016/S1569-3767(06)07004-X

75

76

SEONG-BONG LEE

decreased significantly during the three to four years after the financial crisis (Lee, 2000), and then reversed in 2002, increasing once again. One feature of the current rise in OFDI is that not only are large enterprises engaging in OFDI but also small- and medium-size companies. Therefore, Korean OFDI now shows a mixed investment pattern in which two different types of investment coexist: one performed by big companies, including Samsung, LG, and Hyundai Motors, to secure large local markets such as the United States and the European Union; the other by small- and medium-size companies to ensure low labor costs. This new pattern of OFDI reflects structural changes in the Korean economy as well as the growing trend of reciprocal influences. This paper analyzes new OFDI patterns, OFDI performance after the financial crisis in Korea, and the anticipated impact of these changes in OFDI on the Korean economy. The first part of this paper gives a brief historical overview and talks about recent developments in Korean OFDI policy. The following part examines current trends in Korean OFDI activities from various viewpoints, including the geographical distribution of investments, industry, size, and investors’ equity share. This paper also verifies the relationship between OFDI and the Korean economy through an in-depth analysis of motives, performance, and prospects of Korean OFDI. Finally, two emerging issues regarding Korean OFDI are discussed.

2. TREND OF KOREAN OUTWARD FDI 2.1. OFDI Policy in Korea During the late 1980s, the Korean economy was booming, emerging from the lack of foreign currency that characterized the previous period, and Korean local firms, which matured rapidly under an export-led economic system, started to engage in active market-seeking activities, which directly led to a large increase in OFDI. From the late 1980s until the mid-1990s, Korean OFDI activities showed constant progress. Korea became a member of the OECD in 1996 and, from an institutional perspective, developed a freer market system and executed the switchover from an approval system to a notification system regarding its OFDI rules in August 1997 (Park, 2000). From the 1997 financial crisis until 2001, however, Korea showed an overall decline or stagnation in its OFDI activities. A large number of Korean firms closed or sold off their overseas subsidiaries. A particular

OFDI Performance after the Financial Crisis in Korea

77

characteristic of that time is that many overseas subsidiaries had a heavily indebted capital structure, which required parent companies to feed them in the form of loan guarantees. As a result, the Korean government introduced a system limiting the amount of loan guarantees from Korean parent companies in overseas capital investment to achieve a higher degree of soundness in OFDI activities. The limit for a loan guarantee was set at 95 percent of the outstanding level of guarantee, which only applied to affiliates of the 30 largest chaebols. Since 2002, Korean OFDI has been increasing, with a foreign exchange reserve of over $200 billion. The Korean government has also changed its OFDI policy to be less limiting to promote more OFDI. The limit on loan guarantees will be abolished in 2006. The current Korean policy for supporting OFDI can be categorized into the following three policies: loan programs, information provisions, and the reduction of non-commercial risks. The first policy, loan programs, consists of a loan program of the Export-Import Bank of Korea for financing of OFDI (up to 80 percent of the OFDI amount) and a loan program from the international economic cooperation fund of the Korean government, which is applicable to OFDI in developing countries with a long-term return period (The Ministry of Finance and Economy, 2002). The Export-Import Bank of Korea is a representative of the institutions that provide Korean overseas investors with useful information on the economic, business, and living conditions in host countries. The Korea Trade and Investment Promotion Agency provides, through its branch offices in each country, information to Korean firms that are investing in a particular region. In addition, the Korean Export Insurance Corporation offers export insurance to investors to reduce the non-commercial risks of OFDI, such as losses from war, nationalization, and so forth. The Korean government signed the ‘‘Investment Protection Agreement’’ with more than 70 countries to provide a more secure legal shelter for Korean OFDI. It also signed tax treaties with 62 countries to avoid possible international double taxation on the incomes of Korean subsidiaries abroad. 2.2. Overall OFDI Trends Clear differences between pre- and post-crisis trends in Korean OFDI over the past decade can be observed. Before the crisis of 1997, OFDI, led by Korean firms, was consistently rising. In 1994, it started increasing by nearly $1 billion per year due to the fact that large Korean firms, especially chaebols, began actively engaging in OFDI in accordance with global

78

SEONG-BONG LEE

business strategies. However, a radical restructuring of Korean firms during and after the crisis resulted in a sharp decline in OFDI performance, a trend that would reverse after 2002. The net outstanding invested amount of Korean OFDI is now $43.5 billion (October 31, 2005). Another important characteristic of post-crisis Korean OFDI is the sharp increase in OFDI withdrawal from 1998 to 2001. The annual average amount of OFDI withdrawal during the 1993–1997 period equals about $350 million (Table 1). During the post-crisis period, it exceeded $1 billion in 1998 and $3.3 billion in 2001. Then the amount began to decline in 2002, and eventually fell to $760 million in 2004. The total amount of OFDI withdrawal during 1998–2002, when the firm restructuring process was most extensive, equals $8 billion, which amounts to no less than 64.5 percent of $12.3 billion, the total OFDI withdrawal during the whole period from 1968 to the end of October 2005. This statistic shows that many Korean firms closed their foreign subsidiaries during the restructuring period. The net outstanding invested amount, which is the amount invested subtracted by the amount withdrawn, has demonstrated a recovery since 2002. This can be interpreted as Korean firms resuming active OFDI with the near-completion of the restructuring process. Table 1. Year Until 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Total

Korean OFDI Trends (Thousands of Dollars). Invested Amount

Withdrawn Amount

Net Amount

5,221,230 1,264,637 2,304,511 3,139,923 4,450,375 3,677,605 4,792,703 3,329,688 5,051,696 5,139,834 3,681,759 4,019,355 5,932,860 3,821,552

802,300 245,226 273,031 313,458 652,512 267,082 1,066,117 1,057,325 1,450,319 3,292,283 1,091,841 693,268 766,537 364,932

4,418,930 1,019,411 2,031,480 2,826,465 3,797,863 3,410,523 3,726,586 2,272,363 3,601,377 1,847,551 2,589,918 3,326,087 5,166,323 3,456,620

55,827,728

12,336,231

43,491,497

Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each year. a Until October 31, 2005.

Millions of Dollars


6,000 5,500 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0

Others Latin America Europe North America Asia

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Fig. 1.

79

2004

Korean OFDI by Region (In Years). Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.

2.3. OFDI by Region By 2004, Korean OFDI in Asian countries was $16.8 billion, which accounted for 43 percent of the total net outstanding invested amount of OFDI. North America ranked second with $10.8 and accounted for 27 percent. European countries made up 17 percent with $6.7 billion and Korean OFDI in Latin American countries was $3.4, 9 percent of Korea’s total OFDI. Fig. 1 shows OFDI trends by location. Most interesting is the fact that Korean OFDI in Asia has increased in recent years. This increased investment is closely related to rapidly increased investment in China. Fig. 2 shows figures for Korean investment in China. According to this graph, there has been rapid growth, or 54 percent of annual growth, since 2001: $0.6 billion in 2001, $1.2 billion in 2002, $1.63 billion in 2003, and $2.2 billion in 2004. In 2002, China became the primary recipient of Korean investment, occupying the United States’ former position. Investments in the United States were $0.57 billion in 2002, $1.05 billion in 2003 and $1.34 billion in 2004, exhibiting a widening investment gap between the United States and China. 2.4. OFDI by Industry If we look at the outstanding net amount invested by industry, the manufacturing sector occupies the largest piece of the pie at 53 percent ($21 billion). The second largest belongs to wholesale and retail, which takes up 22 percent ($8.7 billion). The mining industry follows, holding 6 percent

2,200 2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0

04

03

20

01

02

20

20

00

20

99

20

98

19

97

19

96

19

94

95

19

19

93

Total invested projects

19

Fig. 2.

2,200 2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0

Total invested amount

Projects

SEONG-BONG LEE

19

Millions of Dollars

80

Korean OFDI Trends in China. Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.

Fig. 3. Korean OFDI in the Manufacturing Sector by Type and Region. Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.

($2.1 billion), and other sectors including the service industry take up 19 percent of the total. This shows that firms in the manufacturing sector have been the driving force behind Korean OFDI. In other words, Korean investment is concentrated in ensuring overseas production facilities (53 percent) or market expansion to increase sales (22 percent). However, a low degree of investment in the service sector compared to that of the manufacturing sector shows that Korea has relatively weak competitiveness in the world services market. Analyzing investment in ‘‘manufacturing’’ and ‘‘trade and retail’’ by subcategorizing and observing them by region (Figs. 3 and 4) produces


81

Fig. 4. Korean OFDI in the Trade and Retail Sector by Type and Region. Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.

interesting figures. ‘‘Electronics and telecommunication equipment’’ comprises 33 percent of the total net outstanding invested amount in manufacturing sector, with $6.8 billion, and ‘‘motors and equipment’’ follows it taking up 12 percent with $2.5 billion. It becomes clear that Korean OFDI in manufacturing is primarily led by electronics and automobile firms. When divided by region, Asia leads with $11.5 billion, or 55 percent of the total net outstanding invested amount in the manufacturing sector, North America with $4.4 billion or 21 percent, and Europe with $4 billion or 19 percent of total manufacturing investment. Such regional disparity demonstrates that Korean manufacturing firms aim for cost minimization when engaging in OFDI. In other words, the fact that more than half of investments in Asia go to China and South-East Asian countries makes clear the importance of cost minimization through the capitalization of the relatively low wages in those countries. Investment in the North American and European regions can be accounted for by Korean firms’ desire for costeffective measures in those markets, such as tariff and non-tariff trade barrier evasions and the reduction of transportation costs. In the case of ‘‘trade and retail’’ investments by Korean firms, 34 percent of the total net outstanding invested amount ($7.1 billion) was achieved by electronics and equipment firms at the end of 2004, and 21 percent ($4.4 billion) by automobile firms. This means that the aggregate amount invested in foreign sales subsidiaries by electronics and automobile firms takes up 55 percent of the total overseas investment. Therefore, this implies that Korean firms, particularly exporters of electronics and automobile industries, have a clear motive behind developing overseas markets. The geographical dispersal of trade and retail investment shows that 43 percent of the total amount invested ($9 billion) goes to North America,

82

SEONG-BONG LEE

while Europe and Asia receive 25 percent each. This implies that investment in this sector can be fully characterized as a market-seeking investment with the purpose of increasing exports to the target market. 2.5. OFDI by Investing Enterprise

Millions of Dollars

The investing firms can be classified as large enterprises and small and medium enterprises (SMEs). By 2004, the net outstanding invested amount by large enterprises was $27.37 billion, which accounted for 69 percent of the total amount of investments. In comparison, SMEs accounted for 27 percent with $10.7 billion, and other individual investors accounted for 4 percent with $1.5 billion. However, the project basis distribution is different: 61 percent for SMEs, 29 percent for other individual investors, and 10 percent for large enterprises. Large enterprises invested a large amount of money in a few projects, which accounts for the disparity in distribution between amounts and project bases (Fig. 5). This can be observed more clearly when looking at OFDI by invested amount per project. By the end of 2004, the sum of investments over $50 million per project was $14.25 billion, which accounted for 36 percent of the total net outstanding invested amount, and the sum of investments from $10 million to $50 million was $10.29 billion, making up 26 percent of the total. The sum of these two investment types was $24.44 billion, which accounted for 65 percent of total net outstanding invested amount, similar to the 69 percent of the amount invested by large enterprises. The interesting thing is that the share of OFDI by SMEs continues to increase from year to year. From 2000 to 2004, the share of investments by SMEs increased by 33.7 6,000 5,500 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0

Others SME Large enterprise

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 5. Korean OFDI by Investing Enterprise and Year. Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.


83

percent of annual average growth. This growth rate was two times higher than that of 17.5 percent during 1993–1999. It means that SMEs have been actively participating in FDI in the past five years. The amount of investments per project by SMEs is also increasing. From 1993 to 1999, the average investment amount per project by SMEs was $701,000, whereas from 2000 to 2004, it increased to $1.1 million. Increasing SME investment, both in project numbers and monetary value, implies that more and more Korean SMEs are not just trying to secure efficient manufacturing facilities at low costs, but rather struggling to survive by moving most of their value chains to foreign countries. 2.6. Management Control over Foreign Subsidiaries of Korean Firms One of the most used indicators for management control over foreign subsidiaries of Korean firms (Guillen, 2003) is statistics on OFDI by the investor’s equity ratio. On a project basis, wholly owned foreign subsidiaries account for 62 percent, and subsidiaries with 50–100 percent shares by Korean parents were 15 percent of the total number of OFDI projects (Fig. 6). The sum of these two cases is 75 percent and it implies that Korean enterprises are pursuing management control in most OFDI by majority shareholding. On a net outstanding invested amount basis, wholly owned subsidiaries accounted for 66 percent, and subsidiaries with 50–100 percent shareholding were 20 percent. The sum of these two ratios is 86 percent. It is noticeable that the share of majority shareholdings on an amount basis is 11 percent higher than that calculated on a project basis. It implies that there could be more majority shareholding investments from larger enterprises than from SMEs. It implies that large enterprises prefer to have more control over their foreign subsidiaries than SMEs. This tendency could be interpreted to mean

Fig. 6.

Korean OFDI by Investors’ Equity Ratio. Source: The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, each Year.

84

SEONG-BONG LEE

that large Korean firms have their own significant competitive advantages over competitors in foreign markets, and therefore prefer carrying out business themselves. Meanwhile, SMEs are pursuing a cooperative strategy with local partners in foreign markets to a greater degree than large enterprises (Tallman & Shenkar, 1990).

3. MOTIVES AND PERFORMANCES OF KOREAN OFDI 3.1. Motives for Korean OFDI According to an analysis done by The Export-Import Bank of Korea (2004) on motivating factors behind OFDI, written by the investors in the form of a notice to the Korean government, the primary factor was to secure or develop local or third-country markets (about 50 percent of total notices), and a secondary motivator was to take advantage of low labor costs in the host country. Table 2 shows other motives for Korean OFDI. Until 1993, the development of natural resources was the main objective of investment. This statistic is based on amount invested, and a large part of investments during that time was for the development of natural resources. Meanwhile, we can see that the importance of the motive of utilizing low labor costs increased when comparing motives in the period 1997–2001 and after 2002. This corresponds to the fact that OFDI by SMEs has been increasing rapidly since 2002, with intensive concentration in China and other Asian countries. If we look at motives for OFDI by destination and industry (Table 3) from 2002 to November 2004, the trend mentioned above is more clearly Table 2.

Motivation of Korean OFDI by Period (Percent).

Securing or developing local or third-country markets Utilizing local labor costs Avoiding trade barriers Securing raw materials Acquiring advanced technology or management know-how Developing natural resources

1968–1993

1994–1996

1997–2001

After 2002

28.9

50.2

52.4

47.1

14.7 1.7 3.8 1.1

37.2 2.5 4.8 2.6

30.3 2.3 3.9 7.7

38.5 3.1 4.4 4.1

49.9

2.7

3.3

2.8

Source: The Export-Import Bank of Korea (2004).

Motivation of Korean OFDI by Destination and Industry (Percent).

Motive

Destination

Industry

China United States Europe Manufacturing

Securing local or third-country markets Utilizing local labor costs Avoiding trade barriers Securing raw materials Acquiring advanced technology/ management know-how Developing natural resources

Electronics and Telecom Automobiles Equipment

Textiles and Clothes

41.2

65.1

66.7

43.5

50.9

50.7

41.0

47.9 3.2 4.6 1.1

4.7 2.3 3.5 21.2

8.8 0.7 4.8 11.6

45.1 3.4 4.2 2.1

41.7 2.9 0.6 2.5

38.7 4.4 0.9 4.4

52.1 3.2 2.1 0.8

2.1

3.2

7.5

1.8

1.4

0.9

0.7


Table 3.

Source: The Export-Import Bank of Korea (2004).

85

86

SEONG-BONG LEE

identifiable. According to this table, the primary motive for Korean OFDI in China was to take advantage of low labor costs (47.9 percent of total). In contrast, the most important motive for Korean OFDI in the United States and Europe was to secure the local market, weighing in at 65.1 and 66.7 percent, respectively. Meanwhile, acquiring advanced technology or management know-how was the second most important reason for Korean OFDI in the United States (21.2 percent) and Europe (11.6 percent). If we look into OFDI motives by industry, the manufacturing industry’s most important investment motive was to make use of low labor costs (45.1 percent of the total). However, in the case of OFDI in globally competitive industries such as the electronic communications equipment industry and the automobile industry, securing or developing local or third-country markets was the most important motive for investment. Meanwhile, in less competitive industries, like the textile and clothing industries, utilizing low local labor costs was the most important motive for investment. While this analysis of OFDI motives does not include the differences between investments by large companies and those of SMEs, another analysis (The Export-Import Bank of Korea, 2004) of 318 cases of OFDI that amount to more than $10 million in outstanding invested amounts at the end of 2003 – mostly carried out by large companies – revealed that securing or developing local or third-country markets was the primary motive in all destinations, creating a great gap in the motive of utilizing low labor costs. If we reverse these results, it indirectly confirms that most OFDI by Korean SMEs were aimed at utilizing local labor costs. 3.2. Performances of Korean OFDI 3.2.1. Performances of Large Foreign Subsidiaries of Korean Firms For all foreign subsidiaries with an invested amount of more than $10 million, the Korean foreign exchange regulation obliges investors to submit an annual report, including financial statements of foreign subsidiaries, to the Korean government. The Export-Import Bank of Korea has done annual analyses of these reports and financial statements of foreign subsidiaries since 1999 (in other words, for the subsidiaries’ financial performance for the accounting year 1998). In 2005, the Bank’s analysis report included an annual comparative analysis of financial performances from 1998 to 2004. In this paper, we explain some meaningful results of this analysis, including financial performance trends and trade effects. Among the foreign subsidiaries whose invested amount was over $10 million at the end of each year, those with reliable financial statements were analyzed.


87

Therefore, for this seven-year period, sample foreign subsidiaries may differ from year to year. 3.2.1.1. Overall Features of Sample Subsidiaries for Performance Analysis. Table 4 shows overall features of sample subsidiaries for each year. The average amount invested for sample subsidiaries has decreased, but average sales and average net income have clearly increased over the past three years. Among the sample foreign subsidiaries during this period, the portion of wholly owned subsidiaries increased from 60 percent in 2002 to over 70 percent in 2003 and 2004. This figure is based on the number of subsidiaries and amount of equity balance. This shows that large Korean firms have come to prefer controlling all resources for investment and business in foreign markets to collaborating with local partners. 3.2.1.2. Annual Comparison of Financial Performance. Table 5 shows some selected indicators for the financial performances of sample foreign subsidiaries. First of all, the average of the capital stock maintenance ratio, which is calculated by subtracting 1 from the ratio of total equity capital to the capital stock of each subsidiary, was recorded as –38.7 percent in 1998 and – 51.1 percent in 2000. Then the ratio increased to –17.3 percent in 2004. These figures reveal a trend of increasing net income of foreign subsidiaries, while cumulatively remaining in a condition of capital depletion. The financial stability of foreign subsidiaries seems to have improved: the debt ratio was 522.4 percent in 2000 but decreased to 293.8 percent in 2004. Profitability has also been improving, based on the ratio of net income to stockholders’ equity, which reversed to surplus figures and reached 13.3 percent in 2004. Since 2002, subsidiaries’ annual growth (the growth rate of sales and the growth rate of total assets) has remained over two digits Table 4. Year

Overall Features on Sample Subsidiaries for Performance Analysis (Millions of Dollars). 1998

1999

2000

2001

2002

2003

2004

Cases of foreign subsidiaries 290 318 276 311 318 319 377 Investment balance account 11,657 16,395 16,482 17,844 16,941 17,692 18,232 Average investment amount 40 52 60 57 53 55 48 Average sales 174 222 290 226 262 327 383 Average net income 10 6 1 3 0 1 6 Percent of wholly owned OFDI (%) – 61.4 62.6 65.0 61.9 72.4 75.6 Source: Export-Import Bank of Korea (2005).

88

Table 5. Annual Comparison of Financial Performances of Foreign Subsidiaries with over $10 Million Invested (Percent). Year

Source: Export-Import Bank of Korea (2005).

1999

2000

2001

2002

2003

2004

38.7

38.3

51.1

48.5

42.2

35.7

17.3

400.1 98.8

492.4 109.4

522.4 115.5

453.2 121.5

358.7 101.3

329.2 99.5

293.8 91.2

2.8 28.9

0.7 20.4

2.0 5.1

0.6 12.0

1.2 6.2

1.9 4.8

1.8 13.3

1.2 6.5

24.4 7.9

24.4 8.2

9.8 7.4

10.5 4.3

27.5 17.3

33.8 23.8

1.0 8.6 –

1.3 11.2 21.4

1.6 12.0 13.5

1.5 10.2 24.0

1.8 11.0 7.6

2.3 12.2 10.0

2.6 14.2 16.7

SEONG-BONG LEE

Maintenance ratio of capital stock Indicators concerning financial stability Debt ratio Fixed assets to stockholders equity and long-term liabilities Indicators concerning profitability Operating income to sales Net income to stockholders equity Indicators concerning growth Growth rate of sales Growth rate of total assets Turnover of assets (times) Total assets turnover Inventory turnover Return on investment

1998


89

and their sales growth rate is larger than their total asset growth rate. Total assets and inventories turnover, which indicate the utilization of assets, have also been constantly increasing since 2001. From 1999 to 2001, the average annual return on investment was –19.6 percent. In 2002, return on investment figures reversed to positive ratios and has since been rapidly increasing. 3.2.1.3. Annual Comparison of Trade Surplus Effects. The positive effect of foreign subsidiaries in stimulating Korean exports has been well proved by previous research (Lim & Moon, 2001). The effect, calculated by the ExportImport Bank of Korea as the ratio of exports to foreign subsidiaries to the net outstanding invested amount, was recorded as 134.5 percent in 1998 and then continued to increase to over 200 percent after 2003 (Table 6). Foreign subsidiaries’ import-arousing effect, which is calculated as the ratio of imported amount from foreign subsidiaries to Korea to the net outstanding invested amount of them, reached 34.3 percent in 1998 and continued to increase to reach 125.2 percent in 2004. The improving effect of trade balance during the last seven years, calculated by subtracting the import-arousing effect from the export-arousing effect, was 130 percent, on average. This figure reached 141 percent in 2004, perhaps due to foreign investment. 3.2.2. Comparison of Foreign Subsidiary Performance by Size At the end of 2004, The Export-Import Bank of Korea conducted a comparative study of the financial statuses (The Export-Import Bank of Korea, 2005, pp. 99–106) of 646 small- and medium-size subsidiaries that invested less than $10 million and a group of 377 large-size subsidiaries that invested more than $10 million. Those 1,023 foreign subsidiaries were selected from 3,265 foreign subsidiaries whose investment balance was over $1 million, according to financial statements submitted to the Korean government. Table 6.

Annual Comparison of the Trade Surplus Effectsa of Foreign Subsidiaries (Percent).

Year

1998

1999

2000

2001

2002

2003

2004

Export-arousing effect Import-arousing effect Trade surplus effect

134.5 34.3 100.2

158.5 35.6 122.9

188.6 61.0 127.6

154.8 44.0 110.8

156.1 48.8 107.3

229.2 65.5 163.7

266.1 125.2 141.0

Source: The Export-Import Bank of Korea (2005). a Trade surplus effect is the ratio of exports to foreign subsidiaries to the net outstanding invested amount.

90

SEONG-BONG LEE

The statistical overview of this sample (Table 7) accurately represents the overall trends of Korean OFDI as reviewed in Chapter 2. When looking at the regional distribution of sample small- and medium-size subsidiaries in terms of net outstanding invested amount, 74.6 percent of them are located in Asian regions, especially in China (48.2 percent) and in ASEAN countries (13.3 percent). That ratio in the European and North American regions together is just 19.8 percent. For large subsidiaries, locations in Europe and North America together account for 51.5 percent, and 41.3 percent in Asia, mainly in China (20.5 percent) and ASEAN countries (10.3 percent). Table 8 compares selected financial ratios between both groups. According to this comparison, the small- and medium-size subsidiaries had more favorable figures in stability and growth, while in terms of profitability and operation the large-size subsidiaries were identified as better than small- and medium-size subsidiaries. In comparing financial ratios, both groups exhibited negative values in capital stock maintenance ratios. This means that in 2004, subsidiaries of Korean companies were still at a level of capital depletion. The small- and medium-size subsidiaries’ capital stock maintenance ratio was –1 percent, which is relatively high, but the capital stock maintenance ratio of the group of large-size subsidiaries was very low and reached –17.3 percent. In terms of the debt ratio and the stability ratio, the small- and mediumsize subsidiaries showed better values than the group of large-size subsidiaries. This can be interpreted in two ways: first, because small- and medium-size enterprises have relatively less experience in foreign investment, this group tends to avoid potential business risks; second, this result reflects the fact that small- and medium-size enterprises are less able than large-size enterprises to finance funds through borrowing. The profitability of the small- and medium-size subsidiaries group is worse than the group of large-size subsidiaries. The reason is that competition among small- and medium-size subsidiaries is very strong in local markets, so profit margins for sales are low. And due to the lack of experience in running a foreign business, sales and administration costs are high. Among the 646 small- and medium-size subsidiaries, 322 subsidiaries resulted in an operating loss. Meanwhile, in terms of return on investment, large-size subsidiaries were almost three times better than the small- and medium-size subsidiaries. The growth rate of the group of small- and medium-size subsidiaries was higher than the group of large-size subsidiaries, whereas the group of largesize subsidiaries showed higher turnover of assets and capital. Since the difference of these two indicators is between the two groups, it is difficult to find significance in this difference.

Overall Features on Sample Subsidiaries for Comparative Analysis by Size.

Region

Investment of Less Than $10 Million Cases (Number, %)

Investment of More Than $10 Million

Invested Amount (Millions of Dollars, %)

Cases (Number, %)

Invested Amount (Millions of Dollars, %)

Asia China Hong Kong Japan ASEAN North America Europe EU South America Oceania Africa

498 327 25 25 91 85 37 31 13 9 4

(77.1) (50.6) (3.9) (3.9) (14.1) (13.2) (5.7) (4.8) (2.0) (1.4) (0.6)

1,662 1,074 89 95 297 313 129 114 71 34 19

(74.6) (48.2) (4.0) (4.3) (13.3) (14.0) (5.8) (5.1) (3.2) (1.5) (0.9)

215 118 19 10 58 81 50 46 18 11 2

(57.0) (31.3) (5.0) (2.7) (15.4) (21.5) (13.3) (12.2) (4.8) (2.9) (0.5)

7,531 3,729 1,007 338 1,872 6,691 2,703 2,465 586 516 205

(41.3) (20.5) (5.5) (1.9) (10.3) (36.7) (14.8) (13.5) (3.2) (2.8) (1.1)

Total

646

(100.0)

2,227

(100.0)

377

(100.0)

18,232

(100.0)


Table 7.

Source: The Export-Import Bank of Korea (2005, 100).

91

92

SEONG-BONG LEE

Table 8.

Comparison of Financial Performances of Foreign Subsidiaries by Size (Percent).

Average of Selected Indicators

Maintenance ratio of capital stock 100% shareholding subsidiaries Subsidiaries with 50–100% shareholding Subsidiaries with shareholding less than 50% Indicators concerning financial stability Debt ratio Fixed assets to stockholders’ equity and long-term liabilities Indicators concerning profitability Operating income to sales Net income to stockholders’ equity Indicators concerning growth Growth rate of sales Growth rate of total assets Turnover of assets Total assets turnover Inventories turnover Return on investment

Investment of Less Than $10 Million


1.0 1.5 18.6

17.3 19.9 12.0

10.0

12.4

177.9 83.7

293.8 91.2

0.3 1.5

1.8 13.3

56.6 28.5

33.8 23.8

2.0 9.7 6.4

2.6 14.2 16.7

Source: The Export-Import Bank of Korea (2005, pp. 103–104).

Table 9.

Comparison of Trade Effects of Foreign Subsidiaries by Size.

Export-arousing effect Import-arousing effect Trade surplus effect

Investment of Less Than $10 Million


431.8 133.1 298.6

266.2 125.2 141.0

Source: The Export-Import Bank of Korea (2005, p. 104).

One interesting fact is that the trade surplus effects of the group of smalland medium-size subsidiaries are superior to those of the large-size subsidiaries group. Table 9 compares the trade surplus effects between the two groups. The group of small- and medium-size subsidiaries showed 431.8 percent in export-arousing effect and the large-size subsidiaries group showed 266.2 percent. Also, the export-arousing effects of the small- and


93

medium-size group are much higher and the import-arousing effects of the two groups are similar. From this result, it can be concluded that the contribution to trade surplus per unit investments of small and medium size is higher than for large-size investments. This can be attributed to the fact that small- and medium-size enterprises usually procure a large amount of raw material, by-products, and equipment from Korea, manufacture the products locally at lower labor costs, and sell the products to local Korean largesize subsidiaries or export to other countries.

4. SUMMARY AND EMERGING ISSUES REGARDING KOREAN OUTWARD FDI This paper conducts an in-depth analysis on current Korean OFDI and trends, performance, and impact on the Korean economy. Important characteristics of recent Korean OFDI trends can be summarized as follows. First, Korea has increased the amount of its OFDI activities in Asia. In particular, China has become the primary destination of Korean OFDI, relegating the United States to second place. This reflects a global trend in which China is gaining recognition as a favored FDI destination. Second, a large part of FDI is concentrated in the manufacturing sectors, of which some limited industries such as electronics and automobile constitute a large part. This trend reflects the fact that Korea is considerably competitive in these two industries in the global market. Third, it is observed that in the past three years OFDI by small- and medium-size companies has largely increased. This can be interpreted in light of the loss of Korean domestic cost competitiveness, especially for SMEs, which resulted in an increase in OFDI of SMEs to other Asian countries, especially China, in order to ensure lower labor costs (Zhan, 2005). Fourth, investments are motivated by both market-seeking purposes and low labor cost seeking purposes, and the importance of each of the purposes can be differentiated by the size of investment: larger investments are for mainly market-seeking purposes, while small and medium types of investments seek low labor costs. In the case of OFDI to developed countries such as the United States and European countries, the second motive after market-seeking was found to be to ‘‘[learn] advanced technology or know-how.’’ Fifth, overseas legal subsidiaries of Korean firms improved their financial performance remarkably in the past seven years. In the case of small and medium types of investments, however, returns on investments are far lower than those of large-scale investments.

94

SEONG-BONG LEE

Sixth, OFDI actually contributes to the improvement of the Korean trade balance, which is accurate when evaluating data limited to each foreign subsidiary-related export and import. However, this analysis is limited in that it did not include any spillover effects of OFDI toward the trade of other Korean firms. Meanwhile, statistics suggest that small- and mediumsize investments create bigger export-arousing effects. However, as the data include transferred parts of existing production facilities in Korea into export items, it is necessary to evaluate the net export-arousing effect that excludes the transfer of existing facilities. Moreover, there has been no study that clearly and comprehensively evaluates the crowding-out effects of Korean OFDI on domestic investment. Finally, Korean firms perceive their competitiveness in foreign markets as quite positive. However, with the exception of some large-scale investments, a large number of investment activities do not have good profitability, which actually renders the mid- and long-term competitiveness of Korean firms rather gloomy. In spite of all this, Korean firms are expected to consistently expand their overseas investment. This expectation is derived from two facts: first, internationally, the Korean economy is now more strongly linked to the global economic system due to increasing international trade and investment; second, domestically, Korea has been undergoing an industrial restructuring process with a rise in costs of production factors such as labor, land, and so forth. Considering this analysis on current trends, it is estimated that Korean OFDI is about to face two issues. The first issue is the polarization of OFDI between large firms and SMEs. Investment by large firms, which consists of high value-added fields in value chains such as R&D and design activities, are mainly carried out within Korea, while investments for low value-added activities including assembly are made largely in response to the requirements of cost-effectiveness with the goal of reinforcing global competitiveness from the viewpoint of rationalized production. On the contrary, a remarkable aspect of Korean SMEs is that many of them, as a way of survival, have moved their entire production facilities to a foreign country, giving up the Korean market because of a loss in competitive edge due to an overall rise in Korean production costs. The impact of OFDI by SMEs to long-term enhancement of firms’ competitiveness is not considerable. In the case of SMEs, therefore, other changes in the management environment such as increased production costs in the local market are likely to affect the very existence of the company. The second issue is the problem of imbalance between OFDI in manufacturing sectors and that in service sectors. So far, Korean OFDI has


95

mainly focused on manufacturing sectors, and even in service sectors it has been ‘‘trade and retail’’ that has been important. Moreover, Korean investment in this ‘‘trade and retail’’ part is actually for the purpose of sales of electronic and automobile products in overseas markets. This trend clearly reflects the weakness of Korean OFDI in specific service industries, such as banking or distribution. Despite a highly complementary relationship between OFDI in the banking and financial or distribution sectors and that of the manufacturing sectors, actual foreign investment by Korean banks or logistics companies does not meet the demand of large Korean firms for financial and distribution services. From the fact that growth in the service sectors is a key element for sustainable growth in the Korean economy, it is high time for Korea to improve its competitiveness in the service sectors through more positive OFDI activities. Although these challenges should be overcome on the level of the individual firm, there should be specific measures thoroughly examined and implemented on the governmental level. To achieve high quality of OFDI by SMEs, it is crucial to fortify cooperation between large firms and SMEs, which will consequently minimize the negative effects of SME OFDI in reducing their investment in domestic markets. In addition, the Korean government should ease regulations that hinder investment to pave the way for boosting OFDI activities in the service industries.

REFERENCES Bae, C. S., & Hwang, S. (1997). An empirical analysis of outward FDI of Korea and Japan. Multinational Business Review, 5(2), 71–80. Dent, M. C., & Randerson, C. (1997). Enter the chaebol: The escalation of Korean direct investment in Europe. European Business Journal, 9(4), 31–39. Guillen, F. M. (2003). Experience, imitation and the sequence of foreign entry: Wholly owned and joint-venture manufacturing by South Korean firms and business groups in China, 1987–1995. Journal of International Business Studies, 34(2), 185–198. Lee, S.-B. (2000). Korea’s overseas direct investment: Evaluation of performances and future challenges. Korea Institute for International Economic Policy Working Paper 00-12. Lim, S.-H., & Moon, H.-C. (2001). Effects of outward FDI on home country exports: The case of Korean firms. Multinational Business Review, 9(1), 42–49. Park, Y. S. (2000). The Korean chaebols’ dash for globalization: Has it really been driven by government policy? Korea Observer, 31(4), 579–605. Tallman, B. S., & Shenkar, O. (1990). International cooperative venture strategies: Outward investment and small firms from NICs. Management International Review, 30(4), 299–315. The Export-Import Bank of Korea (2004). A Way to Promote Outward Foreign Direct Investment of Korea (in Korean).

96

SEONG-BONG LEE

The Export-Import Bank of Korea (2005). Analysis of Business Performances of Foreign Subsidiaries of Korean OFDI (in Korean). The Export-Import Bank of Korea, Overseas Direct Investment Statistics Yearbook, Each Year. The Ministry of Finance and Economy (2002). Regulations on the Outward Foreign Direct Investment (in Korean). Zhan, X. (2005). Analysis of South Korea’s direct investment in China. China & World Economy, 1, 94–104.

PART III: STRATEGIES AND FIRM PERFORMANCE


THE VALUE CREATION PERSPECTIVE OF INTERNATIONAL STRATEGIC MANAGEMENT$ Reid W. Click ABSTRACT This paper applies the concept of value creation to examine the strategic management of multinational enterprises. ‘‘International strategic management’’ is first defined as the process through which value is created by managers operating across a national border. The domain of international strategic management is thus determined by activities that distinguish international management from domestic management in the process of value creation. This perspective on value creation is used to answer three questions pertaining to international strategic management. First, how important is international strategic management? Simple statistics presented here demonstrate that the international component of value creation is important in the U.S. economy. Second, what is the domain of international strategic management? The paper presents a framework in which international strategic management is the aggregation of value

$

An earlier version of this paper, entitled ‘‘Strategic Management of Multinational Enterprises and Value Creation,’’ was presented at the Eighth International Conference on Multinational Enterprises at the Chinese Culture University in Taipei, Taiwan, March 2006. I am grateful for comments from the conference participants.


99

100

REID W. CLICK

created through international production, marketing, and financial activities, and reveals that the domain of international management is vast. Third, does international strategic management make the whole multinational enterprise worth more than the sum of its parts? Empirical evidence suggests that the answer is yes, at least for U.S. multinationals in the early 1990s.

1. INTRODUCTION Management is the process through which value is created by an individual or a group of individuals, and international management is the process through which value is created by an individual or group of individuals operating across a national border. Since the meanings of ‘‘international,’’ ‘‘management,’’ and ‘‘international management’’ are apparently in dispute, or are at least under-understood (see Boddewyn, 1999, and references therein), these simple introductory definitions require further examination. A dictionary definition of ‘‘management’’ is ‘‘the act, manner, or practice of managing, handling or controlling something’’ (Morris, 1975, p. 192). In recognition of a raison d’eˆtre for the individuals who manage something, a normative version of the definition is ‘‘the act, manner, or practice of managing, handling or controlling something in order to create value or wealth.’’ Although the normative appendage is disputable, managers must recognize that the things worth doing are the ones that create value. This is the overarching message conveyed in modern business school curricula. As Harrigan (1992) points out, ‘‘Business schools exist to improve management practice. All scholarly research, development of teaching materials, testing of theories, and other activities carried on by faculties of business schools should help managers develop tools for coping with the problems of international value creation’’ (p. 251). The rest of this paper tackles the essence of value creation, so the definition being offered is the starting point for discussion rather than a definitive statement. Since ‘‘international’’ is ‘‘relating to, or involving two or more nations or nationalities’’ (Morris, 1975, p. 685), ‘‘international management’’ is ‘‘the act, manner, or practice of managing, handling, or controlling something involving two or more nations.’’ Hence, incorporating the normative element offered in this paper, international management is the process through which value is created by an individual or group of individuals operating across a national border.

Value Creation Perspective of International Strategic Management

101

In attempting to circumscribe the domain of international management, the definition of international management provides guidance for two reasons. First, it distinguishes international management from general management with the phrase ‘‘operating across a national border.’’ Second, it distinguishes international management, being focused on the ‘‘process through which value is created’’ internationally, from international business, which encompasses anything related to international commercial activity (such as elements of the operating environment). The domain of international management – as opposed to the domain of management or the domain of international business – should therefore be identified based on the value created by managerial actions across national borders. Without denying that domestic management is complicated, international management is often distinguished from domestic management by its additional complexities. For example, while domestic management requires a vast knowledge base in accounting, finance, production, human resources, marketing, economics, law, public policy, behavioral psychology, and so on, international management requires a commensurately larger knowledge base as the traditional areas become internationalized (and thus become part of the domain of international business) to account for additional governments, legal systems, cultural orientations, languages, and the like. In addition, international management requires knowledge of disciplines typically ignored in domestic business, such as international political economy, foreign affairs, and geographic area studies, which are important components of international business. Hence, the complexities that distinguish international management from domestic management result from operating across borders and having to deal with multiple economies, societies, and governments. Mastering international management can be a Herculean task if the domain is too broad, so some priorities must be set to focus on the most germane activities of international value creation. The focus on valuecreating activities is in fact what distinguishes international management from the broader concept of international business. This paper adopts a strategy approach to international management by considering a back-to-basics framework in which international management is the aggregation of value created through international production, marketing, and financial activities. The framework is not a simple internationalization of the functional areas of business, but is instead a broadening of the perspective to analyze production, marketing, and financial activities in a larger global context. Other elements of international business can be put into this framework if they create value. For example, foreign languages and geographic area

102

REID W. CLICK

studies are often included in the domain of international business. The framework presented here requires specification of the value created by having such managerial knowledge to be included in the more exclusive domain of international management. If knowledge of a foreign language facilitates management–labor communication at a foreign plant and thereby increases production efficiency, or if the talent improves relations with customers and thereby increases sales revenue abroad, it creates value and is within the domain of international management. However, the importance of the knowledge is directly related to the value created, and foreign languages and area studies are not automatically part of the domain of international management. The concept of value creation is used in this paper to answer three questions pertaining to strategic management of multinational enterprises (MNEs). After this introduction, Section 2 broadly examines value creation and the question, ‘‘How important is international strategic management?’’ Section 3 develops the strategy framework for analyzing international management and asks ‘‘What is the domain of international strategic management?’’ Discussions of international production, marketing, and financial activities suggest that international management requires an outward-looking perspective rather than an inward-looking perspective to maximize the value of the MNE, and reveal that the domain of international management is very broad. Section 4 ties the elements of value creation strategy together by asking, ‘‘Does international strategic management make the whole multinational enterprise worth more than the sum of its parts?’’ Empirical evidence from U.S. multinationals in the early 1990s indicates that the answer is yes.

2. VALUE CREATION AND THE RELATIVE IMPORTANCE OF INTERNATIONAL STRATEGIC MANAGEMENT Production and marketing operations represent the two main areas of business activity. Porter (1985) described these as the ‘‘primary activities’’ in the firm’s ‘‘value chain.’’ The production operations are the ‘‘upstream activities,’’ such as inbound logistics, operations, and the initiation of outbound logistics, which represent the ‘‘cost side’’ or ‘‘supply side’’ of the firm. The marketing operations are the ‘‘downstream activities,’’ such as the distribution activities of outbound logistics, marketing and sales, and after-sales service, which represent the ‘‘revenue side’’ or ‘‘demand side’’ of the firm.


103

Often, the objective of managing the supply side is to minimize the costs of producing goods and services, and this requires analysis of competitors and industry structures. A firm with production advantages, such as a proprietary production technology that utilizes lower quantities of inputs than competitors, is a low-cost producer in the industry and is able to pursue a ‘‘cost-based strategy.’’ The objective of managing the demand side is to maximize the revenues from selling goods and services, which requires analysis of customers and markets. A firm with a marketing advantage, such as brand recognition or unique products, emerges with market power and is able to pursue a ‘‘differentiation strategy’’ or ‘‘revenue-based strategy.’’ Production and marketing decisions clearly need to be mutually consistent to produce an equilibrium. For additional discussion, see Besanko, Dranove, Shanley, and Schafer (2003). Discussion of value creation in international strategic management must also distinguish international production and marketing from domestic production and marketing. In considering the internationalization of the value chain, Porter (1986a) more specifically asserts: The distinctive issues in international, as contrasted to domestic, strategy can be summarized in two key dimensions of how a firm competes internationally. The first is what I term the configuration of a firm’s activities worldwide, or where in the world each activity in the value chain is performed, including how many places. The second dimension is what I term coordination, which refers to how like activities in different countries are coordinated with each other. (p. 17)

Porter then considers various combinations of geographically dispersed or concentrated configurations versus high or low coordination of activities. By the end of the article, Porter (1986a) suggests that ‘‘today’s game of global strategy seems increasingly to be a game of coordination – getting more and more dispersed production facilities, R&D laboratories, and marketing activities to truly work together’’ (p. 36). These ideas are further developed in Porter (1986b, 1990). In a variation on the issue of coordination, Kogut (1985) asserts that international strategic management is distinguished from domestic management by the importance of flexibility: ‘‘the unique content of a global versus a purely domestic strategy lies less in the methods to design long-term strategic plans than in the construction of flexibility which permits a firm to exploit the uncertainty over future changes in exchange rates, competitive moves, or government policy’’ (p. 27). Kogut then considers coordination of activities internationally to reap the benefits of flexibility, not just to manage risk and uncertainty, but to profit from them as well. There has subsequently

104

REID W. CLICK

been considerable attention to pre-planned flexibility and international strategic investments; see, for example, de Meza and van der Ploeg (1987), Kogut and Kulatilaka (1994), and Amran and Kulatilaka (1999). The ideas presented in Porter and Kogut relating to configuration, coordination, and flexibility suggest that strategic management of MNEs creates value. Many other articles have also made this assertion in one way or another, including the seminal works of Hamel and Prahalad (1985), Prahalad and Doz (1987), and Ghoshal and Bartlett (1990). The subject begs the question, ‘‘How important is international strategic management?’’ Although difficult to answer definitively, some simple statistics illuminate the importance. A first way to consider the importance of international management relative to domestic management is to examine macroeconomic aggregates. Net domestic product (NDP), a measure of valued created by all factors in a particular country, amounted to $10.3 trillion in the U.S. in 2004 (U.S. Department of Commerce, 2005). Factor income from abroad, which is predominantly earnings on direct investment, was $53.7 billion, representing 0.5% of NDP. Hence, the pure foreign value creation does not appear very large, but further examination suggests that international activities are in fact important. Table 1 shows the breakdown of NDP into its basic components, and corporate profits (before taxes) amount to $1161.5 billion. (Note that this figure is not a measure of economic value added because part of corporate profits represents the required rate of return to shareholders.) Factor income from abroad thus represents 4.6% of total corporate profits. In addition, these figures exclude the contributions of exports to domestic profits and the contribution of imported inputs in reducing production costs. Export receipts amounted to 11.4% of NDP, and although the value created from these exports cannot be determined, the gross magnitude suggests that international management is indeed important. Some value is also created by using imported inputs to reduce production costs. Imports amounted to 17.5% of NDP, and although the portion attributable to inputs cannot be determined, this gross magnitude again suggests that international management is important. A second way to evaluate the relative importance of international strategic management is to examine firm-level data. Data are widely available for the subset of all firms that are publicly traded, and although this subset likely contains more MNEs than the subset of firms not publicly traded, the available data are useful in analyzing value creation. In the Compustat database (Standard and Poor’s Corporation, 1997), 6,345 nonfinancial U.S. firms report operating profits for 1994 amounting to $503.1 billion. Of these


105

Table 1. Value Creation in the U.S. Based on Net Domestic Product, 2004. Type of Income Compensation of employees Corporate profits Proprietors’ and rental income Net interest and dividends Net taxes on production and imports, net business current transfer payments, and statistical discrepancy Net domestic product (NDP) Factor income from abroad Exports of goods and services Imports of goods and services

Billions of Dollars

Percent of Total

6693.4 1161.5 1023.8

65.0 11.3 9.9

446.0 974.3

4.3 9.5

10,299.0

100.0

53.7 1173.8

0.05 11.4

1797.8

17.5

Source: U.S. Department of Commerce, Survey of Current Business, August 2005, pp. 36–172.

firms, 1,215 report operating profits of foreign operations in addition to total operating profit, and operating profits from foreign operations amount to $74.3 billion from total operating profits of $261.8 billion, or 28.4%. The $74.3 billion in profits from foreign operations represents 14.8% of the aggregate profits of $503.1 billion, although this understates the relative profits from foreign operations because not all firms with foreign operations report profits from foreign operations when they report total profits. In addition, the figure does not include profits from exports, licensing/franchising agreements, or any form of international business other than direct foreign investment. A third way of capturing the importance of international strategic management is to more broadly consider the sales revenues of the U.S. firms. A total of 6,345 nonfinancial firms report sales for 1994 amounting to $5 trillion. Table 2 presents some decomposition of this, indicating that 1,634 firms report export sales and 1,341 firms report sales from foreign operations. Because only 563 firms report both export sales and sales from foreign operations, a total of 2,412 firms report some form of foreign sales. Hence, 38% of the 6,364 firms are labeled MNEs and 62% appear to be purely domestic corporations. Since MNEs are larger, however, they account for

106

REID W. CLICK

Table 2.

Sales of U.S. Corporations, 1994.

Sales Export sales Sales from foreign operations Total foreign sales Total multinational enterprise sales Total domestic enterprise sales

Number of Firms

Millions of Dollars

6,364 1,634 1,341 2,412 2,412 3,952

5,124,123 149,430 899,993 1,049,423 3,059,468 2,064,655

Source: Compustat and author’s calculations.

60% of total sales and domestic firms account for 40%. Foreign sales reported by the 6,364 firms amount to $1 trillion, or 20% of total sales. Among the 2,412 MNEs, the foreign sales account for 34.3% of total sales. Furthermore, the data are again understated because not all firms with exports or foreign operations report segment sales. Of course, the focus on sales highlights the marketing side of the business and ignores the production side, but comparable data on costs are much harder to get than the data on revenues and so cannot be examined here. Financial activities also create value, although these are a little more suspicious than value creation through production and marketing activities since financial management does not obviously contribute either to the sales revenues of the firm or savings in the production costs of the firm. Porter (1985, 1986a, 1986b) implicitly considers financial management as a ‘‘support activity’’ rather than as a primary activity and hardly addresses the issue. However, some of the most interesting questions in management relate to whether value is created through financial advantages, debt capacity, risk management, and so on. In addition, there are firms that exclusively create value through financial activities – such as banks and brokerage houses. Recognizing the difficulty of subjecting these financial firms to the same ‘‘supply side’’ and ‘‘demand side’’ analysis as nonfinancial firms, this paper considers the activities of financial firms in the category of financial activities. Hence, two categories of value created through financial activities are the financial activities of nonfinancial firms and the activities of financial firms. One way to assess the importance of international financial activities is to examine the performance of U.S. financial firms. In the Compustat database, the operating profits of 714 financial firms amount to $137 billion. Of these, 109 firms report profits from foreign operations at $4.6 billion from total profits of $61.1 billion, or 7.5%, which also represents 3.4% of total financial industry profits. Thus, international financial management might


107

be marginally important, but this figure again understates the importance of profits on foreign involvement because it does not capture the profit on international capital flows booked at home – such as profits on foreign deposits in the U.S. or profits on foreign loans made from the U.S., and similar transactions in the bond and stock markets. International capital flows in the form of intermediated investment and portfolio investment are quite important compared to direct foreign investment, and firms managing these international capital flows are likely creating value through financial activities. Based on these simple statistics for the U.S., we can conclude that strategic management of multinational firms is indeed important. Although it is not true that ‘‘all management is international’’ in the U.S., international management is important in terms of its contributions at the margin when compared to domestic management, and is therefore well worth investigating.

3. VALUE CREATION AND THE DOMAIN OF INTERNATIONAL STRATEGIC MANAGEMENT Having discussed value creation in terms of production, marketing, and financial activities, this section combines these three categories into a basic strategy framework designed to examine the domain of international strategic management. Discussion is particularly focused on identifying the additional complexities resulting from crossing national borders that distinguish international management from domestic management. Each one of these areas could constitute a paper (or book) on its own, so the discussion is general rather than exhaustive, and references to additional literature are provided to fill gaps. A framework for analyzing strategic management of MNEs is presented in Table 3. This summarizes the distinctive issues in international strategic management, elements in the domain of international strategic management, and some representative studies. The table suggests that areas in the domain of international production are international trade theory (comparative advantage), international political economy, foreign affairs, and geographic area studies (political risk), international financial theory (exchange rates and risk), and area studies, law, sociology, anthropology, religion, psychology, and language (applied to human resources management). Elements in the domain of international marketing are geographic area studies, sociology, anthropology, religion, and psychology (cultural

108

Table 3.

A Strategy Framework for Analyzing International Management.

Distinctive Issues in International Management

The Domain of International Management

Representative Studies

Production activities

Source inputs domestically or internationally, including supplier relations Produce domestically or internationally International alliance (licensing, franchising, joint venture) versus wholly owned DFI International site selection decisions, including number and size of plants Input mix abroad (capital, unskilled labor, skilled labor, raw materials) Foreign production technology and equipment selection Human resources management in foreign countries Inventory management and crossborder transportation logistics R&D in process technologies

International trade theory, including comparative advantage, analysis of transportation costs, tariffs, and nontariff barriers International political economy and foreign affairs Geographic area studies International financial theory, particularly exchange rate behavior and risk analysis International law Sociology/anthropology/religion Psychology Languages

Stobaugh (1969a, 1969b) Flaherty (1986) Flaherty (1996) Dunning (1988) Ferdows (1997) Karrenbrock (1990) Hofstede (1984) Adler (1996) Quelch and Bloom (1999)

REID W. CLICK

International Value Creation Activities

International market selection, including number of foreign markets Market service and distribution methods (choosing exporting, international alliances, or DFI) Development of product attributes in different markets, including R&D into product technologies Product pricing internationally, including currency effects Promotional activities internationally, including language translation

Cultural studies Geographic area studies Sociology/anthropology/religion Psychology International financial theory, particularly exchange rate determination and behavior Languages

Levitt (1983) Quelch and Hoff (1986) Douglas and Craig (1989) Hout, Porter, and Rudden (1982) Kashani and Quelch (1990) Ricks (1993) Knetter (1994)

Financial activities

Foreign project evaluation (international capital budgeting) International cost-of-capital analysis and capital structure decisions, including capital structure of foreign subsidiaries Risk management for foreign exchange risk and foreign interest rate/inflation rate risk, including diversification and hedging strategies International financing and the debt denomination decision International tax management

International macroeconomics International financial theory, particularly exchange rate and international interest rate behavior and risk analysis International political economy and foreign affairs Geographic area studies International portfolio theory International financial markets International tax law

Shapiro (1983) Kogut and Kulatilaka (1994) Adler and Dumas (1984) Lessard and Lightstone (1986) Dufey and Srinivasulu (1983) Aliber (1989)


Marketing activities

109

110

REID W. CLICK

determinants of product demand, product attributes, and promotional programs), international financial theory (exchange rates and pricing in different currencies), and language (promotional activities). The domain of international financial activities includes international political economy, foreign affairs, and geographic area studies (political risk), international financial theory and international macroeconomics (exchange rates and interest rates), international tax law (focusing on financing), and international portfolio theory and international financial markets (risk management and hedging). Clearly, the domain of international strategic management is quite broad even under the criterion of value creation. Indeed, the diversity of international management is one appealing aspect that draws many managers into the field. References to additional literature are provided since space limitations preclude additional discussion here.

4. DOES INTERNATIONAL STRATEGIC MANAGEMENT MAKE THE WHOLE MULTINATIONAL ENTERPRISE WORTH MORE THAN THE SUM OF ITS PARTS? By this point, the notion that international management creates value in ways somewhat different from domestic management is firmly established. After the discussion of production, marketing, and finance activities covered in the strategy framework of Section 3, a return to the more general issue of value creation is now warranted. Rather than focus on specific tasks that distinguish international management from domestic management, recall that Porter and Kogut discuss the distinction in more generic terms focusing on coordination and flexibility. In addition to indicating that international management creates value, both imply that the whole of the MNE is worth more than the sum of its parts; having n international subsidiaries is worth more as an MNE than the summation of the separate values of the n subsidiaries, as long as they are coordinated and are flexible enough to profit from changes in the environment. Several reasons why the whole may be worth more than the sum of its parts have been developed in the international management literature. Kogut (1985) identifies several ‘‘arbitrage’’ opportunities for international management to create value: production shifting, tax minimization, financial market imperfections, and information arbitrage. Kogut also identifies a


111

couple of ‘‘leverage’’ opportunities: coordination and political risk management. Profits from arbitrage result from ‘‘moving’’ something from one place to another, but profits from leverage result from using a position in one national market to enhance a position in another market. For example, global coordination can build a coalition of suppliers in one country to improve the position of the firm with respect to a rival in another, and with regard to political risk the position of a subsidiary in one country is likely to affect the firm’s bargaining power vis-a`-vis the government in another. A second, more strategic reason for the whole to be worth more than the sum of the parts is due to cross-subsidization of operations. Hamel and Prahalad (1985) provide the example of competition in the tire industry: Michelin attacked Goodyear’s U.S. market by financing an aggressive marketing campaign using the profits from its European market, but rather than confine the competition to the U.S. Goodyear attacked Michelin’s European market by financing market expansion in Europe using the profits from its U.S. market. International cash flows may thus create value, such that Michelin (or Goodyear) is worth more than a simple summation of its operations in Europe and the U.S. would suggest because one operation can be defended by the other. (Without the second operation, the first would have to sink or swim on its own.) Necessarily, the cross-subsidization must not entail throwing good money into an unprofitable situation, but must be a temporary response to a situation that retains a positive present value. Economies of scope from global distribution systems and synergies from knowledge created in different environments provide additional reasons for an MNE to be worth more than the sum of its parts; see, for example Ghoshal (1987) and Douglas and Craig (1989). Economies of scope create value due to an ability to share investments across products and markets, as when two products sharing one distribution channel have a lower cost of joint distribution than the total cost of distributing the two products separately. Synergy is the ability to make two operations work better together than they would separately, and can result from learning; a production operation in one country might be improved by knowledge developed in a second country, and simultaneously the production operation in the second country might be improved by knowledge from the first country. Given all this theoretical discussion of the value of the whole in relation to the sum of the parts, a natural question to ask is whether there is empirical evidence that the whole of an MNE is worth more than the sum of its subsidiary parts. If so, this is further testimony to the importance of studying international management. Unfortunately, the question has not been

112

REID W. CLICK

extensively researched. In a purely domestic setting, empirical evidence in Berger and Ofek (1995) and Comment and Jarrell (1995) reveals that product diversification destroys value because conglomerates lose focus, but literature considering international diversification is relatively scarce (however, see Hitt, Hoskisson, & Kim, 1997; Quian & Li, 1998). One way to approach the question is to consider whether an MNE creates more value than otherwise similar domestic firms. If so, the additional shareholder value is most likely created through international strategic management, the activities that otherwise similar domestic firms do not undertake. 4.1. Data and Empirical Methodology Data are widely available for the subset of all firms that are publicly traded. This study utilizes the Compustat database of publicly traded U.S. firms to compare firms engaged in international business to their domestic counterparts. In particular, it relies on the geographic segment files to distinguish exports from sales of foreign operations and to identify the regions in which sales from foreign operations occur. The Compustat data are likely the highest quality available for U.S. firms. However, they understate the true importance of multinational operations because not all firms with exports or foreign operations report segment sales. Since there is no way to correct for this, analysis proceeds under the assumption that any understatement is small. Of course, the focus on sales highlights the marketing side of the business and ignores the production side, but comparable data on costs are much harder to obtain than the data on revenues and so cannot be examined here. Although there are many possible measures of firm performance, the most common measure in the field of finance is the total return to shareholders (capital appreciation and dividend payments). This return is construed as a summary measure of all the information available on the firm – including all other performance measures such as the return on assets (ROA) frequently studied in the management literature, as in Hitt et al. (1997). In finance, the total return is typically described as an increasing function of the firm’s ‘‘beta’’ from the Capital Asset Pricing Model (CAPM). In a series of papers, Fama and French (1992, 1993, 1995) show that for U.S. data, beta in fact does not contain much cross-section information when firm size and the ratio of book equity to market equity are included as determinants of returns. Financial leverage may have additional explanatory value as well. This paper therefore controls for all of these variables in the investigation of the effects of international business.


113

One way to determine whether an international enterprise has different performance than an otherwise similar domestic firm is to determine whether measures of international involvement are significant determinants of total return, controlling for other characteristics known to affect firm performance. This paper specifically considers five variables measuring international involvement based on exports, sales from foreign operations, and the geographic diversification of sales, and considers their contributions to total returns in cross-section regressions controlling for beta, firm size, the ratio of book equity to market equity, and financial leverage. The measures of international involvement are derived from the Compustat data described above. The simplest type of international involvement is to export some domestic output. Bernard and Jensen (1999) revealed that exporting firms have superior performance when compared to nonexporting firms. Based on this, the empirical analysis here considers the effect of exporting on firm returns by using the Compustat data on export sales to form both a dummy variable for firms reporting exports and the ratio of export sales to the firm’s total sales (the ‘‘export ratio’’). The dummy variable will capture the differential return for exporting firms vis-a`-vis nonexporting firms. If exporting firms have higher (lower) returns than nonexporting firms, the coefficient on the dummy variable will be positive (negative). The export ratio will indicate whether the relative magnitude of exports contributes to differential performance. If exporting firms have higher (lower) returns than nonexporting firms, returns might also be an increasing (decreasing) function of the proportion of exports in total sales. A more complicated type of international involvement is to operate facilities in foreign countries, and this is what is typically studied to help assess whether ‘‘multinationality’’ affects firm performance. The literature in finance is generally inconclusive with respect to the empirical effects of multinationality on total returns. One of the earliest studies, Mikhail and Shawkey (1979), reports that multinational shares outperform domestic shares. However, Brewer (1981) subsequent study reports no significant difference in returns between multinational and domestic firms. After that, Fatemi (1984) reports that returns are identical between multinational and domestic firms except when the multinational operates in competitive foreign markets, in which case shareholder returns are lower for multinationals than for domestic firms. Subsequent research is equally inconclusive, so there is no consensus in the profession on the effects of multinationality on shareholder returns. Empirical analysis here uses the Compustat data on sales of foreign operations to form both a dummy variable for firms with foreign operations

114

REID W. CLICK

and the ratio of sales from foreign operations to the firm’s total sales (the ‘‘multinational sales ratio’’). If MNEs have higher (lower) returns than otherwise similar purely domestic firms, the coefficient on the dummy variable will be positive (negative). In addition, the degree of multinationality might be important, such that returns might be an increasing (decreasing) function of the proportion of sales from foreign operations in relation to total sales. The dummy variable and ratio for sales from foreign operations capture basic information as to whether a firm is an MNE and the overall degree of its multinationality. However, the variables do not capture the degree to which the firm is diversified with respect to its multinational activities. The value of multinationality is often associated with an integrated empire of operations, such that the whole is worth more than the sum of the parts. To the extent that international involvement makes the whole of the MNE worth more than the sum of its parts, it is because there are many parts, in many geographic areas throughout the world. Hence, the degree of geographic diversification is potentially the single most important determinant of shareholder returns due to international involvement. The most common measure of geographic diversification is the entropy measure associated with Theil (1967): ENTROPY ¼

n X

S i ln ð1=S i Þ ¼

i¼1

n X

S i ln ðS i Þ

i¼1

where Si represents the share of operations in the ith region relative to total operations. Sales are typically used as the measure of operations. Compustat reports 1994 sales data broken into seven geographic regions (U.S., non-U.S. North America, South America, Europe, Asia, Oceania, and Africa) for 1,221 firms. With seven geographic regions, entropy theoretically ranges from 0, which represents sales entirely in one region, to a maximum of 1.95, representing sales equally distributed among the seven regions. For the 1,221 firms reporting geographic segment data, entropy empirically ranges from 0 to 1.37 and has a mean of 0.567. A central feature of the entropy statistic is that it contains information not contained in the multinational sales ratio, as distinctions can be made among firms at a particular multinational sales ratio based on the dispersion of the sales from foreign operations. Fig. 1 plots entropy against the multinational sales ratio, revealing of an inverted U shape. A regression estimating entropy as a quadratic function of the foreign sales ratio reveals: ENTROPY ¼ 3:479 RATIO 3:271 RATIO2 ð:260Þ

ð:400Þ


115

1.4

ENTROPY

1.2 1 0.8 0.6 0.4 0.2 0 0

0.2 0.4 0.6 0.8 1 1.2 RATIO OF SALES OF FOREIGN OPERATIONS TO TOTAL SALES

Fig. 1. Entropy as a Function of the Ratio of Sales of Foreign Operations to Total Sales.

with an adjusted R2 ¼ 0.79. This equation implies that on average entropy reaches a maximum at 53% multinational sales. To determine the significance of international business the regressions control for other firm characteristics known to determine shareholder return, at least ex post. The first is the CAPM beta, measuring the risk of the stock against the marketwide portfolio. Compustat reports stock betas estimated by Standard & Poor’s Corporation with five years of monthly data. As an independent variable, beta is expected to have a positive coefficient, such that riskier firms have higher returns. A lagged value of beta is used in order to reduce endogeneity. Guided by findings from Fama and French, several additional control variables are included. Market size is the log of the lagged values of total market equity, which is expected to have a negative coefficient when small firms outperform large firms. The ratio of the book equity to market equity of the firm is the log of the lagged value, and is expected to have a positive coefficient since firms with high book equity in relation to market equity outperform firms with low book- to market-equity. Financial leverage is the lagged value of the ratio of long-term debt to the sum of long-term debt and the market value of equity. The leverage variable is included in both level and squared form, which are expected to have positive and negative coefficients, respectively, suggesting that increasing leverage raises returns but at a decreasing rate and will even lower returns after some maximum.

116

REID W. CLICK

Regressions also control for industry using a set of 13 industry dummy variables based on SIC code. This setup allows the constant in the regressions to vary by industry, since the industrial composition of MNEs is generally shown to be different from that of domestic corporations. The industries covered are agriculture; mining and construction; food and tobacco; textiles and apparel; lumber, furniture, and paper; chemicals; rubber, leather, stone, and glass; metals; machinery, computers, and electronics; transportation equipment; transportation and communications; wholesale and retail trade; and services. 4.2. Results Tables 4 and 5 present cross-section regression results examining determinants of the return to shareholders of nonfinancial firms. Table 4 contains results using the return for 1994 in a sample of 3,678 firms for which data are available. Table 5 contains results using the annual average return for the five-year period 1990–1994 in a sample of 2,891 firms. The independent variables capturing international involvement are based on data for 1994. The dummy variables and ratios for export sales and foreign sales are as described above. Entropy is based on the formula given above; for any firms reporting zero sales from foreign operations, entropy is set equal to zero since all their sales therefore occur in the U.S. To conserve space, Tables 4 and 5 report fairly comprehensive equations containing the main combinations of the five variables representing international involvement. Many other combinations of variables were considered, but are not reported here because the results are consistent with these main combinations. Note that, as is typical in studies of returns, the adjusted R2 of the equations is fairly low – at approximately 11% for 1994 and 6% for 1990–1994. To determine the significance of international involvement, the regressions control for other firm characteristics usually thought to determine shareholder return. Prominent among these variables is beta. For regressions in Table 4 using the 1994 return, beta is based on the period 1990– 1994. For regressions in Table 5 using the 1990–1994 returns, beta is based on the period 1985–1989. The instability of coefficients on beta in crosssection regressions is reflected in the results presented in Tables 4 and 5. Using the single-year returns for 1994, beta is shown to have a negative coefficient, which is an ‘‘incorrect’’ sign according to the CAPM, and is statistically significant. For average annual returns for 1990–1994, however, the sign is positive and significant. According to these results, 1994 was thus

1 Dummy for exports Dummy for multinational sales Export sales ratio Multinational sales ratio Entropy Beta ln(market equity) ln(book/market equity) Leverage Leverage2 Observations Adjusted R2 SEE

8.258 (2.081) 5.899 (2.025)

0.472 (0.133) 4.278 (0.519) 18.182 (1.572) 10.833 (12.815) 25.315 (20.922) 3,678 0.11 50.340

2

3

21.808 (8.417) 10.366 (5.908)

20.870 (8.408)

0.433 (0.135) 4.542 (0.508) 18.531 (1.572) 14.150 (12.932) 29.632 (21.015) 3,676 0.11 50.440

0.412 (0.136) 4.774 (0.477) 18.608 (1.567) 14.822 (12.978) 30.175 (21.074) 3,678 0.11 50.460

Note: Regressions also include industry dummy variables. Significant at 10% level. Significant at 5% level.

4

9.361 (5.905) 0.424 (0.133) 4.530 (0.508) 18.590 (1.573) 11.068 (12.911) 26.623 (20.996) 3,676 0.11 50.487

5

12.531 (3.199) 0.481 (0.128) 4.440 (0.587) 18.831 (1.758) 11.597 (14.246) 30.294 (22.928) 2,974 0.12 50.552


Table 4. Determinants of Shareholder Returns, 1994 (Heteroscedasticity-Consistent Standard Errors in Parentheses).

117

118

Table 5. Determinants of Average Annual Shareholder Returns, 1990–1994 (Heteroscedasticity-Consistent Standard Errors in Parentheses). 1 Dummy for exports Dummy for multinational sales Export sales ratio Multinational sales ratio Entropy Beta ln(market equity) ln(book/market equity) Leverage Leverage2 Observations Adjusted R2 SEE

2

3

4

5

1.177 (0.873) 1.515 (0.886) 4.033 (3.589) 4.340 (2.408) 1.742 0.439 5.063 13.472 25.519 2,891 0.06 19.749

(0.778) (0.199) (0.608) (5.969) (9.095)

1.755 0.459 5.078 14.426 26.743 2,889 0.06 19.744

1.790 0.564 5.137 14.041 26.320 2,891 0.06 19.759

(0.779) (0.185) (0.608) (5.989) (9.112)

4.146 (2.398) 1.794 0.451 5.088 14.074 26.495 2,889 0.06 19.745

(0.779) (0.195) (0.608) (5.967) (9.100)

2.891 (1.149) 1.825 (0.871) 0.478 (0.230) 5.459 (0.696) 12.565 (5.967) 26.399 (10.447) 2,335 0.07 20.002

REID W. CLICK

Note: Regressions also include industry dummy variables. Significant at 10% level. Significant at 5% level.

(0.778) (0.195) (0.607) (5.982) (9.105)

3.630 (3.575)


119

a year in which riskier stocks performed poorly, although the period 1990– 1994 was generally one in which riskier stocks outperformed safer ones. The Fama and French variables have consistent coefficients, but not necessarily conforming to earlier findings. The coefficient on size is positive in both tables, indicating that larger firms outperformed small firms during the early 1990s. This sign contradicts the Fama and French findings, although it is not unusual because the effect of size on firm returns is the subject of some debate in the finance literature. The positive sign on the ratio of book equity to market equity is the ‘‘right sign.’’ This ratio is typically regarded as a more robust determinant of returns in the finance literature, so the result somewhat reaffirms this robustness. The effect of leverage is also as expected: there is a positive coefficient on the variable and a negative coefficient on its square, and the variables are jointly significant. The maximum return occurs in the range of 19–27% leverage. The significance of international involvement as a determinant of shareholder returns is demonstrated by the results provided in the upper rows of Tables 4 and 5. Note that all of the signs are positive, indicating that shareholder returns are consistently positively related to international involvement. Nearly all of the coefficients in Table 4 are statistically significant at the 5% level, and only the multinational sales ratio seems to be a relatively weak determinant. The coefficients in Table 5 are more often insignificant or are only weakly significant (i.e., only at the 10% level). Exporting firms appear to have superior performance when compared to nonexporting firms for 1994, though not for the 1990–1994 period as a whole. For 1994, the dummy variable suggests that exporters had returns 814 percentage points above nonexporters. The export sales ratio suggests that returns were an increasing function of the proportion of exports, such that each percentage point of exports increased returns by 21–22 basis points. MNEs appear to have superior performance when compared to domestic enterprises. The results for 1994 are stronger, both in magnitude and statistical significance, but the results for 1990–1994 still affirm the finding. For 1994, the dummy variable suggests that multinationals had returns nearly 6 percentage points higher than domestics, and the comparable figure for 1990–1994 is 112 percentage points. Although weaker than the results for the dummy variable, the coefficients on the multinational sales ratio suggest that returns were an increasing function of multinationality, such that each percentage point of sales from foreign operations increased returns by about 10 basis points in 1994 or 4 basis points during 1990–1994. Among all the variables capturing international involvement, entropy might be the most important. The coefficient on entropy is statistically

120

REID W. CLICK

significantly positive at the 5% level in both tables, suggesting that multinational diversification increases the total shareholder return. The average firm with an entropy value of 0.567 had a return 7.1 percentage points higher in 1994, and an average return 1.6 percentage points higher over 1990–1994, than an otherwise similar domestic firm (i.e., a firm in the same industry, with the same beta, of the same size, with the same book- to market-equity ratio, and the same leverage). This is strong evidence that the value of international involvement is tied to the degree of diversification in the firm, such that a MNE as a whole is worth more than the sum of its subsidiary parts. Further discussion of this is provided in the conclusion.

5. SUMMARY International strategic management is the process through which value is created by an individual or group of individuals operating across a national border. This definition thus distinguishes international strategic management from general management with the phrase ‘‘operating across a national border,’’ and distinguishes international management from international business by being focused on the ‘‘process through which value is created’’ internationally. This paper answers three questions from the premise that skills creating the most value in cross-border management are the most important elements in the domain of international strategic management. First, how important is international management? The relative importance of international management is determined by the value created internationally compared to that created domestically, and Section 2 presents several statistics that reveal the importance of international activities. Factor income from abroad, which is predominantly profits on direct foreign investment, represents 4.6% of total corporate profits in the U.S. Among publicly traded nonfinancial firms, sales and profits from foreign operations amount to 34% and 24% of corporate sales and profits (respectively) for MNEs, and 20% and 15% of aggregate sales and profits of all U.S. firms. In addition, financial firms create value from their international financial activities. Second, what is the domain of international management? Since the domain is potentially unbounded under the definition provided here, this paper considers a strategy approach in which international management is the aggregation of value created through international production, marketing, and financial activities. Section 3 develops this framework and uses it to


121

define the core domain of international management. The conclusion is that the domain of international management is quite broad even under the criterion of value creation. Indeed, the diversity of international management is one appealing aspect that draws many managers into the field. Third, does international management make the whole multinational corporation worth more than the sum of its parts? Effective management of production, marketing, and financial activities is capable of creating more value in an MNE than in an otherwise similar purely domestic firm. Section 4 presents some preliminary empirical evidence that this is indeed the case for U.S. firms by showing that shareholder returns are positively related to measures of international involvement (and, by inference, the amount of international management). In particular, shareholder returns are positively related to the firm’s geographic diversification; the average firm with an entropy value of 0.567 had a return 7.1 percentage points higher in 1994 (and an average return 1.6 percentage points higher over 1990–1994) than an otherwise similar domestic firm.

REFERENCES Adler, M., & Dumas, B. (1984). Exposure to currency risk: Definition and measurement. Financial Management, Summer, 41–50. Adler, N. (1996). International dimensions of organizational behavior (3rd ed.). Cincinnati, OH: South-Western College Publishers. Aliber, R. Z. (1989). The debt denomination decision. In: R. Z. Aliber (Ed.), The handbook of international financial management (pp. 435–451). Burr Ridge, IL: Dow Jones-Irwin. Amran, M., & Kulatilaka, N. (1999). Real options: Managing strategic investment in an uncertain world. Boston, MA: Harvard Business School Press. Berger, P., & Ofek, E. (1995). Diversification’s effect on firm value. Journal of Financial Economics, 37, 39–65. Bernard, A., & Jensen, J. (1999). Exceptional exporter performance: Cause, effect, or both? Journal of International Economics, 47, 1–25. Besanko, D., Dranove, D., Shanley, M., & Schafer, S. (2003). Primer: Economic concepts for strategy. The Economics of Strategy, 1–37. Boddewyn, J. J. (1999). The domain of international management. Journal of International Management, 5, 3–14. Brewer, H. (1981). Investor benefits from corporate international diversification. Journal of Financial and Quantitative Analysis, 16, 113–126. Comment, R., & Jarrell, G. (1995). Corporate focus and stock returns. Journal of Financial Economics, 37, 67–87. de Meza, D., & van der Ploeg, F. (1987). Production flexibility as a motive for multinationality. Journal of Industrial Economics, 35, 343–351. Douglas, S. P., & Craig, C. S. (1989). Evolution of global marketing strategy: Scale, scope, and synergy. Columbia Journal of World Business, Fall, 47–59.

122

REID W. CLICK

Dufey, G., & Srinivasulu, S. L. (1983). The case for corporate management of foreign exchange risk. Financial Management, Winter, 54–62. Dunning, J. H. (1988). Explaining international production. Boston, MA: Unwin Hyman. Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. Journal of Finance, 47, 427–465. Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33, 3–56. Fama, E. F., & French, K. R. (1995). Size and book-to-market factors in earnings and returns. Journal of Finance, 50, 131–155. Fatemi, A. (1984). Shareholder benefits from corporate international diversification. Journal of Finance, 34, 1325–1344. Ferdows, K. (1997). Making the most of foreign factories. Harvard Business Review, March– April, 73–88. Flaherty, T. (1986). Coordinating international manufacturing and technology. In: M. Porter (Ed.), Competition in global industries (pp. 83–93). Harward Business School Press. Flaherty, T. (1996). Global operations management. New York: McGraw-Hill. Ghoshal, S. (1987). Global strategy: An organizing framework. Strategic Management Journal, 8, 425–440. Ghoshal, S., & Bartlett, C. A. (1990). The multinational corporation as an interorganizational network. Academy of Management Review, 15, 603–625. Hamel, G., & Prahalad, C. K. (1985). Do you really have a global strategy? Harvard Business Review, July–August, 139–148. Harrigan, K. R. (1992). A world-class company is one whose customers cannot be won away by competitors: Internationalizing strategic management. In: A. M. Rugman & W. T. Stanbury (Eds), Global perspective: Internationalizing management education (pp. 251–263). British Columbia, Canada: Centre for International Business Studies, University of British Columbia. Hitt, M. A., Hoskisson, R. E., & Kim, H. (1997). International diversification: Effects on innovation and firm performance in product-diversified firms. Academy of Management Journal, 40, 767–798. Hofstede, G. (1984). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hout, T., Porter, M. E., & Rudden, E. (1982). How global companies win out. Harvard Business Review, September–October, 98–108. Karrenbrock, J. (1990). The internationalization of the beer brewing industry. Federal Reserve Bank of St. Louis Review, November–December, 3–19. Kashani, K., & Quelch, J. A. (1990). Can sales promotion go global? Business Horizons, May– June, 37–43. Knetter, M. M. (1994). Exchange rates and corporate pricing strategies. In: Y. Amihud & R. M. Levich (Eds), Exchange rates and corporate performance (pp. 181–219). Burr Ridge, IL: Irwin Professional. Kogut, B. (1985). Designing global strategies: Profiting from operational flexibility. Sloan Management Review, Fall, 27–38. Kogut, B., & Kulatilaka, N. (1994). Operating flexibility, global manufacturing, and the option value of a multinational network. Management Science, 40, 123–139. Lessard, D. R., & Lightstone, J. B. (1986). Volatile exchange rates can put operations at risk. Harvard Business Review, July–August, 107–114.


123

Levitt, T. (1983). The globalization of markets. Harvard Business Review, May–June, 92–102. Mikhail, A., & Shawkey, H. (1979). Investment performance of U.S.-based multinational corporations. Journal of International Business Studies, 10, 53–66. Morris, W. (Ed.) (1975). The American heritage dictionary of the English language. Boston: Houghton Mifflin. Porter, M. E. (1985). Competitive advantage: Creating and sustaining superior performance. New York: Free Press. Porter, M. E. (1986a). Changing patterns of international competition. California Management Review, 28, 9–40. Porter, M. E. (1986b). Competition in global industries. Boston, MA: Harvard Business School Press. Porter, M. E. (1990). The competitive advantage of nations. New York: Free Press. Prahalad, C. K., & Doz, Y. L. (1987). The multinational mission: Balancing local demands and global vision. New York: Free Press. Quelch, J. A., & Bloom, H. (1999). Ten steps to a global human resources strategy. Strategy & Business, First Quarter, 18–29. Quelch, J. A., & Hoff, E. J. (1986). Customizing global marketing. Harvard Business Review, May–June, 59–68. Quian, G., & Li, J. (1998). Multinationality, global market diversification, and risk performance for the largest U.S. firms. Journal of International Management, 4, 149–170. Ricks, D. A. (1993). Blunders in international business. Cambridge, MA: Blackwell. Shapiro, A. C. (1983). International capital budgeting. Midland Corporate Finance Journal, Spring, 26–45. Standard and Poor’s Corporation. (1997). Compustat PC-Plus. Stobaugh, R. B., Jr. (1969a). Where in the world should we put that plant? Harvard Business Review, January–February, 129–136. Stobaugh, R. B., Jr. (1969b). How to analyze foreign investment climates. Harvard Business Review, September–October, 100–108. Theil, H. (1967). Economics and information theory. Chicago: Rand McNally. U.S. Department of Commerce. (2005). National income and product accounts of the United States. Survey of Current Business, August, pp. 36–172.


OWNERSHIP STRUCTURE, DIVERSIFICATION STRATEGY, AND PERFORMANCE: IMPLICATIONS FOR ASIAN EMERGING MARKET MULTINATIONAL ENTERPRISES Juichuan Chang ABSTRACT The purpose of this study is to provide an integrated framework that conceptualizes multifaceted antecedents pertaining to international expansion of emerging market businesses in relation to firm performance. This paper develops multiple-item measures of multiple dimensions to clarify ownership structure and three diversification strategy relationships to performance. We test how ownership structure and diversification strategy affect emerging market multinational enterprises’ financial performance. The result shows that the relationship between ownership structure and firm performance is a nonlinear relationship (S shape). We also found that excessive international diversification, product diversification, and geographic scope of the expansion process negatively moderate the impact of Asia Pacific multinational enterprises’ performance. Value Creation in Multinational Enterprise International Finance Review, Volume 7, 125–148 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1569-3767/doi:10.1016/S1569-3767(06)07006-3

125

126

JUICHUAN CHANG

1. INTRODUCTION International expansion is accompanied by both opportunities and threats. However, empirical research and theory development have traditionally focused on US and European multinational company data, which may reflect features of the corporate behavior of multinational enterprises (MNEs) in developed countries. However, the rise of emerging market multinational firms is a more recent phenomenon and has attracted limited empirical research attention (Pananond & Zeithaml, 1998). It seems that existing corporate internationalization literature lacks an integrated framework that conceptualizes multifaceted antecedents pertaining to international expansion of emerging market businesses in relation to firm performance. Therefore, the primary purpose of this study is to fill this gap in the context of multinational firms. Asian emerging market MNEs are expected to internationalize and globalize at a faster pace in the future. They enhance their competitive cost advantages and tap foreign markets. Some of Asia MNEs are moving outside their Asian bases to North America and Europe to strategically position themselves for new markets and technologies in the 21st century. Some studies suggest that there are good reasons to believe that firms from lessdeveloped countries differ in their internationalization from firms in developed countries. Therefore, we build a theoretical argument and examine the effects of ownership structure, three diversification strategies – international diversification strategy, geographic scope, and production diversification strategy – of MNEs’ international expansion process in a sample of Asian emerging market MNEs over the period 1998–2002. In particular, our hypotheses were tested on panel data on 115 Asia Pacific MNEs (Taiwan, Singapore, Hong Kong, and South Korea) that expanded abroad. The concept model we test is illustrated in Fig. 1.

2. THEORETICAL BACKGROUND AND HYPOTHESES DEVELOPMENT 2.1. Ownership Structure and Performance Numerous empirical studies have tried to highlight evidence of the relationship between ownership structure and corporate performance. According to agency theory, several categories of shareholders can have an influence on the managers’ efficiency: the managerial shareholders, the financial shareholders, and the institutional shareholders. In this paper, we only focus on

Ownership Structure, Diversification Strategy, and Performance

127

Outside Block Ownership Ownership Structure Managerial Ownership

Diversification Strategy

H1

International Diversification

H2

Product Diversification

H3

Interaction of International and Product Diversification

Performance

H7

H4

Geographic Diversification

H5 H6

Intra-market Extra-market

Fig. 1. Conceptual Model.

managerial ownership. McConnell and Servaes (1990) defined management ownership as equity owned by corporate officers and members of the board of directors. Now three main hypotheses exist to explain the relationship between ownership structure and performance: convergence in interest hypothesis, neutrality hypothesis, and entrenchment hypothesis. The first is Jensen and Meckling’s (1976) ‘‘convergence in interest hypothesis.’’ They show that managerial ownership increases a firm’s value by reducing agency costs. When managers own a large proportion of the firm’s shares, they benefit to a larger extent of the benefits of their effort. The second is Demsetz’s (1983) hypothesis, which shows that corporate performance depends on environmental constraints. This hypothesis is known as the ‘‘neutrality hypothesis.’’ The third is Shleifer and Vishny’s (1989) hypothesis, which suggests that the greater the percentage of shares held by the manager, the less the other shareholders can compel him to manage the firm in their interests. This hypothesis is known as the ‘‘entrenchment hypothesis.’’

128

JUICHUAN CHANG

The nonlinear relation between a firm’s ownership and performance is confirmed by different empirical studies. For example, Morck et al. (1988) indicate that corporate value rises first with increases of internal ownership below 5 percent, decreases between 5 and 25 percent, and finally increases slightly when internal ownership exceeds 25 percent. McConnell and Servaes (1990) show a negative effect of internal ownership between 5 and 25 percent and a nonsignificant one for ownership values exceeding 25 percent. Shleifer and Vishny (1989) argue that the positive relationship between Tobin’s Q and managerial ownership is sustained at higher levels of ownership for small firms than it is for large firms. Empirical research and theory development have traditionally focused on US and European multinational company data; there may also be good reasons to believe that the relationship between ownership structure and performance is different for firms in emerging market MNEs. Therefore, we hypothesize the relationship between the financial performance of Asian emerging market MNEs and their ownership structure as follows: Hypothesis 1. There is a nonlinear relationship (S shaped) between managerial ownership and performance in Asian emerging market MNEs. In this paper, we focus on three diversification strategies: international diversification, geographic diversification, and product diversification. 2.2. The Effect of International Diversification Strategy on Performance There may also be good reasons to believe that the internationalization– performance relationship is different for firms in emerging market MNEs. First, firms in developing countries are generally less advanced in managerial and technological knowledge in comparison with firms in developed countries (Tolentino, 1993). Second, they may face different macroeconomic conditions and institutional environments compared with firms in developed countries (Khanna & Rivkin, 2001). Third, developed countries and developing countries differ in international competitiveness by sector (Nachum, 2004). Some studies have found a positive relationship between internationalization and firm performance (Geringer et al., 1989; Rugman et al., 2000). Some studies have found a negative relationship (Kumar, 1984; Michel & Shaked, 1986). However, recent studies suggest that the form of the relationship is an inverted U-relationship (Geringer et al., 1989; Hitt, Hoskisson, & Kim, 1997). More recently, Ruigrok and Wagner (2003) have found a standard U-shape relationship between internationalization and firm performance.


129

In stage 1, the relationship is a negative slope as costs and barriers to initial international expansion outweigh the benefits. Initially, the emerging market MNEs typically have large learning costs because of unfamiliarity with foreign markets, cultures, and environments. Second, the emerging market early internationalizers often suffer from less efficient production and less internationally competitive knowledge-base industries; they also often suffer from weak or non-existing institutions (Khanna & Rivkin, 2001) as well as underdeveloped capital markets (Singh, 1995). Apart from economic and legal barriers, barriers to emerging market MNEs internationalization include cultural distance (Johanson & Vahlne, 1977) and establishing the firm’s legitimacy abroad (Zaheer & Mosakowski, 1997). For early emerging market internationalizers, the small scale of global operations is insufficient to recoup costs of creating an internationalization (Gongming, 1998; Hitt et al., 1997). In stage 2, the relationship becomes positive as benefits of emerging market MNEs’ international expansion are now realized. Further geographical expansion makes possible efficiencies that improve resource utilization (Kim, Hwang, & Burgers, 1989; Porter, 1985). While market-seeking firms are better able to scan for market opportunities, the incremental benefits of further international expansion are now greater than the incremental costs of a further stage 2 expansion. Other benefits of stage 2 international expansion are the ability of some companies to exercise global market power (Grant, 1987) and to extend the product cycle (Vernon, 1966). Vernon argues that competitive advantages of developing country MNEs do not lie in frontier technology. Rather, they can take the form of adaptation of imported technology, development of products suitable for developing countries, or innovations of small-scale production techniques. For instance, while the manufacturing and commodity sectors have seen a decline in many industrialized countries, the manufacturing and commodity sectors have been growing in many developing countries. Young, Wansley, and Lane (1999) indicate that some unique features of Asia Pacific emerging market MNEs’ internationalization strategies can be highlighted in stage 2: (1) sectoral specialization through vertical integration, (2) diversification into unrelated businesses, and (3) family ownership and management. Therefore, we hypothesize the relationship between performance of a multinational firm and its internationalization as follows: Hypothesis 2. Moderate levels of internationalization are positively related with firm performance, but a high-level internationalization should be negatively related with the firm’s performance.

130

JUICHUAN CHANG

2.3. Product Diversification Strategy and Performance The relationship between product diversification and firm performance has long been studied by management scholars although the results are puzzling, given that most are theoretical arguments. Salter and Weinhold (1981) indicate that a diversified firm can apply particularly skills and knowledge acquired in one business to solve problems and exploit opportunities in other business. Wrigley (1974) distinguished two modes of product diversification: related and unrelated diversification. Then, Hitt and Ireland (1986) found that related diversification facilitates the sharing of activities and the transfer of skills across businesses to increase firm value; therefore, related diversification has potential sources of value increases at both the corporate and business levels. Conversely, unrelated diversification provides few operational synergies and, therefore, must rely on financial synergies for increasing value. Wrigley (1974) suggested that related diversified firms emphasize strategic control to achieve superior performance and that unrelated diversified firms emphasize financial controls to achieve superior performance. However, at some point these efforts are also associated with major costs. For example, Grant, Jammine, and Thomas (1988) highlight the growing strain on top management as it tries to manage an increasingly disparate and less familiar portfolio of businesses. Markides (1992) identifies other costs, such as control and effort losses (due to increased shirking), coordination costs, and other diseconomies related to organization, inefficiencies from conflicting ‘‘dominant logics’’ between businesses, and internal capital market inefficiencies. However, some researchers argue for a curvilinear relationship between product diversification and performance with net benefits increasing to a maximum and decreasing as costs increase. Therefore, we hypothesize that the emerging market MNEs’ moderate levels of product diversification have a beneficial influence on their performance when they did not engage heavily in product diversification, but when emerging market MNEs go far beyond their original industries, then their performance turned out to be negative with further high levels of product diversification. The effect of diversification on business group performance is as follows: Hypothesis 3. Moderate levels of product diversification are positively related with emerging market MNEs’ performance, but a higher product diversification is negatively related with the emerging market MNEs’ performance.


131

2.4. Geographic Diversification of Emerging Market MNEs’ Operation In this study, we examine geographic scope effects on the relationship between internationalization and Asian emerging market MNEs’ performance. There are good reasons to believe that the internationalization– performance shape of the curve may vary with the country of origin of the firm and the markets it enters. The Uppsala model of internationalization (Johanson & Vahlne, 1977) proposed that firms internationalize incrementally from ‘‘psychically close’’ countries to ‘‘psychically distant’’ countries. This would predict a pattern of internationalization in which one would find internationalization in familiar countries in the first stage and internationalization in less familiar countries in the latter stages. By definition, intraregional markets are located at a closer distance. When emerging market businesses first internationalize, they usually enter countries whose culture and business customs most resemble their home (OECD, 1997). If they are in the same time zone, it is also easier to coordinate activities and reduce transportation costs, the costs of coordination, as compared to extraregionally dispersed operations. By concentrating on markets in the same region, firms may also avoid diseconomies of time compression (Vermeulen & Barkema, 2002). Finally, markets in the same region are also likely to be part of the same trade blocs and so benefit from reduced market entry barriers. Therefore, a regional strategy may allow emerging market MNEs to achieve a better balance between the two demands, gaining some of the benefits of globalization while remaining responsive to local market needs (Ghemawat, 2003; Ghoshal, 1987). However, Ghoshal (1987) suggests that as firms concentrate on extra-regional international markets, more exposure to foreign exchange fluctuations systematically increases the foreign exchange risk, and then the agency problem, communication, coordination, and motivation problems stemming from cultural diversity in the MNEs. Therefore, the relationship between internationalization and firm’s performance is a crucial question for researchers. We assume that these effects seem to be confirmed by the preponderance of regionalization strategies among Asia emerging market MNEs. The logic described above suggests the following hypotheses: Hypothesis 4. A higher geographic diversification negatively moderates the relationship between international diversification and the firm’s performance. Hypothesis 5. The internationalization of intra-regional operations will be positively related to the firm’s performance.

132

JUICHUAN CHANG

Hypothesis 6. The internationalization of extra-regional operations will be negatively related to the firm’s performance. 2.5. The Interaction Effects of International Diversification and Product Diversification on a Firm’s Performance It is now widely recognized that internationalization and diversification effects are related. A small number of existing studies have examined the combined effects of product diversification and internationalization. Geringer et al. (1989) found that the effect of the interaction of product diversification and internationalization on a firm’s performance has no significant effects. Franko (1989) indicates that a high level of product diversification in geographically diverse internationalization leads to lower performance. Kim et al. (1989) suggested no effect of global diversification on related diversified firm performance. Hitt et al. (1997) found that high levels of product diversification will raise the cost of governing firms; therefore, excessively high degrees of product and international diversification together should depress performance. The interactive and direct influence of product diversification on international operations has also been elaborated in many previous studies (Geringer et al., 1989; Grant et al., 1988; Hitt et al., 1997; Kim et al., 1989). Overall, these studies indicate that significant interactive effects exist between international diversification and product diversification. We argue that moderate levels of product diversification and internationalization are positively related with firm performance, but a high level of diversification and internationalization should be negatively related with a firm’s performance. Therefore, we hypothesize the interaction effects of internationalization and diversification on business groups as follows: Hypothesis 7. Moderate interaction of international diversification and product diversification are positively related with emerging market MNEs’ performance, but a higher interaction of the expansion process is negatively related with the emerging market MNEs’ performance.

3. METHODOLOGY 3.1. Sample and Data collection First, we identified the Asian emerging market multinational companies from The Forbes 2000, a comprehensive ranking of the world’s top 2000


133

largest companies over the period 1998–2002, measured by a composite of sales, profits, assets, and market value. Second, data on ownership structure, the international operations (foreign sales/total sales and foreign asset/total asset), and return on sales. Returns on asset were obtained from stock exchanges of Taiwan, Hong Kong, South Korea, and Singapore, company annual reports, and supplementary materials comprising group founders’ biographies, governmental statistics, corporation disclosure statements, and business newspaper and journal reports. Firms used in this study had to have a least 10 percent of their sales originating overseas, an implicit criterion used in many of the previous studies (see, e.g., Daniels & Bracker, 1989; Geringer, Tallman, & Olsen, 2000). Companies that did not distinguish between export and foreign subsidiary sales and those that did not provide the figures necessary for calculating the ratio were removed from the sample. Table 1 shows the distribution by four countries and across seven industries. The final sample amounted to 115 multinational firms in Asia Pacific MNEs (Hong Kong, South Korea, Taiwan, Singapore) over the period 1998–2002. The breakdown by country and industry sector is shown in Tables 1a and 1b, respectively. Table 1a.

Sample Distribution of Asia Pacific MNEs across Countries.

Asia Pacific MNEs

Total Samples

Hong Kong Taiwan South Korea Singapore

43 29 33 10

Total

Table 1b.

115

Sample Distribution of Asia Pacific MNEs across Industries.

No. 1 2 3 4 5 6 7

Industry

Asia Pacific MNEs

Conglomerates Consumer goods Food Primary manufacturing Secondary manufacturing Service Utility

7 18 10 15 35 26 4

Total

115

134

JUICHUAN CHANG

3.2. Variables and Measures 3.2.1. Dependent Variable Performance: We measured firm performance by two methods: (1) Return on assets (ROA): income before extraordinary item, divided by total assets. (2) Return on sales (ROS): income before extraordinary item, divided by total sales revenue 3.2.2. Independent Variables Managerial ownership: We define managerial ownership as the percentage of share held by the firm’s board of directors. Degree of internationalization (DOI) was measured using the sum of foreign sales as a percentage of total sales, FSTS (e.g., Grant, 1987) and foreign assets as a percentage of total assets, FATA (Gomes & Ramaswamy, 1999). DOI (degree of internationalization) ¼ FSTS+FATA Intra-regional: Asia Pacific region Extra-regional: outside Asia Pacific region Global region: intra-region+extra-region Product diversification index: We use a Herfindahl-type quantitative index as a measure of firm diversification (Grant et al., 1988). The index satisfies the following important requirements: it varies directly with the number of different products produced; it varies inversely with the increasingly unequal distribution of products across product lines; and it is bounded between zero and unity. The product diversification index takes into account the number of segments in which a firm operates and the relative importance of each segment in sales. X Product Diversification Index ¼ 1 ðsiÞ2 where si is the proportion of a firm’s sales reported in product group i. Geographic diversification: the number of countries in which the firm operates. 3.2.3. Control Variables Size: This was measured by the natural logarithm of number of employees (Gomes & Ramaswamy, 1999) and was used to control for the potential


135

effect of scale economy differences. Logarithmic transformation makes the results easier to interpret, because the changes in the logarithm domain represents the relative (percentage) changes in the original metric and also makes the distribution of data closer to normality. Debt: To control for changes in the firms’ capital structure, we include a leverage variable, measured by total debt over total assets. Table 2 indicates the operationalization of variables, the sources of information, and its previous usage. 3.3. The Empirical Model We use a cross-sectional heteroskedastic, timewise autoregressive model, which makes it possible to reveal individual exploratory variables as well as the dynamics over time. The multiple regression analysis was carried out on five-year (1998–2002) averaged data for internationalization, performance, and other control variables. To test Hypothesis 1, we used a cubic regression model with first-, second, and third-order terms for degree of ownership as follows: Performance ¼ a þ b1 ðmanagerial ownershipit Þ þ b2 ðmanagerial ownershipit Þ2 þ b3 ðmanagerial ownershipit Þ3 þ S b3þm I m þ b3þmaxðmÞþ1 C 1 þ b4 ðoutside block ownershipÞ þ b5 ðSizeÞ þ b6 ðDebt ratioÞ þ it MARS model. In the second stage, we used the Multivariate Adaptive Regression Spline (MARS) methodology developed by Friedman (1991) to determine the turning points in the ownership–performance relationship. The piecewise linear regression model allowing for two changes in the slope coefficient of the degree of internationalization is as follows: Degree of managerial ownership: DMO Performance ¼ a þ b1 ðDMO1i Þ þ b2 ðDMO2i Þ þ b3 ðDMO3i Þ þ b4 ðSizeÞ þ i

136

Table 2. Definition of Variables and Data Sources. Variable Dependent variables Performance

Managerial ownership Outside block Ownership Im m I t

Definition

Adapted From

(1) Return on assets (ROA): income before extraordinary item, divided by total assets (2) Return on sales (ROS): income before extraordinary item, divided by total sales revenue The percentage of shares held by the firm’s board of directors The percentage of shares held by institutions and other external ownership interests Set of dummy variables to control sub-sector effects Knowledge and capital intensive service sector Represents the companies in the study (cross-sectional component) Corresponds to the different years (timeseries component)

Dunning (1995), Grant (1987), Daniels and Bracker (1989)

McConnell and Servaes (1990) McConnell and Servaes (1990) Contractor et al. (2002) Contractor et al. (2002) Contractor et al. (2002) Contractor et al. (2002)

JUICHUAN CHANG

Independent variables International diversification ¼ FSTS+FATA ¼ DOI (degree of internationalization), Sullivan (1994) FSTS Foreign sales as a percentage of total sales Rugman (1986), Haar (1990) FATA Foreign assets as a percentage of total assets Daniels and Bracker (1989), Sambharya (1996) (International diversification)2 Quadratic of international diversification Contractor et al. (2002) P 2 Product diversification Product diversification index ¼ 1 (si) si is the proportion of a firm’s sales in Grant et al. (1988) product group i

Size Debt ratio

The number of countries in which the firm operates Natural log of number of employees Total debt over total assets

Grazialetto-Gilies (1998)

Industries dummies I1 I2 I3 I4 I5 I6

Using the Utility sector as the baseline variable, to reach lowest Cronbach’s alphas 1 ¼ Conglomerates 1 ¼ Consumer goods 1 ¼ Food 1 ¼ Primary manufacturing 1 ¼ Secondary manufacturing 1 ¼ Service

Haveman (1993) Hitt et al. (1997)


Geographic diversification

137

138

If If If If If If If

JUICHUAN CHANG

DMOiolevel one then DMO1i ¼ DMOi DMOi>level one then DMO1i ¼ level one DMOiolevel one then DMO2i ¼ 0 level onerDMOiolevel two then DMO2i ¼ DMOilevel one DMOiZlevel two then DMO2i ¼ level twolevel one DMOiolevel two then DMO3i ¼ 0 DMOiZlevel two then DMO3i ¼ DMOlevel two

4. RESULTS 4.1. The Relationship between Ownership Structure and Firm Performance: A Nonlinear Relationship (S Shape) Table 3 and Fig. 2 confirm Hypothesis 1. The results show that the relationship between a firm’s performance and managerial ownership is a nonlinear relationship (S shape). The results of Models 1 (see Table 3) are strongly significant, with high F-values. The results (Fig. 2 and Model 1) indicate that corporate performance (ROA) rises first with increases of managerial ownership below 8 percent (beta ¼ 0.163, po0.01), decreases between 8 and 23 percent (beta ¼ 1.018, po0.05), and finally increases when managerial ownership exceeds 23 percent (beta ¼ 2.213, po0.001). Hence, the results are in accordance with those obtained by Morck et al. (1988). In the second step, we use a cross-sectional heteroskedastic, timewise autoregressive model and the MARS methodology to determine the turning points in the ownership–performance relationship (Fig. 2). Initially, as the managerial ownership increase, the corporate performance improves. But we also find that the square of the level of managerial ownership is negatively related to a firm’s performance (8 percentomanagerial ownershipo23 percent). As managerial ownership exceeds 23 percent, the managerial ownership is likely to benefit corporate performance because mangers who own enough stock to dominate the board of directors could get entrenched. Therefore, managerial ownership increases a firm’s value by reducing agency costs. As the greater the percentage of shares held by managers, the lesser the other shareholders can compel him to manage the firm in their interest. The results of Table 3 also show that a firm’s performance is positively related to outside block ownership (beta ¼ 0.142, po0.05), indicating that outsider owners are active in monitoring management. The results also show that a firm’s performance is positively related to the firm’s size and negatively related to debt ratio.


Table 3.

139

Regression of Asia Pacific MNEs’ Performance (Samples N ¼ 115) (Cubic Regression Model).

ROA (Dependent variable) (Test Hypothesis 1) Intercept Managerial ownership (Managerial ownership)2 (Managerial ownership)3 Outside block ownership I1 I2 I3 I4 I5 I6 Firm size Debt ratio R2 F

Model 1 1.891 0.163 1.018 2.213 0.142 0.028 0.414 0.136 0.125 0.053 0.079 0.542 0.173 5.394 7.027

Notes: I1: sector effect (dummy, 1 ¼ Conglomerates; 0 ¼ otherwise). I2: sector effect (dummy, 1 ¼ Consumer goods; 0 ¼ otherwise). I3: sector effect (dummy, 1 ¼ Food; 0 ¼ otherwise). I4: sector effect (dummy, 1 ¼ Primary manufacturing; 0 ¼ otherwise). I5: sector effect (dummy, 1 ¼ Secondary manufacturing; 0 ¼ otherwise). I6: sector effect (dummy, 1 ¼ Service; 0 ¼ otherwise). Using the Utility sector as the baseline variable to reach lowest Cronbach’s alphas. po0.05; po0.01; po0.001.

4.2. The Interaction Effects of International Diversification and Product Diversification on a Firm’s Performance In Table 4, Model 3 revealed a clear quadratic relationship between international diversification and performance. The results of Model 3 are strongly significant, with high F-values. International diversification and performance were significantly positively related up to a point in the 0–64 percent range (beta ¼ 1.367, po0.001), after which an increase in international diversification was associated with declining performance (beta ¼ 1.068, po0.01). Therefore, the results support Hypothesis 2. In the case of product diversification, the quadratic term was also similar.

140

JUICHUAN CHANG Performance (ROA)

% 8% 23% Managerial Ownership (The percentage of shares held by the firm’s Board of Directors)

Fig. 2.

Ownership Structure and Performance (MARS Regression Model).

Table 4.

Regression of Asia Pacific MNEs’ Performance (Samples N ¼ 115).

ROS (Dependent variable) (Test Hypotheses Model 2 (Linear Form) 2, 3, 4, and 7) Intercept Internationalization diversification (Internationalization diversification)2 Product diversification (Product diversification)2 Geographic diversification (Internationalization diversification) (Product diversification) (Internationalization diversification)2 (Product diversification)2 Firm size Debt ratio R2 F

1.147 2.692 1.385 2.361

Model 3 (Quadratic Form) 2.636 1.367 1.068 2.183 2.273 2.879 1.363 3.485

0.681 0.219 0.553 6.231

0.724 0.195 0.683 7.248

po0.05; po0.01; po0.001.

The performance is positively related to product diversification in the 0–48 percent range (beta ¼ 2.183, po0.01) and negatively related up to 48 percent (beta ¼ 2.273, po0.05). Therefore, the results support Hypothesis 3. This suggests that the scope of foreign operations may exert great influences on a firm’s performance.


141

In Table 4, Model 3 also shows that the interactive effect of international diversification and product diversification (Internationalization diversification Product diversification) was positively related to performance (beta ¼ 1.363, po0.01), when MNEs did not engage heavily in foreign activities. MNEs’ performance turned out to be negative (beta ¼ 3.485, po0.001) with further high levels of international diversification and product diversification: (Internationalization diversification)2 (Product diversification)2. This suggest that MNEs’ performance will be positive at lower and moderate levels of interaction of international diversification and product diversification, but may be negatively affected with further increase in the degree of international diversification and product diversification. This result supports Hypothesis 7. Model 3 also shows that the geographic diversification is negatively related to a firm’s performance (beta ¼ 2.879, po0.01). This result supports Hypothesis 4. 4.3. Geographic Diversification of Emerging Market MNEs’ Operation In Table 5, we examine geographic scope effects on the relationship between internationalization and Asian emerging market MNEs’ performance, and Regression Models 4 and 5 were used to test Hypotheses 5 and 6. There are good reasons to believe that the internationalization–performance shape of the curve may vary with the country of origin of the firm andP the markets it enters. Model 4 (Performanceintra ¼ a+b1 (DOIit)intra+ b1+m Im+b2 (Firm size)+b3 (Debt ratio)+eit) tests the intra-region international expansion path of Asian emerging P market MNEs and Model 5 (Performanceextra ¼ a+b1 (DOIit)extra+ b1+m Im+b2 (Firm size)+b3 (Debt ratio)+eit) tests extra-region (outside Asia pacific region). Model 6 (PerP formanceglobal ¼ a+b1 (DOIit)intra l+b2 (DOIit)extra+ (b1+m Im+ b2+m Im)+b2 (Firm size)+b3 (Debt ratio)+eit) tests the global international expansion path. The results of Table 5 are strongly significant, with high F-values. The results of Model 4 strongly support Hypothesis 5. The internationalization of intra-regional operations is positively related to a firm’s performance (beta ¼ 0.583, po0.001). On the other hand, the results of Model 5 indicate that the internationalization of extra-regional operations is negatively related to its performance (beta ¼ 0.114, po0.05), the results supporting Hypothesis 6. Why should this be the case? We propose that intraregional international expansion processes reap the positive benefit of internationalization easier, gain the benefits of the multinational network more effectively, easily coordinate activities, benefit from diseconomies of time

142

JUICHUAN CHANG

Table 5. Regression of Performance (Samples: 115 Asia Pacific MNEs). ROS (Dependent Variable) (Test Hypotheses 5 and 6) Intercept DOI (intra-regional markets) DOI (extra-regional markets) I1 I2 I3 I4 I5 Firm size Debt ratio R2 F

Model 4, Intraregion 0.615 0.583 0.317 0.145 0.059 0.113 0.192 0.075 0.248 0.413 9.726

Model 5, Extraregion

Model 6, Global Region

1.891

2.162 0.416 0.106 0.377 0.151 0.001 0.173 0.159 0.182 0.204 0.386 8.326

0.114 0.414 0.136 0.125 0.053 0.179 0.063 0.183 0.372 7.219

Notes: DOI (extra-regional markets): degree of internationalization within intra-region. DOI (extra-regional markets): degree of internationalization within extra-region. DOI (global regional markets): degree of internationalization within global region (global region ¼ intra-region+extra-region). I1: sector effect (dummy, 1 ¼ Electronic/electrical equipment; 0 ¼ otherwise). I2: sector effect (dummy, 1 ¼ Food; 0 ¼ otherwise). I3: sector effect (dummy, 1 ¼ Chemical; 0 ¼ otherwise). I4: sector effect (dummy, 1 ¼ Industrial equipment; 0 ¼ otherwise). I5: sector effect (dummy, 1 ¼ Software and services; 0 ¼ otherwise). Using the transportation equipment as the baseline variable to reach lowest Cronbach’s alphas. po0.05; po0.01; po0.001.

compression, reduce transportation costs and market entry barriers, and allow firms to optimize on the cost and changes in prices of goods, interest rates, labor, and raw materials by shifting operations across nearer nations (Pantzalis, 2001). But extra-regional international expansion processes require a larger global scale before the net benefits of expansion are realized.

5. DISCUSSION This paper contributes to open debate about the link between ownership structure and firm performance. Previous studies considering a nonlinear relation between managerial ownership and a firm’s value report different


143

breaking points probably due to differences in the samples used (Kole, 1995). The results of Table 3 show a statistically significant S shape relationship between managerial ownership and firm performance for the sample of Asian emerging market MNEs. This result suggests that initially, as mangers’ equity increases, their interests coincide more with those of outside shareholders. Initially, as the managerial ownership increases, the corporate performance improves. But we also find that the square of the level of managerial ownership is negatively related to a firm’s performance (8 percentomanagerial ownershipo23 percent) in our samples, the reason is that corporate performance depends on environmental constraints (Demsetz, 1983). This hypothesis is known as the ‘‘neutrality hypothesis’’. As managerial ownership exceeds 23 percent, the managerial ownership is likely to benefit corporate performance because mangers who own enough stock to dominate the board of directors could get entrenched. Therefore, managerial ownership increases a firm’s value by reducing agency costs. The greater the percentage of shares held by managers, the lesser the other shareholders can compel him to manage the firm in their interests. The results are in accordance with those obtained by Morck et al. (1988). Finally, we find that a firm’s performance is positively related to outside block ownership, indicating that outsider owners are active in monitoring management. All of our empirical results of Table 4 confirm Hypotheses 2, 3, 4, and 7. Multiple regression analysis was performed to examine their effects on performance for a sample of Asian emerging MNEs. The results show that international diversification and product diversification will exert both individual and interactive effects on performance. We observed that international diversification had also a curvilinear relationship with performance. This result is fully consistent with Hypothesis 2. Initially (Stage 1), Asian emerging market MNEs typically have large learning costs because of unfamiliarity with foreign markets, cultures, and environments. They are also generally less advanced in managerial and technological knowledge in comparison with firms in developed countries and face different macroeconomic conditions and institutional environments when compared with firms in developed countries (Khanna & Rivkin, 2001). For early emerging market internationalizes, the small scale of global operations is insufficient to recoup costs of creating an internationalization (Hitt et al., 1997). In stage 2, the relationship becomes positive as benefits of emerging market MNEs’ international expansion are now realized; further geographical expansion makes possible efficiencies that improve resource utilization (Kim et al., 1989; Porter, 1985). Asian emerging market MNEs’ internationalization strategies can be highlighted in the second

144

JUICHUAN CHANG

stage: (1) sector specialization through vertical integration, (2) diversification into unrelated businesses, and (3) family ownership and management. Moderate product diversification increases the firm’s capacity to exploit different market opportunities when they engage in foreign activity. In contrast, firm performance will turn out to be negative when MNEs heavily expand their product offerings and geographic market. This is due mainly to the increase in both environmental and technological complexities associated with increase of geographic scope of foreign operations. Movement into more international markets significantly increases managerial transaction costs and information-processing demands. As firms encompassed increasingly broader geographic markets, the costs associated with geographic dispersion began escalating sometimes quite rapidly, thus eroding profit margins (Geringer et al., 1989). This suggests that foreign operations eventually become highly complex and difficult to manage (Hitt et al., 1997). All of our empirical results of Table 5 confirm Hypotheses 5 and 6. The internationalization of intra-regional operations by Asian emerging market MNEs is positively related to its performance. On the other hand, the internationalization of extra-regional operations is negatively related to its performance. The results show that Asian emerging market MNEs’ international expansion paths have significant differences between intra-regional markets and extra-region markets. Why should this be the case? We found several reasons: First, when Asian emerging market MNEs operate in intraregional markets, it could be the case that MNEs may gain a higher return or margins than when they operate in extra-regional markets, because of the greater risk involved in extra-region operations. This would predict a pattern of internationalization in which one would find MNEs’ internationalization in familiar countries in the first stage and internationalization in less familiar countries in the second stage. The Uppsala model of internationalization (Johansson & Vahlne, 1977) proposed that firms internationalize incrementally from ‘‘psychically close’’ countries to ‘‘psychically distant’’ countries. Second, extra-regional international expansion processes require a larger global scale before the net benefits of expansion are realized. Extra-regional subsidiaries in different circumstances ask for different organizational systems and processes (Gupta & Govindarajan, 1991) and companies active in a variety of cultures need to adapt their structures as well (Ghoshal, 1987). Building these systems and structures takes considerable time and attention. Third, as Asian emerging market MNEs typically have more flexibility in intra-regional markets, they are able to react quickly and efficiently to both market and technological changes. Fourth, because extra-regional expansion is accompanied by an increase in organizational and environmental


145

complexity, it is therefore subject to time compression diseconomies. That is, hasty international expansion will over exhaust a firm’s absorptive capacity within a given time frame, thereby causing inadequate structural adaptation that, in turn, triggers negative performance effects. Therefore, Asian emerging market MNEs would be able to choose intra-region markets where the competitive conditions are more suitable for their resource profile compared to being forced to operate in extra-region markets where they might not possess the greatest advantage. Rugman (2000) has shown that most multinational companies tend to serve primarily markets in the intraregion rather than in extra-regions, presumably because of competitive advantages in the home region.

6. CONCLUSIONS This paper develops multiple-item measures of multiple dimensions to clarify ownership structure and three diversification strategy relationships to performance. All of our empirical results confirm our hypotheses. The results of this study suggests that Asian emerging market MNEs’ ownership structure, institutional heritage, industry characteristics, and its own internationalization strategy are likely to create specific preference for operating in differing geographic regions. Although developed countries’ MNEs initially dominated the international competitive landscape, Asian emerging market MNEs have also entered the race owing to the narrowed gap in competitive advantage in international markets (Oviatt & McDougall, 1994). However, Asian emerging market MNEs have certain advantages over developed countries’ MNEs, including greater flexibility, efficiency, quality and advantage-seeking behavior, which allow Asian MNEs to develop capabilities to succeed in the international markets. Finally, because external globalization drivers, internal organization resources, global strategy, and global performance are all multidimensional, we suggest that Asian emerging MNEs must set clear performance goals before formulating their global strategy. It is also important that businesses constantly monitor the external environmental changes and adjust their global strategy accordingly to ensure that the performance goals are being achieved. Hopefully, this study may prove useful to Asian emerging market MNEs facing expansion decisions. However, these findings still require further elaboration. Because of some inherent limitations of this study, several major areas remain unexplored or need further study to verify the conclusions and explore related areas.

146

JUICHUAN CHANG

Moreover, further enlarging the sample may offer a way to secure additional insight that would permit broader, more valid generalization and provide a better base from which to make inferences and predictions.

REFERENCES Contractor, F. J., Kundu, S. K., & Hsu, C. C. (2002). A three-stage theory of expansion of international expansion: The link between multinationality and performance in the service sector. Journal of International Business Studies, 34, 5–18. Daniels, J. D., & Bracker, J. (1989). Profit performance: Do foreign operations make a difference? Management International Review, 29, 46–56. Demsetz, H. (1983). The structure of corporate ownership: Causes and consequences. Journal of Political Economy, 93, 1155–1177. Dunning, J. H. (1995). Reappraising the eclectic paradigm in an age of alliance capitalism. Journal of International Business Studies, 26, 461–492. Franko, L. G. (1989). Unrelated product diversification and global corporate performance. International Strategic Management, 1, 221–241. Friedman, J. H. (1991). Multivariate adaptive regression splines. Annals of Statistics, 19, 1–50. Geringer, J. M., Beamish, P. W., & Costa, R. C. (1989). Diversification strategy and internationalization: Implications for MNE performance. Strategic Management Journal, 10, 109–119. Geringer, M. J., Tallman, S., & Olsen, D. M. (2000). Product and international diversification among Japanese multinational firms. Strategic Management Journal, 21, 51–80. Ghemawat, P. (2003). Semi-globalization and international business strategy. Journal of International Business Studies, 34, 138–152. Ghoshal, S. (1987). Global strategy: An organizing framework. Strategic Management Journal, 8, 425–440. Gomes, L., & Ramaswamy, K. (1999). An empirical examination of the form of the relationship between internationalization and performance. Journal of International Business Studies, 30, 173–188. Gongming, Q. (1998). Determinants of profit performance for the largest US firms 1981-92. Multinational Business Review, 6, 44–51. Grant, R. M. (1987). Internationalization and performance among British manufacturing companies. Journal of International Business Studies, 18, 79–89. Grant, R. M. (1987). Multinationality and performance among British manufacturing companies. Journal of International Business Studies, 18, 79–89. Grant, R. M., Jammine, A., & Thomas, H. (1988). Diversity, diversification, and profitability among British manufacturing companies, 1972–84. Academy of Management Journal, 31, 71–80. Grazialetto, A., & Gilies, R. M. (1998). International diversification: Effects on innovation and firm performance in product diversified firms. New York: Palgrave. Gupta, A. K., & Govindarajan, V. (1991). Knowledge flows and the structure of control within multinational corporations. Academy of Management Journal, 34, 138–152. Haar Jerry., Dandapani, K., & Stanley, P. (1990). The American Depository Receipt (ADR): A creative financial tool for multinational companies. Global Finance Journal, 1, 163–171. Haveman, J. D. (1993). New dimensions in regional integration. Cambridge University Press.


147

Hitt, M. A., Hoskisson, R. E., & Ireland, R. D. (1986). A mid-range theory of the interactive effects of international and product diversification on innovation and performance. Journal of Management, 20, 297–326. Hitt, M. A., Hoskisson, R. E., & Kim, H. (1997). International diversification: Effects on innovation and firm performance in product diversified firms. Academy of Management Journal, 40, 767–798. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm; managerial behavior, agency cost and ownership structure. Journal of Financial Economics, 3, 305–360. Johanson, J., & Vahlne, J. E. (1977). The internationalization process of the firm: A model of market knowledge and increasing foreign market commitments. Journal of International Business Studies, 8, 23–32. Khanna, T., & Rivkin, J. W. (2001). Estimating the performance effects of business groups in emerging markets. Strategic Management Journal, 22, 45–74. Kim, W. C., Hwang, P., & Burgers, W. P. (1989). Global diversification strategy and corporate profit performance. Strategic Management Journal, 10, 45–57. Kole, S. R. (1995). Measuring managerial equity ownership: A comparison of sources of ownership data. Journal of Corporate Financial, 1, 413–435. Kumar, M. S. (1984). Growth acquisition and investment: An analysis of the growth of industrial firms and their overseas activities. Cambridge, UK: Cambridge University Press. McConnell, J. J., & Servaes, H. (1990). Additional evidence on equity ownership and corporate value. Journal of Financial Economics, 27, 315–595. Michel, A., & Shaked, I. (1986). Domestic corporations: Financial performance and characteristics. Journal of International Business Studies, 18, 89–100. Morck, R., Andrei, S., & Robert, W. V. (1988). Management ownership and market valuation: An empirical analysis. Journal of Financial Economics, 20, 293–351. Nachum, L. (2004). Geographic and industrial diversification of developing country firms. Journal of Management Studies, 41, 274–294. OECD (1997). Towards a new global age: Challenges and opportunities, Policy Report. OECD, Paris. Oviatt, B. M., & McDougall, P. P. (1994). Toward a theory of international new ventures. International Business Study, 25, 45–64. Pananond, P., & Zeithaml, C. P. (1998). International expansion process of MNEs from developing countries: A case study of Thailand’s CP group. Asia Pacific Journal of Management, 15, 163–184. Pantzalis, C. (2001). Does location matter? An empirical analysis of geographic scope and market valuation. Journal of International Business Studies, 32, 133–155. Porter, M. E. (1985). Competitive advantage: Creating and sustaining superior performance. New York: Free Press. Rugman, A. M. (1986). Strategies for national competitiveness. Long Range Planning, 20, 92–97. Rugman, A. (2000). The end of globalization: Why global strategy is a myth & how to profit from the realities of regional markets. New York: Amacom. Rugman, A. M., & Verbeke, A. (2000). Six cases of corporate strategic responses to environmental regulation. European Management Journal, 18, 377–385. Ruigrok, W., & Wagner, H. (2003). Internationalization and performance: An organizational learning perspective. Management International Review, 43, 61–84. Salter, M. S., & Weinhold, W. A. (1981). Choosing compatible acquisitions. Harvard Business Review, 59, 117–127.

148

JUICHUAN CHANG

Sambhraya, R. B. (1996). The combined effect of international and product diversification strategies on the performance of U.S.-based multinational corporations. Management International Review, 35, 197–218. Shleifer, A., & Vishny, R. W. (1989). Large shareholder and corporate control. Journal of Political Economy, 94, 461–488. Singh, A. (1995). Corporate financing patterns in industrializing economies. International Finance Corporation, Technical Paper, Vol. 2, Washington, DC. Sullivan, D. (1994). Measuring the Degree of Internationalization of a Firm. Journal of International Business Studies, 25, 325–342. Tolentino, P. E. E. (1993). Technological innovation and third world multinationals. New York: Routledge. Vermeulen, F., & Barkema, H. (2002). Pace, rhythm, and scope: Process dependence in building a profitable corporation. Strategic Management Journal, 23, 637–653. Vernon, R. (1966). International investment and international trade in the product life cycle. Quarterly Journal of Economics, 80, 191–207. Wrigley, L. (1974). Divisional autonomy and diversification. Ph.D. thesis, Harvard Business School, Boston. Young, H., Wansley, J., & Lane, W. (1999). Stock market recognition of internationalization and international events. Journal of Business Finance and Accounting, 12, 263–274. Zaheer, S., & Mosakowski, E. (1997). The dynamics of the liability of foreignness: A global study of survival in financial services. Strategic Management Journal, 18, 439–464.

SUCCESSFUL ADAPTATION STRATEGIES OF MULTINATIONAL ENTERPRISES IN CENTRAL AND EASTERN EUROPE Roxana Wright ABSTRACT The paper explores strategies of adaptation to the environment as employed by multinational corporations in Central and Eastern Europe. Organizations are treated as adaptive systems that have to match the complexity of their environments. The justification of the research lies in the complex nature of the market institutions emerging from transition that emphasizes the need for new managerial frameworks. Adaptive approaches such as vertical integration and/or value-chain development, leveraging autonomy and integration, local knowledge acquisition, and embedding in the social and political environment are explored in their relationship to success in the region.


149

150

ROXANA WRIGHT

INTRODUCTION The present study analyzes the strategic responses of multinational corporations (MNCs) to the relatively idiosyncratic environment that still persists in countries of Central and Eastern Europe (CEE). The environment in the region has gone through periods of rapid shifts through the transition from centrally planned to a market economy. The paper asserts that the MNCs that strive in the CEE economies are the ones that are actively maintaining and renewing their competitive advantages. Successful firms may find ways to limit the impact of institutions that are set or slow changing, take opportunities to turn deficiencies into advantages, and actively acquire and incorporate local knowledge into their operations. This paper analyzes the common patterns of strategic adaptation to the economic environments in CEE. The study takes a dynamic approach with wide applicability to other economies in transition and beyond. The paper starts with a presentation of the theoretical framework that aids in modeling companies’ responses to the complex environment in the region. The patterns of organizational adaptation are classified into strategies of complexity reduction or complexity absorption. The formal hypotheses are derived from the theoretical framework and assert these approaches impact on company performance, such as the potential positive effect of value-chain integration, medium degree of autonomy, active knowledge acquisition, and local embedness. The selection criteria are substantiated in the subsequent section and argue for the inclusion of subsidiaries in sectors with significant foreign investment, growth, and impact from reforms. The performance outcomes measures are presented separately as a combined quantitative and qualitative assessment of company success. The following data and methodology section describes and justifies the use of logistic regression in evaluating the impact of adaptive approaches, measured as quantitative and qualitative variables, on the probability of success. The ensuing results are argued and discussed. It is mainly found that the level of integration may not affect the probability of success due to low internalization potential, a medium level of autonomy is essential for a successful local operation and endogenous learning, that subsidiaries benefit from the general ability of the multinational to process knowledge, and that a strategy of local embedding is not necessarily increasing the probability of success. The closing section reflects on the confirmed effectiveness of complexity reduction strategies as means to anticipate changes toward fully functioning market structures and institutions in CEE. The paper concludes that the patterns of successful adaptation rely on

Successful Adaptation Strategies of MNEs in Central and Eastern Europe

151

responsiveness as well as strong capabilities and learning routines at corporate level.

THEORETICAL FRAMEWORK AND RESEARCH HYPOTHESES The analysis treats organizations as adaptive systems that have to match the complexities of their environments. The nature of transition may mean that multinational firms need new tools to function in dynamic and complex environments. The challenge for the managers is how to model complexity and analyze the implications. The current paper assists academics and managers in meeting this challenge. To handle complexity, it is asserted that two major strategic responses are available to firms operating in the environments of CEE. The first is reduction, translated into strategies that attempt to reduce complexity by bringing it under apparent control such as vertical integration and/or value-chain development, combining existing business models with localized strategies, medium degree of autonomy and integration, and complementary specialization. The second alternative is complexity absorption, defined here as active knowledge acquisition, processing and incorporation, gaining from imperfect market structures/ benefiting from and enhancing imperfect competition, embedding in the social and political environment, and matching local deficiencies with core businesses. Four of these aspects are treated in separate formal hypotheses, as presented below. The rest have been added to the empirical model for completeness and are discussed briefly in the findings. The paper examines the empirical evidence of the impact of such strategies on the firm performance. Each of the hypotheses considers that the complexity reduction and absorption strategies lead to increased performance. The alternative hypotheses are also discussed, suggesting that these strategies could increase transaction costs; engage the firm in a greater level of variance that it is familiar with, which in turn may limit its ability to relate policies and practices in CEE countries to its worldwide system. The research thus explores the possibility that the adaptive approaches considered may have a negative impact on performance outcomes. The hypotheses are complemented by the investigation of potential explanations for either alternative. Hypothesis 1. Primary value-chain organization: Vertical integration has a positive impact on the success of CEE operations.

152

ROXANA WRIGHT

It is asserted that successful MNCs may eliminate inefficiencies in the value chain through vertical integration either by ownership or by building a network of alliances. Eliminating inefficiencies may also mean that MNCs need to be actively developing the capabilities of their local network of suppliers and distributors. We expect that the successful MNC adopts the role of a strategic bridge among organizations that are having difficulties cooperating, thus eliminating inefficient practices in the value chain. The literature1 has shown that the dominant mode of growth in CEE countries has been based on networking and foreign acquisitions. It could thus be hypothesized that companies integrated vertically perform well. There are however challenges in actively building values chains and this strategy may have not been in fact successful. Companies seeking to create or expand a distribution network would encounter difficulties in locating and evaluating distributors – it is thus possible that some MNCs may make poor or expensive decisions. It is also possible that the distribution systems in transition economies in CEE have not achieved their potential costefficiency. The literature (e.g., Dyker, 2001; Pavlinek, 2004) has noted that foreign investors in some of the CEE countries rarely use local firms as firsttier suppliers in their supply networks and develop no or only few linkages with local firms because they find it difficult to acquire the supplies at the required level of sophistication or quality. Research (Radosevic, 2004) has interpreted the predominance of vertical alliances as being driven by unexploited market opportunities and cost differential rather than displacement of competition. We would then expect that as competition intensifies, horizontal expansion and alliances will become more prevalent. Vertical integration may not necessarily have a positive impact on performance. Hypothesis 2. Degree of autonomy and corporate integration: Medium degree of autonomy and loose integration with the corporation lead to successful operation in CEE. It is hypothesized that successful corporations operating in CEE allow for a medium degree of autonomy of their local ventures and a loose integration with the rest of the operations. A medium level of autonomy is essential for a successful local operation: allowing subsidiaries to test strategy/new products and engage in endogenous learning. MNCs may gain from providing subsidiaries with autonomy and refraining from imposing worldwide practices and short-term efficiency targets. A more loosely integrated affiliate has access to the resources of the MNC and local managerial freedom to


153

generate diversity and enhance the corporation’s global capabilities. However, diversity and high degrees of adaptation generate high coordination costs and may result in a continual search within unrelated capabilities. The successful MNC needs to balance the trade-off. A time line may also be associated with the relation between performance and level of autonomy (high autonomy in the beginning creates only medium-term success). Retaining and developing context sensitivity and diversity have been shown to be valuable for the MNC (Lieb-Doczy & Meyer, 2000). From an evolutionary perspective, permitting a variety of managerial practices and organizational structures gives the company a long-term benefit of more options and a greater pool of capabilities. In the idiosyncratic transition environment, loose integration with the headquarters may permit for twoway learning and thus enhance the MNC’s global capabilities. Allowing a local operation to develop capabilities and solutions in the transition context would require not only a reasonable degree of autonomy but also access to resources for developing these capabilities. Literature has suggested (e.g., Meyer, 2000) that foreign acquirers risk losing valuable local capabilities if they concentrate on the transfer of their established best practice and neglect development of variety by fostering indigenous capabilities. Rigid integration of internal routines may increase allocative efficiency, but MNCs may yield more long-term benefits if they allow for some degree of autonomy in local managerial practice, organizational arrangements, and technology. Many local practices may initially be perceived to be inferior, yet they may be better adapted to the environment. Experimentation may be needed to develop new managerial practices that correspond to the local cultural values and resources. New practices developed locally may outperform the established ones. As a caveat, loose integration with the rest of the corporation could sabotage consistency in the MNC’s overall strategy and ultimately have a negative effect on performance. Hypothesis 3. Local capability development: Successful MNC subsidiaries pursue active knowledge acquisition and incorporation in CEE. Successful MNCs may pursue active knowledge acquisition, as well as knowledge processing and incorporation into their operations through specific learning outcomes. MNCs that followed a strategy of using their firm-level capabilities and strong market positions were successful at the beginning of transition. Active generation and development of competitive advantages may be essential in current and future successful operations in

154

ROXANA WRIGHT

CEE. Success may be related to putting together capabilities that build on activation of distinctive local knowledge or technology. Research (e.g., Radosevic, 2004) has found that companies that have acquired essential knowledge about the local environment and clients are in a better position in bargaining with foreign partners. It would thus appear that knowledge of the local market can be effectively traded for technology access. Alternatively, given the fact that local managerial know-how and marketing capabilities are somewhat limited in CEE, MNCs may benefit largely by applying their firm-level capabilities. Successful operation in the transition environments of CEE may imply the use of existing business models and familiar strategy when the MNCs pursue wealthy markets and localized strategies for poor markets. In pursuing top-of-the-pyramid markets in emerging economies, MNCs may rely on proven capabilities to incrementally adapt existing business models and a familiar subsidiary strategy based on controlling resources, extracting knowledge, and leveraging economies of scale and scope. When operating in CEE, it might be more appropriate to develop separate strategies for the wealthy rising class and the poor customers. From a marketing strategy perspective, for example, the emphasis on global brands and products with little adaptation to local demand, distribution structures, and cultures may actually be detrimental in the long term. Although the introduction of global brands may initially take advantage of pent up demand, the mediumand long-run benefits of such strategy are uncertain. Hypothesis 4. Embedding into the local environment: Successful MNCs follow a strategy of embedding their organizations into the local environment. Creating close relations with banks and local governments, lobbying, and pursuing reputation-building tactics (playing an active role as a good citizen in the community) are part of an embedding strategy that many MNCs are following in the CEE region. However, the local embedding may only be a support strategy of diminishing limitations while taking full advantage of opportunities and may only have an insignificant impact on performance.

CHOICE OF FIRMS OPERATING IN CEE For the purpose of the investigation a few selection criteria identify a relevant sample. The focus is on subsidiaries in technology and information


155

technology, consumer goods, transport, metallurgy, and infrastructure companies. These areas are among the sectors with most foreign investment in the CEE; they are rapidly growing sectors in the region and are largely affected by reforms. The selection allows for comparisons across industries and has the benefit of contrasting evidence across successful and unsuccessful companies. These areas encompass not only some of the most successful companies but also among the largest loss-making firms in their respective economies (Political Risk Services, Investment Climate, 2004; Romanian Business Digest, 2004). Only investment projects with foreign equity of over $1 million are selected. UNCTAD (2005) observes a shift of FDI toward services, which can be noticed in CEE countries as well. Most countries have liberalized their services FDI regimes that have made larger inflows possible, especially in industries previously closed to foreign entry, such as utilities. In general, the countries of CEE outside the CIS are characterized by substantial FDI penetration in infrastructure services (e.g., banking, telecommunications, water, electricity). This trend justifies the inclusion of services and IT as areas of investigation in this research. Literature suggests (Radosevic & Rozeik, 2005) that the value creation potential of CEE as a global automotive location has not yet been fully exploited. According to this recent research, finding the patterns of successful strategies of MNC subsidiaries in this sector could bring a significant contribution. The automotive industry in the region is concentrated in Central Europe (Czech Republic, Slovakia, Hungary, Slovenia, and Poland) with vast benefits as regards clustering of supplier network. Potential discovery of clustering in other countries of the CEE and cross-country patterns are possible and could further complement the discussion on the link between business environment and company strategy. Therefore, MNCs’ subsidiaries in the automotive industry are included in the present analysis.

PERFORMANCE OUTCOMES MEASURES The study uses two measures of company performance: numerical measures of profitability (profits and profits as percentage of sales and productivity as revenues per employee are used also in the discussion) that are easy to use in investigating possible correlations between variables, and a qualitative evaluation of performance that classifies companies as successful or unsuccessful. The present research defines a successful company as an organization that

156

ROXANA WRIGHT

fulfills the following conditions: It generates positive profits: As a cautionary note, although we would expect profitability to be a sign of success, there are other factors that need to be taken into consideration – in some cases, losses may be temporal and due to exogenous factors, productive enterprises may show losses and transfer their profits elsewhere, etc. A company may also generate positive profits occasionally, while being a loss maker on average. Calculating an average profit across time may have its limitations due to lack of reliable deflators. It has an output that is not excessively volatile (although possibly fluctuating in response to market conditions, demand, etc.): We would expect that a successful firm shows stable performance not only in terms of profits but also in terms of output. A highly volatile nominal output may be a sign of instability. It invests into fixed assets and has a long-term strategy at the foreign location that reflects a sustainable position in the market: A qualitative interpretation is also necessary for some companies – a firm may not invest because it has spare capacity, it accumulates money for a large project, purchases other firms’ equity, etc.

DATA AND METHODOLOGY The paper treats subsidiaries of MNCs in selected CEE countries as the unit of analysis. The research identifies MNCs’ subsidiaries established in CEE between 1990 and 2004. For statistical investigation, a number of variables have been employed as quantitative and qualitative measures of the aspects considered in the hypotheses. Due to the limited nature of data availability on the variables specified in the model, a combination of a variety of sources has been used. The sources for data used in statistical analysis include the following: Thomson Gale Business and Company Resource Center,2 Mergent Online3 International Company Data, and Kompass,4 USA. Additional information was obtained from The American Chamber of Commerce in CEE – through Mr. Elias Wexler, representative in New York, USA; Arisinvest – Romanian institution promoting foreign direct investment – through Mr. Alexandru Mitroi, representative; AVAS – Romanian Institution for the Valorification of State-owned Assets – through Mr. Viorel Dinu, representative; and the Romanian Chamber of Commerce and Industry – through Mr. Adrian Grecescu.


157

A logistic regression model is used for statistical generalization. The sample of subsidiaries respects the criteria described previously. Data for a total of 570 companies, subsidiaries of MNCs, have been compiled on the variables presented in the methodology description. From these companies, 542 had complete information – this sample has been used for further investigation and in comparative models. The number of companies in various countries across the region generally reflects the amount of FDI existent in each of these countries, although there were some limitations due to availability of data. The representation is reasonably close to expected proportions among manufacturing and services industries in the region. Companies’ representation is illustrated in Table 1. A logistic regression has been used for statistical generalization. The model used for statistical analysis is formulated as follows: pðxÞ ¼ EðyÞ ¼

eu 1 þ eu

where y ¼ success ¼ S u ¼ b0 þ b1 VI þ b2 AD þ b3 AI þ b4 SM þ b5 LC þ b6 MS þ b7 EM þ b8 LD þ b9 RE þ b10 SZ þ b11 IN Probability of success is calculated as pðSÞ ¼

eb0 þb1 VIþb2 ADþb3 AI þb4 SMþb5 LCþb6 MSþb7 EMþb8 LDþb9 REþb10 SZþb11 IN 1 þ eb0 þb1 VIþb2 ADþb3 AI þb4 SMþb5 LCþb6 MSþb7 EMþb8 LDþb9 REþb10 SZþb11 IN

where S – measure of success: 1 – yes, 0 – no (at least one of the conditions for success is not met). A discussion of the conditions considered for successful operation is presented in detail in the methodology description section. Income after tax in equivalent US dollars has been used as profitability measure. The success measure also includes the volatility of output and investment in fixed assets. VI – number of stages in the value chain in which the company is involved. AD – level of adaptation (1 – low to 5 – high). SM – strategic motivation: 1 – market seeking, 2 – efficiency seeking, 3 – knowledge seeking, 4 – hybrid motivation. AI – level of autonomy and integration (1 – low, 5 – high). LC – local capability development (1 – low, 5 – high). MS – market structure: number of local companies at the three-digit level of NAISC industry code.

158

ROXANA WRIGHT

Table 1. Companies Representation in the Sample by Industry and Country. Companies Representation by Industry

Companies Representation by Country

Industry

% of MNC Subsidiaries

Country

% of MNC Subsidiaries

29

Albania

1

19

Belarus

1

Manufacturing, high technology Manufacturing low technology Construction Retail, food, beverages, tobacco Retail, IT Retail, pharmaceuticals, and other Wholesale, food, beverages, tobacco Wholesale, IT, equipment, cars Wholesale pharmaceuticals and other Banking, insurance, accounting, administrative management Utilities generation and distribution Telecom Hotels, transportation, other Total

Source: Author’s own.

3 2

Bosnia and Herzegovina Bulgaria

3 2

Croatia Czech Rep.

2

Hungary

5

6

Estonia

5

4

Latvia

3

23

Lithuania

2

2

Macedonia

Less than 1%

2 3

Moldova Poland

Less than 1% 25

Romania Russia Serbia and Montenegro Slovakia Slovenia Ukraine

8 5 Less than 1%

100

Less than 1% 6 3 22

8 2 2


159

EM – local embedness (0, 1). LD – matching local deficiencies (1–5). Control Variables: RE – regional effects: 1 – EU, 2 – US, 3 – Other. SZ – size: number of employees. IN – industry: 1 – manufacturing, high tech; 2 – manufacturing, low tech; 3 – construction; 4 – retail food, beverages, tobacco; 5 – retail IT, equipment; 6 – retail pharmaceutical and other; 7 – wholesale food, beverages, tobacco; 8 wholesale IT, equipment; 9 – wholesale pharmaceutical and other; 10 – banking, insurance, accounting, administrative management; 11 – utilities generation and distribution; 12 – telecom; 13 – hotels, transportation, farming. The MNCs’ subsidiaries included in the analysis represent successful as well as unsuccessful companies in approximately equal numbers. The average revenues across the sample are of USD 2,160,000, with an average number of 1,700 employees. The majority of the firms come from the European Union. The parametric descriptive statistics show there is significant positive correlation (significance below 10% in a two-tails test) between the success measure and the level of local embedding, and between the success measure and the industry. Nonparametric correlation measures (Kendall’s Tau and Spearman’s Rho) are also significant and positive between the success variable and the local embedding variable and also between the success variable and the industry. Nonparametric correlations are significant and negative between success and size variables. Nevertheless, the correlations between variables do not grant any conclusion of causal relationships or impact on probability of success. The findings of the logistic regression presented below allow for such generalizations.

RESULTS AND INTERPRETATION The statistical results for the complete model (570 observations, with 29 missing cases representing 5.1% of the data) are not significantly different from the model that included only the sample for which all information was available (542 companies, with no missing cases). For this model, results show no inordinately large parameter estimates or standard errors, which means that there is no reason to suspect a problem with outcome groups perfectly predicted by any variable. There is also no indication of violation of the linearity in the logit for the model proposed. The results show no

160

ROXANA WRIGHT

problem with convergence, nor are the standard errors for parameters exceedingly large. No multicollinearity is evident, although there is some correlation between the level of adaptation variable and local capability development variable. No outliers were found. A test of the full model with all predictors against a constant-only model was statistically reliable, with a chi-square of 31.129 with low significance (0.01), showing statistical significance at 1% level. This indicates that the predictors, as a set, reliably distinguish between successful and unsuccessful firms. The variance in level of success accounted for is small, however, with Nagelkerke R square and Cox and Snell R square values low. Prediction of success is reasonable, with an overall success rate of 61%. More specifically, the value of chi-square as a measure of improvement due to the introduction of the independent variables (likelihood ratio as a measure of goodness of fit) and its low p-value show that the outcome might be predicted from the set of variables considered. From the two measures of strength of association (low values for Nagelkerke R square and Cox and Snell R square), we cannot conclude on a strong association between the outcome and the variables. The log likelihood high value may also be an indication of imperfect fit. However, Tabachnick and Fidell (2001) suggest that for large samples differences between models might have no practical importance. For large samples, classification might be good (as is the case here) even though the model deviates from a perfect model. The authors conclude that a statistical difference between a fitted model and the observed frequencies may not indicate a poor model with large samples. The analyst should thus keep in mind both the effects of sample size and the way the test works while interpreting the results. The classification has a cut-off probability criterion of 0.5, and the 61% is reasonable, as a method of assessing the success of the model. The classification evaluates the ability to predict correctly the outcome category of cases for which the outcome is known. The only way to improve the overall accuracy of classification is to find a better set of predictors. The classification is reported for various models, however, and no significant contribution to the correct percentage seemed to be added by other models. Selected results for the logistic regression run with no missing cases are included in Table 2. Based on the original model, the results show (Wald test) that three variables have an impact on the probability of success: level of adaptation (AD), market structure (MS), and industry (IN) at the 20% level of significance. Only the industry variable is significant at less than 1% significance level. These findings suggest that higher levels of local adaptation as


Table 2.

161

Summary of Results for Logistic Model.

Summary of Tests/Results Omnibus tests of model coefficients Measures of strength of association Classification

Variables in the Equation VI AD SM AI LC MS EM LD RE SZ IN

w2 ¼ 31.126

Significance ¼ 0.001

Cox and Snell R2 ¼ 0.056

Nagelkerke R2 ¼ 0.075

Overall percentage ¼ 60.900

Cut value ¼ 0.500

Coefficient

Wald

Significance

0.052 0.179 0.044 0.098 0.078 0.012 0.129 0.018 0.009 0.000 0.114

0.173 1.929 0.384 1.329 0.298 1.692 0.349 0.018 0.003 0.018 22.475

0.677 0.165 0.535 0.249 0.585 0.193 0.555 0.893 0.956 0.895 0.000

Source: Author’s own.

reflected in use of local brands, supply system, local management personnel, business models, etc. increase the probability of success for a multinational subsidiary in the CEE region. More competition in the market also increases the probability of success. The findings regarding industry show that companies operating services have a higher probability of success. The high significance level for the adaptation and market structure variables should be noted. Although these variables could be interpreted to have some impact on success, the set probability of making a Type I error is high (20%). The part of the distribution that remains under the curve for the known population but is beyond critical value in the region of rejections is large. Only at this set probability are the two variables significant and the interpretation above is valid. Such high potential for a Type I error cautions against the limitations of the statistical findings regarding the adaptation and market structure impact on success of CEE subsidiaries. Because some level of correlation between AD (level of adaptation) and LC (local capability development) was found, a model from which the AD

162

ROXANA WRIGHT

variable was removed was run. The results are similar to the original model, as regards goodness of fit and classification. The MS and IN variables were again the only significant ones. The inclusion of interaction terms and the application of stepwise procedure provided no additional information. The model run with the entire sample, including the missing cases, shows slightly better goodness of fit and classification, but the variables found significant in their impact on the probability of success are the same as in the results for the model run excluding missing cases. Reduced models (various combinations of explanatory variables) show low practicality. Results based on statistical inference show evidence supporting the alternative of Hypothesis 1 – vertical integration does not necessarily lead to successful operation in CEE. Although it was suggested that successful MNCs operating in CEE may use vertical integration to eliminate inefficiencies in the value chain, bridging across activities may not necessarily increase the probability of success. Successfully supporting suppliers to reach the level of quality required by their operations, as well as other ways of creating and improving distribution networks have proven to increase the abilities of foreign subsidiaries and the opportunities for profit at the beginning of transition. We must consider, however, that the capabilities of local suppliers, distributors, and other members of the value chain have evolved significantly in recent years. The finding that a company’s level of integration may not affect the probability of success could be related to the fact that foreign firms do not have to ‘‘internalize’’ different levels of operations, as local alternatives have similar potential. On the other hand, in countries and industries where supply and distribution systems are not developed satisfactorily, challenges in actively building an integrated value chain have not been successful, due to high costs, administrative blockages and a lack of experience/ability in creating a local or regional value chain. The statistical findings support Hypothesis 2 – medium degree of autonomy and loose integration with the corporation lead to successful operation in CEE. This result suggests that an intermediate level of autonomy is essential for a successful local operation, in that it lets subsidiaries test strategy and try different local tactics, allowing for endogenous learning processes to take place. A more loosely integrated affiliate has access to the resources of the MNC and local managerial freedom to generate diversity and enhance the corporation’s global capabilities. As discussed previously in hypothesis definition, retaining and developing context sensitivity and diversity can be valuable for the MNC. From an evolutionary perspective, permitting a variety of managerial practices and organizational structures gives the company a long-term benefit of more options and a greater pool of capabilities.


163

In the idiosyncratic transition environment, loose integration with the headquarters may permit two-way learning and thus enhance the MNC’s global capabilities. Allowing a local operation to develop capabilities and solutions in the transition context requires not only a reasonable degree of autonomy but also access to resources for developing these capabilities. Experimentation may be needed to develop new managerial practices that correspond to the local values and resources that outperform the established ones. The results of the model presented here with regard to Hypothesis 3 advocate that MNCs do not necessarily increase their probability of successful operation in CEE countries by developing subsidiary-level capabilities. Although it is conceivable that companies that have acquired essential knowledge about the local environment and clients are in a better position, the ability to acquire the local knowledge may come from the general ability of the multinational to process knowledge. Although networking, establishing connections with national and local governments, and other behaviors of embedding in the local environment are considered by previous literature as essential for success in CEE, the findings here suggest that establishing close relationships with governments and communities is not related to the success of CEE subsidiaries (findings relating to Hypothesis 4). Creating close relations with banks and local governments, lobbying and pursuing reputation building tactics (playing an active role as a good citizen in the community) are part of an embedding strategy that many MNCs have followed in the CEE region. However, the local embedding may only be a support strategy of diminishing limitations while taking full advantage of opportunities, and may only have an insignificant impact on performance on its own. Finally, the level of adaptation is found to have a significant impact on the probability of success. High adaptation as regards local management, local distribution systems, local products, and brands appears to be the key in operating winner subsidiaries. Finally, one of the control variables (industry) is noteworthy in its impact on success. Conversations with industry regulators, managers, and representatives of foreign chambers of commerce have consented on the high profitability in the services sectors. In fact, the representative of the American Chamber of Commerce for CEE noted that most US multinationals are services companies, as the success expectation in the manufacturing sector is very low. The results presented support this expectation. The findings show no evidence to suggest that success may be related to strategic motivation of MNCs operating in the CEE region or to the capability of the firm to match local deficiencies.

164

ROXANA WRIGHT

In conclusion, it can be inferred that the strategies of complexity reduction are most effective in increasing the probability of success of a foreign subsidiary in CEE. The findings suggest low impact of complexity absorption strategies that may reflect the fact that thriving operations in countries in this region do not attempt to ‘‘absorb’’ the environment but rather, they anticipate change and convergence toward features of fully functioning market structures and institutions, adapt to the local customer, and reduce the perceived complexity of the setting. Although they may use localized strategies, they draw from strong multinational and headquarters capabilities. They also find a trade-off between high levels of local autonomy that encourage experimentation (with possible loss of consistency, and continuous testing of unsuccessful strategic avenues) and low levels of autonomy hindering learning and potential synergy feedback. The services sector is more conducive to success. Competitive markets are usually the most dynamic and provide more incentives for strategies that generate sustainable advantages. The results of this analysis show the evolution of winning strategies in CEE in recent years. As the markets progress, competition increases and consumers become more sophisticated, companies need the backing of strong capabilities, high responsiveness to the market, and an adequate level of autonomy that allows for flexibility but does not promote wasteful iterations on losing strategies.

CONCLUSIONS The present study investigates MNCs’ strategies to cope with resilient components of the idiosyncratic transition environments in CEE (such as embedding and drawing advantages from regional deficiencies) as well as strategies to react to dynamic components of the environment (such as knowledge acquisition and building or improving value chains). The research argues the appropriateness of a business model that allows for local responsiveness and relates successful implementation of strategies to local autonomy and combination of resources across company network. The dynamic environment in the region is viewed as an incubator for new and innovative strategies. The assumption is that of ‘‘accumulated wisdom’’ at the location, however, the adaptive strategies are viewed to be either positive (successful adaptation) or negative as a result of complexity and imperfect foresight. The descriptive and inferential statistics using a logistic model show that vertical integration does not necessarily lead to successful operation in the


165

CEE region, whereas a medium degree of autonomy does. One of the main statistical findings is that the MNCs do not necessarily increase their probability of success by developing subsidiary-level capabilities but rather can rely effectively on proven capabilities and a familiar strategy. Establishing close relations with governments is not related to success. It would appear that where companies have fewer options to rely on market mechanisms, they attempt to absorb the local environment. Strategies of complexity reduction are most effective in increasing the probability of success of a foreign subsidiary in CEE. Thriving operations in the region do not attempt to ‘‘absorb’’ the environment but anticipate change and convergence toward features of fully functioning market structures and institutions, adapt to the local customer, and reduce the perceived complexity of the setting. Although they may use localized strategies, they draw from strong multinational and headquarters capabilities. They also find a trade-off between high levels of local autonomy that encourage experimentation (with possible loss of consistency and continuous testing of unsuccessful strategic avenues) and low levels of autonomy hindering learning and potential synergy feedback. The main contribution of the research presented here is that of a clear demarcation of what constitutes successful adaptation to the environment. The strategic approaches analyzed represent patterns of solutions. As a general and final conclusion it is noted that the best performing strategies are those that are based on decisive action ensuring high quality learning and responsiveness. Leveraging strong global and regional capabilities and the knowledge acquisition experience of the firm are prerequisites for success. The limitations of the research presented here are related to the limited access to information and the accuracy of shared information from primary sources. The statistical methodology is based on qualitative data that are drawn from existing evidence. The quality of the data and the results depends on the quality and sometimes the interpretation of this evidence. The statistical results have limitations due to their general nature. The companies included represented as many heterogeneous agents across which the generalization occurred. The future research agenda could be expanded to other transition economies; also as an opportunity to test the validity of the framework in a different region with a similar environment. A comparable model may be applicable to MNCs’ subsidiaries operating in dynamic environments in general. The discussion of data included in the paper includes references to distinctive patterns of adaptation for companies from various home countries. The model can be expanded to include explanations related to cultural and psychic distance. Future research can focus exclusively on

166

ROXANA WRIGHT

multinationals from a particular country, with potential findings’ justifications drawing upon specific relationships between home and host country. The present research reaches across industries and generalizes on overall strategies. Industry-specific subsets of strategies can improve the accuracy of the findings and add essential details. As multinational subsidiaries become established in the region, future investigation can emphasize on the requirements to build sustainable advantage. The findings suggest that some companies with a ‘‘diversified’’ portfolio of knowledge and strategic options might be in a position to catch up from an inferior position. It would be interesting to see whether such cases exist or can be predicted in the near future. Transition toward market structures in the region may have also meant a transition in companies’ strategies – from using their firm-level capabilities to developing firm-level practices of incorporating new capabilities and local innovations.

NOTES 1. A case study of a French MNC (Yoruk, 2001) has shown that the company has been successfully supporting its suppliers to reach the level of quality required by its operations. In a study of marketing obstacles in a number of CEE countries, Batra (1996) has found that MNCs have been at the vanguard of creating and improving distribution networks at the beginning of transition. 2. Business and Company Resource Center is an integrated resource with company profiles, brand information, rankings, investment reports, company histories, chronologies, and periodicals. The information on companies (MNCs’ subsidiaries) located in CEE countries provided by this source includes addresses and contact information, a complete classification by NAISC codes, annual sales and profits, number of employees, productivity, and company management and profile. 3. Mergent Online provides Internet-based access to a global company database, including business description, history, property, subsidiaries, officers, and directors, as well as financial statements of subsidiaries in CEE countries. The International Annual Reports module is integrated into the International Company Data module, Mergent’s global company database. The coverage includes subsidiary contact information, company profile, primary NAISC classification, business summary, and financial highlights. 4. The main source of data for this research is Kompass Database. Kompass provides business information mainly for the international commercial and industrial community. Information collected using this database includes company information, profile, and financial data. This database is particularly useful as it reveals CEE subsidiaries’ associations with local organizations, a list of the MNC’s subsidiaries and their activities, a comprehensive history of the subsidiary, a description of operations owned in the primary value chain as well as a complete financial profile and a discussion of the local activities/operations.


167

ACKNOWLEDGMENTS The author would like to thank Dr. Massood Samii, Chair of International Business, Southern New Hampshire University, Dr. Phillip Vos Fellman, Professor of International Business, Southern New Hampshire University, Dr. Yusaf Akbar, Professor of International Business, Southern New Hampshire University, and Dr. Bulent Aybar, Professor of International Business, Southern New Hampshire University for their support and guidance on this paper.

REFERENCES Batra, R. (1996). Marketing issues and challenges in transitional economies (Working Paper No. 12). Ann Arbor, MI: William Davidson Institute. Dyker, D. (2001). The dynamic impact on the Central-Eastern European economies of accession to the European Union: Social capability and technology absorption. Europe-Asia Studies, 53(7), 1001–1021. Lieb-Doczy, E., & Meyer, K. (2000). Context specificity of post-acquisition restructuring: An evolutionary perspective (Discussion Paper No. 6). London: London Business School, Center for New and Emerging Markets. Meyer, K. (2000). International business research on transition economies (Working Paper No. 32). Copenhagen, Denmark: Copenhagen Business School, Center for East European Studies. Pavlinek, P. (2004). Regional development implications of foreign direct investment in Central Europe. European Urban and Regional Studies, 11(1), 47–70. Political Risk Services. (2004). March 1 release: Investment climate – Romania country conditions. From The PRS Group International Country Risk Guide online http:// www.prsgroup.com/icrg/icrg.html Radosevic, S. (2004). Growth of enterprises thru alliances in Central Europe (Working Paper No. 36). London: University College London, Centre for Study of Social and Economic Change in Europe. Radosevic, S., & Rozeik, A. (2005). Foreign direct investment and restructuring in the automotive industry in Central and East Europe (Working Paper No. 53). London: University College London, Center for the Study of Economic and Social Change in Europe. Romanian Business Digest. (2004). Larive Romania: Direct investments in Romania. From http://rbd.doingbusiness.ro/2004_02/larive2004.pdf Tabachnick, B., & Fidell, L. (2001). Using multivariate statistics (4th ed.). Needham Heights, MA: Allyn & Bacon. UNCTAD. (2005). April 4 press release: World FDI flows grew at an estimated 6% in 2004. From http://www.unctad.org/Templates/webflyer.asp?docid=5700. Yoruk, D. E. (2001). Industrial integration and growth of firm in transition economies: The case of a French multinational company (Working Paper No. 20). London: University College London, Centre for the Study of Economic and Social Change in Europe.


THE RELATIONSHIP BETWEEN ORGANIZATIONAL STRUCTURES AND PERFORMANCE: THE CASE OF THE FORTUNE 500 Nicole Avdelidou-Fischer ABSTRACT This paper investigates the relationship between organizational structures and the performance of FORTUNE 500 companies, which have always been among the most profitable and admired in the world. After a discussion of whether companies should organize regionally, nationally, or globally, the important assumption is made that each structural type utilizes resources differently in generating profit. Performance is conceptualized as Return on Capital Employed (RoCE) and Return per Employee (RpE). A sample of 50 companies was randomly selected. Testing revealed that structural types are positively related to financial performance, calculated as RoCE, with Multidivisional-structured companies outperforming Functional-structured ones; structural types are not related to human resource performance, calculated as RpE.


169

170

NICOLE AVDELIDOU-FISCHER

INTRODUCTION Problem Statement The National Bureau of Economic Research declares that the United States entered recession in March 2001 (BBC-News, 2001b). A slew of companies from various sectors issue profit warnings, following the terrorist attacks on New York and the Pentagon (BBC-News, 2001a). In October 2002, BBC broadcasts that the US economy is still struggling to banish the specter of recession (BBC-News, 2002). For Ramesh (2002) the cause is the overinvestment and over-borrowing spree of the 1990s and the slowdown is a welcome corrective to these excesses. Miller (cited in Harrington, 2003) makes CFOs responsible for using the old accounting rules to pump up revenues, so as to have a higher stock price. A number of scapegoats could be blamed for this depressing state of affairs – terrorism bringing a sluggish economy, 2002’s sagging stock market, and a seemingly unending string of scandals come to mind. Still, given all the stumbles and falls in corporate America over the past years, most of the FORTUNE 500 companies look like Olympic sprinters (Boyle, 2002), considering that Wal-Mart, the top company on the 2002 list made over $6,670 million and even the bottom end, The New York Times, made over $444 million profits – and that in bad times. No wonder that FORTUNE 500 companies have always been among the biggest, most profitable, and most admired companies in the world (Kahn, 1998). Creating a multinational organization like GM or Coca-Cola that can respond quickly and creatively when crises strike or opportunity knocks, requires more than luck. So why do some companies perform better than others?

Possible Explanations Early economic theory depicted the organization as synonymous with the larger-than-life entrepreneur, the ‘‘peak coordinator’’ (Mintzberg, 1983; Papandreou, 1952). But experts agree that old style CEOs who were driving organizations to make a number rather than do the right thing, are out of fashion, or awaiting prison time; a different kind of leadership is called for (Gertner, 2002; Sellers, Mulcahy, & Notebaert, 2002; Stein, 2000). This confession would definitely be a relief for Mintzberg (1983, p. 115), who predicted that converting external influence into internal action effectively,

The Relationship between Organizational Structures and Performance

171

is beyond the capacity of one person. The CEO must coordinate people and align tasks through good organizational design; it is the ‘‘whole’’ that matters. Bower (2003) agrees that a person’s effectiveness, job satisfaction, and zest for work are all vitally affected by the structure of the organization in which he or she works. Back in 1998, Kahn highlighted that global strategy with a vision was what this year’s winners had in common. They were in the forefront of creating the truly multinational corporation, demonstrating abilities to seize opportunity from chaos and commit to long-term plans (Kahn, 1998). But even if the thought global strategy is as smart as P&G’s or Gillette’s, it will not get the firm far if it cannot be implemented. ‘‘You cannot be adaptive if there are 12 levels of decision-making for any given change’’ says Pfau, vice president of the Hay Group, and argues that an effective, flat organizational structure is necessary (Kahn, 1998, p. 208). Whittington (1993, p. 112) concurs that management spends more time and energy on choosing strategies rather than implementing them, since even strategies that are well chosen fail because of poor implementation. Getting the organizational structures right for a particular strategy is thus clearly critical to practical performance. What these possible explanations have in common is that they all come down to one term: organizational structures. As suggested by OECD (1987, pp. 7–9), the way in which multinational enterprises structure and organize their activities is of great managerial importance, raising a general and wellknown issue; governments and labor are notably interested to the extent that these will influence country benefits from direct investment, such as the level and type of investment, employment, or R&D. Equally, employee representatives have a legitimate interest to be informed about the decisionmaking structure within the enterprise. Robbins and Coulter (2002) stress the importance of measuring organizational performance and emphasize how vital an organizational structure is, since it is the vehicle through which managers can coordinate the activities of the various functions or divisions to exploit fully their expertise and capabilities. Business failure frequency is increasing in the high-competitive world industries (Eidleman, 1995). Several studies have concluded that soft keys, such as management and employees’ alignment are key determinants in producing value (CGEY, 2000; Ernst & Young, 1997; KPMG, 1999). And also for Porter (1980, pp. 50–51), organizational structure is one component of competitor analysis that requires comprehensive diagnosis. Although literature underlines the importance organizational structures play in financial performance, and although earlier studies have chosen the FORTUNE 500 list as population of interest (Bethel & Liebeskind, 1993;

172


Capon, Farley, Lehmann, & Hulbert, 1992; Cummings, 2004; Horgan & Garnick, 2000; Moore & Reichert, 1983; Walker & Bain, 1989) there are, to my knowledge, no prior studies that have examined the direct relationship between FORTUNE 500 company performance and structural types. The purpose of this paper is to answer the following question: Is there a relationship between organizational structures and the performance of FORTUNE 500 companies?

REVIEW OF THE LITERATURE Organizational Structures The organization first comes into being when an initial group of influencers join together to pursue a common mission (Mintzberg, 1983, p. 22). The way individuals relate to work towards this goal is determined by the structure of the organization (Duncan, 1981). Chandler defines structure as the design of organization through which the enterprise is administered: This design, whether formally or informally defined, has two aspects. It includes, first, the lines of authority and communication between the different administrative offices and officers and, second, the information and data that flow through these lines of communication and authority. Such lines and such data are essential to assure the effective coordination, appraisal, and planning so necessary in carrying out the basic goals and policies and in knitting together the total resources of the enterprise. (Chandler, 1966, p. 14)

Simply, organizational structure is the process by which organizations formally divide, group, and co-ordinate job tasks. The value creation activities of organizational members are meaningless unless some type of structure is used to assign people to tasks and connect activities (Galbraith, 1973). Coordination of effort within a function and between functions within an organization is by no means an easy task (McKenna, 2000, p. 423). Each function needs a structure designed to allow it to develop its skills and become more specialized and productive. Left to themselves, the functions may have little to say to one another, and value creation opportunities will be lost (Hill & Jones, 2001, p. 384). Accordingly, scholars suggest that organizational design is related to the revenue side of the equation (Hill & Jones, 2001, p. 386; Yunker, 1983, pp. 51–64). New information technologies, especially groupware and client-server, fast changing customer needs, competitor offerings, and more complicated products, require better integration of manufacturing, design, and marketing


173

functions. The workforce is becoming more heterogeneous sexually, racially, culturally, individually, and so on due to changing demographics and globalization of the labor market. This diversity is a source of conflict/ communication problems because of different styles of interaction, presentation, physical appearance, and dress (Borgatti, 2001). Companies cannot afford to ignore the fact that they are now operating in extremely complex environments where survival depends on the ability to understand and respond to multiple demands and opportunities (Ghoshal & Bartlett, 1999, p. 121). As Welch puts it, ‘‘Anytime there is change, there is opportunity. So it is paramount that an organization gets energized rather than paralyzed’’ (Kahn, 1998, p. 207). Organizational structures provide the framework for a social-operational-control system, influencing greatly individual and group behavior, helping organizations actively adapt to their environments. The basic building blocks of organizational structure are differentiation, which refers to the way in which a company divides itself into parts (divisions and functions), and integration, which refers to the way in which the parts are then combined. The two processes determine together how an organizational structure will operate (Hill & Jones, 2001, p. 385). Horizontal differentiation focuses on the division and grouping of tasks to meet the objectives of the business. Perhaps the first person to address this issue formally was Harvard Business historian Chandler (cited in Hill & Jones, 2001, p. 394). Chandler’s famous hypothesis is that structure follows strategy. Companies are quickly recognizing that if they are to retain their flexibility, they need to develop a range of internal perspectives (in order to understand the environment) and diversity of resources and capabilities (in order to respond to it). Restating the general thesis, Chandler perceives strategic growth as resulting from an awareness of the opportunities and needs to employ existing or expanding resources more profitably (Chandler, 1966, p. 15). ‘‘In such a world,’’ advise Ghoshal and Bartlett (1999, p. 121), ‘‘a company’s organizational structure must reflect, not deny, the complexity of its external environment.’’ That is, unless new structures are developed to meet new administrative needs, which result from opportunities and strengths, economies of growth and size cannot be realized. For example, when General Motors (GM) faced bankruptcy during the 1920 crisis, Durant’s retirement made it possible for Sloan and du Pont to create a decentralized management structure to cope effectively with the company’s widely varying products and markets, an approach adopted at DuPont just a year later. Sloan firmly believed that divisional independence encouraged initiative and innovation. By the time du Pont retired from the

174


Board of Directors in 1928, GM became the largest corporation in the world (Chandler, 1966; DuPont Heritage, 2002). So is structure a reflection of strategy? The question ‘‘what is strategy?’’ leads to the work of Whittington (1993). He devised a historical typology of strategy theory, outlining the nature and the assumptions of four ‘‘conceptions’’ within which most of the literature could be placed. His view is that a dominant ‘‘concept’’ is identifiable in each decade: the Classic in the 1960s, the Processual in the 1970s, the Evolutionary in the 1980s, and the Systemic in the 1990s (Whittington, 1993, p. 40). According to this view, Porter is positioned in the Classic approach. He is confident that ‘‘every firm competing in an industry has a competitive strategy. This may have been developed explicitly through a planning process, or implicitly through the activities of the various functional departmentsy the emphasis being placed on strategic planning today in firms in the United States and abroad reflects the proposition that there are significant benefits to gain through an explicit process of formulating strategy, to insure that at least the policies (if not the actions) of functional departments are coordinated and directed at some common set of goals’’ (Porter, 1980, p. xiii). This raises two questions. First, could the best strategy be developed explicitly through a planning process? ‘‘No’’ according to Mintzberg (1994), who concludes that strategy cannot be planned because planning is about analysis and strategy is about synthesis. That is why the process has failed so often and dramatically. Second (also in agreement with Whittington, 1993, p. 111), is strategic planning what unites all individuals who make up an organization around the effective pursuit of a coherent goal? Mintzberg is skeptical: ‘‘There can be little doubt that planning can serve as important mechanism to knit disparate activities together. But to consider this imperative is another matter. Coordination can be effected in other ways too y And even when plans do serve in a coordinating role, it cannot be assumed that planning has created those plans’’ (Mintzberg, 1994, p. 17). The message of the Classical and Evolutionary Schools is obvious: change structure to match changes in strategies. Donaldson’s (1987) statistical analysis of the correlation between performance and appropriate structure among large American corporations certainly confirmed the financial logic of this argument: diversified firms with Multidivisional structures and nondiversified firms with Functional structures performed better in terms of profit margin and return on investment growth than their mismatched peers. One could say, that according to these results and at least in the United States, fitting strategy to structure pays. Porter’s Competitive Strategy has met the needs of both academics and managers who were looking for a


175

strategy theory and models that could be empirically tested, e.g.,the fiveforces analysis. Harfield (1998) finds Porter’s work attractive because it provides the illusion of security, control, and legitimacy in the face of uncertainty, but also problematic because if every business adopted the strategies advocated, none would be able to secure a competitive advantage. The need to match structure with strategy is well accepted but as Whittington, Pettigrew, and Ruigrok argue, the old notions need to be re-examined (FT, 1999). For the Processualists, the direction of causality is reversed. They not only point to how organizational structures in practice fail to fit strategies, but also how they can actually shape strategies (Whittington, 1993, p. 111). The Processual view is that organizations are partnerships of individuals, each of whom brings their own personal objectives and cognitive biases to the organization. For scholars like Cyert, March, or Simon, people are only ‘‘restrictedly rational’’ (Harfield, 1998). Because of this limitation, strategy is a gradual process of negotiation and adjustment of routines to environmental messages, which eventually force themselves on the manager’s attention. Thus, strategies are often ‘‘emergent,’’ their coherence accruing through action and perceived in retrospect, while successive, small steps eventually merge into a pattern (Harfield, 1998). So, is strategy a reflection of structure? In practice, the strategies of most organizations are probably a combination of the intended and the emergent (Hill & Jones, 2001, p. 22). Mintzberg (1994, p. 25) maintains that few, if any, strategies can be purely intended, and few can be purely emergent. One suggests no learning, the other no control. All real-world strategies need to mix these in some way – to attempt to control without stopping the learning process. An organization’s capability to produce them is fostered by the organization’s structure. Still, research evidence seems to indicate that formal planning systems do help companies make better strategic decisions; e.g., the study from Miller and Cardinal in 1994 (cited in Hill & Jones, 2001, p. 23) analyzed in detail the results of 26 previously published studies on the relationship between strategic planning and company performance, concluding that, on average, strategic planning does indeed have a positive impact on company performance and hence is a valuable activity. But should companies organize regionally, nationally, or globally; centralize or decentralize control; build around functions, markets, or products? Taking a look at 21st century organizations, a variety of ways to structure appear: The Simple Structure. This is normally used by the small entrepreneurial company producing a single product or a few related ones for a specific market segment; no formal arrangements regarding organization exist. If

176


the company wants to grow and perform all tasks required by a rapidly expanding market, it needs a more complex form of horizontal differentiation (Hill & Jones, 2001, p. 394). It should be noted that none of the FORTUNE 500 companies is structured this way (Fortune, 2003). The Functional Structure. This is based on the primary activities that have to be carried out, such as production, finance, and accounting, marketing and personnel. When a company enters and competes in an international market, it just adds a subsidiary in the foreign country to handle sales and distribution. The Functional structure is typically found in smaller companies, or those with narrow, rather than diverse, product ranges. It allows greater operational control at a senior level – the CEO is in touch with all areas, and there is a clear definition of roles and tasks. It is a very popular structure; however, there are disadvantages, particularly as organizations become larger or more diverse. In such circumstances, senior managers might be burdened with everyday operational issues, neglecting the overall strategy; they rely on their specialist skills rather than taking a strategic perspective on problems (Johnson & Scholes, 1999, p. 403). In comparison to other forms, such as the Matrix, the Functional structure is inflexible and, thus, less suited for ‘‘ad hoc’’ projects. Perhaps the biggest weakness is the conflict that tends to arise between departments. It is not uncommon for ‘‘demarcation lines’’ to be generated with managers from each department unwilling to work with others. There is a tendency to ignore the ‘‘big picture’’ (RBR, 2001). The Geographic Structure or Territorial Grouping. When an organization is geographically dispersed, groupings are created by region within a country or, where appropriate, by world regions. Frequently, sales or marketing groups are divided by geographic region (McKenna, 2000, p. 433). Significant benefits accruing to this type of structure are that the organization can cater for the specific needs of a given location, achieving better market coverage, and that it can provide training to familiarize managers with operations in the field (Hodgetts, 1991). A study by Berenbeim (1983) exposed a relationship between Geographic structure characteristics and an advantageous approach towards delegation and control. In comparison to other structures, the Geographic is characterized by a greater degree of local autonomy, which fosters entrepreneurship and motivation (Leavitt & Lipman-Blumen, 1995, p. 113; Nayak & Ketteringham, 1997, p. 369; Timmons, 1999, p. 224). Drawbacks could include an increase in administrative expenses, advertising, promotion and sales outlet costs, and the isolation the work groups in


177

the regions feel when distanced from the nerve center of the organization. This could cultivate loyalty to the regional work group that is of greater magnitude than commitment to the larger organization. A difficulty for top management is the exercise of overall control in the face of significant devolution of authority to Territorial Groupings (McKenna, 2000, p. 433). The Product-Team Structure. This has been a major innovation in recent years. Its advantages are similar to those of the Matrix structure, but it is much easier and far less costly to operate because of the way people are organized into permanent cross-functional teams, instead of being assigned only temporarily to different projects. Tasks are divided along product or projects lines to reduce bureaucratic costs and to increase management’s ability to monitor and control the manufacturing process. This results in much lower costs associated with coordinating activities than in a Matrix structure, in which tasks and reporting relationships change rapidly (Hill & Jones, 2001, pp. 404–405). When a company embarks on a global strategy, it locates manufacturing and other value creation activities in the lowest-cost global location to increase efficiency, quality, and innovation. A product-group headquarters (similar to SBU headquarters) is created to coordinate the activities of the domestic and foreign divisions within the product group, providing, at the same time, a centralized control. This way, managers can decide which value creation activities should be performed in which country to increase efficiency (Hill & Jones, 2001, pp. 464–465). The Multidivisional Structure. Many commentators compare the current period of organizational change to the birth of the divisional organization 80 years ago in the great American corporations like DuPont and GM. Consistent with Chandler’s premise, these companies having changed their strategies from focus to diversification, needed the new decentralized structure to match (FT, 1999). This structure is subdivided into units (divisions) on the basis of products, services, geographical areas, or the processes of the enterprise. This way, each division is able to concentrate on the problems and opportunities of its particular business environment. The products and markets in which the company operates may be so diverse that it would be impractical to bring the tasks together in a single body. Thus, a division can be created, which relates closely to an SBU, allowing a tailoring of the product/market strategy to the requirements of that SBU and improving the ownership of the strategy by divisional staff (Johnson & Scholes, 1999, p. 404). When a company enters an international market, it adds an International Division to coordinate the flow of different products across different countries (Hill & Jones, 2001, p. 464).

178


The research that Armour and Teece performed in 1978 (Hill & Jones, 2001, pp. 398–399) suggests that large companies that adopt the Multidivisional structure outperform those that retain the Functional; perhaps, this is the reason why this structure is said to be adopted by more than 90% of all large US corporations. However, as with all of the structural types, there are disadvantages to the Multidivisional structure. Some, mentioned by Hill and Jones (2001, pp. 399–402), are the following: establishing the divisional-corporate authority relationship, distortion of information, competition for resources, transfer pricing, short-term Research and Development focus, and, finally, high bureaucratic costs. Nearly a century after the innovations of DuPont and GM, it has been discovered that the old divisional form needs extending (FT, 1999). Most multinationals are now in the third phase of international development (Kahn, 1998). Firms first regarded international operations as an ugly stepchild – simply, ‘‘one old executive with an airline ticket.’’ Later, they realized the value of internationalization but tended to clone their operations, creating a loose federation of businesses, each of which duplicated the hierarchy of the home office. Now, power has shifted to units responsible for performing a given function globally, and the emphasis is on optimizing processes worldwide. This third phase is the ‘‘Matrix.’’ The Matrix Structure. This was developed by NASA in the United States and based on two forms of horizontal differentiation (RBR, 2001). It could also be said that it is a combination of structures, operating in tandem. Activities on the vertical axis are grouped by function or geographical divisions, while the horizontal pattern is based on differentiation by project or product group. The result is a complex network of reporting relationships among projects and functions. Employees inside the Matrix have two bosses, namely, a functional boss, who is the head of a function, and a project boss, who is responsible for managing the individual projects (Hill & Jones, 2001, p. 402). And this is perhaps its most obvious disadvantage; the classical school as represented by Taylor, Fayol, and Urwick (Armstrong, 2001, p. 183) has been stressing, from early times, the importance of unit of command. That is, each individual employee should have only one supervisor. The division of priorities can lead to conflict and lack of clarity (RBR, 2001). This structure has many strengths, which help in overcoming many problems within the Matrix (Bolton, 2002; Decker, 1996; Hill & Jones, 2001):


179

Team members control their own behavior. This freedom provides the autonomy to motivate employees. Flexibility. Decisions on major projects are made quickly, customer response time increases. Participation in project teams allows team members to monitor other members and learn from each other. As the project goes through different stages, different specialists from different functions are required. Therefore, the Matrix can make maximum use of employees’ skills as existing projects are completed and new ones come into existence. A Matrix structure is flat and requires a minimum of direct hierarchical control by supervisors, leaving top management free to concentrate on strategic issues. This operating reality allows companies to break away from the functionally dominated model institutionalized by the old organizational chart. For example, Intel’s Matrix relationships link products and functional units, shared relationships becoming the norm (Ghoshal & Bartlett, 1999, p. 122). In their study, Backstro¨m and Ladan (2002) introduce a Matrix structure to dysfunctional organizations, also with the goal of enabling the principles of a learning organization. Positive effects like holistic awareness, alignment, flexibility, learning, motivation, cooperation, and shorter time to decision are reported. It is remarkable how each structural type shapes the six elements that divide and group labor (Candea, 1976; Sablynski, 2002): work specialization, departmentalization, chain of command, span of control, centralization vs. decentralization, and formalization. Several examples illustrate how thought-through organizational structures rejuvenated PepsiCo’s Frito-Lay division and raised product quality (Sellers & Barlyn, 1996), supported Kmart’s plan to exit bankruptcy (Howell, 2002, p. 1), and pulled Shell out of its midlife crisis, moving it into profitable new businesses (Guyon, 1997). Managing the strategy-structure relationship can be really difficult; when the number of hierarchical levels in a Functional structure becomes too great, the structure also becomes too expensive; motivational problems and information distortion appear. Depending on a company’s situation, problems can be solved and bureaucratic costs reduced by a decentralized Product-Team Structure. Globalization and the search for unsaturated markets bring growth through Geographical dispersion to administer local field units. As company size increases, however, decentralized structures may

180


cause coordination problems. Careful organizational design can help a firm grow and diversify and increase people performance while economizing on costs without becoming too tall or too decentralized. Does this mean that the type of structure can make a difference? Is there really a relationship between organizational structures and performance?

SELECTION OF PERFORMANCE MEASURES AND FORMULATION OF HYPOTHESES Performance In the previous pages, the imprecise term ‘‘performance’’ was used. The diversity of its possible measures stresses the necessity of a comprehensive framework that relates it to structure and the importance of selecting the appropriate dimension. Agreeing upon how to measure performance is not that simple since there are so many aspects of it. Each academic field offers a different point of view. They range from general financial outcomes to specific subjective ones (Bello & Gilliland, 1997; Campbell, 1990; Chaudron, 2001; Levinson, 1992; Robbins & Coulter, 2002). Nevertheless, accepting the notion that the ultimate objective of a ‘‘for profit organization’’ is the quest for profits and their maximization (Mintzberg, 1983, p. 9) and also accepting that monetary performance is the critical, ‘‘bottom line’’ kind of result that companies must deliver to survive (Chaudron, 2001), then the core of the concept consists of a plethora of outcome-based financial indicators. Absolute performance measures will be excluded because they face a comparability problem and limit the focus of the study. They may be vulnerable to the size of the company, market segment, industry, and current market share, to mention only a few factors. They are also hard to compare in cross-country studies (Styles, 1998, pp. 18–19). This is why accounting numbers will be translated into relative values; the technique to be used for evaluating companies is ratio analysis. A combination of measures will be considered in order to capture the rich complexity of performance (Bello & Gilliland, 1997). Given that there are dozens and dozens of acknowledged financial ratios, one needs to be selective in their use (Kerr, 2001). Specifically, the selection of the measures that will be used to gather the data should be driven by the research problem and the focus of the study (Agathangelou, 2001). In topics like organizational structure and performance, the efficiency with which


181

resources are utilized needs to be carefully assessed, e.g., offices, buildings, and costs from operations made for generating profit. For example, when an organization is Geographically structured, it clones its operations. If this structure is not aligned to the strategy, dividing sales or marketing groups by geographic region and duplicating the hierarchy of the home office will needlessly increase administrative expenses, advertising, promotion, and sales outlet costs. Financial performance might be negatively affected. As noted by Armstrong (2001, pp. 143–153), the role/job situation individuals are in is one of the influencers of human resource performance. A job consists of a group of finite tasks to be ‘‘performed’’ (Armstrong, 2001, p. 350). Since each structural type divides, groups, and coordinates job tasks in a different way, it can be expected that each structural type would have a different impact on human resource performance. For example, the Matrix structure is a complex network of reporting relationships among projects and functions. Its most obvious disadvantage is that employees inside it have two bosses. Without a unit of command, the division of priorities can lead to conflict and lack of clarity. Human resource performance might be negatively affected. As exemplified in the two previous paragraphs, several structural characteristics have power over company performance in terms of financial and human resources. Interpreting these two dimensions with the aim of understanding the performance of a firm in the recent past requires identifying factors believed to be relevant to these changes. The first factor is the efficiency with which the long-term capital (property, offices, plants, machinery, and intangible assets) has been employed in a for-profit organization. This will be measured with the Return on Capital Employed ratio (RoCE ¼ Profit/Capital Employed 100), which is considered crucial for testing profitability and financial performance (Kerr, 2001). It is intended to focus attention on the efficiency with which all of the resources available to the company have been utilized (Pendlebury & Groves, 2001, p. 112). The second factor is the effectiveness of the utilization of labor. A quick glance at a company’s Profit and Loss account shows that there are several levels of profit. Which one of those would make sense to use for the human resource performance test? As mentioned before, the type of organizational structure can be held responsible for the level of distribution costs, cost of sales, and administrative expenses of the company. But it cannot be held responsible for the level of interest payments, as the markets set interest rates, and it cannot be held responsible for the amount of taxation the company pays, as the government sets taxation rates (Kerr, 2001). For this

182


reason, it would definitely make sense to use Operating Profit as the profit measure. The absolute numbers will be translated into relative values to facilitate company comparisons, through the Return per Employee ratio (RpE ¼ Operating Profit/Number of Employees), which gives the profitability of the company per employee (Pendlebury & Groves, 2001, p. 276).

Hypotheses Organizations are now operating in extremely complex environments where survival depends on the ability to understand and respond to a wide variety of forces, some internal to the enterprise, others external. As these forces change over time, and many of them have changed quite dramatically in the last decades, it can be expected that these will often be associated with changes in what is felt to be the most suitable organizational structure (OECD, 1987, p. 42). Unless new structures are developed to meet new functional and administrative needs, company performance cannot be realized. The central question: ‘‘Is there a relationship between organizational structures and the performance of FORTUNE 500 companies?’’ is now developed into testable hypotheses cast in a ‘‘hypothetico-deductive’’ form.

Structures and Return on Capital Employed The Functional Structure is based on the primary activities that have to be carried out, and, as organizations become larger or more diverse, the number of levels increases, together with bureaucratic costs; its local character lifts distribution costs. The duplication of specialist services in the Multidivisional structure is expensive to run and manage. When an organization is Geographically structured, it clones its operations. Dividing sales or marketing groups by geographic region and duplicating the hierarchy of the home office, increases administrative expenses, advertising, promotion and sales outlet costs. All these factors can affect a company’s financial performance negatively. In the Product-Team structure, functional specialists are placed in permanent cross-functional teams and tasks are divided along product or projects lines to reduce staff costs. Employees inside the Matrix structure have two bosses, which raises bureaucratic costs. But the structure’s flexibility makes maximum use of employee’s skills; decisions on major projects are made faster, and customer response time increases. These factors can affect a company’s financial performance positively.


183

Although the degree to which each structure affects a company’s financial performance cannot be defined in advance, it is accepted that some attributes make companies perform better than others. We can also distinguish how each of these attributes ‘‘groups’’ companies under a structural name. Consequently, it is suggested that a significant RoCE difference will be observed at least between two of the structural groups. Hence, HA. IF companies are categorized by structural type THEN a significant RoCE difference will be observed between two or more of the groups. FORTUNE 500 companies are among the largest in America. A centralized organizational structure is counterproductive in this case, because it slows response time and lowers the incentives of those at the local level (Porter, 1980, p. 211). From a practical point of view, Multidivisional structures are ‘‘better suited’’ to be run in a decentralized manner, while Functional structures suggest that a rather centralized management process be enforced (Candea, 1976, p. 4). Accordingly, Armour and Teece’s research from 1978 (cited in Hill & Jones, 2001, p. 399) suggests that large companies that adopt the Multidivisional structure outperform those that retain the Functional. These arguments suggest HB. IF companies have a Multidivisional structure THEN the RoCE will be higher than from companies with a Functional structure. A major innovation in recent years has been the Product-team structure. It enjoys the advantages of the Matrix structure, but it is much easier and far less costly to operate. Tasks are divided along product or projects lines to reduce costs associated with coordinating activities and to increase management’s ability to monitor and control processes. The following hypothesis is, therefore, suggested: HC. IF companies have a Product-Team structure THEN the RoCE will be higher than from companies with a Matrix structure. Structures and Return per Employee In a Functional structure, senior managers are burdened with everyday operational issues, neglecting the overall strategy. Functional structure is inflexible, and there is conflict between departments. The Matrix structure is a complex network of reporting relationships among projects and functions. Its most obvious disadvantage is that employees inside it have two bosses. Without a unit of command the division of priorities can lead to conflict and

184


lack of clarity. All these factors can affect a company’s human resource performance negatively. A significant benefit of the Geographically structured organization is that it can offer Cross-Cultural Marketing and Team Building Trainings, to familiarize managers with operations in the field and better cater for the specific needs of a given location. Work groups cultivate regional loyalty. The Multidivisional structure enables divisional staff to concentrate on SBU problems and opportunities and tailor the product/market strategy to the requirements of that SBU. In the Product-Team structure, permanent crossfunctional teams result in better coordinating activities. Dividing tasks along product or projects lines increases management’s ability to monitor and control the processes. All these factors can affect a company’s human resource performance positively. Again the degree to which each structure affects a company’s human resource performance cannot be defined in advance, but it is accepted that some companies perform better than others. Consequently, it can be suggested that a significant RpE difference will be observed at least between two of the structural groups. Hence, HD. IF companies are categorized by structural type THEN a significant RpE difference will be observed between two or more of the groups. The study by Berenbeim (1983), which examined various decision areas according to the features of the company, found a number of relationships between such company characteristics and the general approach towards delegation and control. In comparison with other structures, companies organized along geographic lines may see themselves more as a federation of local companies, associated, therefore, with a greater degree of local autonomy. Work group loyalty and motivation is a significant advantage of the Geographic structure. The following hypothesis is therefore suggested: HE. IF companies have a Geographic structure THEN they will have a higher RpE than other companies. A relationship between organizational structure and company performance is generally hypothesized. An important assumption that is made, supported by the literature review, is that each structural type utilizes resources differently in generating profit. The hypotheses will be empirically examined and discussed in the next parts of the paper.


185

METHODOLOGY Data The previous sections provided a list of constructs (variables). There are five basic types of organizational structures: Functional, Geographic, ProductTeam, Multidivisional, and Matrix. Although the variable is qualitative, it can be operationalized in a way that allows reducing it to some quantifiable index by assigning numbers to these types using a nominal scale. This is a pure labeling activity; there is no logical/natural order. That is, they are simply listed in categories alphabetically. These categories are mutually exclusive and exhaustive. RoCE is equal to Profit divided by Capital Employed ( 100) and RpE is equal to Operating Profit divided by the Number of Employees. Values are snapshot figures for fiscal 2002; the terms involved are defined as follows: Profit is Revenue minus cost. Capital Employed is taken to be the total assets of the company: property, offices, and plants; machinery, stock, and intangible assets (Dyson, 2001, p. 44). Operating Profit/Operating Income is the revenue from a firm’s regular activities after cost of sales, distribution costs, and administrative expenses, but before the deduction of interest and tax (Harvey, 2003). Number of Employees is the number of permanent or full-time equivalent employees, as reported by the company in its 10-K report, who work in return for financial/other compensation. All terms are numeric and hence, quantifiable. The data were collected from audited Annual Reports and 10-K Forms, prepared under the regulations of the US Securities and Exchange Commission and recognized as reliable sources of financial and other significant information.

Population and Sampling For almost 50 years, Fortune Magazine has been ranking the top 500 USbased corporations with the largest revenue in the past year – calculated using publicly available data. The result is their annual FORTUNE 500 list. This is the population of interest. The basic unit to be examined is each company. The necessary sample size was found to be n ¼ 46:25; and was rounded up to 50, which is 10% of the

186


population. The complete 2002 FORTUNE 500 frame was downloaded and saved in the Statistical Package for Social Sciences (SPSS), keeping the original rank-order by revenues (descending), where each company was numbered from 1 to 500. The random sampling facility generated 50 figures. The corresponding companies were selected and results were stored in one major file. Data Collection The sample was contacted in two phases. In the first phase, copies of the audited Annual Reports and 10-K Forms were ordered electronically. All companies offered in their Web sites several contact possibilities; some provided e-mail addresses to directly contact an expert. This e-mail also functioned as a notification of the research, encouraging the receiver to mail back the contact details of a person – preferably an executive – who is most directly involved in HR or corporate strategy matters. Eleven companies (22%) offered the link and the option to download the entire reports or specific sections; most of them had chosen not to distribute hard copies of the 2002 reports in 2003, and the rest were not yet ready with printing. The requested financial reports were received, on average, 22.31 days after they were ordered (Table 1). It was necessary to conduct several follow-up e-mailings, spaced about two weeks apart. Two companies did not answer; their reports were downloaded. The minimum interval was 0 days (one company answered the same day), and the maximum was 48 days. The second phase was concerned with identifying the structural types that companies in the sample had during fiscal 2002. The instrument to collect the desired primary data was a PowerPoint presentation, which described and showed the five basic structural types that the literature review provided. The pilot was administered to five of the punctual respondents who had participated in phase 1, was corrected according to comments, and was then sent to the rest of the 45 companies. During this phase, the response duration fell remarkably: 23 (46%) of the companies answered within the first week (Table 2). One of them refused to provide the requested information due to its strategic and competitive nature. Nine companies (18%) answered during the second week; five companies, unable to settle on a type, requested a telephone interview to answer specific questions for which they would not feel comfortable mailing/writing answers due to their strategic nature. Seven companies did not answer at all. Although the first letter encouraged receivers to preferably identify executives who are most directly involved in HR or Corporate strategy


Data Collection: Phase I.

Table 1.

Annual reports 48 2

Duration in days Mean Median

22.31 19

Mode

19.00

Std. deviation

12.61

Minimum

0

Maximum

48

10 9 8 7 6 5 4 3 2 1 0

N=48

0 5 10 15 20 25 30 35 40 45 50 Duration in days, shown in midpoints

7

2

20

3

3

Customer service/ affairs

2

1

3

Human resources

2

3

1

6

E-Center/ web site

2

1

2

5

4

1

1

6

21

16

6

43

Media/Public relations Department

Corporate comm./ offices Total

Frequencies

Female Male No Info Total 11

Investor relations

Direct e-mail

N:12

Mean

19.67

Minimum

7

Maximum

36

Mail to IR dept.

N:21

Mean

26.71

Minimum

9

Maximum

48

Web contact form

N:15

Mean

18.27

Minimum

0

Maximum

45

Data Collection: Phase II.

Table 2.

Respondent demographics

Duration by contact style

Duration statistics

Valid Missing

Frequencies

N

187

25

N=43

20 15 10 5

0 1st 2nd 3rd 4th Response duration shown in weeks

matters, only 13 of the respondents were from management. Accordingly, the qualitative data are collected from employees embodying a diversity of departments and job-titles, ranging from the simple Representative of the Investor Relations Department, to the Vice President of Corporate Communications. The majority (46.5%) of the respondents were from the Investor Relations Department. People on various positions may hold different views regarding patterns of interaction and organizational structures. For example, managers derive a sense of identity from the affiliation with a company or their connection to social groups (SBU, function) within the company; in large and heterogeneous organizations, they tend to identify more strongly with their immediate work groups than with the organization as a whole (Houston, Walker, Hutt, & Reingen, 2001, pp. 19–35). As the aim is not simply to save results

188


of which structure the respondents ‘‘believe’’ their organization has, all answers were controlled for reliability. Data were collected independently, and a transcript with results was prepared during the time it took the companies to respond. These were compared with the respondents’ results; disagreements in seven cases were captured but were resolved by discussion. An independent judge was consulted for the cases where no answer was received.

ANALYSIS OF FINDINGS Descriptive Statistics All records were first collected in an Excel file in order to compute the ratios. On average, firms in the sample made $969.96 millions profit, with 3.87% RoCE and about $251,850 thousands RpE (Table 3). Summarizing the categorical variable ‘‘organizational structure’’ in a frequency table displays that the highest percentage of cases falls under the Multidivisional category. This result was expected to be higher, considering the statement from Hill and Jones (2001) that Multidivisional structure has now been adopted by more than 90% of all large US corporations. Since many statistical tests assume data are normally distributed, the first step was to check the data distribution. Inspecting the skewness, kurtosis, and their standard errors (Table 3), it was concluded that none of the distributions is a normal one. Exploring the data (see the appendix) revealed two most extreme scores. After eliminating them, RoCE does not differ Table 3.

Descriptive Statistics.

Case summaries

N

Minim

Profit

50

-1,303.00

7,829.00

969.96

1,732.20

2.09

.34

4.86

.66

Total assets

50

1,095.70

887,515.00

61,589.77

163,348.46

3.88

.34

15.54

.66

Return on capital empl Operating profit

50

-23.19

17.11

3.87

7.19

-.97

.34

3.60

.66

50

-894.90

46,335.00

2,866.37

7,093.32

5.11

.34

29.80

.66

Number of employees

50

2,003

355,421

51,128.56

67,776.50

3.04

.34

10.36

.66

Return per employee

50 -73,678.58 9,653,125.00 251,849.14

1,359,453.95

7.03

.34

49.57

.66

Structures Frequency Valid percent

Functional

Maxim

Mean

Geographic

Matrix

Std. deviation

Skewness Std. E Kurtosis Std. E

Multidivisional

Product-Team

Total

12

8

7

14

9

50

24%

16%

14%

28%

18%

100%

Note: All values in $millions, except RoCE (in %), employees and RpE (in thousands).


189

significantly from a normal distribution. Therefore, RoCE hypotheses are tested with parametric tests and RpE hypotheses with non-parametric ones. The hypotheses are empirically examined and discussed in the same sequence they appeared at the end of the literature review section. Structures and Return on Capital Employed HA. One-way ANOVA tests for differences between group means. The dependent variable is quantitative; the factor variable is categorical. The following null and alternative hypotheses are implied: The null hypothesis states that there will be no observed differences between group means; The alternative hypothesis states that there is a significant difference, at least between two of the means. A report table lists the selected statistics, describing the distribution of the dependent variable for each group. For the Geographic and Product-team groups, the mean is approximately equal to the median, suggesting a symmetrical distribution of RoCE. Plotting the means for the groups illustrates how they differ, but if one conducts a study with two or more groups, the resulting group means will almost never be identical. The task is to decide if the independent variable has influenced the group means; this decision will be based on the outcome of a significance test. The ANOVA table compares the means for the different groups, partitioning the total variation into two components: Between Groups represents variation of the group means around the overall mean, and Within Groups represents variation of the individual scores around their group means, sometimes referred to as error variation. The F test is the ratio between these two measures of variation. In Table 4, the F value is large and the significance level is less than 0.05, indicating that at least one of the groups differs from the others. The null hypothesis can be rejected. The small significance value of Linearity and the high significance value of Deviation from Linearity would indicate that a linear relationship exists between the variables, if the categories of the independent variable were ordered. When the null hypothesis is false, the difference between means is due to error variance plus the treatment effect. How much this effect depends on the independent variable (especially when this is nominal) is tested with the Crosstabs procedure (the word ‘‘depends’’ has no implications for a causal relationship). The continuous numeric RoCE data are converted to a discrete number of categories, each one containing approximately the same

190


Table 4. Mean Difference between Multiple Groups.

0.520

4.943

-2.179

4.937 8

8 5.165

5.010

1.998

-0.531

0.237 6

Matrix

7 7.278

5.910

5.203

0.249

-1.884

Multidivisional 14 4.188

1.815

6.597

0.879

-0.235

Product-team

8.450

5.142

0.496

-1.138

9 8.706 ANOVA table

2 0 -2

F

Sig.

Combined

570.300

4

142.575

5.176

0.002

Linearity

484.484

1

484.484

17.588

0.000

Deviation from Linearity

85.816

3

28.605

1.038

0.385

Within groups

1212.007

44

27.546

Total

1782.307

48

Between groups

Sum of squares

4

PRODUCT-TEAM

11 -1.161

Geographic

MULTIDIVISIONAL

Functional

Mean line-chart

MATRIX

Skewness Kurtosis 10

GEOGRAPHIC

N Mean Median Std. dev

FUNCTIONAL

Structures

df

Mean square

number of cases. The value of 1(low) was assigned to cases below the 25th percentile, 2(medium) to cases between the 25th and 50th percentile, 3(high) to cases between the 50th and 75th percentile, and 4(very high) to cases above the 75th percentile. Selecting Structures as the row variable, and levels of RoCE as the column variable, the crosstabulation shows the frequency of each RoCE level at each structure. If each structure influences financial performance similarly, the pattern of RoCE should be similar across structures. The following null and alternative hypotheses are generated: The null hypothesis states that the two variables are independent; the alternative hypothesis states that the two variables are not independent (they are related). From the clustered bar chart, it is evident that each pattern of bars is not constant across structures; this indicates that the two variables are not independent. The Functional group has RoCE incidents only in the ‘‘low’’ and ‘‘medium’’ categories. The Multidivisional is the only group with a complete set of colored bars and, hence, frequencies in every category. An interesting observation is that the Matrix and Multidivisional appear to have the same frequencies for the ‘‘high’’ and ‘‘very high’’ levels (Table 5). From the crosstabulation and the bar chart, it is impossible to tell whether these differences are real or due to chance variation. For that reason, the chi-square statistic is used to measure the discrepancy between the observed cell counts and what would be expected if the rows and columns were unrelated. Both significance values are below 0.05, indicating that there is some relationship between the two variables. The Likelihood ratio results


Table 5.

Structures Count Observed

Functional Expected % within Observed

Geographic Expected % within Observed

Matrix

Expected % within Observed

MultiExpected divisional Productteam

Clustered bar chart: Count

7

% within Observed Expected % within

low

191

Crosstabulation and Relation.

Asymp.Sig Low Medium High V. High Total Value df Chi-square test (2-sided) 7 4 0 0 11 34.074* 12 0.001 2.7 2.7 2.9 2.7 11.0 Pearson chi-sqre 40.291 12 0.000 63.6% 36.4% 0.0% 0.0% 100.0% Likelihood ratio 49 0 1 6 1 8 N of valid cases 2.0 2.0 2.1 2.0 8.0 *20 cells (100.0%) expected count less than 5 0.0% 12.5% 75.0% 12.5% 100.0% Directional measures Value Appr. sig. 0 2 2 3 7 Symmetric 0.296 0.013 Lambda 1.7 1.7 1.9 1.7 7.0 RoCE dependent 0.361 0.007 0.0% 28.6% 28.6% 42.9% 100.0% Uncertainty Symmetric 0.277 0.000 5 4 2 3 14 coefficient RoCE dependent 0.297 0.000 3.4 3.4 3.7 3.4 14.0 Value Appr. sig. 35.7% 28.6% 14.3% 21.4% 100.0% Symmetric measures 0.834 0.001 0 1 3 5 9 Phi 0.481 0.001 2.2 2.2 2.4 2.2 9.0 Cramer's V 0.640 0.001 0.0% 11.1% 33.3% 55.6% 100.0% Contingency coefficient

medium

high

very high

6 5 4 3 2 1 0

Functional

Geographic

Matrix

Multidivisional

Product-team

(a goodness-of-fit statistic similar to Pearson’s chi-square) are more reliable because 100% of the cells have expected values less than 5. The null hypothesis can be rejected. The direction of the relationship is studied with Lambda and Uncertainty coefficient measures of association. The low significance values point to a relationship between the two variables; but the average values for both test statistics indicate that the relationship is an intermediate one. The Uncertainty coefficient result moreover indicates that knowledge of structural type reduces error in predicting values of RoCE by almost 30%. The values of the three Symmetric measures are medium to high, and all significance values are 0.001. It is safe to say that the relationship is not due to chance and that it is a medium to strong one. Testing confirmed the expectations defined in HA. The findings obtained from the chi-square statistics, are cross-validated with the ones from the powerful ANOVA. But ANOVA alone would not have offered enough information on each group distribution and insight to draw some more

192


conclusions regarding the very first syllogism that led to testing for difference – how some structural attributes can affect a company’s financial performance negatively while others can affect it positively. The Product-Team and Matrix flatter arrangement, which Lawler (cited in Wheeler, 2000) called a ‘‘high-involvement structure,’’ allows for quicker reactions to the business environment, involving individuals throughout the organization in the information flow and decision-making capacities. This approach is consistent with democratic principles and, thus, may be especially appropriate in this increasingly democratic world (Wheeler, 2000). The flexibility that characterizes Product-Team and Matrix structures supports better coordination and makes maximum use of assets, reducing costs. These factors were expected to affect a company’s financial performance positively, and apparently they do, comparing the means in Table 4 and also the RoCE patterns in the clustered bar chart – none of the two types has RoCE incidents in the ‘‘low’’ categories. HB. For the independent-samples t-test the Functional and the Multidivisional group were defined by specifying two values; cases with any other values were excluded from the analysis. The following null and alternative hypotheses are generated: The null hypothesis states that the two means are equal; The alternative hypothesis states that the two means are not equal. The Standard Error of Mean (in the Group Statistics table in Table 6), demonstrates how much the value of the mean may vary from sample to sample taken from the same distribution. Since the independent-samples t-test compares the two group means, it is useful to know what the mean values are. In this case, the mean value of the Multidivisional group appears Table 6.

Structures N Multidivisional 14 Functional 11

Equal variances Assumed Not assumed

Mean Difference between Two Groups.

Group statistics Mean Std. deviation Std. error mean 4.188 6.597 1.763 1.490 -1.161 4.943

Fisher and Yates t table 0.050 0.020 Levels for two-tailed Levels for one-tailed 0.025 0.010 df 23 2.069 2.500

Independent samples test 95% Confidence interval of Levene's test for t-Test for equality of means equality of variances the difference Sig. Mean Std. error F Sig t df (2-tailed) difference difference Lower Upper 1.779 0.195 2.237 23 0.035 5.349 2.391 0.402 10.295 2.317 22.970 0.030 5.349 2.309 0.573 10.125


193

to be higher than the mean value of the Functional group. But is this difference significant? In the Levene test for homogeneity of variances the significance value is high (more than 0.05), so results that assume equal variances for both groups will be used. The t value is significant –equals the value between those shown in the Fisher and Yates’ table. The low significance value for the t-test (0.035) and the fact that the confidence interval for the mean difference does not contain zero indicate that there is a significant difference between the two group means. The null hypothesis can be rejected. The effect of Multidivisional structure on financial performance was positive and significant as predicted. This finding complies with Armour and Teece’s research from 1978 (cited in Hill & Jones, 2001, p. 399), suggesting support for HB. HC. The independent samples t-test is again used to compare the means for the Product-Team and the Matrix groups. The following null and alternative hypotheses are implied: The null hypothesis states that the two means are equal; The alternative hypothesis states that the two means are not equal. After running the test, the mean value of the Product-team group appears higher than the mean value of the Matrix group (Table 7). Is this difference significant? The Levene’s test for equality shows homogeneous variances for both groups, as the F value is small and the significance level is higher than 0.05. To be significant, the t obtained from the data, should have been equal to or larger than the value shown in the Fisher and Yates’ table – here the t is much lower. The significance value for the t-test is high, and the confidence interval for the mean difference contains zero, hence, it cannot be Table 7.

Structures Product-team Matrix

Equal variances Assumed Not assumed

N 9 7

Mean Difference between Two Groups.

Group statistics Mean Std. deviation Std. error mean 8.706 5.142 1.714 1.967 7.279 5.203

Fisher and Yates t Table 0.100 0.050 Levels for two-tailed Levels for one-tailed 0.050 0.025 df 14 1.761 2.145

Independent samples test 95% Confidence interval of Levene's test for t-Test for equality of means equality of variances the difference Sig. Mean Std. error F Sig t df (2-tailed) difference difference Lower Upper 0.039 0.847 0.548 14 0.592 1.427 2.605 -4.159 7.013 0.547 12.966 0.594 1.427 2.609 -4.210 7.064

194


concluded that there is a significant difference between the two group means. The null hypothesis cannot be rejected. Consistent with HA and HB, it was expected that structural type would continue to influence financial performance. But remember that ProductTeam and Matrix structures both enjoy the advantages of flexibility and flatness, while operating in teams – the first one in permanent crossfunctional teams, the second one in temporary project teams. Apparently, this small variation is not imperative for shaping financial performance. This suggests lack of support for HC. Structures and Return per Employee HD. The Median test is used to assess the RpE difference between the groups, if any. The following null and alternative hypotheses are generated: The null hypothesis states that there will be no observed differences between group medians; The alternative hypothesis states that there is a significant difference, at least between two of the medians. The Percentiles Statistic Table (Table 8) gives a numerical representation of the shape of the distribution. Across all 49 subjects, the RpE median is a score below 28. The null hypothesis for the median test is that this particular value is a good approximation of centre for each of the groups. To test this hypothesis, each group is divided into two subgroups: those whose scores fall at or below the median, and those whose scores are above it. The result is a two-way Frequencies Table with two rows and 5 columns (the grouping variable). If the groups do not differ, for each group, half of the scores should be above and half of the scores below this overall median. For the Geographic, Matrix, and Multidivisional groups, the median does what the null hypothesis says it should do: it divides them into two equal subgroups. But for the rest, the null hypothesis does not provide a good approximation of centre. The Median test now determines whether this pattern represents group differences that are significant. From the two-way Frequencies Table, a chisquare statistic is calculated to test row and column independence. In fact, the median test is a chi-square test of independence between group membership and the proportion of cases above and below the median. For each cell, the distance between the observed and expected counts is squared, then divided by the expected value. Finally, these quantities are summed across


Median Difference between Multiple Groups.

Table 8.

Structures Functional

N Mean Median 11 50,555 19,504

Geographic Matrix Multidivisional

Percentiles statistic table

Std. dev Skewn. Kurtosis 112,482

1.692

2.122

N=49

25th 11,105

27,734

78,026

2

3

4

8 68,983

27,181

123,675

2.698

7.439

RpE Structures

7 47,764

44,289

43,971

1.267

1.297

14 65,694

20,551

80,672

1.685

3.210

9 64,142

37,770

67,835

2.468

6.564

Product-team

195

Frequencies table

50th(Median)

Median test table N Median

Structures Functional Geographic Matrix Multidivisonal Product-Team

75th

Chi-square

> Median

2

4

4

7

7

df

Value Creation in Multinational Enterprise, Volume 7 (International Finance Review) (International Finance Review)

International finance

International Finance

Multinational Finance

International Business Finance

International Review of Cytology

International Review of Cytology

Asian Financial Crisis, Volume 1: Financial, Structural, and International Dimensions (International Finance Review)

International Review of Cytology

International Review of Cytology

International Review of Cytology

International Review of Cytology

International Finance: Third Edition

International Review of Cytology

INTERNATIONAL REVIEW NEUROBIOLOGY

International Review of Research in Mental Retardation, Volume 24 (International Review of Research in Mental Retardation)

International Review of Cytology, Volume 88

International Review of Cytology, Volume 83

International Review of Cytology, Volume 77

INTERNATIONAL REVIEW OF CYTOLOGY V130, Volume 130

International Review of Cytology, Volume 62

International Review of Neurobiology Volume 14

International Review of Neurobiology, Volume 23

International Review of Neurobiology Volume 1

International Review of Cytology, Volume 56

International Review of Neurobiology Volume 13

International Review of Neurobiology Volume 5

International Review of Neurobiology, Volume 21

International Review of Cytology, Volume 58

International Review of Cytology, Volume 76

International Review of Cytology, Volume 61

Value Creation in Multinational Enterprise, Volume 7 (International Finance Review) (International Finance Review)

International finance

International Finance

Multinational Finance

International Business Finance

International Review of Cytology

International Review of Cytology

Asian Financial Crisis, Volume 1: Financial, Structural, and International Dimensions (International Finance Review)

International Review of Cytology

International Review of Cytology

International Review of Cytology

International Review of Cytology

International Finance: Third Edition

International Review of Cytology

INTERNATIONAL REVIEW NEUROBIOLOGY

International Review of Research in Mental Retardation, Volume 24 (International Review of Research in Mental Retardation)

International Review of Cytology, Volume 88

International Review of Cytology, Volume 83

International Review of Cytology, Volume 77

INTERNATIONAL REVIEW OF CYTOLOGY V130, Volume 130

International Review of Cytology, Volume 62

International Review of Neurobiology Volume 14

International Review of Neurobiology, Volume 23

International Review of Neurobiology Volume 1

International Review of Cytology, Volume 56

International Review of Neurobiology Volume 13

International Review of Neurobiology Volume 5

International Review of Neurobiology, Volume 21

International Review of Cytology, Volume 58

International Review of Cytology, Volume 76

International Review of Cytology, Volume 61

Recommend Documents