Column 11 - Revisiting the Issue of Cross-Shareholding

OSANO Hiroshi
Professor, Institute of Economic Research, Kyoto University

Three patterns of partial ownership

The pros and cons of cross-shareholding among Japanese companies have been hotly debated recently in the context of their role as a defense against hostile takeovers. Partial ownership is an arrangement in which a nonfinancial company (hereinafter referred to simply as "company") holds a portion of shares in another company. The practice, however, is not unique to Japanese companies and can be observed in countries throughout the world, including the United States.

There are three patterns of partial ownership:
1) Mutual or unilateral shareholding between companies in a strategic partnership. This type of partial ownership is typically observed in cases where one company supplies products or services to the other and, in that context, can be defined as "vertical partial ownership."
2) Mutual or unilateral shareholding between companies in a competitive relationship in the same industry. This type of partial ownership concerns companies that compete with each other within the same industry and can be defined as "horizontal partial ownership."
3) Mutual or unilateral shareholding between companies having no or only a minimal business relationship between one another. This type of partial ownership concerns companies that are respectively engaged in business unrelated to that of the other, and can be defined as "conglomerate partial ownership."

Cases of partial ownership

Now, take a look at some specific cases for each of the three types of partial ownership.

As a case of vertical partial ownership, we can cite cross-shareholding between a final product manufacturer (e.g., Toyota Motor Corp.) and its parts supplier (e.g., Koito Manufacturing Co.) that is typically observed in the automobile and electronics industries in Japan. Meanwhile, in the U.S. biotech and high-tech industries, a large established company with marketing and management capabilities often holds an ownership stake in a small, start-up research and development (R&D) company. Such a case is also classified as vertical partial ownership. This type of vertical partial ownership is also observed in the financial industry, as evidenced by the recent news about U.S. financial giant Merrill Lynch's decision to sell its 20% stake in Bloomberg L.P., a financial data and news company, in a bid to raise capital to cover losses from securitized subprime mortgages.

Examples of horizontal partial ownership include a three-way alliance among Nippon Steel Corp., Sumitomo Metal Industries, Ltd., and Kobe Steel, Ltd.; in which each holds stakes in the other two. There are also cases of cross-shareholding between a major Japanese steelmaker and its overseas counterpart, while ArcelorMittal, the world's largest steel company, holds stakes in many of its rival steelmakers overseas. As such, somewhat unexpectedly, there are quite a few cases of horizontal partial ownership.

Lastly, conglomerate partial ownership is typically observed within the zaibatsu corporate groups in Japan, namely the Mitsubishi, Mitsui, and Sumitomo groups. This pattern of partial ownership is rarely observed in the U.S. or the United Kingdom.

Theoretical background of partial ownership

Some academic essays, published in internationally-recognized journals specializing in finance or industrial organization theories, attempt to offer rational explanations as to why companies enter into these partial ownership arrangements. One theory explaining vertical partial ownership is as follows.

In order to enhance productivity by improving the compatibility of parts and components with the relevant final products, a final product manufacturer may want its parts suppliers to make specific capital investments that would exclusively benefit the final product manufacturer, as opposed to making general capital investments that could be beneficial to any number of final product manufactures. To achieve this end, the final product manufacturer would need to assure parts suppliers that they will be properly rewarded for their efforts.

By the way, when a final product manufacturer holds an ownership stake in its supplier, the supplier's earnings will be reflected in the earnings of the final product manufacturer. Because of this, it is assumed that equity participation by a final product manufacturer in its parts supplier promotes bonding between the two, and ultimately serves as an unwritten guarantee of such reward (Dasguputa and Tao 2000). Meanwhile, another theory has it that partial ownership by an established company in an R&D startup is a rational means of providing assurance that the established company will either not enter the startup's market after obtaining knowledge and technology, or will offer a good price should the startup decide to sell its business (Mathews 2006).

Meanwhile, as a theory of horizontal partial ownership, it has been argued that cross-shareholding between rival companies helps to create a cooperative-competitive relationship, thereby benefiting shareholders. This argument may have significant implications for industries suffering from excessive competition. For other industries, however, this argument cannot justify horizontal partial ownership from the viewpoint of economic welfare because such an alliance is competition-restrictive.

Very few theories are available for conglomerate partial ownership, partly because such arrangements are rarely observed in the U.S. or UK. However, one theory has it that cross-shareholding within a conglomerate serves as an effective disciplinary mechanism to prevent corporate managers from engaging in conduct contradictory to the conglomerate's goals. That is, the possibility that other companies within the conglomerate may ally themselves to remove the top manager of a troubled member company at its general shareholders' meeting or other future occasion, would serve as an effective deterrent to corporate managers who disrupt the cooperative relationship among member companies (Berglof and Perotti 1994).

Another argument is that reducing the possibility of a hostile takeover through cross-shareholding beforehand would result in an increase in corporate value because corporate managers would not need to pursue a myopic strategy for the sake of avoiding a hostile takeover when faced with such a threat (Osano 1996). Both studies show that cross-shareholding would increase overall economic welfare under certain conditions. However, it is questionable to what extent this argument can be generalized.

Is there any rationality for cross-shareholding?

By combining the above arguments, the following can be said at this moment about the rationality of cross-shareholding, specifically from the viewpoint of shareholders. Vertical cross-shareholding can result in higher corporate value and thus is rational, provided there exists a substantial need for a parts supplier to make capital investments specifically for the benefit of a particular final product manufacturer in order to improve the compatibility of parts and components with the relevant final products. Vertical cross-shareholding also adds value in the case where an R&D firm has a strong need to seek cooperation from a large, established company.

Horizontal cross-shareholding should not necessarily have to be ruled out from the viewpoint of shareholders in the sense that such an arrangement would increase corporate value when applied to companies in an industry plagued by inefficiency resulting from excessive competition. Meanwhile, horizontal cross-shareholding between companies in an industry not experiencing such problems appears to be beneficial only to those holding shares in the companies concerned, because such an arrangement increases the value of these companies. However, this is not positive for shareholders investing in a diversified portfolio of companies (e.g., pension funds) because they may hold shares of companies involved in said capital tie-up as well as those using their products. Needless to say, such horizontal cross-shareholding is undesirable from the viewpoint of shareholders as consumers.

Finally, conglomerate cross-shareholding may work to tighten discipline over corporate managers in specific conditions, but this does not apply generally. Rather, such an arrangement is more likely to preclude hostile takeover bids even when they are desirable from the shareholders' point of view. This provides grounds for concluding that conglomerate cross-shareholding should be avoided.

September 3, 2008
  • Berglof, E. and E. Perotti, "The Governance Structure of the Japanese Financial Keiretsu," Journal of Financial Economics 36, 259-284, 1994
  • Dasgupta, S. and Z. Tao, "Bargaining, Bonding, and Partial Ownership," International Economic Review 41, 609-635, 2000
  • Mathews, R.D., "Strategic Alliances, Equity Stakes, and Entry Deterrence," Journal of Financial Economics 80, 35-79, 2006
  • Mathews, R.D., "Optimal Equity Stakes and Corporate Control," Review of Financial Studies 20, 1059-1086, 2007
  • Osano, H., "Intercorporate Shareholdings and Corporate Control in the Japanese Firm," Journal of Banking and Finance 20, 1047-1068, 1996

September 3, 2008

Article(s) by this author