European Corporate Governance Network

Comment on Becht/Roëll, "Blockholdings in Europe: An international comparison", prepared for the ECGN meeting at Milan, Nov. 6/7, 1998. by Theodor Baums (Institut für Handels- und Wirtschaftsrecht, Universität Osnabrück)

I.

In this paper the authors preview empirical work by the European Corporate Governance Network on the size of block shareholdings in Europe. The most salient finding is the extraordinarily high degree of concentration of shareholder voting power in Continental Europe relative to the U.S. and the U.K. From this finding the authors conclude that - citation - "while in the USA the main agency problems seem to stem from conflicts of interest between managers and dispersed, insufficiently interventionist shareholders, in much of continental Europe there are generally large blockholders present who can and do exercize control over management. Instead, the main potential conflict of interest lies between controlling shareholders and powerless minority shareholders " - end of citation - (p. 3).

I would rather - from my perspective as a lawyer - suggest great care in drawing conclusions from this still very thin empirical basis. I would rather recommend to extend the empirical research into various further directions.

II.

First. So far the data in the authors' paper display the largest ultimate voting blocks for all listed companies. But that does not say very much about whether in this sample management or a large owner holds the reins. To give you an extreme example: If in a sample of one hundred listed companies in all cases the ultimate voting block is held by a small group of, say, 5 companies with widely dispersed ownership, we end up with an extreme form of a management-controlled system. Let me give you a more realistic example. Figure 1 [pdf] shows only a small part of the holdings of Deutsche Bank AG which itself is a widely held stock corporation. Now let us assume that all stock corporations which I have marked are listed companies, and let us further assume that Deutsche Bank is always the largest ultimate voting block holder. If that is the case we end up with an extreme form of a management rather than a shareholder-controlled system: For it is of course much more difficult for a shareholder of, say, the Schiffshypothekenbank zu Lübeck AG to control the management of the controlling shareholder Deutsche Bank than its own management. And it is also extremely difficult for the shareholder of Deutsche Bank to know what is going on in Deutsche Bank's subsidiaries. If a company with widely distributed shares breaks its single firm down into a group with lots of subsidiaries and joint ventures with other similar holding companies the principal-agent problem will worsen. The governance structure of the ultimate controlholders is crucial. In their detailed ECGN country report on Germany Becht and Boehmer present some more information on this point. They display the number of share blocks in the German listed companies and the names and types of the major blockholders. Their findings show that holdings by banks, industrial firms, insurance firms and holding companies play the most important role by far. What we still need to know is the governance structure of these ultimate blockholders. To the extent that these are also stock corporations with widely held ownership, foundations, state-owned companies and the like, we probably end up with a much more management-oriented and management-controlled system than one would expect at first sight. As I said, principal-agent problems in a group of companies can be much worse than in a single firm.

III.

Second. The conclusion that - citation - "in the USA the main agency problems seem to stem from conflicts of interest between managers and dispersed, insufficiently interventionist shareholders, [whereas] in much of continental Europe there are generally large blockholders present who can and do exercize control over management" points at another limitation of the study which once more suggests to take care in drawing conclusions from its findings. The study does not take the different regulations and legal institutions into account. Comparing mere numbers of voting stake distributions in different countries does not say very much about the extent of principal-agent problems between shareholders and managers and about conflicts among shareholders. The influence which shareholders may exercize by casting their votes depends completely on the range of issues on which the shareholders' meeting may take decisions under the respective legislation. And the informal influence stemming from a large stake in a company may be diluted by legal provisions like, say, co-determination or other management-friendly rules. On the other hand, there may be also in systems with widely dispersed ownership provisions and regulations in place, like transparency rules, incentive contracts, directors' liability and derivative suits, the threat of takeovers and the role of a strong capital market supervisory authority, which could serve to reduce agency problems. Once again, comparing the sizes of voting blocks without taking the legal and regulatory framework into account does not tell us very much about the reality.

IV.

Final remark. How are we to explain the empirical findings - large blockholdings on the Continent vs. widely dispersed ownership in the Anglo-Saxon countries? Several possible explanations come to mind and should be explored further:

- First. Private investors funnel their money into other investments, for instance, a state-organized social security system. Institutional investors are therefore also underdeveloped. The remaining investors (families, foundations, other firms, the state) may have incentives to acquire and hold larger stakes.

- Second. The regulation of financial intermediaries and the capital markets looks different. German banks may hold equity stakes in industrial firms and vote their client's shares; Anglo-Saxon ones don't. British companies must make a mandatory bid on all shares above a stake of 30%. Or: Until recently small investors on the Continent were not efficiently protected against insider trading.

- Third. There may be tax incentives to build up certain thresholds, e.g., the 10% stake incentive in German tax law. On the other side, there may be tax disincentives to build up corporate groups. Everywhere in the world managers of companies with widely distributed shares will have an interest in building up pyramidal structures with subsidiaries and sub-subsidiaries. Similarly, a pyramidal structure allows the ultimate shareholder to wield comparatively more power with a relatively small equity investment than he could exercize in a single firm. But if you have a tax system which taxes earnings on each level rather than the consolidated earnings of the whole group the incentives for empire-building will be suppressed.

- Lastly, different types of production require different amounts of money. Large firms will more likely have a widespread ownership than medium-sized and small firms. Industries with large firms will have more companies with dispersed ownership than industries with small and medium-sized firms.

Further research should try to corroborate or disprove these impressions.

Last updated: April 29, 1999.