Undisputed empirical studies confirm that horizontal shareholding poses a great anticompetitive threat. What can antitrust enforcers do about it? Quite a lot, in fact.
Editors’ note: In the last few weeks, the Federal Trade Commission has been holding a series of public hearings to discuss whether competition enforcement policies should be updated to better reflect changes in the US economy, namely market concentration and the proliferation of new technologies. The FTC hearings, which will be held throughout the fall and winter, cover topics as varied as privacy and big data, the consumer welfare standard in antitrust and labor market monopsonies. In order to provide ProMarket readers with a better understanding of the debates, we have asked a number of selected participants to share their thoughts on the topics at hand.
You can read all previous installments here.
Horizontal shareholding exists when the leading shareholders of horizontal competitors overlap. (Although often called “common shareholding”, that terminology is imprecise because common shareholding often exists between noncompeting companies.) Such overlaps used to be relatively rare, but has become common because of the growth of institutional investors and index funds. Moreover, their influence is even higher than their shareholdings because they are far more likely to vote than individual investors. Today, institutional investors own 70 percent of shares and cast 88 percent of votes in publicly-traded firms, and those figures rise to 80 percent of shares and 93 percent of votes at S&P 500 firms. The big three index fund families (BlackRock, Vanguard, and State Street) alone own 18 percent of shares and cast an estimated 24 percent of votes in publicly-traded firms, and own 20 percent of shares and cast an estimated 26 percent of votes at S&P 500 firms.
The result is that today the leading shareholders of large competing firms often comprise the same set of institutional investors. Common sense and economic theory indicates that firms with the same leading shareholders are less likely to compete vigorously against each other.
As I detail in my article, a series of empirical studies have confirmed that horizontal shareholding has anticompetitive effects in concentrated markets. Although one of those studies has been disputed, the others largely have not.
One of the undisputed studies is a cross-industry regression analysis by Gutiérrez and Philippon which found that the gap (now historically high) between corporate profits and investment is mainly driven by the level of horizontal shareholder ownership in concentrated markets. Further, this new study found that, within any industry, the investment-profit gap is mainly driven by those firms with high horizontal shareholding levels.
Gutiérrez and Philippon’s undisputed study establishes a link between anticompetitive horizontal shareholding and the economy-wide lack of corporate investment that has contributed to low economic growth in recent decades. Indeed, it suggests that horizontal shareholding is a more important cause than other causes (like general macroeconomic, technological, or policy trends) that would operate similarly across industries and certainly across all firms within any industry.
Another recent cross-industry empirical study by Anton, Ederer, Gine and Schmalz shows that horizontal shareholding makes changes in executive wealth less sensitive to firm performance. This study is undisputed as well. To be sure, critics like to focus on an old dispute that existed about the empirical effect of horizontal shareholding on executive annual pay, but such annual pay captured only 22 percent of executive wealth changes. The natural effect of making changes in executive wealth less sensitive to firm performance is obviously to blunt executive incentives to compete.
Recent industry studies have likewise confirmed anticompetitive effects from horizontal shareholding. Two empirical studies by Jin Xie and Joseph Gerakosz and by Newham et. al show that horizontal shareholding delays and prevents entry into pharmaceutical markets. These studies are undisputed too.
Another empirical study by Azar, Raina, and Schmalz shows that horizontal shareholding harms bank fees and rates. Although critics argue that this study has been disputed by another study that found more mixed results, there is not much of a real dispute yet—the “contrary” study stresses that its results are preliminary because it relies on a data source with known irregularities that they have not yet investigated and corrected. Moreover, this contrary study tries to measure the effects of horizontal shareholding while excluding any information about bank market shares and concentration, which naturally makes the results more mixed. After all, economic theory indicates that even absolute horizontal mergers between some firms in an unconcentrated market are unlikely to affect prices.
If a study of all horizontal mergers (whether or not in concentrated markets) found mixed effects on prices, no one would conclude that it proves that horizontal mergers in concentrated markets have no anticompetitive effect. Likewise, no one should conclude that a study of all horizontal shareholding (whether or not in concentrated markets) that finds mixed effects on prices proves that horizontal shareholdings in concentrated markets have no anticompetitive effect.
The empirical dispute has focused on a study by Azar, Schmalz and Tecu showing that horizontal shareholding harms airline prices. But this finding has been replicated by critics even using their own re-construction of the data and measures of horizontal shareholding. Critics have negated airline price effects only by incorrectly altering the regression, such as by using an instrumental variable that is actually negatively correlated with horizontal shareholding or by incorrectly setting many shareholding rights equal to zero. Accounting for other critiques, like claims of endogeneity, imprecise market definition, or the effects of changing fuel costs, turn out to actually increase the estimated price effects.
The empirical literature is thus not too uncertain to take action against horizontal shareholding. In any event, empirical uncertainty could not support current enforcement practices, which rely on metrics like HHIs that affirmatively assume that horizontal shareholding has zero effect—an assumption that clearly lacks any theoretical or empirical basis.
Critics argue that we need clearer proof of the precise causal mechanisms. But the causal mechanisms are the same as the ones that law and economics scholars have always cited to explain how agency slack is limited: board elections, executive compensation, control contests, the stock market, and the labor market. As I detail in my article, each of those mechanisms is supported by copious evidence.
Critics also argue that three contra-mechanisms will prevent horizontal shareholding from having anticompetitive effects. But those contra-mechanisms are all unpersuasive. First, critics argue that anticompetitive effects will be prevented by accountability to non-horizontal shareholders. But when horizontal shareholding has anticompetitive effects, it affirmatively benefits non-horizontal shareholders because it increases profits at their firm by simultaneously lessening competition at both their firm and rival firms. Thus, non-horizontal shareholders have no more incentive to object to anticompetitive horizontal shareholding than they would to the corporation entering into a legally-permitted cartel.
Second, critics argue that index funds who are horizontal shareholders in the airline industry also own stock in firms that buy from or sell to that industry (such as airplane makers and business travelers), which, they argue, should negate any anticompetitive incentives. But even the large index funds stressed by this argument externalize the lion’s share of anticompetitive effects onto buyers outside the index. Indeed, in the airline example favored by critics, the S&P 500 externalizes 95 percent of the anticompetitive effect outside of firms in the index.
Third, critics argue that index fund families lack incentives to exert the effort needed to influence corporate behavior in anticompetitive ways to increase portfolio value. But those critics are mistaken, because the anticompetitive gains are vast and the incremental effort costs are generally zero or negative, given that the index fund families have to decide what position to take on votes or management interactions either way and taking a position that does not pressure managers to compete more vigorously is likely to please them.
Index fund families also have strong incentives to compete with active funds for investment flow, and index fund families have many active funds of their own. The critics also ignore the fact that what matters is relative effort, and however much index fund incentives to exert effort may fall short of the optimal, the incentives of other shareholders with smaller shareholdings are even less. In any event, any theoretical claim that index fund families lack sufficient incentives to influence corporations is clearly disproven by empirical study after empirical study showing that, in fact, index fund families are hugely influential in influencing corporate behavior.
Moreover, index fund families are not the only or main horizontal shareholders. Claims that horizontal shareholders lack incentives to influence corporate behavior conflict not only with the above-noted empirical studies showing anticompetitive effects, but with dozens of other empirical studies showing that common shareholders influence corporate behavior on many other dimensions. At some point, theoretical claims that it is implausible that common shareholding could affect corporate behavior must give way to the dozens of empirical studies showing that it does just that.
What can antitrust law do about horizontal shareholding? Quite a lot, in fact. The Clayton Act bans any stock acquisition that may substantially lessen competition, and Supreme Court case law makes clear that continuing to hold stock is an “acquisition.” Even critics acknowledge that the plain meaning of the Clayton Act would ban horizontal shareholding—if they were convinced that it empirically had anticompetitive effects. Indeed, the Clayton Act has one provision for cross-shareholding between commercial entities and another, separate provision that expressly extends the Act to stock acquisitions by any entity in commercial entities: i.e., to horizontal shareholding.
To be sure, the Clayton Act has a solely-for-investment exemption, but it exempts a stock acquisition from liability only if it both (1) confers no influence over the corporation and (2) fails to actually create anticompetitive effects. The exemption thus does not apply if anticompetitive effects are actually proven. Contrary claims that the Clayton Act applies only if a stock acquisition confers control or influence conflict not only with the statutory text, but also with six federal court opinions and the agency guidelines on cross-shareholding.
In any event, horizontal shareholding necessarily involves contractual agreements, so when those agreements have anticompetitive effects, they necessarily constitute agreements in restraint of trade. Thus, the Sherman Act bans any horizontal shareholding that has anticompetitive effects. Indeed, the historic trusts attacked by the Sherman Anti-Trust Act were horizontal shareholders.
EU merger law cannot remedy all anticompetitive horizontal stock acquisitions, but it could remedy the subset of those acquisitions that potentially give horizontal shareholders decisive joint influence over corporate activities. In any event, EU Article 101 on agreements and concerted practices is at least as broad as the Sherman Act and can thus condemn any horizontal shareholdings that have anticompetitive effects. EU Article 102 also bans anticompetitive horizontal shareholdings because such shareholdings create a collective dominant position and abuse that position with excessive pricing.
Anticompetitive horizontal shareholding is thus also illegal in all the other nations (including China) that ban anticompetitive concerted practices and/or abusing a collective dominant position via excessive pricing.
Even if we fail to directly tackle horizontal shareholding, its effects have important implications for traditional merger analysis. In particular, it lowers the concentration levels that traditional merger analysis can tolerate. Continuing to allow unimpeded horizontal shareholding would thus provide strong support for those who currently argue that antitrust law should be far more aggressive about preventing market concentration.
Horizontal shareholding can also mean that what now look like non-horizontal mergers should be treated as horizontal. Horizontal shareholding thus turns out to also support current antitrust movements that decry our increasing levels of national industrial concentration, even if those concentration figures do not correspond to relevant antitrust markets.
For more on this, check out the following episode of the Capitalisn’t podcast:
Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.