In many markets, large institutional investors like BlackRock and Vanguard own significant stakes in most, if not all, horizontal competitors. Legal scholars have suggested that this pattern of ownership might reduce the incentives of horizontal competitors to engage in aggressive competition (horizontal shareholding hypothesis). In this vein, they have argued that horizontal shareholding above a certain threshold should be considered (quasi) per se illegal, or even that institutional investors should be allowed to hold shares only in one horizontal competitor in each market.
In my recent working paper ‘Horizontal Shareholding: The End of Markets and the Rise of Networks’, I argue that legal scholars have misunderstood the consequences of such ownership structure. In particular, the horizontal shareholding hypothesis implies that horizontal shareholders no longer attempt to maximize the value of each firm separately. Instead, because they own stocks in horizontal competitors, institutional investors are assumed to be ‘market-value maximizers’ and attempt to separately maximize the net present value of the firms in each market in which their portfolio companies operate. In other words, within the horizontal shareholding hypothesis institutional investors engage in m separate maximization problems, where m is the number of markets in which their portfolio firms operate. But this assumption is implausible. As markets are interconnected, higher prices in one market might affect firms operating in other markets. For instance, higher prices in the airline industry have a negative impact on airlines’ suppliers and consumers. And, because most large institutional investors own stakes also in these firms, they have every reason to account for these inter-market spillovers and act as ‘portfolio-value maximizers’. Once inter-market spillovers are incorporated in the analysis, it can be shown that the main proposals advanced by the literature could backfire and further reduce the level of competition among horizontal competitors.
Nevertheless, there is empirical evidence suggesting that the current patterns of ownership might be reducing the level of competition in some markets. If one believes that the existing empirical evidence is sufficiently compelling, then one might also believe that structural reforms are necessary. I remain agnostic on whether such structural reforms are desirable and how invasive they should be. However, I argue that if any such reform is to be implemented, it must account for inter-market effects, and therefore must be grounded on the insights of network theory.
A ‘network-sensitive’ policy could be structured as follows: Whenever the level of horizontal shareholding in one market is above a certain threshold, the Federal Trade Commission informs the horizontal shareholders that there is a rebuttable presumption that the current ownership structure of the horizontal competitors has anticompetitive effects. Using tools already developed by network theorists and data that the US Bureau of Economic Analysis kept collecting until 2007, the horizontal shareholders will be allowed to show that higher prices in the market of interest create negative spillovers onto other markets in which they have stakes. Two scenarios are then possible. First, these spillovers would cause losses to other firms in the portfolio of the horizontal shareholders that are larger than the possible gains accruing to the horizontal shareholders due to the higher prices in the relevant market(s). In this case, the presumption that high horizontal shareholding creates anticompetitive effects is rebutted simply because horizontal shareholders would have no reason to prefer higher prices. Alternatively, the potential gains from higher prices in the market of interest are not offset by negative spillovers. In this case, horizontal shareholders will have either to divest partly from the market of interest or to buy additional stakes in firms that would be negatively affected by high prices in the market of interest.
The proposed approach is significantly less intrusive than the reform proposals advanced in the literature. Moreover, it has the key advantage that it allows policymakers to account for the specific structure of each market and the portfolio composition of each investor. In fact, not all markets are the same, as some hub-like sectors create disproportionally large spillovers compared to other industries. It is likely that a reduction in the levels of input and output in these central industries (eg wholesale trade) has a negative impact on a large number of other sectors. Therefore, higher prices in these industries might be against the best interests of institutional investors. Similarly, not all horizontal shareholders are the same. Both BlackRock and Berkshire Hathaway have acquired significant stakes in many of the largest US airlines. However, while the former has stakes in virtually all US companies, the latter concentrates its investments in a much smaller number of firms. As a consequence, BlackRock will internalize a larger fraction of the input-output spillovers from the airline industry.
Alessandro Romano is a researcher at Yale Law School (JSD).