AdobeStock_232692449.jpeg
VoxEU Column Competition Policy

Revisiting the anticompetitive effects of common ownership

The rise of common ownership of publicly traded companies has important implications for competition in product markets. This column argues that it is important to distinguish inter-industry from intra-industry effects of common ownership. Using data from the airline industry, it shows that intra-industry common ownership is positively associated with prices, while inter-industry common ownership is negatively associated with prices. Antitrust regulation should take into account the procompetitive inter-industry effects of common ownership.

As is by now well known, common ownership of publicly traded companies has increased rapidly in recent years. A debate has emerged over whether this can affect competition, with a special focus on product prices. The theory is allegedly simple enough: if companies in the same industry have the same owners, and they act in the interest of their shareholders, they will compete less aggressively in product markets (Rotemberg 1984, O’Brien and Salop 2000).

However, this theory misses an important point, which is that the recent rise of common ownership is not an industry-wide phenomenon but an economy-wide one, driven to a large extent by index funds who are close to ‘universal owners’ and hold every publicly traded firm in the economy. In fact, our recent theoretical work (Azar and Vives 2021) shows that, in a general equilibrium oligopoly model, common ownership covering the whole economy implies lower markups for consumers, not higher. The reason is that, in general equilibrium, when an industry expands, it creates positive externalities for firms in other industries, and therefore inter-industry common ownership increases the incentive for firms to expand, reducing prices in their industry relative to the price level. It turns out that this effect, in a standard model, is stronger than the intra-industry effect that common ownership of firms in the same industry generates. Thus, the total effect is to reduce product market markups.

The empirical literature, however, has so far mainly focused on measuring intra-industry common ownership and its effects. Therefore, inter-industry common ownership is a crucial missing variable in the analysis. In this column, we address this problem by measuring both intra-industry and inter-industry common ownership and reassess the evidence on its competitive effects in the airline industry. Although the theory is not specific to the airline industry, we use it as an empirical example because it allows us to directly compare the results with those of Azar et al. (2018), and thus see which of the results in that paper change when taking into account general equilibrium effects.

Our main finding is that, while it is still the case that intra-industry common ownership is positively associated with airline prices, inter-industry common ownership is negatively associated with airline prices. The overall predicted effect of common ownership on prices is positive in some routes and negative in others. The average effect is positive, but only because some shareholders are concentrated in airlines, and therefore this failure of complete diversification implies that intra-industry common ownership is still somewhat higher than inter-industry in practice. Although measures of common ownership are positively correlated, a LASSO variable selection model, with the penalty parameter chosen using ten-fold cross validation to minimise out-of-sample prediction errors, suggests that both variables should be included in the model.

Furthermore, we conducted a panel vector autoregression (VAR) analysis and found that both intra-industry (lambda intra) and inter-industry (lambda inter) common ownership Granger-cause prices, in the sense that past values of the lambdas have significant predictive power for future prices, while past values of prices do not significantly predict future changes in common ownership.

Figure 1 Impulse response functions from panel VAR of airline prices, intra-industry common ownership, and inter-industry common ownership

 

In addition, we separate intra-industry common ownership into two measures – one measuring intra-industry common ownership by the ‘Big Three’ asset managers (BlackRock, Vanguard, and State Street), and one measuring intra-industry common ownership by other shareholders that are not the Big Three. We find that, while intra-industry common ownership by shareholders other than the Big Three is positively associated on airline prices, common ownership by the Big Three is negatively associated with airline prices (although the negative effect on prices is not statistically significant in all specifications). When controlling for inter-industry common ownership, the effect of intra-industry common ownership by the Big Three becomes positive. However, we show that the overall effect of the Big Three on prices is negative.

Figure 2 Distribution of the total effect of common ownership on the linear prediction of log price: Big Three versus other shareholders

 

One of the main methodological criticisms of Azar et al. (2018) is that its measure of the impact of common ownership, the MHHI delta, depends on the market shares of the firms in the market, which are endogenously determined. The MHHI is an augmented version of the HHI taking into account overlapping ownership between the firms in an industry. The MHHI delta is the difference between the MHHI and the HHI. However, our economic model suggests that a share-weighted average of a firm's lambdas is a better measure of that carrier's common ownership. To address the endogeneity of market shares, we use unweighted averages of the objective function weights. Using pairwise objective function weights to measure of common ownership was proposed by Azar (2012: Chapter 7).

In addition to the panel regression analysis, we conducted an event study based on mergers of financial institutions, following He and Huang (2017) and Lewellen and Lowry (2021). These mergers generated variation in intra-industry and inter-industry common ownership across airlines. Estimating the effect of lambdas on prices based on variation from these acquisitions only, we confirmed our main result that increases in intra-industry common ownership lead to increases in prices, while increases in inter-industry common ownership lead to lower prices. Bindal and Nordlund (2022) use a similar methodology and find that the positive effect of intra-industry common ownership on margins is more pronounced for firms with more similar products.

Figure 3 Effect of intra-industry and inter-industry common ownership on prices, based on an event study of asset manager mergers

 

There is a debate over which is a plausible mechanism that connects overlapping ownership with firms' decisions (e.g. Hemphill and Kahan 2019, Elhauge 2021, Tzanaki 2022). We are agnostic on this question. However, it is worth pointing out that inter-industry common ownership is in the radar of large asset managers when they are considering their corporate governance strategy. For example, recently Barbara Novick wrote an article for the Harvard Law School Forum on Corporate Governance (Novick 2019) addressing the issue (emphasis in original):

The ‘common ownership’ theory relies on the assumption that all ‘common owners’ benefit from lessened competition, as it is derived from theories of oligopolies and ‘cross ownership’ (e.g., where a company buys a stake in its competitor). While lessened competition might benefit certain concentrated investors, broadly diversified investors, like index funds, own the whole market and do not benefit from lessened competition. This is because broadly diversified investors are subject to inter–industry effects—meaning that what happens in one sector affects the performance of the fund’s holdings in other sectors.

Our results are potentially important for the recent debate on the antitrust implications of common ownership (see Elhauge 2016, Posner et al. 2017, and Rock and Rubinfeld 2017). Our general equilibrium analysis shows that anticompetitive effects in product markets are driven by intra-industry common ownership while inter-industry common ownership is procompetitive. The result is that, because of inter-industry effects that were ignored in earlier empirical work, common ownership by diversified shareholders like the Big Three is actually predictive of lower product market prices.

References

Azar, J (2012), "A new look at oligopoly: Implicit collusion through portfolio diversification", Doctoral dissertation, Princeton University.

Azar, J, M C Schmalz and I Tecu (2018), “Anticompetitive effects of common ownership”, The Journal of Finance 73(4): 1513-1565.

Azar, J and X Vives (2021), “General equilibrium oligopoly and ownership structure”, Econometrica 89(3): 999-1048.

Bindal, S and J Nordlund (2022), “When does common ownership matter?”, SSRN Working Paper.

Elhauge, E (2016), “Horizontal shareholding”, Harvard Law Review 129(1267).

Elhauge, E (2021), “The causal mechanisms of horizontal shareholding”, Ohio State Law Journal 82(1).

He, J J and J Huang (2017), “Product market competition in a world of cross-ownership: Evidence from institutional blockholdings”, The Review of Financial Studies 30(8): 2674-2718.

Hemphill, C S and M Kahan (2019), “The strategies of anticompetitive common ownership”, Yale Law Journal 129, 1392.

Lewellen, K and M Lowry (2021), “Does common ownership really increase firm coordination?”, Journal of Financial Economics 141(1): 322-344.

Novick, B (2019), “Diversified portfolios do not reduce competition”, Harvard Law School Forum on Corporate Governance.

O'Brien, D P and S C Salop (2000), “Competitive effects of partial ownership: Financial interest and corporate control”, Antitrust Law Journal 67: 559-614.

Posner, E A, F M S Morton and E G Weyl (2017), “A Proposal to Limit the Anti-Competitive Power of Institutional Investors”, Antitrust Law Journal 81: 1-2.

Rock, E B and D L Rubinfeld (2017), “Defusing the antitrust threat to institutional investor involvement in corporate governance”, NYU Law and Economics Research Paper, 17-05.

Rotemberg, J (1984), “Financial transaction costs and industrial performance”, MIT Sloan Working Paper.

Tzanaki, A (2022), “Varieties and Mechanisms of Common Ownership: A Calibration Exercise for Competition Policy”, Journal of Competition Law & Economics 18(1): 168-254.

315 Reads