Research and articles

Research

  • The anti-competitive effects of common ownership in airlines

    Azar, Schmalz, and Tecu (2018, “AST”): “Anti-Competitive Effects of Common Ownership

    Journal of Finance

    • Common ownership (“CO”, airlines being more jointly held by the same asset managers) leads to airline ticket prices being 3-7% higher than the would be under separate ownership

    • Uses “MHHI Delta” measure of O’Brien and Salop (2000) to gauge CO. This measures if an investor holds large stakes in multiple companies with high market shares

    • To identify causation, the authors study BlackRock’s acquisition of Barclays Global Investors as a shock to CO. They argue that this merger increased CO, as BlackRock’s stake in each airline rose. Ticket prices also increased
  • Methodological concerns with the above study

    Dennis, Gerardi, and Schenone (2018): “Common Ownership Does Not Have Anti-Competitive Effects in the Airline Industry

    • MHHI Delta depends on firms' market shares by sales. Sales depend on prices. Thus, AST regress prices on something highly correlated with prices, mechanically leading to a significant relationship. When this is accounted for, the results are substantially weaker

    • Standard regression analysis treats all observations equally. AST give more weight to high-traffic markets. When removing these weights, the results are substantially weaker

    • AST failed to take into account that many airlines had filed for bankruptcy. Thus, management’s fiduciary duty was primarily towards creditors, so shareholders didn’t have effective control. After accounting for this, the results are substantially weaker

    • AST lump all tickets together, ignoring the difference between business and economy flights, and flights with and without stops. When applying the standard filters in the literature (e.g. focusing on direct economy flights), the results are substantially weaker

    • See here for AST’s reply

     

  • Incorrect measure of common ownership

    Gilje, Gormley, and Levit (2019): “Who’s Paying Attention? Measuring Common Ownership and Its Impact on Managerial Incentives

    • An investor’s stake in a firm is an incomplete measure of her incentives to monitor it. Her investor’s time is limited – thus, even if she has a large stake in a firm, she may devote little monitoring effort if she has large stakes in many other firms

    • AST’s use of the percentage equity stake essentially ignores all other firms within the investor’s portfolio. When this is taken into account, the results disappear

    • AST assume that BlackRock’s acquisition of BGI increased CO because its stakes in each airline rose. However, its stakes in many other firms rose as well, so BlackRock’s incentives to monitor airlines actually fell in many cases, which further questions the interpretation of AST’s findings

     

  • Incorrect application of the common ownership theory

    Kennedy, O’Brien, Song, and Waehrer (2017): “The Competitive Effects of Common Ownership: Economic Foundations and Empirical Evidence

    • Authors argue that AST misuse the MHHI Delta proposed in O’Brien and Salop (2000, “OS”)
    • The OS theory stresses that the MHHI Delta is an outcome, not a driver. Rather than MHHI Delta driving increases in prices, other factors may drive both the MHHI Delta and also prices. When conducting empirical tests tightly linked to the theory, the results disappear.
    • See here for AST’s reply

       

  • Results may be driven by the financial crisis

    Lewellen and Lowry (2019): “Does Common Ownership Really Increase Firm Coordination?

    • The BlackRock-BGI merger took place during the financial crisis, so any results could be due to the financial crisis’s effect on the airline industry

    • To identify causal effects, the authors study mergers of other financial institutions, many of which occurred outside the crisis. Using these mergers, they find no evidence that increases in CO raised cooperation. They also show that the Blackrock-BGI merger can produce spurious evidence of such improvements

     

  • Results do not extend to other industries

    Koch, Panayides, and Thomas (2019): “Common Ownership and Competition in Product Markets

    • While AST consider a single, highly-regulated (and thus potentially atypical) industry, the authors study all industries

    • CO is neither positively associated with industry profitability or output prices, nor negatively correlated with measures of non-price competition, as would be the case if CO had anti-competitive effects

     

  • The anti-competitive effects of common ownership in banking

    Azar, Raina, and Schmalz (2019): “Ultimate Ownership and Bank Competition

    • CO is associated with lower branch-level deposit account interest rates, higher maintenance fees, and lower fee thresholds

    • Results are robust to focusing exclusively on the increase in bank ownership caused by the growth of index funds

     

  • The effects of common ownership on managerial incentives

    Anton, Ederer, Gine, and Schmalz (2018): “Common Ownership, Competition, and Top Management Incentives

    • CO is associated with weaker managerial incentives to improve own-firm performance

     

Articles

  • Misguided fears of common ownership (Tom Gosling)
    - The evidence for the anti-competitive effects of common ownership is weaker than frequently portrayed
    - But there may be other important effects of common ownership, such as increasing proxy advisor influence