The common ownership hypothesis suggests that, when large investors own shares in many firms within the same industry, those firms may have reduced incentives to compete. Empirical contributions document the rising importance of common ownership and provide evidence to support this hypothesis. However, because managers rather than investors control firm operations, scepticism that common ownership affects product market outcomes will be warranted until a mechanism for this process is identified. This has fuelled a vigorous debate on whether existing evidence on common ownership has a plausible causal interpretation.
This paper, available here, develops a theoretical model that explains how managerial incentives can serve as a simple and plausible mechanism that links higher common ownership and softer product market competition. Under this model, firm-level variation in common ownership causes variation in managerial incentives across firms as well as variation in product prices, market shares, concentration and output across markets—all without communication between shareholders and firms, coordination between firms, or knowledge of shareholders’ incentives or market specific interventions by top managers. The paper also seeks to confirm this model empirically, by evaluating whether there is evidence in favour of the model’s prediction that top management incentives are less performance-sensitive in firms where large investors possess greater ownership stakes in competitors.
Section 2 presents the theoretical model.
The model embeds a canonical managerial incentive design problem with moral hazard into a conventional model of strategic product market competition with diversified owners. The model captures the agency conflicts that exist between those who manage firms (managers) and those who own them (investors), and recognises that large investors routinely hold ownership stakes in several firms in the same industry.
The central driving force of the model is that performance-sensitive managerial compensation encourages productivity-improving managerial effort. This, in turn, has two effects. First, in a setting in which product prices are assumed to be fixed, productivity-improving managerial effort increases firm profitability and is thus desirable for all owners. Second, with endogenous product prices, productivity enhancements also increase how fiercely the firm competes in the product market. The latter channel indirectly reduces the profitability of competing firms, and thus stands in conflict with the interests of common owners who hold shares in other firms in the same industry.
As a result, the model predicts a negative relationship between common ownership and the sensitivity of top management incentives to firm performance. This finding does not rely: (i) on owners having access to sophisticated market-level incentives or communications to steer product market behaviour in different markets, (ii) on top managers’ knowledge of the ownership structure of either their own firm or their competitors, (iii) on top managers making detailed market-specific strategic choices (e.g., setting prices), nor (iv) on explicit or tacit collusion between managers, firms or shareholders. Instead, the mechanism relies solely on managers exerting firm-wide productivity-improving efforts solely based on their own explicit incentives (as in any standard corporate finance or organisational economics model), and on market-level specialists making product market choices solely based on market demand and firms’ cost structures (as in standard industrial organization models).
By allowing for asymmetries in firm-level common ownership in a multimarket industry, the model also generates additional firm- and market-level predictions. First, within the same industry more commonly-owned firms have weaker managerial incentives and compete less aggressively (i.e., set higher prices) than less commonly-owned “maverick” firms with stronger managerial incentives. Second, even though top managers can only exert productivity-improving effort over a single firm (rather than several market-specific ones), commonly-owned firms compete less aggressively in markets in which they face other commonly-owned firms than in markets in which they face maverick firms.
Section 3 details the data and presents a number of common ownership measures.
In this section, the authors test empirically the model’s main prediction about the structure of top management compensation, since this is how the causal link between common ownership and product market outcomes is established in this model. The authors rely on data on different types of executive compensation, ownership and industry.
However, to identify how common ownership is related to managerial incentives, one first requires a measure of common ownership. The authors identify (and use) a number common ownership measures derived from the industrial organisation literature, such as equal- or value-weighted average of the weights on the profits across industry competitors; the average fraction of competitor shares held by the firm’s top five shareholders; the total value of stock held by all the common shareholders of two industry competitors scaled by the total market capitalisation of their stocks; and the modified Herfindahl-Hirschman index delta.
Section 4 presents the empirical results.
The paper documents a strong and robust negative association between a firm’s common ownership and the performance sensitivity of its top management’s compensation. The authors estimate that an interquartile range shift (25th to 75th percentile) of the firm-level degree of common ownership is associated with a 6.6% reduction of CEO wealth-performance sensitivity. To put this incentive-reducing effect of common ownership in perspective, it is comparable to the effect of a one-standard deviation reduction in firm volatility on managerial incentives.
The authors control for industry structure, firm- and manager-level characteristics, as well as (time-invariant) firm fixed effects and (time-varying) industry-year fixed effects. This empirical design choice ensures that spurious inferences from industry-wide trends or time-invariant firm compensation policies are avoided. The result remains robust if various alternative measures of managerial incentives, common ownership, and industry definitions are used. Across all dimensions of the full matrix of robustness checks (i.e. managerial wealth-performance sensitivities, common ownership measures, and industry definitions), the results remain consistently negative, with similar (or even larger) economic magnitudes and statistical significance levels
Section 5 concludes.
This paper proposes a model of unilateral incentives arising from managerial compensation as a mechanism through which common ownership can influence product market competition. Although the empirical results are consistent with this mechanism being important, they do not prove that compensation is the primary mechanism, let alone the only mechanism that causes market-level correlations between common ownership and product market outcomes.
Nor do the authors claim that the compensation channel is the most effective way through which common ownership can affect product market competition. In fact, there are several examples of common shareholders tying compensation to explicit output targets. The anti-competitive effects will also be more pronounced if owners can design managerial incentive compensation schemes that focus on firm output, or on measures more directly linked to output than profits such as profit margins or relative performance.
Instead, this paper merely shows that the mechanism through which common ownership affects product market competition can be very subtle, and therefore provides a conservative lower bound for the anti-competitive effects of common ownership.
This is a long paper (almost 90 pages, if one counts the annexes). One can tell that the authors went to great lengths to ensure that the model and its results are persuasive – the model, and the choices it entailed, is discussed through 17 very technical pages (and a 12-page appendix to boot). The authors’ care is also apparent in how they acknowledge the paper’s limitations. For example, the authors expressly acknowledge that the paper raises additional questions about the source of the proposed mechanism: ‘Do shareholders (or rather board members) explicitly think about the effect of competition when they set managerial incentives contracts? Are board members proposed (and elected) who tend to support more optimal incentive contracts if elected to the compensation committee? Do the contracts themselves make explicit mention of these considerations, or are the incentives only implicit in the payoff structure, as documented in the present paper?’
In short, this is an interesting, serious and ambitious paper which, as is often the case with economics papers, is well beyond my pay grade.