Supreme Court Justice Louis Brandeis famously proclaimed in 1914 that “sunlight is the best of disinfectants.” Today, scholars, policymakers, and regulators commonly hold a related belief that mandatory disclosures can improve markets’ transparency and effectiveness. Typically, these thinkers believe disclosure-based regulation is superior to rules-based alternatives (e.g., Dalley 2007; Etzioni, 2010). The advantage of a disclosure-based approach over a rules-based approach to regulation is that it allows regulators to be more hands-off; rather than being prescriptive, it leaves it up to the recipient of the disclosures to interpret the information and propose desired changes. It thereby avoids ‘one-size-fits-all’ regulation and lowers the burden on regulators in their cost-benefit analyses needed for promulgating regulations.
Yet, a disclosure-based approach may also lead to the proliferation of voluminous disclosures that can hamstring investors and asset managers, who have a limited capacity to process all firm disclosures. These voluminous disclosures may then heighten the role of financial intermediaries such as proxy advisers, who may step into the regulatory void to provide closer guidance on desired behavior. In doing so, proxy advisors may establish their own one-size-fits-all approaches, which, unlike regulatory rules, do not undergo a formal cost-benefit review. A disclosure-based approach can thus result in shareholder governance that is based on proxy advisors’ voting recommendations, which are in turn based on their views and assessments of governance policies. Additionally, mandatory disclosures may lead to externalities in product markets as it may undermine the competitive position of the announcing firm.
Much research in finance and accounting is concerned with the direct effects of disclosure regulation on firm outcomes. In contrast, considerably less attention is paid to regulatory externalities, or spillover effects, of disclosure rules on other participants in financial markets. This comes as no surprise: it is already difficult to provide causal evidence of a new policy on firm outcomes; spillovers, or peer effects, are yet another step removed and are substantially harder to pin down causally (e.g., Manski, 1993). This article describes three examples of such governance externalities based on the authors’ recent work about disclosure regulation in the corporate governance field.
Network effects from compensation disclosure rules
In ongoing work titled “Why Have CEO Pay Levels Become Less Diverse?”, we document a precipitous decline in the variation of CEO pay levels among publicly listed U.S. firms over the past decade. This decline is found within industries, industry-size groups and across the entire economy. This convergence is disconcerting, as prior literature has documented the importance of pay variation among executives within and across firms – so-called tournament incentives – for encouraging risk-taking and firm performance (e.g., Coles, Li and Wang 2018). In our paper, we find strong evidence consistent with one explanation for this decline in pay variation: a network effect that we refer to as “reciprocal benchmarking.” Specifically, it deals with the mutual consideration of executive pay packages paid by similar firms that started after a 2007 SEC disclosure rule on executive compensation policies.
To understand this development, we first need to recognize how public firms and their boards arrive at pay policies for their executives. Aided by compensation consultants, boards typically benchmark their own executives’ pay to the disclosed pay of a set of comparable firms with whom they compete for executive talent. This makes sense: if one persistently underpays executives, they may depart for better outside opportunities. Once a firm has settled on the set of compensation peers, it then targets the level of its own pay at some percentile – typically around the 50th to 60th – of said peer group (e.g., Murphy and Sandino, 2020).
While the pay of compensation peers justifies the level of executive pay to investors, the specific set of peers that boards select long remained a well-guarded secret. As a result, investors worried about powerful CEOs inflating their own pay by surreptitiously adding much larger and higher-paying ‘aspirational’ competitors to this set of peers (Faulkender and Yang, 2013). To address these concerns, the SEC released new disclosure rules that took effect in 2007, mandating that firms publicly disclose their compensation peer groups (17 CFR § 229.402 section l). Confronted with this new information, yet unsure which compensation peers should qualify as a firm’s legitimate competitor in executive labor markets, investors turned to proxy advisers for guidance, who filled the void by advocating peer selection based on industry affiliation and firm size (e.g., ISS 2012).
In our paper, we show that since the SEC mandated disclosure of compensation peers, firms have been converging on common criteria for peer selection based on industry and size. One consequence of such convergence is the increase in “reciprocal benchmarking,” wherein similar firms reference each other in their compensation peer groups, leading to the emergence of ‘pay clusters.’ What has gone unnoticed, however, is that this reciprocal benchmarking reduces pay variation across firms. Intuitively, if all mining companies select similar-sized mining companies as their compensation peers, and they then target the pay at the median pay level of this peer group, then all mining companies converge over time to the same level of pay.1 In reality, the overlap in compensation peers across firms within a specific industry is less than complete and not all firms target their CEO pay to the median of their peer group. Notwithstanding, we find strong evidence of a decline in pay variation within industries (as well as within industry-size groups) as a consequence of the increase in the reciprocal benchmarking among firms.
We find further evidence that the move towards reciprocal benchmarking is strengthened by two parallel developments: i) the growth in passive ownership as hands-off institutional investors routinely follow proxy advisers who advocate universal industry-size selection criteria, and ii) the introduction of mandatory “say on pay” voting in the U.S. that provides investors with an explicit vote on firms’ compensation policies.
Investigating the consequences of pay convergence among similar firms, we find weakening external tournament incentives that lead to reductions of corporate risk-taking and declines in stock performance. Furthermore, we find increasing stock price synchronicity among firms in the same industry, the extent of which is proportionate to the rise in reciprocal benchmarking and the formation of pay clusters in industries’ respective compensation peer networks.
Peer effects from disclosures on “say on pay” voting
In a second paper, joint with Diane K. Denis, titled “Shareholder Governance and CEO Compensation: The Peer Effects of Say on Pay”, we document that low shareholder support in advisory votes on executive compensation packages in public U.S. firms (the so-called “say on pay” vote) have wide-ranging consequences in public markets via governance spillovers.
Starting January 2011, the SEC mandated recurring say on pay shareholder votes for publicly listed U.S. firms (17 CFR § 240.14a-21). Previously, discontent about executives’ pay was typically voiced in non-public talks between institutional investors and management. In most cases, these remained confidential. This changed in 2011, when the public disclosure of vote outcomes of say on pay votes allowed market participations to observe the level of shareholder discontent at any public firm.
In our paper, we document proactive changes in executive pay policies at firms (henceforth “focal firms”) whose compensation peers experience weak support on their say on pay votes. This makes sense as focal firms learn from their compensation peers about the appropriate level and structure of executive pay. A signal that shareholders oppose the executive pay of a compensation peer should thus carry value for the focal firm too. We find empirical support for two channels that lead to those spillover effects: i) deliberate board learning from the disclosure of their peers’ say on pay votes and ii) pay spillovers due to the targeting of executive pay at the median of one’s peers (which exhibit pay reductions after receiving weak say on pay vote support).
Supporting the board learning channel, we find evidence that focal firms’ pay changes reflect the concerns of proxy advisers about their peers’ compensation contracts. In addition, focal firms reduce executive pay more when their weak-vote peers exhibit above-median performance. Hence, focal firms reduce pay when peers’ weak say on pay votes are more likely to reflect shareholder discontent regarding pay design rather than discontent with peers’ general firm performance. Pay spillovers are also larger when primary firms compete more closely in the executive labor market with their weak-vote peers (i.e. when peer pay is most likely to be a relevant benchmark). Conversely, spillovers are lessened when executives have more outside options. These results suggest that the compensation committees of focal firms deliberately weigh the information they receive from the disclosure of shareholder discontent at their compensation peers.
Supporting a second channel for those spillover effects – the compensation targeting channel – we find evidence that focal firms’ pay reductions are proportional to those needed to retain their relative positions in their respective peer groups. Furthermore, they make pay changes only after their peer’s revised pay becomes public. In summary, this suggests that focal firms’ compensation changes are responses to peers’ revised pay, as well as to the information contained in peers’ weak votes.
Shareholder support in say on pay votes overall is quite strong: the median vote support since 2011 is 95 percent. Thus, if weak say on pay votes affect compensation only for the relatively small set of firms that experience low vote outcomes, the overall effect of say on pay is arguably limited. However, our findings imply that – due to these previously unobserved spillovers – say on pay regulation in conjunction with the public disclosure of shareholder discontent on executive pay policies affect pay policies in a much broader set of firms than previously thought.
Disclosure regulation and competitive interactions
Understanding the market-wide effects of firms’ disclosures is crucial to regulating the production of corporate information. A common concern regarding disclosure regulation is that peer firms’ reactions to mandated disclosures undermine the competitive position of the announcing firm. Yet there is little evidence on whether firms’ mandated disclosures impose significant proprietary costs on the announcing firm by affecting competitive interactions.
In a third paper, joint with Marc Badia, Miguel Duro, and Bjorn Jorgensen, titled “Disclosure regulation and competitive interactions”, we study the effects of mandatory disclosure on competitive interactions in the setting of oil and gas reserve disclosures by North American public firms.
The first set of analyses focuses on whether competitors react to reserve disclosures. Consistent with the disclosures conveying information about the competitive position of peers, we find that firms experience lower stock returns when their competitors report increases in reserves. Consistent with firms’ disclosures eliciting peer reactions, we observe that firms undertake higher capital expenditures when competitors disclose larger increases.
We also explore more directly whether competitors’ reactions to reserve disclosures impose significant proprietary costs on the announcing firm. We document that the higher a competitor’s investment relative to the announcing firm’s investment, the higher the competitor’s relative year-end reserves and sales. These results are driven by peers’ investment associated with announcing firms’ reserve disclosures.
To sharpen identification, we analyze several sources of cross-sectional variation in these patterns, the degree of competition, and the sign and the source of reserves changes. We also exploit two plausibly exogenous shocks: the tightening of the reserve disclosure rules and the introduction of fracking technology, which increased recoverable reserves from firms’ existing leases. Additional tests more directly focused on the presence of proprietary costs confirm that the mandated reserve disclosures result in a relative loss of competitive edge for announcing firms. Our collective evidence highlights important trade-offs in the market-wide effects of disclosure regulation.
These results suggest that the mandate on firms to disclose their reserve information imposes proprietary costs on North American firms by shaping competitive interactions, an interpretation consistent with criticisms of the recent regulatory changes tightening the disclosure requirements. However, our evidence need not imply that the regulation was inefficient, as proprietary costs could be outweighed by potential benefits of the regulation, such as higher market liquidity and lower cost of capital.
Taken together, the above research highlights considerable externalities from disclosure policies on executives’ incentives and firms’ competitive interactions. While Justice Brandeis may have been right about transparency’s power to expose and mitigate conflicts of interests, regulators are well-advised to consider potential spillover effects that may end up shaping underlying market governance mechanisms.
Torsten Jochem is an Associate Professor of Finance at the University of Amsterdam.
Gaizka Ormazabal is an Associate Professor in the Accounting and Control Department of IESE Business School.
Anjana Rajamani is an Associate Professor of Finance at Rotterdam School of Management, Erasmus University.