In the paper, Shareholder Votes and Proxy Advisors: Evidence from Say on Pay, which was recently made publicly available on SSRN, my co-authors (Yonca Ertimur of Duke University and David Oesch of the University of St. Gallen) and I examine the analyses underlying the voting recommendations issued by Institutional Shareholder Services (ISS) and Glass Lewis & Co. (GL), the two most influential proxy advisors, for the non-binding vote on executive pay mandated by the Dodd-Frank Act, also known as “say on pay” (SOP). We then investigate the effect of these recommendations on shareholder votes, stock prices and firm’s behavior. Due to the complex and highly firm-specific nature of executive compensation, mandatory SOP votes provide an especially powerful setting to examine the analyses performed by proxy advisors and their impact.
Our analysis of the SOP-related part of the ISS and GL proxy reports for S&P 1500 firms in 2011 shows that both advisors provide a quantitative and qualitative examination of the executive pay plan (structured around certain categories, e.g. pay for performance, disclosures), assign a rating for each category and issue a final voting recommendation (For or Against). ISS issues Against recommendations for 11.3% of the firms and GL for 21.7%, suggesting a more aggressive stance by GL. The difference also reflects the different approaches ISS and GL follow in assessing the “pay for performance” category, a key driver of the final recommendation, with ISS focusing its analysis of pay practices mostly on poorly performing firms. Firms receiving an Against from ISS are not a subset of those receiving an Against from GL. Rather, among firms with potentially questionable executive compensation practices (i.e. firms with an Against from at least one proxy advisor), ISS and GL agree on which firms warrant an Against only in 17.9% of the cases. We interpret this as evidence that the complex nature of SOP has allowed proxy advisors to differentiate themselves from each other. Additional analysis suggests that neither proxy advisor applies a “one-size-fits-all” approach in evaluating the compensation plans for the 2011 proxy season. Specifically, there are numerous cases where the proxy advisors identify similar controversial provisions yet issue different recommendations, based on firm-specific circumstances and other elements of the pay plan.
With respect to the SOP voting outcome, while compensation plans are voted down only at 2% of the sample firms, votes against the plan exceed 20%, a threshold viewed as an indication of substantial dissatisfaction, at more than 15% of the firms. Consistent with earlier work on shareholder voting, proxy advisors’ recommendations are the key determinant of voting outcome—a negative recommendation from ISS (GL) is associated with 24.7% (12.9%) more votes against the compensation plan. When both proxy advisors issue an Against recommendation, voting dissent is higher by 37.9%. The influence of each advisor declines only slightly when controlling for the recommendation issued by the other, suggesting that ISS’s and GL’s recommendations capture different factors and/or appeal to different sets of investors. While our estimate for ISS influence is in line with prior studies, the estimate for GL is higher. This suggests that mandatory SOP, by requiring costly analyses of thousands of different compensation plans, has caused more investors to rely on proxy advisors, thereby increasing their influence (as predicted by Gordon, 2009). However, it may have done so by spurring more competition, giving investors with different preferences greater choice in making voting decisions. Another key result that emerges from the analysis of SOP votes is that not all Against recommendations have the same effect on shareholder votes—the severity and type of concerns underlying the recommendation matter. That is, investors do not mechanically follow the advisors’ recommendations but seem to use the information in the reports to decide which recommendations to support.
We also examine the market reaction to the release of proxy advisors’ SOP recommendations, taking advantage of the fact the subjectivity involved in assessing complex executive compensation plans leads to greater uncertainty regarding proxy advisors’ recommendations and analyses for SOP proposals than for other items on the ballot. We document small but significantly negative mean abnormal returns (-0.5% to -0.7%) around the release of ISS reports with a SOP-related Against recommendation driven by the cases where the negative recommendation was less expected. As for firms’ responses, more than one third of the firms receiving a negative recommendation publicly question the proxy advisors’ methodologies, but this protest has no effect on the recommendation and the voting outcome. The few firms that change their compensation practices obtain a revision in the recommendation and avoid voting dissent.
Our study is the first to examine the analysis underlying proxy advisors’ recommendations, to document stock market and firms’ reaction to the release of proxy advisors’ “routine” reports, and to provide a direct estimate of influence of management recommendations on shareholder votes. As such, it provides an important contribution to the literature on shareholder voting, and in particular, on the role of proxy advisors, at a time when policy reforms are empowering shareholder votes. Our findings are also of interest to policy-makers that are considering whether to subject proxy advisors to greater regulatory oversight.
The full paper is available for download here.