The proxy advisory industry: Influencing and being influenced

Investors, except for the largest ones, are generally apathetic towards corporate governance, reluctant to spend resources to effectively monitor their portfolio firms (Berle and Means, 1932). This apathy paves the way for proxy advisory firms to bridge the gap by exploiting economies of scale in information collection. However, this seemingly efficient solution becomes problematic as the industry consolidates around two major players: Institutional Shareholder Services (ISS) and Glass Lewis. This consolidation raises concerns about the diversity and quality of their advice. The role of proxy advisory firms thus emerges as both a solution and a subject of scrutiny.

Unsurprisingly, the proxy advisor industry has attracted considerable attention from both academics and policymakers. A challenge in analyzing this industry stems from the opaque nature of the advisor-fund relationship. While you may know where the massive BlackRock obtained its proxy advisor (answer: both ISS and Glass Lewis), it is difficult to ascertain the source of proxy advice for every other fund. Without being able to identify each fund’s proxy advisor, it is challenging to draw definitive conclusions on many questions about the industry, ranging from basic issues such as the industry’s concentration to more textured inquiries concerning the objectives of proxy advisors.[1]

In a paper recently published in the Journal of Financial Economics, I provide two methods to infer the identity of each mutual fund’s proxy advisors through its regulatory filings.[2] Mutual funds are required to disclose their votes to the SEC by submitting Form N-PX. I demonstrate that, based on the formatting of the form and the descriptions of issues, one can determine which proxy advisor files the form, thereby providing a concrete link between funds and their proxy advisors. The inference of proxy advisors from N-PX forms can be further corroborated by mutual funds’ statutory prospectuses, which often disclose the identities of their advisors.

By linking investors to their proxy advisors, I offer a more comprehensive overview of the evolution of concentration in the proxy advisory industry during the last decade. As of 2021, ISS and Glass Lewis command significant portions of the market, with ISS holding 48 percent and Glass Lewis 42 percent, respectively, of the assets under advice. This concentration underscores the critics’ concerns about the market’s concentration. I also find that during the last decade, the market share of ISS has remained stable while that of Glass Lewis has increased.

The concentration itself does not necessarily indicate that proxy advisors wield undue influence. Studying the influence of proxy advisors is challenging because it is hard to discern whether the correlation between their recommendations and investor voting patterns is due to investors and proxy advisors perceiving the same merits in a proposal or because proxy advisor recommendations directly influence voting. By comparing the votes cast by subscribers of a specific proxy advisor with those of other investors on the same proposals, I address this omitted variable bias by effectively controlling for the merits of each proposal.

My results reveal a significant decrease in support for proposals opposed by a proxy advisor among its own clients, compared with the votes of other investors who do not subscribe to its proxy advice. For example, when ISS recommends voting against a director’s election, its customers are 20 percentage points more likely to vote against that director than investors who are not ISS subscribers. Similarly, when Glass Lewis recommends voting against a director, its customers are 13 percentage points more likely to oppose that director compared to other investors. I observe a similar pattern for say-on-pay proposals and other shareholder-sponsored ESG proposals.

The format of N-PX forms allows me to deduce the voting systems used by each fund, in addition to the source of its proxy advice. By examining the interplay between voting platforms and sources of proxy advice, my paper provides additional insights into why proxy advisors influence votes. For instance, I find that among investors who subscribe to both proxy advisors for voting advice, those using ISS’s voting system show a 13 percentage point higher agreement with ISS’s recommendations, and those using Glass Lewis’s voting platform tend to vote 19 percentage points more consistently with Glass Lewis’s advice. This finding suggests that the convenience provided by these platforms may contribute to the influence of proxy advisors. In extreme cases, investors vote in lockstep with their proxy advisors’ recommendations. I find that this practice, often referred to as robo-voting, is increasingly prevalent, especially among investors with smaller assets and those offering index funds.

Despite proxy advisors’ massive influence and investors’ reliance on automatic vote execution, my paper uncovers a silver lining: the adaptability of proxy advisor recommendations to evolving investor preferences. Anecdotally, through annual surveys and policy updates, proxy advisors exhibit responsiveness to investor feedback. My findings indicate that proxy advisors are more likely to change their recommendations on specific proposals (such as electing a director) and broad issues (such as climate proposals) when a significant number of investors disagree with their prior recommendations. This adaptability offers a glimmer of hope for a more nuanced and responsive proxy advisory industry.

Endnotes

1Numerous studies cited a statistic indicating that ISS and Glass Lewis together dominate approximately 97 percent of the proxy advice market. However, ISS challenges this 97 percent figure, stating that they have not verified it and cannot confirm its accuracy. See ISS’s letter to the Senate Banking Committee, https:// www.issgovernance.com/file/duediligence/20180530-iss-letter-to-senatebanking-committee-members.pdf.(go back)

2Shu, C. (2024). The proxy advisory industry: Influencing and being influenced. Journal of Financial Economics, 154, 103810. https://doi.org/10.1016/j.jfineco.2024.103810(go back)