Securities lending has proven the most challenging aspect of shadow banking for regulators to bring under a regulatory rubric. One of the most vexing aspects for regulators has to be how to make securities lenders' decision processes about whether to recall lent securities to vote proxies more transparent to investors and the regulators themselves. The calculus about whether to forego lending income in favor of exercising the right to vote a proxy relies on such myriad factors that regulators are hesitant to second-guess. Nonetheless, investors in lenders like mutual funds have a right to know how investment managers make these decisions. And now, the enormous demand for ESG investing is driving an intensifying interest in how funds are managing governance decisions.[1]
In its latest effort to shed light on fund proxy voting decisions, The Securities and Exchange Commission is expanding the scope of Form N-PX, a disclosure form it now finds never met its potential.[2] But will beefing up N-PX be meaningful to ESG investors, and will it provide any significant evidence about potential conflicts of interest? Or is just more disclosure without context or analysis? Perhaps there is another way to meet these needs, based on technologies not available when N-PX was dreamed up?
A Brief History of Form N-PX
Over the past two decades, the Securities and Exchange Commission has tried to shed light on the proxy voting behavior of regulated investment funds. In 2003, the SEC approved rules requiring mutual funds to report their proxy voting data annually on Form N-PX.[2] Currently, Form N-PX requires disclosure of proxy voting information "for each matter relating to a portfolio security considered at any shareholder meeting held during the period covered by the report and for which the registrant was entitled to vote." However, the Form does not require disclosure of the number of shares for which proxies were voted, nor does it require disclosure of portfolio securities on loan when, as is generally the case, the fund is not entitled to vote proxies relating to those securities.[3]
Based on requests from investors and fund advisers alike, the Commission included potential changes to Form N-PX in its July 2010 Proxy Concept Release soliciting comment on expanding the data funds must provide to include the number of shares voted and data on the number of shares out on loan. Also, in November of 2010, the Commission proposed amendments to Form N-PX under Section 951 of the Dodd-Frank Wall Street Reform and Consumer Protection Act requiring certain institutional investment managers to report how they voted on executive compensation matters. However, the concept release resulted in no changes to the content of Form N-PX, and the SEC never finalized the rule proposal.
Incomplete Proxy Data Leads Some to Mistaken Conclusions
"Don't pay attention to critics. Don't even ignore them" -- Samuel Goldwyn
Since the advent of Form N-PX proxy disclosure, academics and commentators (most with incomplete or faulty data) have scrutinized mutual fund proxy voting practices, turning up the pressure for the SEC to act. For example, in 2006, a team of academic researchers claimed to find evidence of "vote buying" and manipulation of corporate governance in the U.S. equities securities lending markets. Their studies claimed that spikes in equities lending activity on proxy record dates proved unequivocally that abuse by share borrowers was "widespread" and, further, that control of voting rights could be acquired at no cost." This led corporate advocacy groups to call on regulators to force the disclosure of "hidden ownership" by activist hedge funds.
A study conducted by the RMA in conjunction with CSFME, however, concluded the data supporting the claims of borrowed proxy abuse was likely incomplete, and failed to account for some key aspects of how the securities lending market and proxy process work (like substitution effects, and proxy allocation controls at the broker level). Lacking proper data, and without considering structural and operational controls currently in place, RMA and CSFME argued that there was no basis on which to conclude from loan volumes around record dates that widespread and abusive empty voting was in fact taking place. In addition, RMA and CSFME called into question the assertion that votes can be acquired for virtually no cost.
More recently, academic researchers published papers concluding that passive index funds may operate to the detriment of corporate accountability – and on ESG matters, in particular, pointing the finger at the SEC's rules which have not kept up with these new investor demands.[5] Others claimed to have data-based findings that, faced with economic pressures to lend out their shares or not recall shares, instead of voting, index funds were taking advantage of loose SEC guidance to vote at all hardly. If true, this would potentially violate the duties of these funds to vote proxies and certainly frustrate the ESG desires of shareholders.[6]
In Response, the SEC is Beefing up Form N-PX
The SEC has turned back to Form N-PX in response to this intense scrutiny of securities lending versus proxy voting. Acknowledging that the current proxy reporting form is not meeting the needs of investors, the SEC proposed to amend the Form to specifically require funds and managers to disclose how their securities lending activity affected their proxy voting.[7] This 2021 proposal expands the disclosures on Form N-PX well beyond what was considered in the never-finalized 2010 proposal. These statistics are intended to provide transparency about whether a fund chose to recall a security and vote the accompanying proxy or keep the security out on loan in favor of the loan income.
What investors can do with the new information and whether it creates the kind of transparency these decisions require remains to be seen. Will this expanded Form N-PX be just more out-of-context disclosure and meet the same fate as version one from 2003?
New Technology can Provide Better Disclosure
Rather than static disclosures on NP-X, investors and regulators need active metrics to monitor funds' proxy voting and identify potential conflicts of interest. In the context of proxy voting versus securities lending decisions, what would be more meaningful is an analysis of the timing of blocking or recall decisions against the significance of the proxy issue and how those correlated against moves in the rebate rate (i.e., the potential profit from the loan) for the security to be voted. Such an analysis was not easily possible back in 2003. But in 2021 with data liberated from isolated silos by cloud computing, distributed ledgers, and blockchains, these kinds of statistical analyses become readily possible given the application of the right technologies.
Perhaps it is in the fund industry's best interest to look into this kind of dynamic analysis to avoid what could be more needless and ultimately unhelpful disclosure to shareholders.
[1] According to one recent Morgan Stanley survey, 95% of millennials and 85% of all investors are now interested in sustainable investing strategies.
[2] The stated goals of Form N-PX were to:
- "enable fund shareholders to monitor their funds' involvement in the governance activities of portfolio companies."
- "illuminate potential conflicts of interest."
- enable shareholders to "evaluate how closely fund managers follow their state proxy voting policies."
[3] The Form arose from a 2002 proposal to require mutual funds to disclose proxy voting policies in their SAI's and their proxy voting records in their annual and semi-annual reports. However, in the final release, the SEC chose to create a new form, N-PX, and a standardized non-calendar annual cycle for proxy vote disclosure rather than in the fund's periodic reports (N-CSR). The 2002 proposal drew 8,000 public comments, a record at the time for any such rulemaking. The N-PX reporting period runs annually from July 1 to June 30 and is still filed on EDGAR in the same unstructured format as in 2003.
[4] Thus, for example, if a fund lends out 99% of its portfolio holdings of XYZ Corporation and therefore votes only 1% of its holdings of XYZ, Form N-PX would disclose that the fund voted proxies for shares of XYZ, but would not also disclose that the fund did not vote 99% of its holdings of XYZ because they were on loan or in a fail-to-receive status.
[5] See Index Funds and the Future of Corporate Governance: Theory, Evidence, and Policy, Lucian Bebchuk and Scott Hirst (Dec. 2019), available at https://columbialawreview.org/content/index-funds-and-the-future-of-corporate-governance-theory-evidence-and-policy/ (finding that an agency-cost analysis shows that index fund managers have a strong incentive to underinvest in stewardship and defer to corporate managers); see also The Agency Costs of Agency Capitalism, Ronald Gilson and Jeffrey Gordon (May 2013), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2206391; see also The Future of Corporate Governance Part I: The Problem of Twelve, John Coates (Sept. 20, 2018), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3247337.
[6] Hu, Edwin and Mitts, Joshua and Sylvester, Haley, The Index-Fund Dilemma: An Empirical Study of the Lending-Voting Tradeoff (December 22, 2020). NYU Law and Economics Research Paper No. 20-52 , Columbia Law and Economics Working Paper No. 647, Available at SSRN: https://ssrn.com/abstract=3673531 or http://dx.doi.org/10.2139/ssrn.3673531
[7] This includes:
- The number of shares that were voted (or, if not known, the number of shares of which votes were instructed to be cast); and
- The number of shares that the fund did not recall.