In Limits of Disclosure recently posted to the SSRN we examine the shortcomings of disclosure. Claire Hill and I do so by exploring two areas where disclosure arguably failed, albeit for very different reasons: synthetic collateralized debt obligations (CDOs), such as Abacus and Timberwolf, sold in the years immediately leading up to the financial crisis, and executive compensation.
One big focus of attention, criticism, and proposals for reform in the aftermath of the 2008 financial crisis has been securities disclosure. Many commentators have emphasized the complexity of the securities being sold; some have noted that disclosures were sometimes false or incomplete. What follows, to some commentators, is that whatever other lessons we may learn from the crisis, we need to improve disclosure. How should it be improved? Commentators often lament the frailties of human understanding, notably including those of everyday (retail) investors: people do not understand or even read disclosure. This leads, naturally and unsurprisingly, to prescriptions for yet more disclosure, simpler disclosure, and financial literacy education.
That disclosure may not work as expected for the average investor is not a new notion. But the securities at issue in the crisis were generally not sold to retail investors, who might be expected to not read or not understand disclosures. Rather, they were mostly sold to sophisticated institutions. Securities laws rely on the assumption that sophisticated investors read and understand securities disclosures. If they do not understand, as some commentators have suggested was the case, they should know that they do not; they should not buy until they conclude that they know enough to do so.
We trace through the disclosure in CDO offering memorandums. While the disclosure available to investors was not perfect, it raised several red flags. The disclosure should have made them warier than it did or otherwise spurred additional investigation. The investors had ample opportunity to engage with the sellers of the securities, doing due diligence until they were satisfied that they knew enough to make sensible investment decisions. But these extremely sophisticated investors appear to have done only cursory or inadequate due diligence in many instances.
We use the case of CDO disclosure to observe that the role of disclosure in investment decisions is far more limited, and far less straightforward, than is typically assumed. Many investors, even sophisticated investors, do not start their consideration of an investment decision with cautious or neutral presumptions about a security and then, after carefully reading the disclosure and appraising it on its merits, make the decision to invest or not. As has been extensively discussed in the literature, investors may be eager to buy “the hot new thing” that their peers are buying. But our aim is not to explain what motivated investor behavior; our aim is to point out what disclosure did not sufficiently motivate investor behavior.
Our argument is not just about the present crisis. Indeed, the complex role that disclosure plays in an investor’s decision as to whether to buy a security is just one example of disclosure’s limits. Those limits reflect the complexity of human decision-making. Why should disclosure work? The obvious answers are that better information should make for better decisions, and that the specter of disclosure should constrain behavior. There are many things people might do if they think they will not be found out. But these answers are importantly incomplete. Better information should, in principle, lead to better decisions, but other factors may be far more important. Such was the case with disclosure regarding the CDOs at issue in the financial crisis.
Another example of disclosure’s limits is in the case of executive compensation disclosures. Executive compensation is high by many metrics. Those paying the compensation are company shareholders. Many argue that providing them with more detailed information would cause them to seek to curtail it in some way, perhaps by pressuring their companies or selling their stakes in companies with high compensation. Companies, anticipating shareholder reaction, would curtail their compensation pre-emptively. That was the theory behind more expansive executive compensation disclosures, but it was apparently not the reality: while measurement difficulties preclude a definitive assessment, it appears that compensation may have gone up rather than down. Our explanation is that the incremental information was insufficient to prompt shareholder action, but was sufficient to prompt action by peer CEOs to put pressure on their boards to raise their pay.
The strong allure of the disclosure solution is unfortunate, although perhaps unavoidable. Solutions are hard to come by, and as hard, if not harder, to agree upon. A solution emphasizing disclosure can give the appearance of “doing something” when nobody can agree on anything else. It is not just policymakers who benefit from a disclosure “solution.” The emphasis on disclosure affirms a too-comforting worldview. People losing money on their investments can tell themselves and others that if they had been told enough, they would not have bought the securities. Market participants generally can tell themselves and others that their purported modus operandi, reading disclosure documents and making investment decisions based thereupon, is viable.
The admittedly nebulous bottom line is this: disclosure is too often a convenient path for policymakers and many others looking to take action and hold onto comforting beliefs in the face of a bad outcome. Disclosure’s limits reveal yet again the need for a nuanced view of human nature that can better inform policy decisions.
The full article is available for download here.