Hedge Funds Voluntary Disclosure
INSEAD
University of Chicago – Booth School of Business
Jeremiah Green
Pennsylvania State University
John R. M. Hand
University of North Carolina Kenan-Flagler Business School
Matthew Neal
affiliation not provided to SSRN (deceased)
Abstract:
We study the determinants of voluntary financial disclosures using a proprietary database of 4,632 letters sent by 435 hedge funds to their investors. We find support for some but not all theories of voluntary disclosure. Consistent with there being proprietary costs to disclosure we observe that better performing funds disclose less about their investment portfolio. In line with agency costs leading to self-serving behavior by hedge fund managers, we find that riskier funds disclose less about fund risk, and that good fund performance is more likely to be attributed to manager skill while bad fund performance is blamed on external factors. And in agreement with precedent setting theory, we show that outside of textual descriptions about fund performance, disclosure choices are stable over time. Our detailed dataset also enables us to determine whether using a hedge fund’s choice to report its returns to a commercial database is a reliable proxy for the level of voluntary disclosure a hedge fund makes to its investors. We find that it is not because the choice to report fund returns to a commercial database is both positively and negatively related to the disclosures made in investor letters, with the sign of the relation depending on the category of information that is disclosed.
Hedge Funds Voluntary Disclosure – Introduction
In this paper, we study the determinants of voluntary financial disclosure using a handcollected, proprietary dataset of 4,632 letters sent by 435 hedge funds to their investors over the period 1996–2011. Our setting differs from others in the voluntary disclosure literature—such as management forecasts, press releases, and discussions in 10-K filings – because hedge funds are exempt from the securities regulations that mandate that investment vehicles make disclosures about their performance, operations and risk to their investors and/or the public (Oesterle, 2006).
As such, the disclosures that hedge funds make to their investors are entirely voluntary.1 They are also rich, spanning a wide array of information about current and past return performance, fund risk and investment positions, along with discussions of fund performance and the investment environment.
A key feature of our research setting is that the economics of the hedge fund industry serve to amplify the degree of information asymmetries between managers and investors, and the proprietary costs of making disclosures. Because hedge funds face no requirement to ?level the playing field? and are not subject to Regulation FD, they can engage in selective disclosure with current and prospective investors, opening the door to acute information asymmetries (Cassar and Gerakos, 2010). The proprietary costs of making disclosures are high because hedge fund managers and current investors can incur losses if competitors learn the fund‘s investment strategy, performance and positions, since such data would increase the ability of competitors to mimic the fund‘s strategy and/or trade against the fund (Aragon, Hertzel and Shi, 2013; Agarwal, Jiang, Tang and Yang, 2013).
Using our large dataset of investor letters, we develop and test the predictions made in our setting by multiple theories of voluntary financial disclosures. Consistent with proprietary cost theory we find that better performing funds disclose more about their performance but less about their investment portfolio. In line with agency costs leading to self-serving behavior by hedge fund managers, we observe that riskier funds disclose less about fund risk, and that good fund performance is more likely to be attributed to manager skill while bad fund performance is blamed on external factors. In agreement with precedent setting theory, we show that the vast majority of disclosure choices are stable over time. We do not, however, find evidence supporting the hypothesis from agency cost theory that hedge funds engage in voluntary disclosure in order to reduce agency costs by providing investors with better monitoring or by signaling fund quality.
Accessing and coding into machine readable form the detailed information that hedge funds actually disclose to their investors is time consuming and costly because the data is not publicly available. As such, prior hedge fund research typically uses whether a hedge fund chooses to self-report its returns to one or more commercially available databases such as HFR and Lipper-TASS as a proxy for the level of voluntary disclosure the hedge fund makes to its investors (Aiken, Clifford and Ellis, 2013; Agarwal, Fos and Jiang, 2013). Noting that returns are just one of the 30 disclosure items contained in our database, we therefore examine the validity of the conventional commercial database proxy by correlating it with the total information that we code as being disclosed in a fund‘s investor letters, and with the information contained in each of five economic subcategories.
See full PDF below.
