Print
E-MailYouchang Wu is Associate Professor of Finance at University of Oregon Lundquist College of Business. This post is based on a forthcoming article by Professor Wu; Russ Wermers, Professor of Finance at University of Maryland Robert H. Smith School of Business; and Josef Zechner, Professor of Finance at Vienna University of Economics and Business.
There are two different organizational forms of mutual funds: closed-ended and open-ended. Closed-end funds (CEFs) issue a fixed number of shares that trade on secondary markets, while open-end funds (OEFs) issue and buy back shares at a price equal to the underlying net asset value on a daily basis. When the Investment Company Act, an overarching law governing the mutual fund industry, was enacted in 1940, the dominant form of mutual funds in the U.S. was the CEF. However, by the end of 2014, the total value of net assets managed by U.S. CEFs amounted to less than 2% of the value of net assets managed by OEFs ($289 billion versus $15.9 trillion). The predominance of the open-end structure is puzzling because, from a portfolio management perspective, the closed-end structure has many advantages. For example, it allows fund managers to take illiquid positions without worrying about money flows into or out of their funds.
Another long-standing puzzle is that shares issued by a CEF typically trade at a discount (or a negative premium) relative to the value of the underlying portfolio held by the fund. While such a discount can potentially be rationalized by investor expectation of poor fund performance, past studies generally do not find a significant relation between the discount and the performance of the fund portfolio (i.e., the NAV performance).
In our article forthcoming at The Review of Financial Studies, Managerial Rents vs. Shareholder Value in Delegated Portfolio Management: The Case of Closed-End Funds, we tackle these two closely-related puzzles by examining the relative control of fund managers versus shareholders over potential rents generated by CEFs. We focus on the dynamics of the total assets under management (AUM) and management fees at the portfolio manager level. Previous studies have focused on the individual fund level. These two perspectives can be very different, because a manager often manages multiple funds, or switches from one fund to another, and a fund often has multiple managers.
Specifically, we investigate how manager-specific AUM and fees are adjusted in response to past NAV performance, and to investor perceptions about future performance reflected in fund premiums. An expansion of AUM increases the fee income collected by fund managers, but is likely to erode fund performance due to diseconomies of scale. Therefore, like a fee increase, it can represent a wealth transfer from investors to fund managers. We further investigate the relation of manager tenure with NAV performance and fund premiums, the time-series properties of NAV performance and premiums, and the response of fund premiums to past NAV performance.
Using a comprehensive sample of 1,137 managers managing 679 CEFs over 1985 to 2010, we test two competing hypotheses. The “Shareholder Surplus Hypothesis” postulates that market frictions, such as the cost of launching a new fund, allow shareholders to extract rents generated by skilled managers, and that effective governance prevents unskilled managers from destroying shareholder value. The “Managerial Rent Hypothesis’’ postulates that market power allows skilled managers to capture the rents they generate, and that unskilled managers are entrenched due to weak governance.
Our results strongly support the Managerial Rent Hypothesis. Importantly, we find that managers perceived to be skilled, and therefore in high demand (for rational or irrational reasons), as proxied by a high fund premium, subsequently capture rents through expansions of AUM (usually achieved by taking on additional CEFs) and increases in management fees. However, managers generating a high discount are not penalized accordingly. The asymmetric response of the AUM of CEF managers to fluctuations in investor demand for CEF shares is fundamentally different from the situation in the OEF industry. For an OEF, the relation between investor demand and fund growth is inherently symmetric, because demand translates automatically into money flows into or out of the OEF. An OEF manager cannot exploit the benefits of a strong demand without bearing the consequence of a weak one.
The response of AUM and management fees to past NAV performance also supports the Managerial Rent Hypothesis. Good NAV performance increases the probabilities of AUM expansions and management fee increases, but poor NAV performance does not increase the probability of AUM contractions or fee decreases. While poorly-performing CEF managers are more likely to leave the industry, this disciplining effect is significantly weaker for managers with long tenure. Correspondingly, for managers whose tenure is above average, tenure negatively predicts NAV performance and fund premiums. Furthermore, NAV performance and fund premiums are more persistent when they are below average than when they are above.
We also find interesting relations between NAV performance and fund premiums at the portfolio manager level. Consistent with the conjecture that the relation between fund premiums and past performance is contaminated by anticipated endogenous AUM and fee adjustments in response to extreme past performance, we find that fund premiums respond positively to past NAV performance only when the performance is not extreme. Furthermore, while premiums do not predict the NAV performance of short-tenure managers, they have statistically stronger predictive power for the NAV performance of long-tenure managers, suggesting that investors are able to form better forecasts of performance for managers with long track records.
Finally, we find strong evidence of decreasing returns to scale in CEF management. Both NAV performance and premiums are negatively related to the size of a manager’s AUM. In addition, for managers who oversee both CEFs and OEFs, the size of their OEF AUM negatively predicts their CEF performance, even after controlling for the size of their CEF AUM. These results suggest that managerial skills are subject to scale diseconomies, and support the idea that asset expansions represent a transfer of rents from shareholders to managers.
Together, our results suggest that CEF shareholders have little capacity to extract rents from skilled managers, or to discipline unskilled ones. This may be an important reason why CEFs are less popular than OEFs, and why CEFs usually trade at a discount.
The full article is available for download here.
