What drives closed‐end fund discounts? Evidence from COVID‐19

This empirical study investigates the rationale for the United States (US) closed-end equity fund discounts using investor sentiment approach of C. Lee, Shleifer, and Thaler (1991) for the period from 2004 to 2013. The result of this study suggests that discounts on closed-end equity funds decrease when small stocks return increase. The closed-end fund discounts have the significant stronger correlation with small capitalization as compared to large company’s stock returns. The results indicate that similar noise trading risk generated by retail traders explains the fluctuations in closed-end fund discounts and small capitalization equity returns even after controlling for fundamental factors. The results validated the existence of noise traders in market producing stochastic demand and supply based on their belief, subsequently affecting closed-end equity fund price in the market.

RICHARD SPURGIN is an assistant professor of finance at Clark University in Worcester, Massachusetts. R ecent studies on the performance of closed-end country fund performance have concentrated on analyzing the levels and changes in fund discounts, the sources of excess volatility of closed-end funds relative to net asset value, and the pricing of closed-end country funds, as well as the implications of the introduction of closedend funds on capital market integration Ž and market efficiency Lee et al. 1991 ; Bailey and Lim 1992 ; Johnson et al. 1993 ; Diwan et al. 1993 ; Hardouvelis . et al. 1995 ; Malkiel 1995 . Most of these analyses, however, concentrate on closedend country funds traded in the United States. Few studies exist on the relative performance of comparable closed-end country funds traded in the United States and other international capital markets Ž . e.g., United Kingdom . Differences in fund clientele and market structure may explain some of the variation in the price of the fund relative to the net asset value Ž . Ammer 1992 . In this article, a paired set of comparable closed-end country funds traded in the United States and the United Kingdom are analyzed for their relative fund performance as well as their relative discount or premium to net asset value Ž . NAV . The next section reviews previous evidence on the factors that determine fund return and the discount premium of closed-end funds. The data and methodology used in the study are presented in the third section. Shares in closed-end country funds are issued in one country, referred to herein as the host country, and the fund makes investments in another country, termed the local country. If investor sentiment plays a role in the value of the fund, then there are two possible sources of changes in sentiment. Investors could be responding to changes in the local country or responding to changes in the host country. As in previous studies, we compare closed-end country fund price and discount returns with an index from the Ž host country the United States and the . United Kingdom and also the local market index. Using funds that have the same local country but different host countries allows us to gain insight into the role of the host country’s index on the discount. These results are discussed in the fourth section. Results indicate that, although the discount return may be related to investor sentiment impacts, the actual process that drives the discount is complex. Results indicate that there is a higher correlation between the discount returns and the host country index than

In open‐end mutual funds (unit trusts), there is a strong positive cross‐sectional relationship between net inflows to individual funds and past performance, as if investors attributed performance to managerial skill. Performance shows only very weak persistence, however, so at first sight investors do not appear to gain anything by responding to past performance information. This behaviour can be explained by the fact that past performance is effectively unpriced in the unit trust market, since management fees are unresponsive to demand. If investors believe that there is a non‐zero probability that future performance will turn out to be positively correlated with past performance (i.e. that there is an element of managerial skill in performance), but a zero probability that this correlation will be negative, it is rational to prefer funds with better past performance when performance is not priced. In other words, it costs nothing to insure against the possibility of some managerial skill effect.If this explanation of the flow–performance relationship in unit trusts is correct, one would expect the relationship between investor demand and past fund performance to be much weaker if past performance were to be priced. We test this hypothesis in the market for closed‐end funds (investment trusts). Because closed‐end funds do not trade at net asset value, but at a price determined in the market, strong demand will raise the ratio of price to net asset value (known as the premium). Since it is well established that premiums are mean‐reverting, future shareholder returns on funds currently on high premiums tend to be depressed by the reversion of the premium to the mean. In the closed‐end fund market, as for open‐end funds, there is little evidence of performance persistence, and therefore, to the extent that funds with good past performance are pushed to higher premiums, the expected return on them is less than on the average fund. This implicit pricing mechanism should mean that demand is a declining function of the premium, so that, even if demand is an increasing function of past performance for a given premium, any effect on the premium itself will be muted.We test this hypothesis for closed‐end funds traded in the US and the UK. We find that there is a statistically significant effect of past performance on the premium in both countries. However, consistent with the hypothesis, it has limited economic significance, since it represents only a small component of premium variability. Copyright © 2008 John Wiley & Sons, Ltd.

Prior Performance and Closed-End Fund Discounts

Possibly the earliest discussion of investor sentiment, albeit unnamed, occurred during the run-up in U.S. stocks in the spring and summer of 1929 when closed-end funds mostly sold at substantial premiums to their net asset value (see De Long and Shliefer (1991)). These early writers were involved in brokerage or portfolio management, and they were warning investors that paying premiums of 100% or more was to pay twice or more the market value for the fund’s assets. The writers were puzzled by the demand for closed-end funds and were quick to point out the superior performance that fund managers would have to achieve in order to justify the lofty premiums.

Managerial Ability, Open-End Fund Flows, and Closed-End Fund Discounts

Empirical evidence from the UK market is examined in the light of recent theories about closed-end fund discounts. Market pricing of skill, relative to the fees charged for it, accounts for a significant portion of discount variation, but cannot explain the rarity of index funds or why they trade at a discount, since fees tend to be lower than on open-end funds. Index funds have lower discount volatility, consistent with the skill hypothesis. The results imply that managerial skill, relative to the fees charged, does not wholly account for the tendency of closed-end funds to trade at a discount.

Closed-End Funds: A Survey

Exchange‐traded funds (ETFs), like closed‐end funds (CEFs), are managed portfolios traded like individual stocks. We hypothesize that the introduction of an ETF in an asset class similar to an existing CEF results in a substitution effect that reduces the value of the CEF's shares relative to that of its underlying assets. Our event studies show that upon the introduction of a similar ETF, CEF discounts widen significantly and relative volume declines significantly. Single‐equation and systems estimation models show that the widening in discounts and reduction in volume are related to returns‐based measures of the substitutability of ETFs for CEFs.

Investment selection and individual portfolio management : Investor sentiment and the closed-end fund puzzle, by Charles M.C. Lee (University of Michigan), Andrei Shleifer (Harvard University), and Richard H. Thaler (Cornell University)

The Limits of Arbitrage and Stock Mispricing: Evidence from Decomposing the Market to Book Ratio

Investor Demand for Leverage: Evidence from Closed-End Funds

Cross-correlations between individual investor sentiment and Chinese stock market return: New perspective based on MF-DCCA

Closed-end funds are thought to have negligible fire sale risk as they have stable funding. However, I show that embedded covenants can generate price pressure in collateralized loan obligation (CLO) funds, even though such funds are closed end. Loans held by constrained CLOs report significantly lower cumulative returns than loans held by unconstrained CLOs. This can be explained by contractual arbitrage, a practice by which CLOs exploit loopholes in the design of covenants to mechanically loosen their covenants and avoid covenant breaches. Covenant breaches are associated with significant pecuniary and nonpecuniary costs, affecting CLO compensation, reputation, and career prospects. I show that when covenants breaches are imminent, managers fire sell distressed loans. Hence, I demonstrate a channel by which closed-end funds can also create fire sale risk, akin to their open-end counterparts. This paper was accepted by Lukas Schmid, finance. Funding: This research was funded by the John and Serena Liew Fellowship Fund at the Fama-Miller Center for Research in Finance at the University of Chicago Booth School of Business, the University of Chicago PhD Program Office, the Stigler Center for the Study of the Economy and State, and the University of Chicago Library. Supplemental Material: The data files and online appendices are available at https://doi.org/10.1287/mnsc.2023.4708 .

Mean-Reversion in Closed-End Fund Discount: Evidence from Half-Life