Dilip Patro, Louis R. Piccotti and Yangru Wu authored a working paper titled “Exploiting closed-end fund discounts: The market may be much more inefficient than you thought.” (although Ben Graham style value investors likely will not be too surprised). They concluded that closed end fund premiums or discounts relative to net asset value (NAV) exhibit mean reversion. They also determined that an arbitrage strategy using closed end funds was profitable over time. Such strategy consisted of purchasing the quintile of closed end funds with the highest estimated returns and selling the quintile of closed end funds with the lowest estimated returns.

Closed End Fund
Closed end fund expected returns
Source: Exploiting closed-end fund discounts: The market may be much more inefficient than you thought; July 18, 2014

Correlation of closed end fund market and the share price

If the closed end fund market was efficient, the share price will be equal to NAV. However, share prices for closed end funds differ from NAVs exhibiting a discount often. Patro and his team also found that the closed end fund market did not incorporate prior information to set prices. There was no statistically significant difference in mean returns in the first time period half of the sample relative to the second half.

Prior research covering closed end funds with a price discount outperform the market, according to Patro, Piccotti and Wu. They also postulated that closed end funds trading at a NAV premium produce negative returns relative to market.

Closed End Fund AUM sample size

Outperformance of discounted closed end funds cannot be explained by beta

Patro and his team used closed end funds that existed for ten years through 2011. They listed categories from Morningstar that included fixed and equity in both domestic and international markets. Mean reversion of spread relative to NAV speeds differed by fund type. Fixed income closed end funds reverted faster than equity closed end funds. Within equity, international closed end funds reverted to their mean spread faster than their domestic peers.

Closed End Fund Templeton Emerging market fund

Outperformance of Patro’s arbitrage strategy of buying the quintile of closed end funds with the largest estimated return and selling the quintile of closed end funds with the smallest estimated return cannot be explained by market risk. Patro’s long short strategy generates an annualized return of 18.2% and a Sharpe ratio of 1.918. The alpha of the strategy is 17.4%. Calculating returns within fixed or equity oriented closed end funds showed results are robust and not driven by closed end fund investment objective differences. However, the alpha for fixed closed end funds is larger than the corresponding measure for equity closed end funds. International closed end funds also have higher alpha relative to their domestic counterparts.

Closed End Fund
RADF model, or Patro’s long short model, incorporates information provided by spreads between NAV and closed end fund price
Source: Exploiting closed-end fund discounts: The market may be much more inefficient than you thought; July 18, 2014