What causes closed-end fund discounts

Several recent papers suggest that closed-end fund discounts are caused by noise traders and that an annual abnormal return in excess of 10% could have been earned using the discount information. This dissertation presents new empirical evidence on these discounts and proposes a new theoretical explanation for the discount puzzle.; For closed-end funds that endured at least ten years in the period 1965-1990, I find that fund discounts are non-stationary. This finding contradicts previous studies which suggest that discounts are stationary and caused by mean reverting investor sentiment. Systematic risks of fund shares are found not to be significantly larger than those of fund net asset values, indicating that comovements among discounts cannot be regarded as systematic investor sentiment. Abnormal returns from the discount information are not significant when a reasonable 2 percent transaction cost is imposed. These results indicate that a noise trader model cannot explain the puzzle and that inefficiency in the closed-end fund market is exaggerated.; I argue that the absence of redemption, which is a fundamental characteristic of closed-end funds, causes the observed discounts. When the risk-return efficient frontier is unknown and a small number of "bad" funds exist, non-redemption causes discounts because investors demand discounts for compensation when the performance of a fund is inefficient and for hedging against possible future inefficient states. I demonstrate that hedging is the main function of discounts. When investors have to estimate the probability that inefficient states occur, noise is generated by the estimation process. This noise is independent of the number of noise traders in the market and it enters the pricing function of closed-end funds. Thus, closed-end funds do suffer from additional noise, but not for the reasons given by the noise trader model.; In conclusion, although there are large observed discounts for closed-end funds, this fact should not be taken to imply that the market is inefficient. Future research should focus on the fundamental non-redemption feature of closed-end funds to completely resolve the discount puzzle.