An Agency Explanation of the Closed-End Fund Puzzles

Closed-end funds have been a topic of interest among academics and professionals on five counts. First, they tend to sell above their net asset value at their commencement. Second, they start selling at a discount within a year. Third, this discount is volatile but mean-reverting. Fourth, the news of open-ending a fund reduces the discount although some discount persists till the actual event of open-ending. Fifth, the volatility of the return on a fund’s price is higher than the volatility of the return on its underlying net asset value. In this paper, we develop a dynamic agency model with a market friction that helps explain these empirical regularities. The agency conflict is that the manager does not want to return money even if he runs out of good investment opportunities because his compensation is proportional to the total assets under management. The market friction is the extraneous fund policy restriction on the manager’s trading strategies so that he may have to make suboptimal non-informational trades. Depending on the dominance of his informational gain and non-informational loss, we predict a systematic pattern of premiums and discounts. Without invoking a specific solution form, we first derive some general results of the model and relate them to the closed-end funds regularities. Additional results and intuition are obtained from numerical solutions.