This article investigates the statistical relationship between the share prices and NAVs of closed-end funds. The empirical analysis shows that the share prices are a function of NAVs which follow Brownian motion independently without regard to the share prices of closed-end funds. Based on this result, this study develops a regression model incorporating time variable additionally to explain the times series behavior of share prices. The empirical results report that this model has a good prediction power for the future share prices. This study contributes to explain the relationship between the share prices and NAVs and to predict the pattern of share prices and discounts of closed-end funds. A useful investment strategy can be developed with this model.