We advance entropy portfolio model with return forecasting.Short-sales, trading costs, minimal threshold for the change of weights are modeled.We evaluate ex post performance.The entropy model outperforms the MV model and 1/N portfolio.Return forecasting improve effectiveness of the entropy portfolio during bearish market. Traditional portfolio models such as the meanvariance framework lean to yield a concentration in a small amount of assets and uncertainty in outcomes due to its sensitivity to the bias in estimation. We develop a portfolio model that synthesizes the techniques improving investment diversity with return forecasting. Specifically, we integrate Yager (1995) entropy with autoregressive integrated moving average (ARIMA) model and evaluate the ex post performance of various models by using the data of the S&P 500 stocks. The results show that the entropy model yields higher performance, lower trading costs, and higher portfolio diversity than the corresponding MV model. Adding return forecasting in the entropy model increases portfolio efficiency to respond the market dynamics, especially during the market downturn. Due to its linearity, Yager entropy is more computationally efficient than the MV model and can be useful to asset management.