Active Portfolio Management Research

This article investigates the possibility of active portfolio management and derives the key difference between consensus expectation model, defines the information ratio and value added by active management which are key variables of active portfolio management, then we come to general law of decision making. After that, we examine the model we set under the condition of China’s capital market. We introduce Bayesian Averaging of Classical Estimates (BACE) Approach to improve the accuracy of forecasting of excess return. In the end of this, we reexamine the model under different conditions as allowing short selling and bull or bear markets.The result we get shows that:first, active portfolio constructed by the model we set does outperform the benchmark; it’s necessary to take prediction accuracy into consideration, the improvement of prediction accuracy is beneficial to excess return and may reduce the tracking-error also; the excess return tend to be lower under the condition of allowing short selling unexpectedly, the short position may be very high, which cause leverage and great instability of asset, we propose that manager should not hold too much short position in active portfolio in spite of the consideration of borrowing cost. In the examination of bull and bear periods, we find that model we set tend to have better prediction power in bear market, which indicates us to pick different variables to set average model. When considering transaction cost, the excess return stays significant, but the real time tracking of transaction cost is out of this article’s reach, which then is the next step of our research.