A Behavioural Approach to Efficient Portfolio Formation

In this paper we investigate the portfolio performance of subjective forecasts given in different forms. In constructing the efficient frontier, the expectation formation processes based is on subjective forecasts and human behaviour, rather than past prices. The efficient portfolios are first constructed using point, interval and probabilistic forecasts. Next their performance is compared to those constructed using the standard approach of time series data. The subjective forecast are given by actual portfolio managers who forecast the prices of stocks actually traded on the stock exchange on a real time basis. The first contribution of the paper is to show that the portfolio performance of subjective forecasts are much more superior to those of standard time series modeling. The next contribution of the paper lies in the fact that it employs experts, professional fund managers with substantive expertise, as forecasters. Third, in this research, point, interval and probabilistic forecasts of expert subjects are investigated and therefore, findings are robust to the task format