Using Investment Portfolio Return to Combine Forecasts: A Multi-objective Approach
National Chung Cheng University - Department of Finance
Cornell University - School of Applied Economics and Management
Mark T. Leung
University of Texas at San Antonio - Department of Management Science and Statistics
Forthcoming in European Journal of Operational Research
In recent years, there has been a growing trend of using multiobjective techniques. The primary advantage of using multiobjective techniques in decision making is, as stated in Spronk (1981), "that most of these (single objective) models and methods are unsuitable for decision situations in which multiple and possibly conflicting objectives play a role, because they are concentrated on the optimal fulfilment of only one objective." Given this notion, we attempt to explore the possibility of taking the multiobjective approach to solve a typical problem encountered by many financial and investment managers, namely, making investment trading decisions based on a set of potentially incompatible forecasts supplied by different analysts. In our experiment, several series of financial analysts' forecasts are generated by different forecasting techniques. The approach examines the historical performance of the various series of forecasts and combines them based on the average, variance, and skewness of investment returns. Through the use of a goal programming model, an investor can construct a portfolio which matches his or her preference. This portfolio-based approach also adds the benefits of diversification in trading. We test our proposed approach with three widely traded broad market indices, S&P 500, FTSE 100, and Nikkei 225. Improved performance of the multiobjective portfolio approach relative to those of individual forecasting techniques and some previously suggested forecast-combining models is measured The empirical results indicates that the performance of the proposed approach statistically outperforms the others at a significance level of 0.05. Moreover, we find that the benefits of our approach becomes more apparent when the market exhibits higher volatility and instability.
Note: This is a description of the paper and is not the actual abstract.
Keywords: Investment analysis, goal programming, combining forecasts, multiobjective decision analysis, trading strategiesAccepted Paper Series
Date posted: September 18, 2001
© 2013 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo2 in 0.343 seconds