A Risk Budgeting Approach Improved by Genetic Algorithms

The risk budgeting approach has been applied to manage and monitor the portfolio risk of large and sophisticated institutional investors. How to decide proper parameters to obtain the optimal or near-optimal solutions for the risk budgeting approach is a problem. In this paper, a risk budgeting approach improved by genetic algorithms is proposed. Experiment results on real financial data demonstrate that, compared with some traditional methods, the proposed algorithm is capable of generating a set of parameters which can attain near-optimal return on investment with high probability and computational complexity of searching process also has been reduced.