Cardinality-Constrained Fofs Selection Model with Time-Varying Higher Moments

No abstract available.This paper proposes a portfolio selection model for Fund of Funds(FoFs) with the framework of mean-variance-skewness-kurtosis. Our model considers the time-varying relationship between realized higher moments and subsequent FoFs. The cardinality constraint is also included to control the number of constituent funds avoiding over-diversification and high costs of FoFs. We compare the performance of five models and find that the optimal model differs by market and market states, which are forecasted by the Hidden Markov Model. The results suggest that adjusting the selection model according to markets and future market states is better than consistently keeping one model