How effective is the tail mean-variance model in the fund of fund selection? An empirical study using various risk measures

Abstract In this paper, we study the tail mean-variance (TMV) model, which incorporates variation and tail risks and allocates the capital corresponding to the asset’s risk, by using several risk measures including the Value-at-Risk (VaR) and a non-linear weighted (NLW) risk measures. We also use a 5-fold cross-validation algorithm and carry out empirical investigations. We find out that the VaR-measured TMV fund of fund dominates all the other funds of fund using several Chinese funds and US’ funds.