A test of performance persistency of Chinese closed-end fund

In this paper we utilize data of Chinese closed-end fund to estimate performance persistence, considering various estimates of returns and relative and absolute benchmarks. One improvement is to split the estimation or holding horizons into 7-8 segments. In this way, we can distinctly discern performance consistence across various history and future horizons. In regression test, two methods are employed to estimate returns of different horizons, one is monthly moving sampling and the other is to enlarge the sampling frequency to lessen the data overlapping. Reverse persistence is found for short holding horizons. When estimation horizons are longer than 1year, findings is' less regular The most significant persistence appears when estimation and holding horizons are 6-12months. Results are affected by method and benchmarks to be chosen. Enlarging moving frequency decreases discrepancy across benchmarks and measures.