An evaluation of Chinese securities investment fund performance

This study evaluates the performance of open-end securities investment funds investing in Chinese domestic equity during the period May 2002 to May 2014. A dataset of 419 funds is examined. High frequency weekly data are employed. This study implements a wide range of risk-adjusted performance measures which are divided into three broad classes, (i) unconditional models (Jensen, 1968, Fama and French, 1993 and Carhart, 1997), (ii) conditional-beta models, in which factor loadings are allowed to vary conditional upon publicly available information (Ferson and Schadt (1996) and (iii) conditional alpha-beta models where alphas are also time-varying conditional on economic information (Christopherson et al., 1998). Estimation diagnostics are then applied to select one ‘best-fit’ model within each of the three classes. Findings from all performance measures suggest no evidence of either statistically significant stock selection skills or market timing ability on average. Based on the statistical significance of the individual parameters and the Schwartz Information Criterion (SIC), a single ‘best-fit’ model from each of the classes is selected. In the case of the Chinese fund industry, this study is the first paper to evaluate fund performance using higher frequency weekly data, is the first to estimate conditional versions of performance attribution models and is the first to examine the time-varying parameters augmentations of the Treynor and Mazuy (1966) and Merton and Henriksson (1981) models.