Mutual Fund Investment Style Empirical Analysis of China: Based on SOFM Algorithm

In this paper we apply self-organizing feature map (SOFM) to classify the mutual fund investment styles in empirical analysis of China. SOFM simulates human brains´ clustering, self-organization, self-education functions in information management process. Compared to other traditional statistical cluster methods, it has unmatched advance in processing large samples, complicated index system; and it also the overcomes deficiency of lacking time series data as most mutual funds in China has not set up long enough. From the empirical results, we get three conclusions: first, most of the mutual funds in China did not adhere to their claimed styles in prospectus language. Second, the accurate classifications based on SOFM give investors and stock supervision organizations credible information to identify equity investment styles effectively. Finally, the empirical classification performances summary can guide the money managers achieve better performances in the future