Data Mining in High-Frequency Financial Data - Long Memory Test in Chinese Agriculture Futures' Market

摘要

An important issue in the study of financial markets is the evaluation of the stochastic memory of market returns. This paper examines Chinese agricultural futures' high frequency returns for evidence of persistent behavior, using a biasedcorrected version of the Hurst statistic. Results on this method provides strong evidence of long memory behavior for five dailyreturn series and two realized range series, but no evidence for one studied series. This finding adds to the developing research papers on persistent behavior in financial markets and suggests the use of new method of forecasting returns, assessing risk and optimizing portfolios in futures' markets.