Autocorrelation Remedy Help

[ardinea]

Hi everyone,

For my bachelor thesis I need to regress historical stock price with earning per share. My data consist of eleven years period with yearly observation in 100 companies in China. However, my data suffer from autocorrelation. I tried to search on autocorrelation removal and I found that it can be done through transforming the data either into log, first difference or using NeweyWest (this only corrects my standard error not making the coefficient to be BLUE) . Unfortunately, both of the steps did not work on my data. LM test on five lags still reject the null hypothesis. Does anyone ever experienced the same problem or know how to solve the problem? Please kindly reply to this post, I really appreciate your help. Thank you :)