returns = pd.read_csv('returns.csv', index_col='Date', parse_dates=True) ff = pd.read_csv('ff.csv', index_col='Date', parse_dates=True)

Use the single stock returns in returns. You may select years and +1, but only use stocks with complete returns data for years and +1.

1) Do mean returns in year predict mean returns in year +1?

2) Does volatility in year predict volatility in year +1?

3)Do Sharpe Ratios in year predict Sharpe Ratios in year +1?

4)Do CAPM betas in year predict CAPM betas in year +1?

5) Does volatility in year predict mean returns in year +1?

6) Does CAPM beta in year predict mean returns in year +1?