In a recent script I worked on the sharpe was significantly higher as the time period tested period tested for Sharpe increased. i.e. 3month sharpe <6 month< 9 month< 12 month. The sharpe ratio was 2.1 according to the metrics displayed after running backtest, but when I analyze the graph of sharpe with lower time intervals it appears the ratio is consistently below the average.
What changes in the script or interpretation of sharpe would help normalize return of the model ?