Momentum Calculation

Course Name: Trading Alphas: Mining, Optimisation, and System Design, Section No: 12, Unit No: 10, Unit type: Notebook

I want to know why we calculate momentum on the cumulative sum of returns of the Closing price? I am just interested in the rationale behind it. What would happen if we calculated the momentum on the returns of the Closing price instead of the returns of the cumulative sums of closing price? Thanks a lot

For example use:

 

def momentum(df, feature="Close", period=40):
   mom=df[feature].apply(lambda x: talib.MOM(x, period))
   return mom 

This doesn't call cumsum() on the Closing column returns 

Hi,



The rationale for using the cumulative sum of returns is that on its own, the returns of the close price are (not perfectly) mean-reverting. If the asset gave a return of 2% daily, and ou are looking at returns only, then you feel that the asset does not have momentum or is not trending.



This is the reason why we are using cumulative sum of returns.



Hope this helps.