Incorrect computation of SPY returns in the code?

Quant Trading Strategies & Ideas
1

Course Name: Trading Alphas: Mining, Optimisation, and System Design, Section No: 6, Unit No: 9, Unit type: Notebook

 

In the last cell why was spy returns were divided by 16

here is the code for your reference:

((((spy.Close.pct_change()+1).cumprod())-1)/16).plot(label='SPY Returns')

 

 

2

Hi Kshitij,



We will check this and get back to you

3

Hey Kshitij,



Here's a reply from the author:

The factor "16" in question is just a scaling factor for plotting to make the two graphs comparable. If we did not do that, the backtest chart would be extremely small and not recognizable in the plot. It could be argued that this means that the total return of this strategy is really low. However, it is important to understand that we are dealing with alpha mining and such returns only make sense in the context of a larger collection of micro-strategies or micro-alphas. 



Thanks

Rushda