In this article, we will discuss the methodology used for back testing a screen or a signal. Since we have already covered broad methodology and bias avoidance, we will just focus on the construct.
Back testing a screen is basically tracking the historical performance of a portfolio that is built using this screen. Every quarter, the screen is run to produce a new set of stocks and the portfolio is rebalanced to these new stocks (old stocks are sold and new stocks a bought). Each stock has an equal weight in the portfolio.
Note that if the number of stocks in the screen on a particular rebalancing date is more than 50, the top 50 stocks by market capitalization are picked in the portfolio.
If the number of stocks in the screen (new screened stocks) is 0, then stocks in the existing portfolio (old screened stocks) are sold and the entire position is moved to cash until the next rebalancing date.
Back testing a signal is basically tracking the historical performance of a portfolio that is built using this signal. Every quarter, the signal is applied to produce a new set of stocks and the portfolio is rebalanced to these new stocks (old stocks are sold and new stocks a bought). Each stock has an equal weight in the portfolio.
Note that if the number of stocks in the signal on a particular rebalancing date is more than 50, the top 50 stocks by aggregate signal score (sorted in descending order) are picked in the portfolio.