Shorting systems slippage

Research
1

Hi all,
I was hoping some members could clarify a question that has come up a few times. When creating a shorting system backtest, should you add the slippage as a negative percentage, or positive? In previous threads, people have advised to use negative slippage in shorting systems.

However, when testing different variations (negative, positive, no slip) on a shorting system and reviewing one specific transaction in the backtest, I see that the negative slip actually produces the highest return, whereas the positive slip produces the lowest.

So I wanted to be sure which method to use so that i am not skewing my own backtest results. This system focuses on shorting within the S&P500, so slippage should be somewhat minimal.

INTU $ open $ close Return$ Return%
No slip 32,398 31,648 750 2.3%
-0.15% slip 44,094 42,945 1,149 2.6%
0.15% slip 23,640 23,162 478 2.0%

Thanks in advance.

2

Apologies for the messy formatting on the chart of the difference in returns. From left to right it shows (at no slip/negative slip/positive slip): $ amount sold short; $ amount covered; $ return; % return.

3

You should not use negative slippage, not even if you’re doing a shorting system backtest. Slippage always diminishes your returns, long or short, by adding fees to your transactions. Negative slippage will give an extra boost to every transaction by subtracting fees from it. The only case I can imagine for using negative slippage would be if someone were to pay you a commission for every trade you made.