Backtest

In reality, trading costs vary by stock, time, and the specifics of your brokerage. These factors can and will significantly impact your overall strategy.

At Equities Lab, we model trading costs as slippage, allowing you to customize them with a formula that can reference any attribute of the security being traded. You can set market cap thresholds, use average trading volume, or create a composite formula tailored to your specific needs—just like the Kini Trading Costs model.

The backtest tilt in this case shows an alternative look at a long/short portfolio. The dense clustering of points around the y-axis suggests that the strategy made significant moves—either up or down—even when the market was relatively stable.

The points clustered around the x-axis indicate periods of drifting while the market was in strong motion. As with many long/short strategies, market correlation is minimal, with a beta of just 0.007. This low correlation allows the strategy to achieve an alpha of 2.25%, despite only rising half as much as the broader market.

Are you concerned that your portfolio might be too large? You can easily limit the number of stocks in your portfolio. In the above example, the max holdings has been set to 5, with the ranking factor being Income Statement Score – Balance Sheet Score.

While stop-losses can help reduce volatility, they can also affect overall performance. In the graph above, the green backtest line utilized a 20% stop-loss until 2010.

After 2010, the strategy became more conservative, implementing a 10% trailing stop-loss. The green backtest line closely follows the orange baseline up to 2010, but then it seems to just lose momentum. Could the tighter stop-loss be the reason?

Investors are human, and their behavior often reflects that. Typically, they won’t jump out of a position they have just entered, and they may be reluctant to repurchase a stock they sold earlier that year.

Do these human tendencies make a difference or impact performance? Find out! The grid above illustrates a list of trades where the investor trades on a quarterly basis, is unable to sell for at least one year, refuses to hold a position for more than three years, and never repurchases a stock they have sold. As expected, this approach underperforms compared to the base strategy, which has outperformed since 2010.

This strategy combines an ordinary Beneish/Piotroski/Momentum combination with a seasonal twist—avoiding holdings during the summer months and certain unfavorable holidays.

The particularly intriguing part here is that the screen rebalances every time we go from a vacation period to a non-vacation period. This occurs several times a year, at irregular intervals. The red line on the graph represents the screen’s performance, which consistently outperforms the market.