Backtesting can be a splash of cold water in the face when you first test your ideas. If the results are less than stellar, you have the ability to alter your strategies and test these changes to see if they improve your results.

Setups

Conditions that must occur on the day(s) before you will consider trading the stock

• If the ratio of winning to losing trades is insufficient, then you can alter your Setup criteria to improve this ratio.
• You may want to consider adding a test that looks at market factors. As an example, you may want to only consider trades when the S&P index is above a 21 day moving average or that a Worden Market indicator is reflecting an improving picture.
• By comparing your best winners to your mediocre winners, you may find additional criteria that you want to incorporate into your strategy to be more discriminating as to when you'll consider taking a trade.

Entries

Conditions that must occur that will establish your purchase price on day of entry. If the Backtest shows that many trades are quickly being stopped out, then consider testing to see if:
 

• Entry only when the stock hits or clears the previous day's close.
• Entry only when the stock hits or clears the previous days high.
• Entry only when the stock hits or clears the previous days high by a fixed amount or percent.
• Entry only if the stock does or does not gap up at the open.

Exits

Probably the most important part of Backtesting, the maximum chance to improve your overall performance will be made with testing and tweaking your exits. Following are ideas that can be explored with a good Backtest product to help improve Exit strategies.
 

• Frequently exiting the trade quickly. This can be a sign that Setup or Entry strategies are picking the wrong stock/date too frequently and that the Setup/Entry strategies are what needs tuning and not the exit.
• By looking at the days immediately after an exit, you see frequent examples of the stock immediately reversing; your stops may be too tight.
• Look at the favorable trades and notice the maximum drawdown percentages. This may suggest an adjustment to your stoploss percentage is needed.
• Notice when the maximum drawdowns on your favorable trades occur on average. This may suggest exit strategies based upon time in the trade as well as percentage. For instance, if the bulk of your winning trades are profitable after 3 days, you may want to test exiting all trades after 3 days if they aren't profitable.
• If your trailing stops aren't capturing sufficient profit, you may want to consider exits when a profit target has been met.
• If you have multiple exits and one of them is the predominant trigger, try altering it to see if greater profit may be captured.
   

In all cases it is a good idea to Backtest with just a portion of the stocks available. Upon completion of what you think is the best strategy, test with a full sample to be certain that the results are reasonably consistent and that you haven't curve fitted the data.

The 8-13 moving average crossover example mentioned at the beginning of this section has not been a winner for this year. Using the S&P500, it would have generated 1296 trades of which 934 were losers. On average you would have lost 1.6% per trade plus commission and slippage cross.