Testing Exit Methodologies of Trading Systems

In part 1 of this series on testing trading strategies, we looked at how to test different entry methodologies for a trading system to determine if our entry technique is actually effective. This breaking apart a trading system into it’s components (Entries, Exits and Stops) allows us to test the effectiveness of each component, and therefore improve upon a lacking component as opposed to throwing out an entire system which may have a great entry methodology, but lacks in the profit taking and stop loss methodologies. This article is to outline the testing of profit taking exit methodologies.

Testing exits is far more difficult than testing entries. Different exit methodologies will have different results based upon the entry methodology used. There it is usually best to choose and test your entry methodology first before you go about testing your exit strategies.

If you would like to test exits without having chosen your entry methodology, choose a generic entry methodology based on the trading style you have chosen. For example:

Trend Following

  • Moving Average Crossover
  • Channel Breakout

Mean Reversion

  • Bollinger Band Reversal
  • Keltner Channel Reversal

Scalping

  • Bid/Ask Imbalance

Once you have your entry methodology, to test your exits, you are going to test each exit methodology separately by entering with a single contract or share for every entry signal given, over a specific period of time. Ideally you would like to choose a period of time that contains as many different market type as possible (Bullish Normal, Bearish Volatile, Neutral Quiet) or, even better, test each market type separately to determine what exit works best in each market condition.

Some example of exit strategies include:

  • Static Profit Targets: A exit is set a certain number of ticks (or currency) away from the trade entry price. The idea is to get in and get out with a profit.
  • Trailing Stop: The trailing stop is designed to lock in profits. As a trade moves in your direction, the stop moves to reduced your risk, but never moves against you. This allows a trade that is going in your direction to continue and a trade that goes against you to be exited.
  • Parabolic SAR: While it was designed as Exit and Reversal system that would always keep you in the market, the Parabolic SAR can be a good exit methodology. The idea behind the Parabolic exit is to give the trade enough room at the start, and then as it moves in your favor, accelerate the stop to capture as much profit as possible.
  • Volatility Exits: Volatility exits close a position if the market has a volatile move in either direction, signally an emotional market which can be signals for turning points. If the market has a bar that is greater than 2 * ATR(14), that would signal a volatile move and thus exit the position entirely.

When comparing the results from different exit strategies, the most important metric is net profit as that is the most important metric in the end. However you will want to watch the win/loss % and Maximum Adverse Excursion (MAE). Some of your trades may go way against you and then come back for a profit, trades that would have been stopped out of with the use of a stop loss. It’s a way to keep track of the exit’s effectiveness.

You are also going to want to look at the Exit Efficiency percentage. The Exit Efficiency number, developed by Chuck LeBeau, takes the number of bars you have been in the trade and doubles it. You then find the ideal price you could have exited and then divide the trade’s profit by the ideal trade’s profits. Here’s an example:

Trade’s Entry Price: $100

Trade’s Exit Price: $125 (After 10 bars)

Ideal Trade’s Exit Price after 20 bars: $150

Exit Efficiency: ($125 – $100) / ($150 – $100) = 25 / 50 = 0.5 or 50%

The exit efficiency gives you an idea if you are closing your trade too quickly. A low exit efficiency % means you are close it too quickly, while a higher percentage means you are doing well capturing the available profit. You can then tweak your exit to attempt to capture more profit with your exit methodology.

As an example, I have taken a SMA(5, 20) crossover entry method on 1 minute bars, and then added in the following exit strategies separately, and then tested them on the ES over 2012. The exit strategies tested are:

  • Static Profit Target
    • 4 tick profit target
    • 8 tick profit target
    • 16 tick profit target
  • Trailing Stop Loss
    • 4 tick trailing stop loss
    • 8 tick trailing stop loss
    • 16 tick trailing stop loss
  • Parabolic Stop

Here are the results:

Exit Research Results SMA (5,20) ES 2012

 

We didn’t collect Exit Efficiency results as this is not built-in to NinjaTrader. However I’ll be looking to at this to my Additional Strategy Output in the near future.

Looking at the results, we can see a few things. First, the larger the space we gave it, the better the exit did. Secondly, the Parabolic SAR had the best % Win rate out of all of them, but also had the worst Avg. Win to Avg. Loss ratio. The best exit out of these tests were the static profit targets, but the trailing stop had a better increase in net profit between the parameter values.

At this point,, you should have a great exit methodology for your entry. All that it is left at this point is to test a stop loss methodology, which we will look at in part 3 of this series.

Advertisements

One response to “Testing Exit Methodologies of Trading Systems

  1. Pingback: Testing Stop Loss Methodologies in Trading Systems | @ShaneBrewer·

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s