The exit strategy for your trading method — not the entry signal — determines whether you make money. Finding the right exit strategy is where traders should spend most of their design effort, rather than searching for the elusive “perfect” entry signal, which really only determines trade frequency, not profitability.
The one thing that is misunderstood most often when developing trading systems is how exits affect overall system performance. Understanding exits begins with defining the four main types: discretionary, inactivity, risk management and profit taking.
By using a simple trend following system as a base, we can demonstrate how, just by changing the exit strategy, a system’s performance can be adjusted to suit risk-tolerance, profit objectives and requirements (see “Six systems,” below). Here, we will discuss the basic mechanics of the four types of exits and introduce the six systems we will use for our test. Next month’s, we will examine historical hypothetical results from each system to see how the exit strategy affects the overall performance of the system.
DISCRETIONARY EXITS
Before you examine how exits affect mechanical/systematic strategies, we’ll first look at situations where positions should be closed for other reasons. These are discretionary exits.
These exit rules are about anything that should, or could, cause you to exit a position (or all positions) that are not due to one of the other types of exit rules. Rules may be at the system level, or individual position level.
System level examples would be: you are unable to monitor your system for new entries, the technical trading environment that your system depends on is not functioning as intended (computer, internet connection, etc.) or you have met or exceeded your parameters for maximum drawdown.
A position level rule would be a specific unpredictable event that may cause you to exit a position immediately but is not directly price-based. Examples would be that liquidity
dries up in the equity for some reason, the company has been purchased by another one, or you are alerted to a new possible trade that has better risk/reward than a current open position and you need the capital to enter the better opportunity.
Discretionary exits are designed to reduce errors by suspending trading when you have a lesser chance of accurate implementation.
INACTIVITY EXITS
If a position is range-bound for a pre-defined period, it means you are taking risk in the position but receiving no reward. This position should be exited so that your capital may be allocated to a new, better opportunity. This is especially important if you are trading a system that can take multiple simultaneous positions, or if you are trading multiple systems in the same account.
For example, the length of your inactivity period should be proportional to your defined average trade duration and have some kind of threshold for the profit or loss that you would consider a non-performing trade. This should normally be proportional to the volatility of the instrument you are trading (that is, a high-volatility instrument may have a larger non-performing threshold).
RISK-MANAGEMENT EXITS
These rules are often called stop loss rules. Getting rid of losing trades within your pre-determined risk parameters is a very important type of exit. These exit points should be:
• Based on price only, in most cases.
• Outside the normal noise of the instrument (that is, adjust for volatility).
• At a point that clearly indicates that the setup has failed and is unlikely to become a winning trade.
The actual amount of risk you are taking with a particular exit is determined by your position-size in combination with your exit price. Always remember that sometimes it may not be possible to exit within the parameters set by your stop (on a gap at the open for example).
Generally these kinds of stops should not be loosened, only tightened, even if volatility increases after you enter the position. Remember: cut your losses short. This is what keeps your average losing trade smaller than your average winner, which is a big contributing factor to a positive expectancy system.
PROFIT-TAKING EXITS
These exit rules are for profitable positions only. Not surprisingly, many of these exit points share many characteristics with risk-management exits. They are:
• Based on price only in most cases.
• Outside the normal noise of the instrument (that is, adjust for volatility).
• At a point that indicates the chance of further profit is clearly diminished.
Generally, these types of exit points should trail from your entry and may be loosened if volatility increases to give winners more room to breathe. In other words: let profits run. A simple volatility-based stop, such as one based on a multiple of the average true range trailed from the high, low or close is effective.
Note that any particular position can, and often does, switch from winning to losing and vice versa. This simply means that the exit rules that are currently in force for the position may change through time.
Usually, having pre-defined price targets is not effective unless your setup includes something that can determine the likely extent of a winning trade. Profit-targets are usually an example of cutting your profits short and are generally counter-productive because they actually reduce the average size of a winning trade.
Profit taking stops should be designed to keep your average winning trade large (compared to your average loser), which is another big contributing factor to a positive expectancy system.
OTHER FACTORS
It is typically a bad idea to have a rule that attempts to move your stop to breakeven as soon as possible. This is a psychologically comfortable thing to do, is arbitrarily based on your entry price, and is therefore common to everyone trading a similar system/setup to yours, and who has a similar entry point. Therefore, it is common for a position to return to your entry point after a few days simply because a lot of breakeven stops have been placed there.
Because most exit rules are price-based, you should always view the last price as the real price, regardless of how much or little volume there was at that particular level.
That said, you should trade liquid instruments exclusively so that you are participating in an orderly market that never suggests the possibility that the displayed price isn’t the real price. You should never have to wonder whether the price you see is a bad tick, anomaly or market manipulation; this gives you the confidence to always stick to your exits.
SYSTEM COMPARISONS
The systems used for our example are all based on a basic trend-following entry that trades both long and short in the S&P 500 cash index. Although we will use the index for this demonstration, this approach is valid to any individual stock or stock index in your portfolio.
Much of the Tradestation code for each system will be available in next month’s article. The system does not take more than one position at a time, and in our testing we used weekly bars going back 10 years.
The basic system has an entry that is the bar’s closing price crossing above or below the 12-bar simple moving average (SMA) of the close.
So,
• go long if the close crosses the SMA (12) from below.
• go short if the close crosses the SMA (12) from above.
This is about as simple an entry that can be created, so we know all our system traits and characteristics come from the way we change the exit strategy, not from anything to do with having a sophisticated entry.
Please note this is not meant to be a profitable or rational entry for a standalone trading system. It is simply one that will provide a reasonable frequency of trades to see the effect of our exit strategies while keeping the entry criteria the same for each sample system.
To compare the value of each exit strategy, we will use an adaptation of the system value concept. The system value is a relative measure of performance that combines profit (or loss) per unit risked, the variability of profits and losses and the number of trades per period. It will be adapted because we are using fixed position-sizing.
We’ll also look at average trade size and other standard performance statistics, but system value will be our primary measure.
Next month, we will look at each system and the corresponding code in depth and determine which exit strategy performs the best.
Paul King is a trader, trading coach and independent financial advisor. His company, PMKing Trading LLC is based in Vermont. This article was adapted from his first book, The Complete Guide to Building a Successful Trading Business. He can be reached at www.pmkingtrading.com.
