When to Abandon a Trading Strategy

No matter what method you use to trade the markets, be it fundamental analysis or technical analysis, there is one common element all must master if they hope to be successful, the art of discipline. Every day a traders discipline is tested as they watch the price of financial instruments fluctuate and feel the emotional urge to enter or exit a position, even when they know from their knowledge and experience that it is best to suppress such urges. However, there is another type of discipline that is required for a trader to become successful, which is especially applicable to system traders, but applies to anyone who uses a predetermined trading strategy. That is the discipline to know when to abandon or change a trading system's rules because they are resulting in poor performance versus when to continue or give the technique more time in its current form as it may just be that the system is experiencing an inevitable period of drawdown.

A technical trading system can be defined as a set of rules, based primarily on the movement of price and volume of a financial instrument(s), which are followed to make decisions on when to enter a position and when to close or exit a position. (More on developing technical systems can be found in this article.) Most system traders will inevitably have encountered the dilemma of developing a great set of rules which they believe to be thoroughly tested and have great profit potential, only to begin trading with real money and experience a serious string of losses. This process not only effects ones capital level, but also seriously affects the psyche of the trader. They begin to doubt the system, doubt their ability, and wonder if they have what it takes to be a successful trader. In other, albeit more rare cases, a trader will develop a system, start trading and find that it performs as expected. However, eventually this trader will also reach a time where the system stops producing strong profits and instead enters a prolonged period of losses. In both cases, the trader reaches a crossroads of whether to change the rules, stop trading the system completely, or continue in the hope that the performance is just an aberration.

Before a trader decides how to proceed when faced with this dilemma, there are a number of things they should examine which will increase their chance of taking the correct course of action. First of all, it must be realized that all trading systems, or even more broadly, all traders and investors, no matter what method they use, experience periods of drawdown. Luckily for system traders, due to the fact their rules for entry and exit are usually objective and developed based on testing with historical data, they can determine what constitutes a normal amount of drawdown by examining how the system performed in the past. Obviously historical levels of drawdown will not correlate exactly to future levels, but if a system is robust enough to be profitable over an extended period of time, it should not see levels of drawdown that far exceed the amount or duration experienced in the past. So, for example, if historically the average drawdown for a system was 5%, and the trader is in the midst of a 15% drawdown, this should catch the trader's attention. Traders who encounter situations like this should begin to seriously consider changing their trading rules, or developing a new system entirely, as possibly the system was over fit to historical data and is not robust enough to work on future market movements.

In relation to drawdown a trader can get an even better idea of whether the current drawdown is reason enough to change their rules by measuring the drawup in profit prior to when the drawdown began. (Drawup being the profit from trough to peak on one's equity curve, versus drawdown being peak to trough loss.) The most recent drawup should then be compared to what the average drawup was over the same historical period that was used to calculate the average drawdown. Now if the trader started trading and the losses began right away then they would examine the most recent drawup on the historical data prior to when they began trading. This can be useful in the following manner; by continuing with the drawdown example in the previous paragraph of 15%, let's say the drawup prior to this drawdown was 30%, while the historical average drawup was only been 10%. In situations of this manner what might have occurred is rather than the system rules not working in a useful manner any longer, instead the volatility in the market may have increased substantially leading to relatively larger gains and larger losses. Or in other words, while the drawdown was larger than normal, the amount of profit earned prior to the drawdown was also much larger. As such the rules may not need to be changed, assuming ones trading capital allows them to cope with the larger amounts of drawdown.

A final few stats that are helpful to examine are the percentage of trades which are profitable, as well as the profit-loss ratio (i.e. the average profit per trade on the winners compared to the average loss on the losers). By comparing these to the historical averages, in addition to drawdown and drawup, one can further determine whether their system requires changes. For example, if the percentage of trades profitable was on average 50%, but has recently shrunk to 20%, then there may be some changes required to the rules of the system. This same logic would apply to a significant decrease in the profit-loss ratio.

In addition to observing the stats just mentioned, a trader can better determine whether changes to their system are need by examining whether the condition or state of the market has changed recently. In other words, has the market moved from a trending market, to a ranging market, or possibly even changed to a market state characterized by extreme levels of volatility (i.e. late 2008). Attempting to determine this can be very helpful as often one will develop a trading system that works very well in trending markets but very poorly in ranging market. This situation is in fact the case with many systems which use moving averages as the primary entry and exit signals. One can attempt to determine if the market state has changed by examining the price charts of the instruments they trade, as well as some of the broader market aggregates, to see if price movement is showing trending conditions, such as higher highs and higher lows in a trending bull market, or lower lows and lower highs in a trending bear. Or conversely, if price is range bound and experiencing neither continued up movement nor continued down movement. Furthermore, with respect to the condition of the market, a trader can also examine measures of volatility such as the VIX. If one observes that there has recently been a significant increase or decrease in the level of volatility, this could have a serious impact on the performance of one's trading system (more on this here).

To conclude, one of the toughest decisions a system trader will face is when they encounter sharp drawdowns in their trading capital and must decide if it is best to stop trading the and change the rules or continue with the system as is. On the one hand, changing the rules may be the correct course of action as possibly the system is not as robust as had been anticipated. While on the other hand, one must be careful not to jump the gun, realizing that drawdown is inevitable and profits may be just around the corner. A trader can improve their chance of making the correct decision by examining the recent performance of the system relative to past performance. With the idea being that if the more recent performance has diverged significantly from historical levels then it is more likely that some changes are needed. One should also attempt to determine if the state of the market has changed, as a move from a trending to ranging market or vice versa may warrant a change in trading methods. Furthermore, through undertaking analysis of this type it is less likely that an individual will make the decision based on the one emotion that drives too many traders actions, fear.