Here’s a question for you. If trend traders talk about following price rather than predicting it, then why do they feel that backtests are important?
Well, as counter-intuitive as it might seem, a trend follower does not backtest to estimate their future profitability but rather conducts backtests to determine risk management thresholds before stepping into uncertainty . It assists in answering the following questions:
- Is the strategy curve fit?;
- Does the strategy stand up with varying market conditions without too much capital deterioration?;
- Is your position sizing appropriate in relation to past unfavourable historic market conditions to give you a better chance of surviving future uncertainty ?;
- Is the past drawdown and series of consecutive losses within my risk tolerance levels?;
- What is the result from the past if I change variables chosen in my strategy or have a different start date?;
- How do I know if my performance is improving or not without a roadmap of how the strategy has performed in the past?
Notice that none of these criteria have a view in relation to future outcomes. Each is related to dealing with risk and uncertainty. A backtest provides you a roadmap of the past and a benchmark to assess your future performance. We use a backtest to understand our risk tolerance thresholds. For example let’s say you have just taken your 6th loss in a row. How confident are you of taking that next trade setup? A backtest gives you a basis to understand if this string of consecutive losses is a frequent occurrence or not. If your favorable backtest includes a string of 15 consecutive losses in a row, then the chances are good that you will be able to commit to that next trade after 6 consecutive losses. Without a blueprint of the past, I am not so sure you would have the willpower to do so.
Undertaking a backtest is akin to steering a big ship into an uncertain future with a risk management rudder as opposed to listing about aimlessly in uncharted waters.
Let’s put it this way. If your backtest concludes that your strategy didn’t stack up over a historic time horizon and your forward test concludes that it did, then this tells you that your strategy has a finite shelf life and is dependent on a particular form of arbitrage that is not enduring. In other words your strategy is curve fit for a particular class of market condition.
Now as trend followers, we know that you can’t predict the future so if you have a strategy with a finite shelf life, then the chances are good that you are walking around with a time bomb that will probably fail in the longer term with different market conditions. So if this is the nature of the strategy you are dealing with you then have to decide when to turn it off and worse still, to be able to survive over the long term it requires you to have other strategies to take it’s place when it fails? This will determine your longevity in the game. Not the success of a single system but rather your ability to decide when to turn off an under-performing strategy and your ability to find other successful replacement systems to choose from when this occurs. This introduces selection bias.
So the backtest is a method to save you this grief and prevent you from entering this pit of despair that so many traders face that like to predict.
Choosing strategies with a short shelf life can lead to strategy hopping which is a major reason why so many traders fail over the longer term as they frequently shift to a new system without being able to determine what is a natural drawdown inherent in every system versus a drawdown that is simply on the way to complete risk of ruin.
If your backtest stacks up and your forward test stacks up and is within your risk tolerance range, then this tells you the strategy is robust over the test range. If a short test range, then it tells you little. It really only tells you about more recent market conditions and how well your strategy fared.
If your long range backtest of 20 years plus stacks up and your forward test stacks up, then this tells you that your strategy is robust over a 20 year plus period and can navigate a variety of different market conditions without having to second guess current market conditions. You don’t know what the future will bring, but if it rhymes with market conditions over the past 20 years, then there is a good chance it will deliver a more favorable outcome than without having conducted a backtest and running blind into the future.
Wouldn’t you like to know this before committing your hard earned dollars towards your wonderful creation?
If you ask a blind man what he would prefer, then he would probably say that he would have preferred to see in the past before he became blind as opposed to being blind from birth as it gives you a richer context of how to possibly navigate an uncertain future with less bruises along the way.
Trade well and prosper