What is the biggest worry for a systematic trader?
If you said, Loss, you are wrong. Yes, a loss does hurt a trader, but unless he has no clue as to what he is doing, he knows that losses are inevitable, even big losses. What he worries more is the risk of his system going into a losing steak.
Lets assume your system is doing fine and you suddenly find markets moving against you and by the time your system comes with its signal, you are on a non leveraged basis looking at a loss of 8%. A loss that big hurts and for some one who is leveraged say 3 times, he would have just seen 24% of his capital vanish into thin air.
But what if the same trader sees a draw-down of say 4.75% but not in one big swoop but one that comes as a streak of 9 consecutive losses and one that lasts over 44 calendar days. Which one do you think shall be more painful?
In case of the 1 big negative trade, While the pain will be immense, it will quickly pass as the trader moves to his next trade (with a much lower position size) and hopes that he can at some point of time recover from the hole he has fallen into.
In case of the 9 small but consecutive trades, the pain is much more dragged on as the trader spends nearly 1.5 months with every trade he has taken going for a toss. Its very tough for even those with strong hearts (not to mention a nice bank balance) losing their sanity and deciding that pulling the plug from the system may be the only way out.
But losing streaks are fairly common and as I shall showcase later, something that you shall have unless your system has a win rate of 100%. When you run a back-test, most reports do provide you with the number of losses that have come in a streak.
But what if your system is more mechanical driven and hence cannot get such a number. The easy way out is to just measure the win or loss % of trades and then check the following table.
The above table lists the probability of you seeing a draw-down of X number of trades based upon your systems winning / losing percentage.
Most trend following systems have a win percentage that ranges between 35 and 45 and if you can check out the same in the above table, you can see that there is a very strong probability that your system at some point of time shall witness a streak of losses ranging from 6 – 9 and that can be said with near 100% confidence.
So, what can be done about that you may like to ask. Unfortunately, there is nothing much you can do other than to keep trading while reducing your position size at every drop of equity. This way, by the time you hit your 8th or 9th losing trade, your position size will be considerably smaller and hence the account will still be able to trade without having to witness fresh capital deployment from outside.
What is interesting though is the fact that even if you have a streak of 9 losing trades, the loss in money terms is generally lower than the biggest losing trade. For example, I was just testing the same for a simple Moving Average Cross over system (3 by 5).
The biggest loss (period being 1st June 2009 to today with trades taken on Close) was a loss of 346 points (8.07%) which took just one day (Entry on 30-07-2009 & exit on 31-07-2009). On the other hand, the streak of 9 losses (loss in point terms being – -40, -9.4, -39.95, -22.45, -44.55, -39, -65.6, -13.2, -20) stretched from 20-07-2010 to 02-09-2010).
While the system may still be positive, that 45 days of bleeding is something that will not pass off in a hurry and like a deep wound would remind the trader every time he encounters a loss.
What is also noteworthy is that a longer streak may not actually be the most painful (in terms of loss). While the 9 trades in total lost around 4.75%, the same back-test data provides evidence of smaller streaks having a bigger impact (in terms of losses).
While there is no escape from such streaks, having a good position sizing algorithm would ensure that the pain in terms of loss of capital would be lower than what it would be if a single position size was maintained throughout.
Use a position size that is neither too aggressive or one that is too meek. In case of the former, you risk blowing up your account while in terms of the later, you never can take advantage of the good runs. But whatever you do, remember that unless the system is faulty, streaks do not stretch to infinity and hence one needs to keep his cool and take all the trades as and when they are signaled.