I bet if you came across a trading system which generated profits 75% of the time, you’d be pretty keen to give it a go. With that kind of win rate, the odds should be firmly stacked in your favour.
All you’d have to do is allocate a fixed amount of capital for every trade. Bang out trade after trade, and you should be able to sit back and watch your capital grow.
You might even be tempted to put a bigger chunk of capital into each trade. After all, with those kinds of odds, what’s there to lose?
The problem in doing so is that you’re just taking a punt. Sure, the next trade has a 75% chance of winning. And so, too, does the one after that. But there’s also the one-in-four likelihood that the next trade could be a loser.
And what’s to say that you don’t lose three times in a row? Put too much money into those three losing trades and your trading system with a 75% success rate has just demolished your account…
Half the picture
By looking only at the win rate, you ignore a vital part of the equation. What you also need to know is how much the trading system generates on average per winning trade, and the average size of the loss on the losing trades.
You can see that, even if a trading system had a 70% or 80% win rate, it could be undone by two basic factors.
First, if the average loss greatly exceeds the average win. And second, ‘randomness’. That is, despite a high winning percentage over a vast number of trades, you could still lose multiple times in a row.
It’s this last part that fund managers spend so much time assessing. They attach probabilities to different outcomes — like successive losses, for example — and adjust their position size to suit.
Of course, there’s also the possibility that you might win a number of times in a row. But there are always two things at play: the win rate; and the average won (and lost) on each trade. You need to determine both before putting a trading system into the market.
Calculating your own performance
If you’re a high-frequency trader who’s in and out of the market all day, it won’t take long to develop your own track record. Day-trade for a month or two, and you’ll quickly amass a high number of trades.
But even if you trade much less than that — maybe just a few times per month — you can still track how you’re performing over the year to see if your system is working.
The first thing to do is calculate your win rate. That’s simple enough; work out the number of successful trades as a percentage of your total trades.
Once you’ve done that, work out the average size of your winning trades. You then do the same with your losers.
Many traders get bogged down in the first part — the win rate. All they want to see is how often they’re right. The higher the percentage, the smarter they believe they are.
But it’s also the second part — the win/loss amounts — that ultimately dictate how successful a trading system will be.
Even if the average size of your winners is greater than your losers, you can still be undone if your win rate isn’t high enough. And the opposite also applies.
A trading system where the average loss exceeds the average win can still be profitable if the win rate is high enough, although it makes it more challenging. You can see that there is always a relationship between the two.
And this is where randomness also comes back into the equation. With the markets, there is always a chance that an outside event could exacerbate your losses. What might be an average loss (or losses) under normal trading conditions could spiral if the market takes a big tumble.
The goal of trading is to always try and maximise two factors: a higher win rate and the average size of the winning trades over the losers. The better both these numbers are, the more profitable your trading system will be.
However, you also need to account for random events — like multiple losing trades and outside events — before deciding how much capital to put into a trade.
For Markets & Money, Australia