April 2024
Public enemy number #1 for your algorithmic or discretionary trading and investing is losing trades. To the extent that eliminating losses should be as great a, if not even higher, priority than additional winning trades. Why is this? Because returns are not arithmetic, or linear, but geometric, or logarithmic. If you lose 10% and win 10% you’re not back to 100%, Mr Market has taken his 1% and left you with only 99% of your initial capital.
And these break-even percentages increase exponentially. 20% drawdown requires a 25% gain to reach equilibrium, whilst a drawdown of 25% demands over a 33% profit recovery; and so on all the way to infinity...
However, losses are a necessary evil that need to be balanced harmoniously within any strategy destined for long-term success. Not all losses are created equally, yet losing is guaranteed, therefore minimising losses, and particularly excessive ones, is essential. Most of our live quantitative trading bots possess a sub 50% win rate; worse than flipping a coin. Yet it doesn’t really impact our bottom line as our winning trades are considerably larger than our losing ones.
That doesn’t mean we do not experience negative returns periods, far from it – everybody both has and will. We just ensure to never exceed our maximum allowed drawdown. This comes in 2 parts, our actual maximum drawdown at which point trading must seize, which we can thankfully say we’ve never reached and sits at 30%; and our roughly enforced point of somewhere between 20% and 25%, dictated by the tolerance to risk at that period, and has been reached just a handful of times... Beyond those points break even becomes an increasingly difficult mark to aim for, and the mental impact of being at a substantial loss is arguably harder to manage than the monetary situation. Even the most stoic of market participants will feel the psychological impact of having maths working against you!
We can manage this due to our approach to the market. We stick to a plan that makes it incredibly unlikely our upper (or I suppose lower) limit would ever be breached. Calculate your expected value and allow the law of averages to come into effect. If I have a 40% hit rate for winning trades, yet my average win outweighs my average loss to a 2.0 ratio, maths dictates that with enough time and trades it’s overwhelmingly probable the strategy will be profitable in the long run.
Although it gets a little more complex than this; you have to calculate what is the probability of near consecutive losses plummeting your account capital to beneath your maximum allowable drawdown? Frequency also comes into effect, we don’t want so few trades that it takes months or years for the maths to work in your favour. How about break even trades and the associated costs and fees to account for? It’s easy in principle yet incredibly difficult in reliable execution.
Restricting your winning trades is not the answer either. Eliminating trades that could be winners or cutting them short to lock in gains is the best way to prevent reaching maximum potential for profitability. You may have heard statistics about the detrimental impact of missing out on the S&P’s best days, and how ‘time in the market beats timing the market’. To some extent this is true, looking back to 1871 and the performance of the 500 largest US companies over that 153 year period, the mean average daily return was 0.039%; which doesn’t sound phenomenal but would have compounded to 109,767.3% from then until today!
Now looking back at the previous statement, if you fine-tuned your DeLorean and played the role of anti-investor and withdrew your investments on 5% of the market’s best days you’d have only returned 36,361.50%. Which is ~66.87% worse. A fairly remarkable performance decrease, requiring a 3.02x multiple on your portfolio just to make it back to where you could have been. However, if you’d instead decided to hold through those best days and instead opt to eliminate the worst 5% days of the market’s returns, you’d have obtained a monumental 548,956.06%, or slightly exceeding 5x gain compared to the simple baseline buy & hold comparison. Reiterating this blog post title and potentially altering the popular above trading mantra to ‘Yes, good days are good; but bad days are worse.’
Moving your limit exits to break even, trailing your stop losses, eliminating your biases, diversifying into uncorrelated assets and strategies or cutting your losers before substantial damage is done etc; are all valid methods of restricting loss impact. And in theory it’s all so simple. Yet, unfortunately, there’s no one size fits all solution. There are considerable factors to consider, one approach may work for our PnL yet be harmful to your PnL. The only certainty is the elimination and minimisation of losing positions should be everyone’s primary focus in trading, investing and even life - from now until the end.