You can’t have everything

May 15, 2015 12:00 PM

Inventing a mechanical trading system is much like solving a Rubik’s Cube. More often than not, you’ll match one side of the cube at the expense of another. In system development, as you improve one aspect of a performance summary, you’ll almost inevitably see deterioration in a different area.

Traders are especially enamored of the percent profitability, or accuracy of a system. In practice, you should not target a given win/loss ratio per se; the goal is the creation of a suitably profitable numbers game. It doesn’t matter whether you get there via relatively few but comparatively large winners, as is the case of trend-following systems, or smaller wins but enough of them to more than offset the losses. It also could be a combination of the above; the totality of the system is all that matters.

This is not to profess insensitivity to percent profitability. All things being equal, more winners are easier on the psyche than less. Each win reinforces our faith in the system while every loser is a potential confidence shaker. This is not the way it should be, of course, losers are an inevitable cost of doing business. Still we’re only human after all.

But again, a high percentage of wins vs. losses is not the only road to a profitable system. Several luminaries have owed their successful careers to unadorned trend-following methodologies. These systems tend to produce winning trades 50% of the time or less. And there is always some giveback of profit before a winning trade is closed out. A position’s equity high is never anticipated; profit targets are not used. It’s therefore easier for a trade to wind up a loser. The overall success stems from how large the winners become when the runaway breakouts hit. Cut your losses short and let your winners run is the mantra of these traders. 

The concept is illustrated in “Runaway winner” (below) of daily bars in the Japanese yen. Note the three smallish losses leading up to a hugely lengthy and profitable short. The performance report shows the results. The 40% profitability is about normal for this type of system. That’s fine considering that the average profitable trade ($2,401.33) is nearly double the average loss (-$1,282.67). The largest win is $19,532.50, roughly four times the maximum loss (-$4692.50).

“Finding an edge” (below) shows results of the day-trading system for the major U.S. stock related index futures. It’s called “2 to 1” because its profit targets are twice as large as the stop losses.

The results are consistent throughout the field. All five markets yield five- figure net profits on a single mini contract over a 10-year period. The sheer number of trades makes it unlikely that we’re seeing a mere statistical fluke. The maximum drawdown column (Max DD) shows the worst equity pullback over the period. It occurs when an account equity reaches a new high followed by a series of losses. The worst level then reached — determined after the fact when the equity again rises to a new high — is the maximum drawdown. Conventional wisdom says that you should be bankrolled for one and a half times that number under the theory that your worst drawdown is always ahead of you.

The last column, return on account (ROA), is the number that this trading system was tested for. It gives historic profitability in relative terms, not just what you theoretically made, but how much pain you had to endure in the process. It assumes your startup bankroll was the max drawdown amount—in the top S&P case, $6,448. The ultimate $47,617 net profit figure is 7.3848 times that amount, or a 738.48% increase over the 10 years.

The percent profit column also is impressive. Because the profit targets are twice as far as the stop losses, you’d expect one winner for every two losers, or a 33.33% result. We’re several points higher across the board.

System building is a constant process of tweaking and retro-fitting. When you do it to excess, you’re curve-fitting and creating a model that won’t likely perform in a similar fashion in the future. The adjusting-vs.-over-adjusting process represents a fine line. There are prudent guidelines for staying on the correct side.

If you must have higher percent profitability for the sake of your personal comfort level, the easiest solution is to re-configure the win-loss ratio. “Hitting for numbers” (below) shows what is theoretically possible with the same system. It just no longer targets a two-to-one profit-to-loss ratio.

Page 1 of 3
About the Author