We can apply the run-length statistics to specific strategies as a guide, whether discretionary or automated trading is practiced. In the fade strategy, which shows a 60% winning percentage over 216 trades, the simulation data predict an average of 0.78, or fairly close to one, run of five losing trades more than 50% of the time. That is, in the recent three-year test period of the fade strategy, half the time we should expect to encounter a five-in-a-row losing streak.
Actual backtest data are consistent with this estimate, and there is no reason to pull the strategy when a losing run of five occurs at this frequency. Later, if we encounter a losing run of, say, eight trades in a row, we might retire the strategy because the simulation shows there is just a 5% chance of this occurring. A reasonable conclusion considering the unlikeliness of that occurring is that market conditions fundamentally have shifted, making the strategy obsolete.
In general, the run length statistics can be used to guide how to structure a trading program that it is consistent with a trader’s psychology, risk tolerance and expectations. There are swing trade practitioners who argue modest (25%-30%) winning trade percentages are the norm while making sure to let winners run and cutting losers quickly (see “Diary of a Professional Commodity Trader” by Peter L. Brandt, John Wiley & Sons, 2011). In this approach, the trader must realize there will be lengthy runs of losing trades. A 30% per-trade winning percentage means that approximately 75% of the time there will be a losing run of eight trades over a 250 trade cycle. If the trader finds this hard to endure, then trading with a higher per-trade winning percentage is needed. The math cannot be beaten, really.
Finally, the run-length statistics offer a clue as to why we often read about even the finest practitioners going fallow, perhaps with a sense that they have lost their touch or have somehow gotten out-of-sync with the market (see Jack Schwager’s “Market Wizards” series). In fact, what may be occurring is the inevitable losing run has been hit — from which the professional learns the importance of managing losses or is retired from the endeavor by the unrelenting nature of the market.
For 20 years Michael Gutmann was a software engineer and manager at Intel Corp. He recently published the second edition of “The Very Latest E-Mini Trading: Using Market Anticipation to Trade Electronic Futures.” Michael can be reached via www.anticipationtrading.com.