CHAPTER 5 Testing and Evaluating a Trading System

If you are going to design trading systems, it is critical that you know how to evaluate the performance metrics of a trading system. This is not always as easy as it sounds. To show you what I mean, take a look at Figure 5.1. This equity curve is for a futures trading system, typical of what you might produce if you tested strategies yourself, or one you might find publicly available on the Internet. This curve was produced with TradeStation software, but results from other popular programs, such as MultiCharts and NinjaTrader, are basically the same. All good trading software gives you a variety of important (and, in my opinion, many unimportant) metrics to review and examine. Most of the time, the equity curve and performance report bring up more questions than they answer. Are the results good or bad? Are the results believable? Do the results have any predictive value? Finally, how do you separate the wheat from the chaff? I’ll answer these questions and more in this chapter.


FIGURE 5.1 Trading System Results—Is This Believable?

The first point to realize and understand when looking at performance reports, equity curves, or trade data is the old adage “if it is too good to be true, it probably is.” As a general rule, future performance of a trading system is almost never as good as its historical performance. In fact, ...

Get Building Algorithmic Trading Systems: A Trader's Journey From Data Mining to Monte Carlo Simulation to Live Trading, + Website now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.