MultiStage (v7) System Results - Part Six

The first several parts of this series have presented results mainly in the area of CAGR, or overall net returns. While returns are VERY important, there are many other "results" worth taking note of. This post outlines several non-CAGR metrics that I pay attention to during the system development process.

This little picture is a clip from the AmiBroker results produced by the current system. There are a few interesting statistics to observe, which I've made note of below:

Number of trades: This is the number in the first red box. During the ten year test period - in this case 1/1/1999 to 12/31/2008 - the system took 4554 trades. I like to know that number because it tells me two things. First, that I have plenty of trades to make this statistically significant. And second, it gives me some idea of trade frequency. About 450 trades a year, or about 9 per week is average. For more information on typical trade frequency check out this post.

Long vs. Shorts: Some people ask why my system does well during bear markets. Is it because I employ a lot of short trades? Not really. As you can see, the MultiStage system takes long trades more than 80% of the time. I do take short trades from time to time, but longs tend to perform better in all markets, oddly enough.

Avg. Profit/Loss %: One of the metrics I pay most attention to in system development, is the average return per trade. This statistic takes into account all winners and losers, as well as all longs and shorts. It basically says, that every time I put $100 at risk during the test period, I made $3.34. That's important! It defines and quantifies the edge we have in the market with this system - at least during the test period. If I can find systems that produce an edge like this, then producing greater return is simply a matter of producing more trade frequency.

Avg. Bars Held: This is also an interesting number to me, because it gives me some idea of what to expect in the longevity of a trade. The MultiStage system holds trades for about 3 days. I'd say this is pretty typical of a swing trading system, and just perfect for satisfying my short attention span and desire for immediate gratification.

In the next post I will add a few more important metrics that I keep track of, so stay tuned.

Good trading...