I was intrigued when I first heard from a fellow trader (thanks Chris!) about a method for measuring the quality of trading systems. The System Quality Number (SQN) is a proprietary measure of system performance developed by Van K Tharp, author of the book Super Trader and one of the world’s top trading coaches. A simpler measure of system quality provided in his Super Trader book is the ratio of Expectancy to Standard Deviation of R, where R is the initial risk of the trade. For example, if you buy a stock for $40 and you have a planned stop loss of $10, then your R would be $10. If you make a profit of $20, then your profit in R multiples would 2R. I have touched on system expectancy in my April 2014 post. It is essentially the mean performance of a trading system in terms of R. Lets call this measure the System Quality Ratio or SQR for short.
I decided to calculate the SQR of two SPX 20 delta iron condor strategies which were back tested by DTR Trading. Both strategies managed winners at 50% of premium collected, but each had a different loss threshold. The first strategy had a loss threshold of 100% of premium collected. The second had a loss threshold of 200% of premium collected. The results are shown in the table below:
It was interesting to see that although the second strategy had a higher winning percentage (85% vs 79%), it had a lower SQR than the first strategy. The following are the guidelines to System Quality provided by Van Tharp in his Super Trader book:
Based on the above guidelines, the first strategy (which is the one we use to trade the XJO Mark III model portfolio) would be considered a good to excellent system. The second strategy would be considered very hard to trade. This is consistent with what I have found from trading different iron condor strategies in the past two years. Our first XJO model portfolio which had a very high winning percentage but large maximum loss was very hard to trade.
One of the trading strategies that I enjoyed trading was the Iron Butterfly. The SQR based on the actual trades in our XJO Iron Butterfly Model Portfolio was 0.49 which is excellent! However, this was based on 10 trades which not enough to be conclusive. Van Tharp suggests calculating the SQR with at least 30 trades.
Another point to note is that most trading systems perform well under certain market conditions. For example, we know Iron Condors do well when there is high IV i.e. when market is slightly bearish. They may not do as well when the IV is low i.e. when the market is very bullish. The SQR for our XJO Mark III model portfolio using the 8 completed actual trades was 0.37. This is higher than the 0.30 that we calculated based on the 100 trades done by DTR Trading using the same strategy on SPX. This could be because we were trading this model portfolio when IV was favourable.
I have found the SQR to be a very helpful measure for evaluating trading systems. It has helped reaffirm that my two favourite iron condor strategies (the Iron Butterfly and Mark III Iron Condor) are indeed high quality trading systems.
Note: Post updated on 23/3/2016 to replace SQN with SQR.