Here's one way of looking at the results and they are actually quite provocative.
Jeff was interested in what happens the following day and those results have to stand on their own as they are the complete inverse of the results here posted.
I was more interested in examining a slightly longer horizon with an eye towards collecting some short term premium by selling calls against rallies or puts against plunges.
Below are the results of my simple code for 5 years and 5 months (just an arbitrary look back period).
While the overall performance stats look pretty benign, things actually start to pick up around trade # 90, which was on 12/07.
Here's the TradeStation 2000i code for those that want to try this out on other ETFs/indices.
Be aware that current versions of TS require slight modifications to the entry/exit command language.
And here's the equity curve for the shortened study with the same Len1, Len2 settings of 2 & 6.
While more work is needed to make the system a more robust. . .perhaps a % change filter or MA crossover confirmation, these initial results certainly appear to merit additional study.
.
And here's where things start to get really interesting. . . . .
This is the same study, but the look back period has been abbreviated to start in 12/07.
This coincides with the high of the 10/02 - 11/07 bull run and the ushering in of the bearish paradigm.
While the long side chugs along at pretty much the same ho-hum rate as the original longer study, the short side results should grab your attention.
And here's the equity curve for the shortened study with the same Len1, Len2 settings of 2 & 6.
While more work is needed to make the system a more robust. . .perhaps a % change filter or MA crossover confirmation, these initial results certainly appear to merit additional study.
No comments:
Post a Comment