anyone have the reversion strategy for tradestation, that is being sold at eministrategies.com? charging 4 grand for this!!! anyone have it?? Liz Merrill
You see a website designed as a huge advert, run like hell. Not to mention that in their FAQ they have a Q:"What are futures?" ... lulz... They say they have a hypothetical track record but i'm betting you that if you look at all the Sells they will be on the High of a bar, and all the Buys at the Low. Very typical for people selling e-mini strats, however, we all know selling the spread isn't realistic unless you're an HFT.
There is software you can buy for about $1K that will build hundreds of such strategies in about one hour. No need to pay for a system that much. But they all fail in real trading because they are selected from a large population of candidate strategies to fir the data best.
My wild guess is he was referring to "Adaptrade builder". You'd have to have a very robust backtesting method for making any of those strats work though, as the fills can often be (as is usual with any advertised strat) unrealistic with such software. It would also take a significant amount of time to find anything worthwhile as that software is kinda slow.
There's an "advertorial" for something very similar called 'Genetic Builder' in the Apr-Jun '12 'FX Trader Magazine. (The online version is free to read after registration.) Regards, Peter
I tried this program and all it does is some really weird curve-fitting. I mean really weird stuff. For example, 99% of the strategies I got had one of the entry conditions in EL set to true and some really weird exit function associated with it. Obviously, this was an attempt to fit the exit to the price series. Then, each time I tried to build strategies I got different ones. Some failed oos, some passed oos, random stuff mostly.
I just saw the webinar. It's very interesting he uses 1/2 tick range bars for the reversion (trend reversal) strategy, but time or tick bars for the trend strategy. Also, he did not answer my question about it's performance in other symbols like the NQ ! A good strat should work with multiple symbols, right ? What does seem intriguing is his "no stop" approach...a daily loss limit is the only stop in place. This is likely the only reason his results look good. I'd love to back-test these with different daily loss limits....as well as to run both strats simultanious.