Tweet
Is there a way to change the parameters on the backtesting scenario in terms of money management? For instance, I want to test a strategy against single indices and equities however, e-signal presumes that I'm not investing the entire amount on a single position and the performance returned is thus not true performance, but solely based on the programs version of money management.
-
-
The language on my original question may have been confusing and I waited a bit too long to edit, so I'll give this another go. My question regarding backtesting is specifically... How can I test a strategy solely on it's merits looking at an individual index or equity?
Currently, if I attempt to backtest within e-signal, the assumption is that I'm buying 1000 shares from within a $100,000 portfolio. Using this methodology, you cannot truly measure the alpha generated by the trading strategy against a buy and hold strategy or virtually any other strategy.
For example, my strategy signals me long at 1069.48 (12/2/03 $spx) and out at 1134.17 (1/30/04). The real return on that trade should be +6.049%, e-signal reports it as 136.28% and causes all of the other strategy related data to be misreported. I’ve gotten around this by typing all of the individual trades into an excel spreadsheet and calculating the numbers myself, but this largely defeats the purpose of the program. Hopefully this is articulated better than my last question. Any thoughts. -
I have the same related issue, as i have a portfolio rating system that when evaluating trades it rewrites certain weightings. It also uses a product substitution algorithim to decide which trades to actually take in a portfolio. Hence we use vba to rank this.
Is there an easy (!??) way to get the trades generated from running a strategy in Esignal, into excel so that we can then rank them and truely test them from a portfolio point of view?Comment
Comment