|
PureBytes Links
Trading Reference Links
|
>Continuous contracts aren't "real," but they're a good approximation
>for testing purposes.
I trade like I test and I test like I trade....exactly. The results of my
backtesting on a back-adjusted (or forward adjusted, doesn't matter) series
simulates reality exactly, since I roll in real life on the same day that
back-adjusted series roll. The only difference is actual slippage vs.
estimated slippage. What I'm saying is if I had traded the backtested system
in real time during the backtest period, my results would exactly match
backtesting, except for slippage and error in exection.
This all works because all my signals depend on relative price changes, not
absolute prices. Negative prices are fine. The day to day marked-to-market
changes in equity are based on price *changes*, not absolute price.
Scott Hoffman
Issaquah, WA
|