Question:
Do you know of any way to backtest without buying software, a trading plan, or a system?
Larry from Nomoneyville
Answer:
Larry, I suppose you could manually overlay your strategic data over select timeframes and see how it would have performed. I guess you could do that. If you are a software writer, you could write your own, I guess. If you know someone who is a software writer, you could ask that person to design a backtesting program for you, I suppose. Other than these suggestions, the only other way I know is to buy software or find a trading platform that provides backtesting as part of the package.
Larry’s question raises an interesting question—does backtesting really work? Think about it for a moment. If you have a current trading strategy that would have produced profit, say, in a 1999 timeframe, does it follow, necessarily, that it will produce profit in today’s trading environment?
I am of a mixed mind on backtesting. Logically, defining what a current strategy might have done in a particular past environment to extrapolate what it might do today seems, well, inconclusive, at best. All markets are specific unto themselves and specific unto the environment in which they exist. One cannot discount or ignore the specific influences that affect both the market and the market environment in any timeframe. So, in order for backtesting to be valid, the market as it existed in that former timeframe and the influences that affected that market would have to be similar to a great degree, no?
I don’t know, maybe I’m all wet here, but I suspect if I am, someone out there will let me know. Just keep in mind this caveat that appears on just about every investment advertisement—past results do not guarantee future returns. Why do they say that?
Trade in the day; invest in your life …
