Honestly I have no idea what fancy systems you come up with that you think need to be proven as working for the last 25 years before they will work today. Markets go up, they go down, they go sideways. I just trade that with a mix of watching price and using quantitative filters to keep me out of certain markets. That's it.
But some do rely on forward/paper testing systems before committing money. Testing/systems are not necessarily bad as long as you're not simply permuting rubbish.
High frequency trading is the hot new thing right now, and every trader and his cousin will try to convince us that they have a guaranteed system that will "beat them at their own game". Yeah, sure, we are going to beat super fast computers connected directly to their liquidity providers via T3 network access lines with our oh-so-fast Pentium 4 computer located in some basement in Kansas City, from our E-trade account and with our DSL connection
Is that so? Take a look at these trading systems, then come back and tell me that a backtested trading system cannot perform for decades and decades, with the same exact original parameters: http://media.futuresmag.com/futuresmag/historical/Issues/2006/7/Documents/Top-10-Trading-Systems.pdf Note from the article : "... a trader who takes the time to correctly develop a trading plan with a solid money management overlay, has a much greater chance of success. Starting with a trading system that has done well throughout time is step one"
You seriously either have a comprehension problem or are desperately looking for validation of your ideas. Read all I have said here. I accept that backtesting is the way some people go, no problem with it, I don't do it and nor do I buy that is the only way to go. I asked why a system needs to test OK for 25 years before it will work today, I didn't say that it can't so why are you wasting your time providing links I won't bother looking at? It would appear you think you have found the Holy Grail of trading. Wonderful, why don't you run along and make some money instead of trying to convince people who have no interest in your ideas that yours is the only way? At the very least then you wouldn't be such a crashing bore.
I never said that, all I am saying is that trading a "system" immediately with real cash before testing it with historical data (to at least make sure it is not a losing system!) is insane. If you have a problem with that simple fact that's your problem. No, I did not find any holy grail, playing the odds and limiting risk is all we need to consistently extract money from the market, assuming our system has a real statistical edge to begin with. And again, there is no way of knowing if a system is profitable unless we test it with at least 10 or 15 years of historical data and under all kinds of market conditions (bear, bull, sideways). I am already making money with a long GBP/JPY trade I initiated this morning around 1 AM (EST) but thanks for the advice.