acrary, if you think this is not the right medium then please post it in the right medium. although i disagree that this is not the right medium - you just need to ignore the noise and reply only to constructive posts.
Problem - how can you tell that an edge is not just "previous market character", given that all your testing and assumptions are based on past market behaviour?
Have you included slippage & commissions? According to my test, this system didn't made any money in years 1994 to 2000 with $30 /RT slippage&commish. Without adding slippage, only commissions at $5 /RT -your system is making money every year, except 2003. I had similar results to wdbaker's test using data from 7 Dec 2000 to 12 Jun 2002.
Well, it's 2005 and I'm curious how this trading system is doing (see first post, and the Wealth Lab posts on page 9). My backtesting computer is down and will be for the week, so I wonder if someone could test it and post the results here. I'm willing to bet that the system went ka-poop since the end of 2002...
i think it stopped working a year ago. tested well on a diversified portfolio, but dried out, probably due to vola crunch in financial futures ...
I saw this thread today and tested the system on the swiss market. it performed quite well in the smi future although the volatility is quite poor. more testing tomorrow.