20 year back-test any idea why the profit curve just levels out?
-
Im pretty new to using fxdreema but not exactly new to trading or programming, so first up a huge congratulations and thank you to the creator of this site, it literally takes five minutes to do what would have taken me a week back when I started out.
So I've been working on a system with the overall aim of avoiding bad trades, accepting that losses will always happen and getting out of bad trades before they get too costly. No martingales, no grids, no chasing losses, just accepting the markets will bite you and that the market doesn't care one bit about any of us.
Now it all seems to be going fairly well with modest but consistent returns from the start of the back-testing in 1999 through till around September 2014, after that it's pretty much just going sideways and I'm failing to understand why. Has anyone else run into this issue? Has the market simply changed behavior since that time or am I just totally missing something else obvious?
If anyone is interested I can upload all the trades made, but it's a 20 year back-test so it would be many pages long.
-
@justmatt
I'm not sure, but one explanation could be trading costs (spread, commission, swap). If your backtest assumes today's standard trading cost to the history prices, it is essentially exploiting trading opportunities that, in fact, were never possible.
When the test approaches this day, the assumed trading costs also approach the actual costs, and "profit" disappears. -
Ahh I hadn't thought of that. A 2 pip spread these days seems reasonable for basic back-testing, but yeah things were a tad different back then. I might see what effect a bigger spread has on it.
I'm currently retesting it at higher time frames and so far seem to be getting more of the expected gentle upwards waveform so you may have hit the nail on the head.
Thanks
