Traders' Roundtable

I have hourly data for the Major FX instruments and I have attempted to backtest from 2001-2008 the system that Leonardo is using on his blog. I have done this on an excel spreadsheet as I do not have the programming skills to do it on any other platform. It has been quite challenging for me.

My initial results are not encouraging. I have not used any position sizing algorithm whatsoever (i.e. when i look at the P&L I am simply looking at the difference between entry and exit prices for the whole period) and I think that this is the reason i have a loss at the end of the period.

Can anyone offer their views on this test? Personally I think that for an OVB system position sizing is paramount.... I'm going to continue the test and incorporate this idea. I will try to normalize all positions by using 1% of capital (fractional capital technique) max risk on every position and call this R similar to Tharp's ideas.

This experiment seems to suggest that without position sizing any trading system is bound to fail.... thoughts and comments?

I suspect that the leading expert on the performance of leonardo's system, is probably leonardo himself. Why not ask him whether your backtests are getting generally the right answers, and whether or not his trading system is bound to fail without position sizing. He might spot an error in your work that's obvious to him but invisible to people unfamiliar with his system.

I'm not a forex trader but I think I've seen copies of Futures Truth magazine which test vendor-sold forex trading systems without position sizing and find them profitable. Maybe you could call up their research director, George Pruitt, and have a chat. (link)