Following up on my my post last Monday of the results of my portfolio of EA's in demo testing, here is an update including the figures for last week:
So the EA that looked really good up until last week - the BRI - had an awful week last week, losing almost all the pips it had gained over the previous 4 weeks. Of itself that's not a concern to me, but what is a concern is that this model tested with a max drawdown of 9.5% over 3 years of back-testing, and the above loss now represents almost exactly a 10% drawdown, so its reached its drawdown limit, within 5 weeks of demo testing. That doesn't seem right so it requires further investigation from me, but at the very least I have decided this model won't go live with the others at the end of the month, and won't until I decided that it re-validates itself.
A 2nd EA (MAV) also did not have a good week, but in context of its back-testing statistics, it remains well within acceptable parameters. The other 3 EA's are performing well and despite the BRI performance the portfolio remains over 4000 pips in profit for 5 weeks, so still ok from a portfolio perspective. What it does do is imo validate the portfolio approach and not relying on a single model/EA.