Hi everyone. I hope you’re all managing to make pips in this rough market. I’ve been rather busy lately so I haven’t had much time to update this blog. Among other things I’ve been rather busy working on and backtesting expert advisors (forex robots for the uninitiated.) I’ve been struggling to get my MT4 modeling quality beyond the max 90% capable with standard MT4 history data and I finally succeeded thanks to Dukascopy.
I found out that Dukascopy offers high quality tick data for free – thanks Dukascopy! After the rather extensive work converting the tick data to .CSV format and then that to the .FXT format necessary for the MT4 strategy tester you would not believe how joyous I felt when I first say the “99% modeling quality” when I ran my first backtest. Oh, by the way, I will post a “How-To” so you too can get this setup.
Now you may ask what does this have to do with Forex MegaDroid. Well, you see the first thing that came to mind when I finally got the history data all setup and ready for use was why not backtest one of my favorite EAs, which so happens to be Forex MegaDroid. This hard working EA has been consistently making money for me on two live accounts, but a bit of doubt hit me when someone posted on the Forex Nirvana forum a few rather nasty backtests of Forex MegaDroid trading from 2009.03.01 to 2010.03.01 (here is a link to the actual post).
I wanted to know if with better history data I’ll get the same results. Sadly though I could not get the Dukascopy backtest to run all the way to 2010 03 01 🙁 I don’t know whether it’s cause I did something wrong when I setup the Dukascopy tick data for MT4 or if it’s because of gaps in the tick data, but at best the backtest runs only from 2007 04 05 to 2008 03 05. I’m hoping that later on in the future I’ll be able to get more recent tick data working, and if I do I will backtest MegaDroid yet again and see if live account performance is equivalent to the backtest. But for now here is my backtest with proudly and painstakingly achieved 99% modeling quality:
Profit factor: 2.75
Expected payoff: 35.94
Maximal drawdown: 1753.29 (12.57%)
Relative drawdown: 12.57% (1753.29)
Short positions: (won %) 71 (94.37%)
Long positions :(won %) 60 (96.67%)
Profit trades: (% of total) 125 (95.42%)
Loss trades: (% of total) 6 (4.58%)
As you can see the results are very good but since I could not backtest the period showed in this forum post I can’t reassure you nor myself that MegaDroid won’t behave as in the backtests shown on that thread.
For now though live account performance continues to be satisfactory for me, but as you should know past performance does not guarantee similar future performance. Anyways, I’ll keep you up-to-date on the performance of my live Forex MegaDroid accounts. Wish me luck.
Time to wrap up this post. Thank you for your attention, and I wish you profitable trading, be it manual or (as in my case) with forex robots.
1 thought on “Forex Megadroid Backtest on Dukascopy Tick Data – 99% modeling quality”
Comments are closed.