CHRISTMAS PROMOTION
LTOptions at a 33% discount during the Year End Holidays.
Tell me More

Sunday, January 5, 2014

Analysis of Forex trading results in 2013

I didn't trade Forex with real money in 2013. I must admit, so far, it is the Forex arena where I find the toughest challenges and where I haven't been able to come up with a consistently profitable method. I designed the LT Trend sniper system back in December 2012 and had it in demo mode for all of 2013. The results were not good.

As a refresher, LT Trend Sniper buys EURUSD at the open the next day after a 70 day new closing high was made. Once in the position the system will ride the trend and will only exit when a new 8 day closing low is made. Or if during 8 consecutive days no new high is reached. That means the momentum has stalled. The Stop Loss level is calculated dynamically based on twice the value of the Average True Range (ATR), and the position size is also dynamic based on the stop loss distance, so that the same specified risk per trade is maintained across trades. For short positions, the logic is the same. After a new 70 day closing low, the system sells EURUSD and will attempt to ride the trend for as long as possible. The exit will be made on a new 8 day closing high, or if 8 consecutive days go by without a new low being made.

The performance for the year was -9.16% which was totally unexpected and very far from the historical results of the system. However, the worst case scenario, calculated with Monte Carlo simulations, suggests that -17.54% is the maximum admissible draw down, that if reached would mean it is not safe to trade the strategy anymore because chances are, the inefficiency that it exploits has dissipated. So, although results were not good, it cannot be said that the system "doesn't work". Not yet. I'll leave it in demo mode for 2014 as well and I would expect to see positive returns. If the worst draw-down scenario takes place, then I will forget about this system.

2013 has been in general a very tough year for Forex algorithmic trading systems. In fact the last three years: 2011, 2012 and 2013 have been very tough years for automated Forex trading at the institutional level as evidenced by the Barclay Systematic Traders Index with results of -3.83%, -3.20% and -2.25% respectively. I would expect an eventual regression to the mean of these results.

I would like to point out however, that discretionary professional Forex traders did better in 2013 than the automated systems. Looking at the Barclay Discretionary Traders Index, we can see the results for 2013 was -0.33% vs -3.83% for automated trading. The result the last three years for manual professional Forex trading is +2.75%, +1.02% and -0.33%. As I have discussed before on this site this doesn't mean that Automated Forex trading doesn't work. 25 years of results tracking at the professional level reflect that both methods, manual an automated have been able to return a compound +8% per year. In my case, I only want to deal with mechanical approaches when it comes to Forex as I have no energy or time to deal with the 24 hour a day Forex madness.

Hopefully I come up with new ideas in 2013 that allow me to find a profitable path trading currencies. Until them, it is necessary to be patient and not to trade without a rock solid proven edge.

Related Articles:
Analysis of Options Trading results in 2013
Analysis of Investing results in 2013
LT Trend Sniper - a Forex strategy that works
The Turtles Trading system automated (Expert Advisor for download)


Go to the bottom of this page in order to see the Legal Stuff

2 comments:

  1. Hi LT,

    Interesting strategy, inspired by the Turtles? My investment strategy for stocks is very similar, though a bit more elaborate.

    Why exactly did you choose 70-day high and 8 day low? and how much is the ATR in comparison?

    Regards,

    boersianer

    ReplyDelete
  2. Hey boersianer,

    Well I ran several simulations and the 70 donchian breakout with the 8 day bar low/high exit yielded the less variable drawdowns through the years. So I went with that one. As for inspired by the Turtles, I don't know, maybe. But there are many important differences. The Turtles would scale in as positions moved in their favor, which this system doesn't do. Also the Turtles would leave at a 20 day low or high, but they didnt have this extra rule of "no progress". If for example in a long position, my system exists not only if there is a new 8 day low while in the position but also if there is no new 70 day high after 8 consecutive bars. That ls, if 8 consecutive bars go by and there is no progress in favor of the position. That's a rule the Turtles didn't have. The typical ATR is simply the ATR(20), that is the average true range of the last 20 days. You can add the ATR indicator in a daily timeframe to one of your charts and you will see the magnitude of those numbers.

    Regards,
    LT

    ReplyDelete