Wednesday, May 23, 2012

Weekly Meantrades

The past few weeks have been nothing short of frustrating. Automation with constant range bars on Metatrader 4 seems futile. With a single disconnect from the brokers server, the constant range bars are immediately scrambled and thus any reversal patterns are suspect. Somehow this had me thinking about how I can get this very decent system called meantrades running without a hitch. I came up with a version which uses 240 minute and 60 minute bars.

The only major difference (there is one minor variation however) between this system and the constant range bars version is that it is restricted to only 1 profitable trade a week. "Profitable trade" as in it will enter up to 2 unprofitable trades in the same direction per week. This is similar to the intraday meantrades which only takes two consecutive unprofitable intra-day trades in the same direction.

The results thus far are quite respectable. Considering the time invested versus return on capital, the weekly version of meantrades is a better deal for smoothing out the all important equity curve.

I realize these numbers probably seem absurdly high and I am just as skeptical as anyone else would be. But it's been backtested for several years and all trades are based on a closed bar basis so it is pretty hard to fudge the results.

Currently the XAU/USD weekly version of meantrades is running at 75% for the year with a total of 698 pips employing a trailing stop. With a fixed 100 pip target it's coming in at 871 pips. Let's say we are in June already, since we cannot take another trade until next week, that is a very tough return to beat of 145 pips per month trading no more than 1 winning trade per week.

The results for the EUR/USD market are also quite respectable. Coming in at 69% for the year trading only once a week the supertrend exit works out to 963 pips thus far while a fixed exit of 63 pips has yielded a return of 335 pips. Winning percentage is based on a total of 26 trades so far this year.

I initially set out on this discovery using the AUD/USD. My thinking was that somehow if I ever got around to getting meantrades automated I should probably try it out first on some less volatile pairs. I have never traded AUD/USD seriously as it seems rather dull market. In fact it's almost as profitable as any other market I looked at. The results came in as follows: 70% winners out of 27 trades taken. Most impressive are the drawdowns to achieve profitability. The market just seems to trend better than others when applying the lens of meantrades. Average losing trade was 67 pips while the average winning trade was 55 pips. This is the only serious issue for me trading less frequently. The average winner is normally larger than the average loser in a robust system. However when you are applying a trailing stop methodology for exits, the out sized returns of just a handful (or even a single yearly trade) can outpace the averages quite easily. As the Black Swan movement has attested to, the market does not reward those who seek the average return on equity as a benchmark, it obliterates them every few years with massive crashes and bubbles. I'm not a fund manager and I'm not seeking an average return so I will keep doing what I'm doing until something comes along which makes more sense.

For now, meantrades with trailing stops seems like the best way for me to trade these volatile and incredibly profitable markets for the foreseeable future.

Some of the inner workings for setting up the weekly meantrades method:

  • The ratio levels need to be set off of the first 4 hour bar of the week or the second bar. This is slightly subjective in that I prefer to use the smaller of these two bars, unless it is just too small a range. In that case I will use the first 4 hour bar regardless of how large it is.
  • Stops are usually an hourly close beyond the swing high of the trade entry. However in some cases this bar closes inside the Keltner bands (even though it is outside the highest high or lowest low for the trade setup). It is best not to take the loss until price has also violated the Keltner bands. There have been many cases where price makes a minor new high or low inside the Keltner bands only to reverse immediately. 
Here is the only trade currently open with Weekly meantrades. An EUR/USD short currently sitting at +110pips. Stops are already at break even since price has already touched the opposite Keltner band from where the trade originated (as per the original rules of meantrades).