I haven’t given up on them forever. To be honest my plan is to let Nick and Metalhawk do the testing and then start trading it again when their results start showing positive.
I figure it’s a better use of my time to let them fix the problem.
I have seen Metalhawk’s results, but I’m scared that because of the low sample size (about 1 year) it might just be curve fitting and have no bearing on the future. You have to be cautious of optimization when it’s over a short time period, which is why I threw in that disclaimer about my new Sunday breakout EUR/JPY results.
Anyway, After Nick and MH settle on some new rules and use them successfully for a few months I’ll start back up.
I’ve got plenty of my own strategies to trade in the meantime.
Phil, thanks for sharing this, I am checking it out and starting my run of backtesting! IBFX just moved into 5 decimals, so will adjust the indicator (guess it is just moving to 100 instead of 10) in the coding.
Hi Phil, I ran through EJ from July 08 based on old GU rules. Good results (was too lazy to detail the exact profits though :p).
But I observed a fair bit of correlation between EJ and GU. It is probably not a cross correlation, but a volatility correlation that is causing the losses i.e. markets are volatile, all crosses are volatile, and stops get hit.
This is seen in the rather similar losses for both GU and EJ in 21/7/08-28/7 weeks, 18/8/08 week, 22/9/08, 13/10/08, 09/2/09 to 09/3/09 (the century’s disaster week) weeks, 15/6 to 6/7 weeks.
Thanks for this Phil. It doesn’t get better taking time going through manually week by week to gain confidence and knowing there is no way to make a quick buck! Your indicator helped a big deal!!
Yes. I’ve done some research into this and for a statistical system like this one minor holidays have little to no effect. The only holidays that I won’t trade it are the big ones where major bank traders take entire weeks off, like Christmas.
After I read that I went and compared EJ with my new rules to GU with the old rules. For 2009 there’s only a 52% correlation between them with the different rule sets.
I’d have to test it further, but that might be enough difference for me to consider trading a full lot size on each pair.
What if we could develop slightly different rules for about 5 different pairs and then trade .5% risk on each trade?? Each rule set will be profitable, but different enough so we don’t hit all losers on all 5 pairs at the same time… At .5% risk x 2 trades per pair x 5 pairs the max we could lose in a week would be 5%, but this would be very rare at a 50% correlation rate. That lets us still follow good money management rules, but ups the profit potential of the system tremendiously.
It’s just a thought… Plus it’s assuming that different rules can be profitable (and not too correlated) over a number of years. That might turn out to be false anyway.
Here is one thought phil, it might interest u. as u said u increase/decrease ur trade size after every profit/loss to keep it @ 2% level.
when i back tested diff systems… i found dat most of d systems go for a loosing streak only for few weeks or months… so when v find such a streak… for a few weeks, increase ur trade size…(d increase which v wanted to do after every profit) or v keep on increasing after every loss and decrease d lot size a bit after every profit… this adds up d annual profit considerabley…though a bit risky in d way… dat if its d time for d system for fail completely … it will die sooner… but for most of d time in between this reverse strategy increase d over all profit.
2nd … u said u will stop trading this system lets suppose if it gives u 10 loosing trades in a row… i was wondering dat since 2005 u dint have any 10 straight loosing trades?
I liked ur idea about Nick B system… dat v should not through it away… just wait and c.
I traded ur sunday system first time last week, and luckily I placed my short trade according to d fxcm candle and it was a straight profit of 240 pips:) or something… though very lil lot size as i had not read ur system in detail at dat time.
Nope, the worse losing streak I’ve had so far is 7 in a row.
And I wouldn’t even consider that a true losing streak because some of the other trades on here had winners during that period. That means the system’s not really losing, but that I just got unlucky and got stopped out by a pip or two.
Hey I’m new to the community (I saw this particular thread and it made me join… Well done so far! This looks incredible).
It’s coming to the time that you’d suggest buying in. I was wondering if you could fully explain all that’s happening… Currently my graph looks like the graph attached and I have no idea what each line means. Could you help a newbie?
Those are the lines from last week, at 1 second after the hour the lines on the indicator should update to this week’s values, then they will probably be a little easier to understand because they will line up with the Sunday candle.
Basically the lines are marking your stoploss, entry, set to breakeven point, and takeprofit… in that order. You’ll have two sets of lines, blue for long trades and red for short trades.
I wish I had time to take some screenshots and label them for you, but it’s only a few minutes till we start trading… If you’re still confused after the lines update let me know and I’ll post a picture for you.
I had a bit of a problem trying to trade GBP/USD and EUR/JPY this week. Too much math in too little of a time span, plus I’m already in the GBP/USD long trade due to an indicator screwup.
I hope it goes up!
The good news is that I might have a functional EA for this. I built one using S05’s indicator as a base. It entered the trade with a 1 pip buffer instead of 10 due to a decimal problem, but that’s easily fixable.
If not can someone please post a labeled screenshot? I’d do it but my charts are covered with lines from the EA I’m testing and there’s no way to get rid of them.
They’ll just confuse the heck out of her if I try and make a screenshot…
I’ll post it after this week’s trades end. I can’t recompile it (I found another minor bug) while it’s running or it will reset everything.
I also want to add a timezone offset for those people not using a GMT broker.
Also, it’s not 100% automated. You have to shut it off manually once you get to the point you don’t want the second trade to happen, but I don’t think that’s a big deal.
