JT - Been looking at the gbp using your system for the euro. Has worked everyday for the past week with one condition. That is a limit on it, if it hits 80 points im out for the day. This would have returned 80 points a day for the past 7 days. Thats 560 points!!! I haven't further backtested this but the past 7 days data is outstanding. Lets see if it holds next week. If it holds on Monday with the same limit condition, im in on Tuesday!
Neil - can i ask why you went long at 10am? Seems like a really good decision as now your up 100 points.
If i had gone long it would have been at the double bottom so i would have gone long at about 12:15 as i mentioned in the euro-gbp where next thread.
... and we aren't happy!
Forgot to remove the auto stop on CS's order, so got stopped out at -120. So have reentered with another short to simulate to the end of day - well, if you're going to make a loss, may as well make it a big one - something to tell your mates about in the pub eh!
Anyway, it might recover something in the next couple of hours ...
Looks like we have our third incidence of 2 losers in a row!
Still, averaging 26 pips a day only means 130 a week, which is pretty much where we are since Tuesday, so actually it's an "average" week.
But pmaa, I agree - we must look further at tweaking eh?
I have to say I'm struggling a bit with the concept of just letting a single trade run all day - part of me likes it because I can set and forget but part of me doesn't like the fact that you only need one "bad" announcement and a few days hard work can be undone ... particularly as the US day (and I assume announcementsand such like that can influence the rest of the trading day) doesn't really start until UK lunchtime. There feels like there can be fundamental changes over those few hours that could be taken into account ...
I know JonnyT says don't use stops, but maybe there's something in looking at a reversal rather than just a stop so that if there's a large reverse trend for the rest of the day you've at least limited damage or at best actually still made something.
Or perhaps there's another strategy to look at - similar but reassessing on the 1-3pm bars maybe, or around that time?
Aaah, so many systems to try ... who has time to code and back test all these! I can feel a "working" holiday coming on!
Yes your data looks about right give or take a few pips.
I've been following it on a CS simulation account too, as well as on my spreadsheet to confirm since Feb 20th - am confident with my results as I'm usually within one or two pips of the actuals on CS on excel.
The data I got from JonnyT's other excel file (the 7-7:05am one) and then just rewrote the VBA to simulate this strategy.
Jan netted about 800 pips, Feb about 200, tried some other data from Dec, and it only made 140 pips. But I'm a little dubious about that data. I also tried it with data from earlier in the year and it lost, but again, I've not had time to really examine the data.
So far March 04 is about 400 (exc. Friday)
I've tweaked it a bit ... it seems there are a few where the trend hasn't really caught until a bit after 10am, and then you get a catastrophic loss as it turns the other way and stays that way all day.
However, lengthening the timeframe to even just 10:30 makes the overall profit much less, so better to leave it at 7-10.
2 alternatives then ... put a stop loss in or do a stop and reverse (SAR)
Tried the stoploss first - putting it at 85 pips seems optimal to stop the really big losers without preventing the natural big swings intraday, but I'm a bit suspicious about "over-optimising", especially on only 2 or 3 months data.
The alternative is the do a SAR, which is optimal at about 70 pips. I.e. once you've lost 80 pips in the day, exit and reverse and take what you can, given that the trend has really gone against you earlier in the day. I found that this elicited a few more pips overall with the data we have. In one trade it made it worse (i.e. took 70 loss, then another 60 odd loss I think as it swung round again), but in others it really helped (one took the full hit end almost completely cancelled it out by the end of the day).
Bottom line ... it seems to work at the moment - Jan was a fantastic month, Feb OK (but hey, isn't 50 pips a week OK?), Dec looks disappointing (but have to check data) and March is more promising again.
I've just started trading this live today on CS ... wouldn't you know it ... 30 pips down at the moment, but the key to this is just set and FORGET! Particularly with forex intraday swings ... it's a lively old place out there!
Questions for everyone:-
Has anyone got some data, maybe 5 min data that we could analyse that would be really useful ... My 1 min data is fine all well and good, but you only get about 3 months into 65000 excel rows!
Am I in danger of "over-optimising" with the SAR value or stop loss values? If so, how do you develop a strategy that needs a figure without using backtesting to calculate the optimum?
Has anyone else been following this one? If so, how are you doing?
Is anyone trading it live?
Has anyone else taken this any further with refinements?
Thanks again to JonnyT for this one. If this is one he throws out for general consumption that's doing so well, what must he be actually trading live with? ... now if we can just persuade him to post THAT system ...
I have also been following JT system with great interest,JT addresses the problem of stops and states that a stop loss will only hurt this system ,maybe JT could comment on this ,but all very interesting none the less ,i would be interested to know if anyone has the data ,what was the biggest loss in a day and also for a month .many thanks .
As I mentioned in the other thread, I've done Dec twice now, and am still not sure about it!
Have a look and let's see what we can put together and share.
I know somewhere on eurostoxx50 there's a set of massive files someone's already prepared from Jan 01 to Sept 03 ... you can extract a goodly amount from that, but the last few months are missing.
Each day since Feb 20th I've downloaded, extracted and added to the spreadsheet, so I have the data from Jan5th to present day, so don't bother doing anything with that. I'll post it when I get time in the next day or two.
That's the data I've been running on Capitalspreads.com simulation account and getting very accurate correlation with pips won or lost, so I'm quite confident of it.
NOTE The data on forexite is 1 hour ahead of GMT, Jonny's spreadsheet uses US data 5 hours behind(?) GMT, so I make it that we have to subtract 6 hours from the forexite data to get the US times for the spreadsheet.
Does anyone have time to write a bit of VBA code to concatenate the 1min massive files into 5 minutes or something like that. Shouldn't be too hard - I just don't have time at the moment.
Probs best in Access, as you have virtually unlimited number of records, and the 2 1/2 year files have 750,000 records in them!
By the way ... don't count on today! I didn't trade today because I decided to start trading this LIVE on Monday and it's record so far this week has been ... well ... ! But just looking at the hourly data, it would've gone short some time between 2-3pm, and now look at it - would be 50 down (7:30pm).
So by 9pm, that might be another 3 in a row losses, a grand total of 220 down this week. Glad I didn't trade it today!
My slightly modified version (unlimited profit target, 85 stop loss, 3 pips spreadloss per trade) results:-
Mar (to date, roughly) +189
Biggest drawdown before this week was 4 in a row in Feb, -220
Now this week (if today stays about -50) -240, 3 in a row
Let's hope something nice happens tomorrow!
The trend is supposed to be our friend