I was also wondering, what time frame did you run it at? Was it 00GMT...if so i wonder what it would say if you ran it at 23:45GMT and 22:00 GMT, so we can compare it to our forward testing results posted in the thread.

Thanx,

Graham

In the interest of curiousity, i entered in the trades based on what the XBot spit out from Dec 18th to April 27th. We had 2 3rd level trades (2 consecutive losses). Outside of that, a total of 180 trades and using $10/pip lots on FXDD and using the 7x method proposed by Timmy on the parent site, i show a yield on a $100,000 account of almost $20,000... I did my best to retain accuracy in transposing the profitable and losing trades and overlaying them with the results based on the core system.

UPDATE: I thought it was good that out of 90 trading days, only twice did it have 2 consecutive losses. That's 2.22%. If our average is 85% accuracy, then the odds of that should be 15% * 15% = 2.25%.... not bad, eh?

Enclosed is my spreadsheet showing the results.

Enjoy,

Graham

Attached Files

xbot text.zip (4.8 KB, 107 views)

Hi Robert
You still alive I did try to contact you a few times.

I thought you where dead.
But it is nice to know you’re still alive and kicking.

Did you move into the mountains?

Take care

Smeden

Ok, i was able to use the v1.1 on fxdd to run a backtest of the EurUsd on the 1 hour chart back to Jan 2005. I believe the timeframe is 2pm, as on FXDD the 2pm trades are executed at 00:00 log time and the summary contains trades at 00:00 times. I also kept the original 50 s/l and 10tp
I enclosed the original .htm summary as well as the modified spreadsheet including martingale changes. Here are the stats i discovered:

Of the total 340 trading days and 680 trades we had:

82.3% winning trades, 17.7% loss trades
1.6% 2 losses in a row (11)
0.4% 3 in a row (3)
0.14% 4 in a row (1)

and a yield of $66600 based on $10 lots, or 6660 pips.


Running the script with 5TP and 2 lots, immediately the results change:
88.4% winning, 11.6% loss
1.3% 2 losses in a row (9)
and no 3 or more losses in a row.

and yield is $66900 based on $10 lots, or 6690 pips.

I guess MoneyQuest was onto something with these results. Hmm...

Now for the progression of martingale multipliers, the 10TP can weather a deeper loss, however it seems that with the 2lot 5TP, the same test virtually removed the any loss deeper than 2. Good thing cuz here is my calculated progression.

1 Lot 10TP: 1, 7, 42, 253, 1519
2 Lot 5TP: 2, 24, 266

I chose these numbers because by doing them you recoup all losses, plus make your 10 pips a day still. The reason the 2lot 5tp progresses so much faster is because the differential between the TPs and SLs makes it progress exponentially.

Enclosed is the original summary html, as well as the spreadsheet i modified. Comments?

Graham

UPDATE:
For information sake, having been able to get the EA somewhat working, i decided to see what the accuracy % was on the other pairs:
Here ya go:
GbpUsd 10TP 1Lot = 81.3% 5TP 2Lot = 87.3%
UsdJpy 10TP 1Lot = 83.6% 5TP 2Lot = 89.4%
UsdChf 10TP 1Lot = 82.1% 5TP 2Lot = 89.3%
EurJpy 12TP 1Lot = 76.7% 5TP 2Lot = 86%
GbpJpy 15TP 1Lot =66.7% 10TP 1Lot = 71% 5TP 2Lot = untestable
EurGbp 10TP 1Lot = 66% 5TP 2Lot = 79.6%

Now i havent, nor do i plan on manually doing martingale to get yields here, esp considering i don't know for sure how accurate this EA calculation is. It takes too long to do manually on 12 more summaries. It does seem to suggest though that for the major US pairs above and possibly the EurJpy, we sure have consistant longterm consistancy that lines up with numbers we have discussed. The higher the %, the better the chance of avoiding losses in a row.

Enjoy

Attached Files

2200GMT FXDD 16month results.zip (198.8 KB, 167 views)

Hi Graham,
Nice work. Have you included spread in your calculations?
So is the 10 pip T/P actually 13 pips?
Mike4X.

All numbers always include spread. FXDD does have a 2pip spread on the EurUsd, so i guess results could be different on other platforms, as i have noticed even by my demo results on FXDD and my live trading on Interbank. Both are more/less the same, but FXDD has lower spreads on alot of the major pairs.

When i do my calculations, i try to make it net, cuz i don't like hidden stuff. The only thing i never included was interest, but that shouldn't be much of a problem as interest is low on the majors, and almost all trades are closed in under 24 hours. The only pair i would include that this might be an issue is the eurgbp as the interest is bigger and most trades are open up to 72 hours.

Graham

hi Robert...

Can i try your Xbot robot?

Thank you

Do you anticipate differences between platform with regards to profitability
of this system? Which will be the best platform? Oanda have continuous
interest system. How may that affect the profit of this system.
Any imput will be appreciated.

Personally i don't plan on trading this with anything other than a MetaTrader-compatible broker. So this would be Interbank, FXDD, Northern Finance(sp?), etc... We hope to get an MQ4 script designed here soon.

As for "continuous interest", the trades at the time go in just after daily interest is calculated (22:00GMT) or 2pm PST for me, so that gives those trades a full 24 hours to complete before interest is calculated, otherwise if they close early, it is "free", in that there was no interest. The other timeframe 00GMT gives us 22hours, which is still more than enough time, as most trades close in less than 18 hours.

Hope that helps,
Graham

Thank you. Another question. To trade this system it appears one would need an account with a lot of "cushion" to withstand the rare instances of consecutive losses. From your experience, assuming one is trading all 7 pairs
with TP 10, SL 50 at 00GMT, what would be a reasonable account balance assuming such account is exclusively for this system for say trading
a) 1 lot ie $1000
b) 1 minilot $100
c) 1 microlot S10?

Also do you see any advantages to trading multiple pairs?