Hi phil, i’ve got pre 2004 records on forex.com mt4. i’m trying to go through them, but one thing - how did you estimate the ATR(14) when you backtest?
btw, the 2004 results in your excel are hair-raising! And my initial entries in 2001 are too. I think it the discretionary none-entry saves a lot of skin, so I want to examine and perhaps refine it.
will report results.
What do you mean by hair-raising? is it good or bad?
For the weekly ATR(14) I used the ART indicator that comes with MT4. Because the value on the indicator is the ATR at the close of Friday, and we actually want the ATR at 0000GMT on Monday I rounded all the numbers down to the previous “25” level… Meaning 238 would round down to 225, 269 down to 250, 295 down to 275, etc…
I figured that would give me a value pretty close to the Monday morning ATR, because the Sunday candle is always small and slightly lowers the value from what it was on Friday.
I’ll be excited to see the values from pre Oct '04. I didn’t have any data from before then and I’d like to know how the system performs that far back. 
Bad. My backtesting started in Oct 2004 because that’s as far back as my charts would go. The system lost money from Oct '04 until Feb '05. In the two months after that it recovered all the losses and as been making money ever sense, but we still don’t know what happened in early 2004 because I’ve never had the data.
But apparently Pirateboy found it somewhere so we will know shortly!! 
I got the data from ODL’s MT4 (not forex.com, sorry), whcih gives from 4/2001 (but there is a data gap in 2007-2008 in ODL’s system). Phil, I’m very cautious about considering them as comparable to yours as I am not using the same technique as you in your backtesting from oct 04 onwards, and just missing one good run results in a big difference in results. In any case, I will spend time back testing, once I figure out the ATR(14) as you explained below.
Hi, did prelim analysis on april to dec 01. results are horrid, I shan’t disclose here 
BUT, the takeaway is that the BE/ATR ratio during that period was about 0.35. With TP at 0.5ATR, we are virtually home by the time we BE.
Whereas for 09, BE/ATR ratio is 0.22. It seems that BE may be too rigid if linked to sunday range, on the pretext that sunday range is a good indication of forward week’s volatility.
But perhaps we should link BE to ATR instead, and I feel perhaps about 0.2x. What do you think Phil.
Back testing is backbreaking and eye sight distructive…I’m tired…taking a break. need a robot to do this… 
At the same time, sample tests show me that using ODL, my aggregate results for 09 are similar to yours notwithstanding that our winners and losers were sometimes different.
I downloaded a copy of ODL’s MT4 and I can tell what the problem is.
ODL’s data feed is starting at 23:00 on Sunday, so it’s only showing 1 hour’s worth of Sunday data. Comparing my IBFX and ODL charts it look like the ODL Sunday data only has around 50-75% the range of the IBFX data.
Your similar results to mine in 2009 are because of different market conditions. 2009 has weekly ATRs between 450-900 pips, but 2001 only had a 250-275 pip range, The Sunday candles on ODL in 2009 are still pretty big, but those tiny 1H Sunday candles in 2001 killed the system!
That’s what’s causing the losses. You can’t trade a Sunday breakout system without Sunday data! 
Phil, you are very sharp! ODL’s result is similar to yours in 2009, because since 2009, the feeds include 4 hrs of sunday data…
only data pre 2009 was 1 hr data!
But the argument you mentioned, which is that ODL’s candles are tiny for 2001 - I didn’t think so…i think they are too big, that’s why they resulted in a large BE / ATR ratio…
I have also reconfirmed your 2008 results on Forex.com MT4. They are virtually the same except for:
I’ve had something come up and I probably won’t be around to post the trades tonight. Could one of you guys or gals that have been trading this a while post them after midnight GMT so the newbies know they’re getting the numbers right?
Hopefully the price won’t move too quickly and I can jump in when I get back home, even if I am a little late and lose a few pips off the trade…
If I don’t make it back in time that’ll be good for everyone else, because if I miss a trade it never fails to be a winner!
Phil, I will be around… I will post the numbers from my chart… IBFX…
Based on IBFX chart, following are the entries.
Long: 1.6457
SL = 1.6404
TP:1.6681
Short: 1.6394
SL = 1.6447
TP:1.6170
20:00 GMT candle High & Lows
H:1.6447
L:1.6404
Weekly ATR:448
Please check your charts before placing orders.
Short already triggerred for me in Daily fXCM.
Actually way past already.
Wonder how the h@#!l this happened so fast.
Hmmmmm, going to wait for Trevpick overnight system.
Got the short too, and the same numbers using Questrade.
Hello Folks,
I really liked the system… Im using IBFX and as long as I cant place and pending buy and sell orders at the sime time, Is there a way to make things easier??? like an expert advisor that opens the orders by itself???
Thanks!!
How about igmarkets?
ETA:
typo in the image, morale = moral
Close for a lost in the short end.
Is it the “win ratio” thread?
but thats not until 3-4 hours later?
Correct, but I set it up usually early since after ~10:30PM EDT it will stay in the range till ~2AM before breaking out.
And sleep till the morning.
