London Opening Range Breakout

tomorton

Guest Author
Messages
8,839
Likes
1,574
Can't resist any longer, I've got to go back to this system and try it again.

Obvious, old and simple rules.
Use H1 chart, wait for 0700-0800 to print.
Set buy at high with stop at low. Set take profit order at r:r of 1:1.
and:
Set sell at low with stop at high. Set take profit order at r:r of 1:1.
(do allow for spreads when setting orders)

Targets -
Major forex pairs - start with the biggest, avoid AUD, JPY and NZD
Major indices - Dow or S&P or Nasdaq100, FTSE100 (cancel untriggered orders on FTSE by 12)
Gold
Brent

Trades to avoid -
pass by if 0700-0800 bar several times larger range than usual for this time of day

Bit of practice this week was encouraging - 2 out of 2 yesterday, 7 out of 10 today. Looking back few months on charts suggests overall win rate of around 2 to 1.

Planning a nice weekend and get on with trialling this again next week.
 
Initial review of results so far is not anything like as good as I had hoped. Just about random, though slightly positive. However, watching how price actually behaved, I now think I understand why chart look-back results were so much better and what I might do about it to continue the trial next week.

6 days of trial
103 trades
89 trades closed at pre-set SL or TP
Of which, 48 winners, 41 losers, giving a win rate of only 53%
Winning days = 3; losing days = 2; draw days = 1

Several factors have downgraded my win rate during actual trading, including -

1. fast moving prices at 0800 - obviously, its impossible to instantaneously enter 2 dozen orders at exactly 0800. Initially, where price exited the 0700-0800 range with a lot of energy I was shying away from entering until it pulled back, and many of them didn't so I missed these excellent opportunities for quick wins.

2. margin - not being used to daytrading I had actually not considered this. When looking back over charts, I had not allowed that the account would run out of margin for additional trades. Therefore the pool of trades taken was smaller than in the look-back, and those taken were biased towards triggers early in the session. This would bias the trial sample pool of trades and undermine the statistical parallel. (Its also very irritating when you see 5 entry orders trigger all at once and all fail to execute because of insufficient free margin.)

3. news events - the look-back did not allow for these events (or absence of). Keeping a much closer eye on news releases in live trading.

Also hoping to collect more evidence on day of the week as a factor.
 
Additional observations re winners and losers.

Of the 48 winners which hit their TP, only 16% had initially gone halfway to their SL.
Of the 41 losers which hit their SL, only 17% had initially gone halfway to their TP.

It starts to look like what people anecdotally report with short-term trades - the losers are in the red all the way from the point of entry, while the winners are winners all the way.
 
A challenging day running these trades today.

5 winners, 10 losers, which is at least better than last Monday. The financial result was break-even as now using half the depth of the set-up bar's range as the SL, and this immediately moves r:r to 1:2.

But hoping for better tomorrow.
 
Today was another challenging experience. 4 winners, 18 losers. Even at 1:2 r:r this left me down on the day financially obviously. Hoping for better tomorrow.
 
Wednesday and Thursday brought very disappointing results. Great majority of early break-outs reversed and were stopped out.

Friday I introduced what might be an obvious second leg to the strategy which evened results out financially. This simply uses an opposing order in case of a reversal and stop-out of the initial break-out trade. The secondary break-outs even appear to be both more reliable and to make more extended moves, which argues for a TP further out than the initial BO.

Hoping to take this test further next week.
Cheers all.
 
Wednesday and Thursday brought very disappointing results. Great majority of early break-outs reversed and were stopped out.

Friday I introduced what might be an obvious second leg to the strategy which evened results out financially. This simply uses an opposing order in case of a reversal and stop-out of the initial break-out trade. The secondary break-outs even appear to be both more reliable and to make more extended moves, which argues for a TP further out than the initial BO.

Hoping to take this test further next week.
Cheers all.

Tomo

If the great majority of break-outs were reversed then looking for failed break-outs might be the better main strategy.

Remember the Turtles? Although they were operating 20 day break-outs I believe the bane of their lives was the high number of failed break-outs. Linda Raschke later exploited this by developing a failed break-out strategy - called Turtle Soup as I recall.
 
Wednesday and Thursday brought very disappointing results. Great majority of early break-outs reversed and were stopped out.

Friday I introduced what might be an obvious second leg to the strategy which evened results out financially. This simply uses an opposing order in case of a reversal and stop-out of the initial break-out trade. The secondary break-outs even appear to be both more reliable and to make more extended moves, which argues for a TP further out than the initial BO.

Hoping to take this test further next week.
Cheers all.
Fascinating stuff – keep it going. I've always found it interesting to play with "systems" and although I nearly always come back to using successful basics, it's a very interesting ride and you learn an awful lot (especially from what doesn't work) on the way.
 
Tomo

If the great majority of break-outs were reversed then looking for failed break-outs might be the better main strategy.

Remember the Turtles? Although they were operating 20 day break-outs I believe the bane of their lives was the high number of failed break-outs. Linda Raschke later exploited this by developing a failed break-out strategy - called Turtle Soup as I recall.


Indeed it was Turtle Soup barjon. As I half suggested, this could have been an obvious necessary second leg to the London ORBO strategy from the very start but I wanted to gather my own data, and specifically from forex, which is a field not explored in many 20th century "standards" of TA and trading. Plus there's always the possibility that what worked in the 70's, 80's or 90's won't work now (though I'm sceptical that basic market dynamics, like trends and break-outs, can ever truly die out).

Onwards and upwards.
 
Indeed it was Turtle Soup barjon. As I half suggested, this could have been an obvious necessary second leg to the London ORBO strategy from the very start but I wanted to gather my own data, and specifically from forex, which is a field not explored in many 20th century "standards" of TA and trading. Plus there's always the possibility that what worked in the 70's, 80's or 90's won't work now (though I'm sceptical that basic market dynamics, like trends and break-outs, can ever truly die out).

Onwards and upwards.

Phil Newton did a forex one years ago, but that was overnight range break-out https://www.trade2win.com/threads/phil-newtons-range-break-out-strategy.31855/
 
Fascinating stuff – keep it going. I've always found it interesting to play with "systems" and although I nearly always come back to using successful basics, it's a very interesting ride and you learn an awful lot (especially from what doesn't work) on the way.


Thank you for the encouraging words. I do intend to wring this out for all its worth.

So far some general principles can be concluded -
B/O's between 0800 and 0900 are very common but only about half then continue in this direction to 1200, few later:
most of the failed B/O's will breach the other extreme of the 0700-0800 range by 1100:
very few B/O's breach both extremes of the 0700-0800 range and then breach the first extreme for a second time.

Should be humanly possible to define an edge from this approach.
 
The very first trading system I ever tried (circa 2005) was an ORB on the E-mini S&P500. I spent months and months...and months developing and back testing the system with years of purchased historical tick data and Excel spreadsheets. I wrote a VB program to automate testing so that I could try different trading rules, reversals, times, thresholds etc. I traded a finalised system with my live account and it did perform according to plan, but I had to top up my account on a few occasions over the few months I tried because it hit an expected drawdown period.

Tested over the very long run (approx 20 years) it proved profitable. I used a simple scaling rule that increased/decreased the number of contracts traded depending on equity in the account. But when I analysed the actual trades I knew that there was no way I could stick with it in reality.

There were some very long periods between equity peaks, sometimes up to 2 years. Also, there were long periods of consecutive losses (more than 10) and missing the one or two ‘big’ trades that made up for them could put the system years behind. This means you couldn’t afford to miss those important big trades for any reason i.e. Loss of internet connection or Broker issues.

The only way to effectively trade ORB with the discipline required to get through the many consecutive losses and long drawdown periods was to automate it, constantly monitor it, and have a back-up system for any internet/broker issues.

This was my conclusion and what ultimately caused me to abandon not only this trading system, but any mechanical/threshold system. I am glad I did because it forced me to learn Tape reading and the ‘art’ of Technical Analysis and I never look back with regret...over 13 years now.
 
Thanks for your experiences with this approach, looks like you thoroughly explored how to make it work.

I tried a London ORBO strategy on the FTSE100 some years ago and though my experience was dabbling compared to your trial, the result was the same. The causes of my poor performance were though a little different -

firstly, I never developed a reaction to fake B/O's - I am now building this in to every trade with a secondary trade in the opposite direction to the initial B/O. I am also not setting a TP on the secondary trade as these often run dramatically for the rest of the session, so I want to not throttle the big winners which you also mention.

secondly, I over-sized my trades ridiculously as soon as they started to show a run of winners. This went way over any sensible % of account capital limit so when the run of losers came, I took fright and pulled out.But that was my personal arrogance and failure to plan, won't happen again.

On other points -
negative equity performance - I hope to smooth out this sort of risk by trading off multiple forex pairs, not just the FTSE like I used to. I accept as ever trading a basket of mixed performance instruments rather than one strongly performing one does cut off winnings, but I'll be happy this time to forego some profit for much reduced loss.

trading stock indices via ORBO's - I hope to discover that trading forex pairs is more reliable over the long run than any individual index. Obviously, this is still only a hope, too soon to be sure.

internet/broker issues - years since I've seen a platform shut-down or internet cut-off. Certainly noticed delayed execution of orders during Friday's NFPR's, but ordinarily I wouldn't carry any trades or orders into NFPR time, that was just a learning exercise. One thing I do need is a second internet device in case the PC fails, and already pricing laptops.

So far this exercise has been a real challenge, day-trading has never been my preferred field, but I'm actually hoping for modest profits within only another week. Fingers crossed.
ATB.
 
A more satisfying day today, running both primaries (initial break-outs) and secondaries (reversals / failed break-outs). Actually made a profit.

38 trades ran to TP or SL.
Of which:-
22 primaries
16 secondaries

5/22 primaries were winners at 1r each (r being the respective 0700-0800 range) = +5r
17/22 primary losers at -0.5r each = -8.5r

9/16 secondaries winners at +2r each = +18r
7/16 secondary losers at -1r each = -7r

Net result +7.5r.

So, predominantly a day of failed early break-outs, followed by reversals into strong continuing trends. GBP/JPY is the classic example of the day: prints a strongly bullish set-up bar 0700-0800, falls to make the low of the day at 0915, followed by a consistent uptrend for the next 8 hours. the initial fall triggered my primary short, which was stopped out, and the secondary triggered around midday, running to its TP by 5.

They say that either the low or the high of the day is statistically most likely to be printed within the first 2 hours.

Up and at 'em.
 
The end of another week of this trial, and it certainly was a trial in both senses of the word.

The system generated 200 ORBO trades, 180 of which closed at their pre-set SL or TP price levels. Despite the positive r:r per trade, neither the primary (initial break-out) trade set nor the secondary (reversal) trade set made a net profit over the week. The position sizes are so small that the £ loss is OK if its just a statistical variation, but obviously unsustainable long-term.

Next week I'll continue to use the same basic plan but must gather more info on the historic and future trades. This must include -

* cancelling sleepy orders - surely, if a B/O from the 0700-0800 range doesn't occur quickly, this trade should not be run anyway?

* but what about USD, commodity and US index trades? - would they be better run as NY market ORBO's rather than London?

* duration of sleepy trades - if a trade is still running hours later but hasn't reached its TP, how soon to manually close it and avoid the risk of the maximum loss when it hits the SL?

Will get some more info on historic trade times etc. out from the trading platform over the weekend. We shall never surrender.
 
Also noted two personal errors in the mid-week -
1. when geo-political events hand over an unexpected windfall, this outcome is outside the normal strategy, so forget the rules, just take it and bank it

2. avoid the temptation to be selective over which pair from a base currency set to trade - its a dumb system and data for filtering targets is not there yet, so set orders on them all and take them all.
 
* but what about USD, commodity and US index trades? - would they be better run as NY market ORBO's rather than London?

I used the cash market timezone of the instrument I was trading when I tested my system. The e-mini S&P500 trades almost 24hrs M-F but I used the cash S&P500 open times.
 
An easy statistical conclusion from last week -
of the 14 secondary (reversal) trades which opened after 1300, 12 hit their SL
 
An easy statistical conclusion from last week -
of the 14 secondary (reversal) trades which opened after 1300, 12 hit their SL

Isnt the 1300 time really close to the US Opening Range, and the behaviour is now closer to USs actions than Londons?

Good thread, by the way. I think being aware of each session behaviours / tendencies is useful.
 
Isnt the 1300 time really close to the US Opening Range, and the behaviour is now closer to USs actions than Londons?

Good thread, by the way. I think being aware of each session behaviours / tendencies is useful.


Yes, and even more so than usual as currently New York is on Eastern Daylight Time and only 4 hours behind London. London time advances on 31/03.

The London ORBO should still be valid principle for forex as the volume increase when London opens is much more significant than the volume increase when NY opens.

My US index trades were a mixed bag across the week, slightly winning in £ but no strong conclusions possible after just one week.
 
Top