Big Ben Strategy - optimised

Thanks trendie, this is going every bit as well as I had hoped.

I did dabble with this on the USD/JPY around the NY open a month back but its not anything like as reliable - too many double false breakouts, too little price travel.

As far as I can tell, there is a great key to finding a suitable candidate for breakouts at the same regular time of day. See the M30 chart for GBP/USD pre the virus emergency, let's say start of the year to 20/02. Add ATR20. The indicator shows extremely sharp peaks, regularly spaced - lows are at about 0500-0630, highs are at about 1700-1800. The ratio between the ATR's daily high and daily low is typically between about 2.5:1 to 3.5:1 Its is not often 2:1 or lower. Its like a heartbeat. I believe this is a pointer to what makes for a great candidate for entries around the London open.

Compare cable's regular sharp up-down, up-down pattern with USD/JPY and it just doesn't have the same pattern. Its pretty irregular, to the point where it looks very possibly random.

EUR/USD does show the heartbeat trace. The ATR peaks and troughs are around the same times as for GBP/USD, so I suspect its again the London open which is the key, not NY.

If someone wanted to trade NY opening range breakouts, I believe they should look for the same ATR heartbeat but with ATR highs and lows 4-5 hours later in the day obviously.
 
. . .As far as I can tell, there is a great key to finding a suitable candidate for breakouts at the same regular time of day. See the M30 chart for GBP/USD pre the virus emergency, let's say start of the year to 20/02. Add ATR20. . .
Any chance of a chart please Tom? A picture paints a thousand words etc. Ta very muchly.
Tim.
 
Any chance of a chart please Tom? A picture paints a thousand words etc. Ta very muchly.
Tim.


Sure - here is GBP/USD

1587208097785.png


But now see the random ATR trace for USD/JPY -

1587208159139.png


:)
 
For a little bit more clarity (and also to show off a bit...) here is the BB trade set-up from this last Wednesday. One of the best BB trades I've seen.

The rectangle shown is drawn off the 0700-0730 M30 bar. Its a pretty narrow range so taking this is aggressive. I set a buy at the top of the rectangle and a sell at the bottom. The long was not triggered, the short triggered in the next half-hour and it is exited manually at 2pm, shown by the oval lower down in the pic.

1587215312778.png


As happens so often, right after 2, price action re-sets. The 1330 bar's close ended up being the lowest of the day. Good result all round even though I say so myself.
 
Last edited:
For a little bit more clarity (and also to show off a bit...) here is the BB trade set-up from this last Wednesday. One of the best BB trades I've seen.

The rectangle shown is drawn off the 0700-0730 M30 bar. Its a pretty narrow range so taking this is aggressive. I set a buy at the top of the rectangle and a sell at the bottom. The long was not triggered, the short triggered in the next half-hour and it is exited manually at 2pm, shown by the oval lower down in the pic.

View attachment 278273

As happens so often, right after 2, price action re-sets. The 1330 bar's close ended up being the lowest of the day. Good result all round even though I say so myself.
 
I am interested in giving it a go just a couple of questions pls
1) I take it the RR would be based on the 0700 candle close from wick to wick not the body. So if 30 points that would be used to work out your risk.
2) what length of candle would you consider too large

cheers jeff
 
I am interested in giving it a go just a couple of questions pls
1) I take it the RR would be based on the 0700 candle close from wick to wick not the body. So if 30 points that would be used to work out your risk.
2) what length of candle would you consider too large

cheers jeff
 
The Wednesday trade explained it thanks


Yes, r in pips is the bar's full height, from high to low.

An extra-long bar would be, for example, double or more the previous London session's average. Leaves it very unlikely price would move by multiples of r pips by 2pm, and quite possible that price would end at that time within the bar's range, between the high and the low, for a minor loss. On the other hand, an extra-long bar is unlikely to see both its high and low breached, so a double loss day would be unlikely. Not worth the trouble/risk in my mind.

An extra-short bar is also a problem - although multiple r rises or falls are probable, so is a double loss day, with both high and low of the bar breached very quickly. A pragmatic work-around would be to combine the 0700 bar with the previous bar, which I have had to do once or twice.
 
Could just answer what range of candle would you consider to wide
Ball park figure would be fine
Jeff
 
The trade you took on Wednesday 8th April was approx 50 point candle which you took on
By my calculation you got about even is that correct.
Would that have been one you should probably have avoided
Cheers
 
The trade you took on Wednesday 8th April was approx 50 point candle which you took on
By my calculation you got about even is that correct.
Would that have been one you should probably have avoided
Cheers


No, the 0700 bar on the 8th was a bit long but not excessive compared with the previous day's London bars - it just looks big compared to the overnight bars immediately before it.

My long on the 8th made about +0.5r. This wouldn't have been a bad result but I had got an immediate short at 0740 when price just barely breached the 0700 bar low: this short of course was stopped out for -1r, so net result for the day was -0.5r. Hey ho.
 
Good, solid, returns there, tomorton!

Are certain days more profitable than others?
Also, given its a session breakout, have you applied your concept to the US markets?


Maybe I will find certain days of the week are better (or unsuitable) for this strategy, but it looks on the face of it unlikely to make a big enough impact though I follow your logic. My pessimism is based on -
1. its very difficult to identify the days of the week from either price moves, close to close, or daily ranges, high to low: a strongly directional price move with a big extent is of course good for this strategy but if this happens much much more on say a Tuesday, so far it isn't obvious.
2. the same lack of differentiation between days of the week is also seen on the ATR indicator chart.

But, I am logging the data so it will be easy to summarise this here in a little while.

Looking forwards to the open tomorrow!
 
I've been running this strategy for April so far, and thanks to Tom for starting the thread. I've experimented in the past with various breakout strategies, including a weekly version which tried to capture the week's move from the price either at 10pm Sunday night or 7am Monday morning. I found it hard to get that reliably past break even, and think the daily breakout is the way forward.

One point Tom made which I think is critical here is the trade-off between having a straddle (i.e. distance between buy and sell orders) which is too small, so that it gets whipped out for -2R losses continually, and a larger points distance which whilst (theoretically at least) should get whipsawed less often, consumes too many negative pips which limits profitability and reduces trade size.

The only difference in my own trading from what's happening here is I'm using the 0600 H1 bar (i.e. at 7am) as my trigger bar. I was using stops of ~30 pips, but found this limited profitability since the loss on a -2R day was 60 pips, and the trade size was less than it could have been (given the 30 pip stoploss). I'm going to aim for a stoploss in the region of ~20 pips which from my back testing, although it will result in a greater number of -2R days, overall profitability looks better.

Cheers, David.
 
.......Are certain days more profitable than others?...........


Hmmm, thinking this through, identifying best/worst days of the week for this strategy is not going to be quite so straightforward.

The further price travels by 2pm, the more return is available on a winning trade. So a wide range day is a good start. Conversely, the narrower the daily range, the less possibility there is for an extraordinary return.

Have looked over the last 11 months for inside days - the largest group prints on a Monday, but that doesn't mean Mondays should be written off or the other days are completely safe -
Mon = 12
Tue = 7
Wed = 9
Thur = 4
Fri = 10

In addition, is every fully comparable to every other Monday? Some of those would have been public holidays in the UK and/or the US and/or the Eurozone. Some would have been the first full trading session after NFPR's. Some would have been in the same weeks as interest rate decisions. Some would have been the first session back after a 3 or 4 day weekend.

Going to need a lot more data to make a decision on this basis and it isn't going to be quick or easy.
 
This table appears here and there on the web and it also suggests Mondays would be the narrowest range days on average for GBP/USD and everything else too -

Currency PairSundayMondayTuesdayWednesdayThursdayFriday
EURUSD69109142136145144
GBPUSD73149172152169179
USDJPY4165829112498
AUDUSD588411499115111
NZDUSD2881988710096
USDCAD4392112106120125
EURJPY19133178159223192
GBPJPY100169213179270232
EURGBP357481797591
EURCHF355555648776

Maybe it might be wise to take Mondays off.......
 
OK, so what about Outside Days? These have good potential for extraordinary price movements during the day, increasing profit potential.

I looked back again over the last 11 months on GBP/USD to see on which days Outside Day bars printed -
Mon = 1
Tue = 7
Wed = 1
Thur = 8
Fri = 5

Again, Monday shows up as the poorest prospect on this measure, and again Tuesday and Thursday are pretty good.

There doesn't seem to be much to be gained by excluding Fridays from these Big Ben trades (a bit surprising as many experienced daytraders swear off Fridays).
 
Top