Big Ben on the FTSE100

[splashy]

Thanks for your thoughts, Splashy. The set & forget aspect is the basis of Big Ben and this thread - instigated by those who have to go to work I believe. (Terrible). Managing might or might not improve the s/r. I agree with you generally about the 20 pip stop & I'm often to be found using 40 to 45 pip stops on Cable when managing longer term strategies - hopefully for more reward. As for BB as you say there are many false breakouts & these frequently result in a reversal of 30 or 40 pips. But also many of these false breakouts will breakout 25 to 30 pips before retracing, thus the 20 pip breakeven or take profit scenario. So the 20/20 is very low risk & also has a good s/r. In the period I've studied, a 40 pip stop would lose horribly & the trade often drifting for hours. I believe this is supposed to be a short tem strategy.

I've gone back 16 months (that's all I have on 5min bars), & although I also must agree that markets are always changing, to some extent this is covered by using the 10 day average BB range. It can change fairly quickly, from 52 in May 2010 to 77 about 5 weeks later for example so reflects changing conditions to some extent. Over the period the BB range has been 37 up to 77.
I don't claim to be any kind of an expert, & just hope my input will be food for thought on this thread And comments like yours are very welcome. Basically we are gamblers, like ALL traders. Cheers Splashy.

Nice one gelly. If we don't try new things, we never learn anything...
I wasn't referring to managing the trade - quite the opposite. Just that slow-burning sentiment will regularly get you where you want to go after lunch when short-term momentum fizzles out at elevenses (even if i'm at work). Having said that, i haven't found gbpusd 8-10 worth trading either.

Incidentally, don't dismiss the second signal of the day. Having 6 months data from big ben, I haven't seen enough to warrant trading any of the first signals from the instruments i've measured but I'm trading reverse signals with some reasonable success, and presumably because they trigger later in the day they tend not to turn round often and hence risk is lower. They don't go on all night - you can cut them off whenver you want. There have been many, many weeks (like this week) when the first signals got hammered but the second signals recovered much of the loss. See what you think.
cheers and good luck

 

[Gelly]

Thanks for results Gelly, I will have a look tomorrow but will in any case immediately benefit from your hard work by cutting my sl/tp on this pair to 20pips.

I think 0800-0900 very worth looking at as opposed to 0800-1000, in fact it's amazing how often the BB range I use as reference period on FTSE100 and GBP/USD is derived from the 0800-0830 bar/candlestick, while the 0830-0900 bar is an inside bar.

Yes this does appear often - maybe we should take the inside bar as a hi/lo entry? In meantime I am now analyzing cable 0800-0900 range to compare with 8 to 10. We do need more profitablity! (First 4 months of 2011 show good improvement, getting some of the big moves but still a lot of false b/outs. Will always get these I guess).

 

[Gelly]

Hi all, will be checking 9 months of 2010 for BB on gbpusd for 8 to 9am range, but meantime here's what I've found for 2011 so far.
The percentage of average range factor is not significant for this period, in fact large ranges usually produced winning trades. So I will abandon this calculation for 2010, which will speed up my backtesting a lot. Last month was very bad, which you already know, and I'm sure bad months will always be with us. However we'll see what results I get from 9 months of 2010. Very many of the losing trades result in false breakouts of only a couple of pips. Depending on your trading platform, some of these you won't get in but results will always show the worst if any doubt about price levels. (Well that always happens to me). Hope this is of interest. Oh, also I noted that if traded, all 7 public holidays would result in a loss.

 

[tomorton]

Weekly BB summary for FTSE100, GBP/USD and EUR/USD -

FTSE actually made a winning week, 2-1, certainly a welcome result and raising its 4 week performance to break-even at 7-7, though still only 9-11 over the last 20 trades.

GBP/USD was a small loser, 2-3, but its 20-trade score is 8-12.

EUR/USD made 2-2 and has the only positive 20-trade score at 11-9.

I have been trying to match up the regular price charts with BB performance to see if a simple rule can be devised to indicate when BB is not worth the risk. So far the best I could suggest would be to avoid BB if the 14EMA is below the 50EMA. This would have curtailed trading through March, June and most of July, particularly weak BB periods. I will track this from here on and go back over charts to confirm BB results with this relationship.

 

[tomorton]

Using the position of the 14 and 50EMAs over the last 11mths or so would seem to have been a simplistic but fairly good rule of thumb when to take a Big Ben trade and when to stay in bed -
BB trades when 14EMA above 50EMA = 61% win rate
BB trades when 14EMA below 50EMA = 54% win rate
I personally wouldn't get out of bed for a 54% win rate, though it's disappointing the 61% win rate isn't higher.

 

[Gelly]

Using the position of the 14 and 50EMAs over the last 11mths or so would seem to have been a simplistic but fairly good rule of thumb when to take a Big Ben trade and when to stay in bed -
BB trades when 14EMA above 50EMA = 61% win rate
BB trades when 14EMA below 50EMA = 54% win rate
I personally wouldn't get out of bed for a 54% win rate, though it's disappointing the 61% win rate isn't higher.

Will also try this out Tomorton. We really do need some indicator, judging by my latest backtest. The 8 to 9 range does work better than 8 to 10, but still not terrible encourging to trade as a set & forget strategy without some corroboration. Will check out eurusd sometime, meanwhile here's 17 months of gbpusd.

 

[brewski1984]

I see you've moved onto pin bars tomorton? Are you going to give up this system now?

 

[tomorton]

I'm not giving up on it brew but it just seems to be a system for positive markets. Pins are more universal but I will run BB's again when prices stabilise and start creeping upwards and take the MAs positive again. Will keep up the weekly observations to make sure I'm not missing a trick of course.

 

[welshguy]

a wise man gives up when he s in front a stupid man continues when he going backwards

guess i m a pretty STUPID man

 

[MajorDutch]

a wise man gives up when he s in front a stupid man continues when he going backwards

guess i m a pretty STUPID man

don't be too hard on yourself this is just a stage of learning. if it doesn't register as a stage of learning then that's when you got problems.

IMO if you want to trade in a set and forget way you should trade daily bars on a small set of instruments.

 

[welshguy]

i understand learning i was 3900 uk up on the week and just wiped 2200 of it i m more pissed of with myself everything i done went against me and my thought s but i still done it

 

[tomorton]

We've passed a milestone this week, 52 weeks since the start of the system.

I make this week 2-2, with a 20-trade score of 9-11, 7-8 over the last 4 weeks.

2011 to date looks like 59-50, a win rate of just 54% if you'd kept trading through the recent doldrums. Better than the FTSE and almost everything else but not great.

I continue to set BB aside for now.

 

[splashy]

We've passed a milestone this week, 52 weeks since the start of the system.

I make this week 2-2, with a 20-trade score of 9-11, 7-8 over the last 4 weeks.

2011 to date looks like 59-50, a win rate of just 54% if you'd kept trading through the recent doldrums. Better than the FTSE and almost everything else but not great.

I continue to set BB aside for now.

Out of curiosity, i checked ftse for last week. if anyone had traded using the specified ranges as target, they would have made an absolute killing... but they would have to have been very brave (and not use adr)...

w/e 12/08
ftse 8-9 1st signal 479pts, 2nd signal 59 pts
ftse 8-10 1st signal 554pts, 2nd signal 59 pts

Incidentally, I measured eurusd, gbpusd, eurgbp and ftse on 8-9am and 8-10 am ranges since feb-end july so you could call that 8 instruments (at a stretch). The sum of the average daily targets, and therefore risk, is around 450 pips. At a £1 per pip you would need around £3000 to start off with if each instrument/trade is around 2% of capital. You would have doubled your money around a month ago (6 months).
Problem is, most of it is on eurusd which could have been just on a roll, and 2nd signals which tend to get overlooked...

 

[tomorton]

Good bit of research, thanks splashy, keeping my hopes up for Big Ben through all the difficult recent times. I'm going to keep focused on the FTSE and expect I'll stay away from it until price exceeds 50EMA again - it will do one day.

 

[Bartlby]

I've been keeping an eye on the ftse. I had been using a % of ATR to set targets and stops but that seems to have been blown out of the water by the last 2 weeks. ATR is now so large that the system doesn't work - without the extreme volatility of the past few weeks targets are too far out to be reached.

I could change the % of ATR I'm using but the whole point of trading a mechanical system is that it doesn't need constant adjustment - at that point its become much closer to discretionary trading.

 

[splashy]

I've been keeping an eye on the ftse. I had been using a % of ATR to set targets and stops but that seems to have been blown out of the water by the last 2 weeks. ATR is now so large that the system doesn't work - without the extreme volatility of the past few weeks targets are too far out to be reached.

I could change the % of ATR I'm using but the whole point of trading a mechanical system is that it doesn't need constant adjustment - at that point its become much closer to discretionary trading.

IMHO, any results from last week should probably be ignored. They are not representative and would taint any set of data...

 

[Bartlby]

Well I agree the last few weeks have been more volatile than things have been for a while but that volatility is hardly a unique event, periods of it recur with reasonable regularity. Personally I feel that anything that happens on the ftse is part of the data set of prices on the ftse. To say its not representative is false - it happened and therefor represents price action on the ftse - the idea must be to find systems that fit the data and not to fit the data to the system. Unless the system has a clear set of rules about when data will and won't be included its not a system.

 

[splashy]

Well I agree the last few weeks have been more volatile than things have been for a while but that volatility is hardly a unique event, periods of it recur with reasonable regularity. Personally I feel that anything that happens on the ftse is part of the data set of prices on the ftse. To say its not representative is false - it happened and therefor represents price action on the ftse - the idea must be to find systems that fit the data and not to fit the data to the system. Unless the system has a clear set of rules about when data will and won't be included its not a system.

Agreed, that ftse is ftse and all that happens makes the market what it is. And agreed that the idea is to find and trade a system with a clear set of rules that preclude certain undesireable trades.
You can include the extreme data if you have a system that can trade it but if the integrity of the system is stretched by such levels of volatility then you're including data that you couldn't (or wouldn't) trade. I don't really see the benefit in that.

 

[Bartlby]

I see your point but one only knows its extreme data after the fact. So last monday for instance one has placed one's BB trade and the day turns out to be volatile. You couldn't know in advance how volatile the day would be and so one couldn't know that one wouldn't trade those conditions. Moreover we have no idea if changes in volatility represent a short term change or are the beginnings of a longer term change in behaviour.

The BB system does have some protection against unusual volatility by not trading if the BBr is greater than x% of the ATR. But if push comes to shove change the system not the data.

 

[tomorton]

Personally I don't like to exclude data after the event: if the data represent the market, that's our world and there's no point navigating your way around it using a map that excludes the perilous areas.

However, if I was a bit more certain that the area beneath the 50EMA was untradable via BB, I would be happy to one day exclude the BB data from that zone of the chart, and scrap it as unrepresentative of where I'm going to trade. But it's a forward-looking rule, not applied in retrospect of course.