SteamRoller
Member
- Messages
- 70
- Likes
- 15
Yes it's been a poor couple of weeks for both the 0700 M30 bar entry and also 0600 H1 bar entry.
I've been thinking about this, and have recently traded with wider stops which has given a better outcome than the narrow ranges that these entry methods (above) dictate.
I'm not necessarily doubting the long term profitability of these aggressive entry methods, but I am finding in practice even though you get some high multiples of R profits, the high incidence of -2R days can erode these gains fast.
On many occasions, with tight stops, even when the first breakout proves to be in the right direction for the day's trend, you're stopped out and stopped in to the opposite trade before the market moves in the original direction without you, and you've banked a -2R loss. This has happened a lot recently.
By running bigger stops, if the first breakout proves to be the day's trend, you're much less likely to get whipped out. If the first breakout is false, and ultimately the second entry triggers, price needs to go further to cover the loss from the first trade. Also, there are fewer big % gains of course, because your stops are bigger, but this has to be compared against the fact that -2R double loss days are rare (as opposed to at least 50% of the time with the smaller stops).
It's all very well getting the odd +10R day, but if you have to go through 5 or more -2R days for everyone you see, you've spent all the profit in the big gain on a series of -2R double losses.
So what I've settled on after a great deal of thought and back testing, and have been doing this last week, is trading with stops that are about double what I was doing. Draw downs are less, double losses are far less common, there are many more days where only one order triggers, % of winning trades is greater. Of course the size (+ multiple of R) of the winners is less, but this is viewed against far fewer -2R days to have to cover.
Whether or not it is long term as profitable in all conditions, is hard to say, but I find it psychologically easier to trade.
Last edited:
I like this, we should be constantly pushing at the boundaries of the same basic system, wringing whatever we can from it.
This is an age-old problem though - narrow stops are often hit but profits can more commonly be multi-r (risk): wide stops are less commonly hit but profits will rarely be multi-r. Its a balancing act, there literally can't be a right answer, only answers that work. Keep at it.
This is an age-old problem though - narrow stops are often hit but profits can more commonly be multi-r (risk): wide stops are less commonly hit but profits will rarely be multi-r. Its a balancing act, there literally can't be a right answer, only answers that work. Keep at it.
SteamRoller
Member
- Messages
- 70
- Likes
- 15
Agreed, although we can work in from both extremes to what is a fairly clearly 'correct' range of stop sizes.
Clearly 2 pip stops would be far too tight, you'd go bust before you got a massive winner (probably).
100 pip stops would be too wide - you'd need too much of the day's move to occur before you even got triggered in.
If you work in from both extremes, I think you can identify a range that suits the individuals risk profile. For me, the 20 pip stops I was using previously are into the too narrow range.
There is also the psychological aspect - I personally am better with a higher win rate, lower drawdowns than a very low win rate but with occasional big winners, which is why it comes down to the individual's attitude to risk etc.
trendie
Legendary member
- Messages
- 6,875
- Likes
- 1,433
I think you have already established this is a robust system.
Important to be careful not to be spooked by one bad week, and try to correct for a possibly transient phenomena by changing the rules.
But, by all means keep researching, as should we all.
Its been a tough start to May so far, I've marked the strategy down at a net loss of -9r so far, with 5 double-loss days (though one may or may not have been a no-trade day due to a wide 0700 30-minute bar).
But is 5 double-loss days unusual? Statistically its not unusual for any month. Of the last 12 months, 10 of the 12 months had at least 5 double-loss days. 3 of those 10 months ended up as net losers, but the other 7 of the 10 ended up as winners.
On top of which, I can't see the GBP/USD chart as looking like anything exceptional recently. Its possible there is more hesitancy in the markets, which translates into more hesitant moves and more failed moves, though May has seen only 2 inside days this month, which is about average for where we are.
So I'm intent on just ploughing on, hoping to get a good result tomorrow.....
All the best.
But is 5 double-loss days unusual? Statistically its not unusual for any month. Of the last 12 months, 10 of the 12 months had at least 5 double-loss days. 3 of those 10 months ended up as net losers, but the other 7 of the 10 ended up as winners.
On top of which, I can't see the GBP/USD chart as looking like anything exceptional recently. Its possible there is more hesitancy in the markets, which translates into more hesitant moves and more failed moves, though May has seen only 2 inside days this month, which is about average for where we are.
So I'm intent on just ploughing on, hoping to get a good result tomorrow.....
All the best.
SteamRoller
Member
- Messages
- 70
- Likes
- 15
Thanks for this. With regard to your comment above, today was a fairly good day too wasn't it?
Also, assuming you're still triggering off the 0700 M30 bar, given that UK employment data is released at 0700 tomorrow, it's likely that the 0700 M30 bar is of large range.
Are you planning to do anything else in such situations to avoid being left behind if by 0730 the bar has a big range? Or just wait and see?
I'm just planning to go off the bar range - if really wide compared to today's daytime M30 bars I'll let the day go buy and no-trade it. If its medium, I'll just treat it as normal - set a buy and a sell and go out cycling.
trendie
Legendary member
- Messages
- 6,875
- Likes
- 1,433
Hope this week was good.
I eyeball it as being as a good Monday, maybe 3-4R;
Pass on Tues and Weds,
Thurs seems touch and go on Stop-and-Reverse, depending on broker, otherwise maybe 2R.
And today, another 2-3R??
With regards to outsize days, have you determined whether they are actually winners or losers?
If the opening 30-min bar is outsize, could you use 1/2R as risk?
If the outsize is a winner, might as well eke something from it.
EDIT: just remembered, Mon is a pass!
I eyeball it as being as a good Monday, maybe 3-4R;
Pass on Tues and Weds,
Thurs seems touch and go on Stop-and-Reverse, depending on broker, otherwise maybe 2R.
And today, another 2-3R??
With regards to outsize days, have you determined whether they are actually winners or losers?
If the opening 30-min bar is outsize, could you use 1/2R as risk?
If the outsize is a winner, might as well eke something from it.
EDIT: just remembered, Mon is a pass!
Slightly negative week, my table shows this as a net -1r. Disappointing but better than the weeks before.
Yes, there must be rational ways to trade off a very wide 0700 bar, but my obs show price doesn't advance much after a big big bar. They typically tend to look like Tuesday - tall flagpole with a never-ending streamer from the top - a 50 pip 0700 bar, which before 1400 never made better than another 18 pips.
Actually the US open on Tuesday was again the turning point in GBP/USD price action: in this case price went higher, just to annoy me of course. But the take-away lesson is still the same - whatever signal applied at London breakfast-time is history when New York gets to work.
Yes, there must be rational ways to trade off a very wide 0700 bar, but my obs show price doesn't advance much after a big big bar. They typically tend to look like Tuesday - tall flagpole with a never-ending streamer from the top - a 50 pip 0700 bar, which before 1400 never made better than another 18 pips.
Actually the US open on Tuesday was again the turning point in GBP/USD price action: in this case price went higher, just to annoy me of course. But the take-away lesson is still the same - whatever signal applied at London breakfast-time is history when New York gets to work.
Chad Seven
Well-known member
- Messages
- 280
- Likes
- 41
I am interested to know what happens if you apply a directional bias to the back-testing of this strategy. For example trade in the direction of the trend using a moving average as a filter.
SteamRoller
Member
- Messages
- 70
- Likes
- 15
Interesting, I've thought about the same thing.
However you define a trend (pretty straight forward), either moving average, higher high higher low etc, or a combination, you could trade in the direction of the trend with just one order. You could also run a larger position size as you would not be having two orders in. I guess you could use the H4 chart for bias, or a zoomed out H1 chart.
Theoretically, I would have thought the return should be greater doing this - think about it, if you thought it wouldn't make any difference, then what you'd effectively be saying is that there is no edge in trend following(?), as the counter trend breakouts would be as profitable as the with trend breakouts.
Suppose you only took the with trend breaks, but with twice the position size? If the move went against your direction, and you were not in a trade, doesn't matter. it would just give the opportunity to be in at a better price tomorrow, assuming according to the trend bias rules you were using, the trend was still valid.
Also, to update, I had a lacklustre week too, I have been running bigger stops (though not much bigger - 35 to 40 pips) and this week had no second trades triggered out of 4 trades. the returns were small this week: Monday +1R, Tuesday break even (slightly +), Wednesday break even (slightly -), Thursday didn't trade, Friday +0.5R, making about +1.5R for the week. (I'm calling anything between -0.5R and +0.5R a break even trade). Running the slightly wider stops is definitely better for me and I'm sticking with it.
Caution anyone using this strategy or a variant.
Tomorrow is a UK public holiday: though forex and indices quotes might be available to us price action will be thin and atypical.
Tuesday is therefore the first London trading session of the week and maybe its going to be like a Monday - high probability of a narrow range day, possibly an inside day bar, with weak and hesitant price action. I will trade Tuesday but no great expectations.
Tomorrow is a UK public holiday: though forex and indices quotes might be available to us price action will be thin and atypical.
Tuesday is therefore the first London trading session of the week and maybe its going to be like a Monday - high probability of a narrow range day, possibly an inside day bar, with weak and hesitant price action. I will trade Tuesday but no great expectations.
Late update.
May is finished - thank the lord!
Trading Tuesdays-Fridays only I make this a net result of -14r, the worst of the last 13 months. Actually, including Mondays (apart from the UK public Holiday this month) would have made the net result slightly better, improving it to -9r. Still pretty bad. Think I might as well go back to trading 5 days a week, what's to lose?
I suppose there are consolations - at least the worst month recorded was also the most unique - the national economies of both the UK and the US are crippled by the respective responses to the covid crisis. So perhaps it would have been a worryingly anomalous indicator if the strategy had done well in May. Maybe every strategy should be under-performing right now.
But the worst thing is that there is no concrete indication on the long-term charts for May that this should have been any worse a month than April or March. And that's worrying because although current unique conditions might have led to the month's poor performance, how will we know when these conditions are passed?
There might be some things to be done to optimise poor performance under current conditions -
1. as ever, keep r to a small percentage of the account capital (should be a constant rule, but maybe reducing r might be prudent)
2. price moves are failing, whether long or short: consider setting a TP at a conservative level, say +3r or even +2r: right now the rule is to conserve capital rather than maximise profits
3. if feeling a bit more aggressive, use a third entry order, set only after the first has been stopped out: so if the long is triggered first and is then stopped out (at which point the short is triggered), set a second new long at the original long entry level: if the short is triggered first and then stopped out, set a new short at the original short entry level. I don't have stats for this but a look-back over charts suggests this would work.
Onwards and upwards.
May is finished - thank the lord!
Trading Tuesdays-Fridays only I make this a net result of -14r, the worst of the last 13 months. Actually, including Mondays (apart from the UK public Holiday this month) would have made the net result slightly better, improving it to -9r. Still pretty bad. Think I might as well go back to trading 5 days a week, what's to lose?
I suppose there are consolations - at least the worst month recorded was also the most unique - the national economies of both the UK and the US are crippled by the respective responses to the covid crisis. So perhaps it would have been a worryingly anomalous indicator if the strategy had done well in May. Maybe every strategy should be under-performing right now.
But the worst thing is that there is no concrete indication on the long-term charts for May that this should have been any worse a month than April or March. And that's worrying because although current unique conditions might have led to the month's poor performance, how will we know when these conditions are passed?
There might be some things to be done to optimise poor performance under current conditions -
1. as ever, keep r to a small percentage of the account capital (should be a constant rule, but maybe reducing r might be prudent)
2. price moves are failing, whether long or short: consider setting a TP at a conservative level, say +3r or even +2r: right now the rule is to conserve capital rather than maximise profits
3. if feeling a bit more aggressive, use a third entry order, set only after the first has been stopped out: so if the long is triggered first and is then stopped out (at which point the short is triggered), set a second new long at the original long entry level: if the short is triggered first and then stopped out, set a new short at the original short entry level. I don't have stats for this but a look-back over charts suggests this would work.
Onwards and upwards.
Volatility back to "normal"? (pre-corona levels)
www.tradingview.com
www1.oanda.com
BPVIX Index Charts and Quotes — TradingView
View live CBOE CME FX British Pound Volatility Index chart to track latest price changes. CBOE:BPVIX trade ideas, forecasts and market news are at your disposal as well.
Trading Tools | Online Forex Trading Tools | OANDA
Discover the range of powerful analysis tools and partner API's contained within the OANDA trade platform
trendie
Legendary member
- Messages
- 6,875
- Likes
- 1,433
Unfortunate May.
How badly off was May compared to other losing months?
If you have established that Monday is a nett loser, I think you should hold onto that, and stick with passing on Mondays. Maybe, review in another month or two before deciding.
We are in June now. You are trying to trade May again in June.
We all do this from time to time. "If only my indicator was X instead of Y. I would have made money" etc.
Obviously its up to you, but by jumping a rule after one month, is to ignore your 12-month analysis.
Good luck, and kudos for showing all your trades, good and bad; a very rare trait.
How badly off was May compared to other losing months?
If you have established that Monday is a nett loser, I think you should hold onto that, and stick with passing on Mondays. Maybe, review in another month or two before deciding.
We are in June now. You are trying to trade May again in June.
We all do this from time to time. "If only my indicator was X instead of Y. I would have made money" etc.
Obviously its up to you, but by jumping a rule after one month, is to ignore your 12-month analysis.
Good luck, and kudos for showing all your trades, good and bad; a very rare trait.
Unfortunate May.
How badly off was May compared to other losing months?
If you have established that Monday is a nett loser, I think you should hold onto that, and stick with passing on Mondays. Maybe, review in another month or two before deciding.
We are in June now. You are trying to trade May again in June.
We all do this from time to time. "If only my indicator was X instead of Y. I would have made money" etc.
Obviously its up to you, but by jumping a rule after one month, is to ignore your 12-month analysis.
Good luck, and kudos for showing all your trades, good and bad; a very rare trait.
Without Mondays, May was the worst loser in the last 13, but not by much margin, and there were only 3 other losers. If the 3 available Mondays in May had been traded, the month would have still been a big loser but smaller than two of the other 3.
I actually don't have statistics to show Monday is a loser, only that its the weakest day of the 5 for the strategy. Yes, I am going to bring it back in except for UK or US Public Holidays of course (Monday is the preferred option for these in both countries). I'm not convinced its going to make much difference either way, except that running the trades with a fixed SL and no TP always leaves the door open to an unexpected big winner. Well, we'll soon know.
My greater worry is that the charts in both short and long time-frames, and their various indicators, don't show May as being anything special.
