i think its impossible to make an expert for the system cause it depends on the brain indicator
Spud's Predictive MTF Stochastics
- #842
- Nov 24, 2008 6:42am Nov 24, 2008 6:42am
- Joined Feb 2008 | Status: Trader | 6,736 Posts
Time hides Nothing
- #843
- Nov 24, 2008 7:23am Nov 24, 2008 7:23am
- | Joined Nov 2008 | Status: Trader | 1,344 Posts
__Thanks__ MJ
- #844
- Nov 24, 2008 7:31am Nov 24, 2008 7:31am
- Joined Feb 2008 | Status: Trader | 6,736 Posts
Time hides Nothing
- #845
- Nov 24, 2008 7:48am Nov 24, 2008 7:48am
- | Joined Nov 2008 | Status: Trader | 1,344 Posts
__Thanks__ MJ
any ideas of parameters or such are welcomed (also brain indicators ;-)
(I`ve done about 100trades)
These are what I found.
( These backtesting is from May and June during Asian/London Session)
1. simple 5m/15m Entry / Exit at Must Exit Point mechanically
- about 45% w/l ratio
- but still profitable (average 6pips per trade)
2. exclude trade which no other TF (30M 1H 4H) agree
- about 60% w/l ratio
- average 10pips per trade
3. exclude trade which 15m is started at middle (30 50 60)
- not exactly but more profitable than No.1
I think some useful parameters can be
- choice of 30M 1H 4H TF agreement (independently)
- choice of where 15M start from
- choice of each TF`s position user want to exclude (ex:15M over 20 etc..)
- few method which can identify twitches on 15M stoch
(I think agreement of other TFs are good method to do that)
- Fixed TP and SL
- choice option of Take Profit Point at each 5m trigger candle
(in my backtest, aiming 10th trigger was best result)
Also I`m doing some backtest with 5m cross 15m method.
- I enter trade when 5m is closed under/above 15m when 15m and 5m is on same direction
- SL : Swing High or Low, Fixed SL, Fixed amount of Pips on Signal
candle`s high and low
- TP : Fixed TP, 1:1 TP, use Average True Range (X) for TP
Since I have no knowledge on programming, I just try to make rules and Filtering through manual backtesting. I found this is very valuable to understand MTF mechnism.
But I found it`s very difficult to do trade mechnically for forward testing because of time consuming (some trade takes 7Hours only for 8pips profit ;p)
Automatic method (maybe EA?) for backtest purpose can be very good for understanding and analysing which situation can be whipsaw and which one is not.
I hope this can help
also you undertand my bad english
- #847
- Nov 24, 2008 8:27am Nov 24, 2008 8:27am
- | Joined Nov 2008 | Status: Trader | 1,344 Posts
__Thanks__ MJ
- #849
- Nov 24, 2008 9:24am Nov 24, 2008 9:24am
Looking to go short unless some news comes along and messes it up 
Thanks Spud for your explanation. I'm out of town this week so I am limited on how much trading I can get done but best of luck to the Spud team!
***EDIT: Can't attach the chart for some reason.
Thanks Spud for your explanation. I'm out of town this week so I am limited on how much trading I can get done but best of luck to the Spud team!***EDIT: Can't attach the chart for some reason.
Looking to go short unless some news comes along and messes it up
***EDIT: Can't attach the chart for some reason.
No problem on the chart....we're all looking and waiting on the same thing.
- #852
- Nov 24, 2008 9:36am Nov 24, 2008 9:36am
- Joined Feb 2008 | Status: Trader | 6,736 Posts
Looking to go short unless some news comes along and messes it up
***EDIT: Can't attach the chart for some reason.
Here ya go...
Time hides Nothing
- #853
- Nov 24, 2008 9:45am Nov 24, 2008 9:45am
- Joined Feb 2008 | Status: Trader | 6,736 Posts
today's range is already 300+ pips! Did anyone manage to catch this train when it left around 5am? Just curious. As usual, I was asleep during the good trade
Time hides Nothing
- #855
- Nov 24, 2008 10:57am Nov 24, 2008 10:57am
- Joined Jan 2007 | Status: MTF Stochastics and Volume/Price | 1,133 Posts
I think there are some components that can fit well into the EA format that rinfg the alarm bells in our head. Whether we could fully automate and take out our instincts and views of movement will undoubtably cause the worst results of the system. Having said that, we could certainly try to piece together the different MTFS and PMTFS indentifiers and signal what we are in and what we should trade. It may be under that process we can define parameters to twiddle with to try and devlop a more automated system. I've actually been toying with that on a simpler scale with which developed my Lab Accident thread but it still needs a few refinements I think...but certainly that is more balck and white.
The Escalator to Pips is a graduated occurance and typically it can start with M1 and m5 going high with M15 and M30 (even H1,H4 and Daily) very low or even well below 20. The best case scenerio is when each earlier time frame starts to go in an overbought market and the longer time frames are sucked up to the top....a glorious experience. I think this is very very difficult to transform into an EA since there are so many judgements required along the way to stay in the trade.
The M5/M15 or M1/M5 crossover is much better for EA'ing. Here we can define say that we need the 100,3,3 stochastic to be above 85 or below 15 (variable to test) and the crossover in the right direction. Or that the 14,3,3 is below 10% (variable) and price above 90% (variable) and trade the M5/M15 crossover to get the 14,3,3 stoch to move to the 100,3,3.
In itself the M5/M15 crossover is quite a good tool but it does need some filters to go with it...what the best filters are is where the EA would be most helpful. Creating an EA with a wide variety of filters like the one in the above paragraph can certainly be tweaked on a whole host of different stochastic properties.
I may simply workout that if we have a chart full of MTF conditions (say M5 crossing M15 at xx%, 100,3,3 at xx%, all MTF stochastic levels last crossed 20 or 80, 14,3,3 stochastc at xx%, index channel 6,1,1,to 13,1,1 spread or converged..peaked when and where, 15,1,1 to 24,1,1 spread or converged peaked when and where, and so on and so forth)...this type of one screen knowledge would probably lead us to better define an EA overall...with some work I'm sure many of you have detected or found certain Stochastic indications that work for you best or have modified what I have done to fit more black and what trade decisions..these can beexamined and added as well.
Obviously a dauniting proposition, but I'm certainly willing to spend time working through each parameter if someone(s) is willing to suffer and program it. As well as having people who use the systems to throw in their findings and advice.
The Escalator to Pips is a graduated occurance and typically it can start with M1 and m5 going high with M15 and M30 (even H1,H4 and Daily) very low or even well below 20. The best case scenerio is when each earlier time frame starts to go in an overbought market and the longer time frames are sucked up to the top....a glorious experience. I think this is very very difficult to transform into an EA since there are so many judgements required along the way to stay in the trade.
The M5/M15 or M1/M5 crossover is much better for EA'ing. Here we can define say that we need the 100,3,3 stochastic to be above 85 or below 15 (variable to test) and the crossover in the right direction. Or that the 14,3,3 is below 10% (variable) and price above 90% (variable) and trade the M5/M15 crossover to get the 14,3,3 stoch to move to the 100,3,3.
In itself the M5/M15 crossover is quite a good tool but it does need some filters to go with it...what the best filters are is where the EA would be most helpful. Creating an EA with a wide variety of filters like the one in the above paragraph can certainly be tweaked on a whole host of different stochastic properties.
I may simply workout that if we have a chart full of MTF conditions (say M5 crossing M15 at xx%, 100,3,3 at xx%, all MTF stochastic levels last crossed 20 or 80, 14,3,3 stochastc at xx%, index channel 6,1,1,to 13,1,1 spread or converged..peaked when and where, 15,1,1 to 24,1,1 spread or converged peaked when and where, and so on and so forth)...this type of one screen knowledge would probably lead us to better define an EA overall...with some work
I'm sure many of you have detected or found certain Stochastic indications that work for you best or have modified what I have done to fit more black and what trade decisions..these can beexamined and added as well.Obviously a dauniting proposition, but I'm certainly willing to spend time working through each parameter if someone(s) is willing to suffer and program it. As well as having people who use the systems to throw in their findings and advice.
- #856
- Nov 24, 2008 11:11am Nov 24, 2008 11:11am
- Joined Jan 2007 | Status: MTF Stochastics and Volume/Price | 1,133 Posts
I would agree that I can sometimes run on many tangents. I have and will continue to work on simplifying a single solution that can carry us into profitability. You have to remember this is my mind on paper and the many ways I use Stochastics....I also try and present the same ideas in different visual forms so that those that don't see it one way might see it the otherw ay. As such, what you see as many different "systems" or strategies are in their root foundation all the exact same thing in how the stochastics move together. I take the approach...
Just focus on the pips. The rest is insignificant.
- #859
- Nov 24, 2008 12:23pm Nov 24, 2008 12:23pm
- | Joined Nov 2008 | Status: Trader | 1,344 Posts
__Thanks__ MJ
