PipMaker v1 - Price action based EA - page 202
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
EAs cannot work for such a long period of time. You NEED a manual system. its the only way.
still cannot sleep using PM v17_3
I try using PM V17_3, this is the great EA but, when The direction change/ reverse I must close manually all open order, this ea cannot do that (if Some body considered to add it)
the point is EA must close all open order when the direction change/ reverse
(lihat gambar) ketika arah berubah dari bearish ke bullish, ea tidak close semua posisi open sell yang tersisa (no 1) sehingga ketika arah berubah ea tidak open buy dan ea nunggu arah bearish lagi yg itu merupakan bearish palsu (no3)
intine EA kudu close semua posisi ketike direction berubah arah dan open order tunggu filter & direction confirm semua
Buat para master yang bersedia, di modif lagi EA PM v17_3 ini benar2 dasyat
thank ya
Im a little confuses why this ea let a trade run almost 6 months, with one trade losing nearly 14000 and the other 'hedged' trade gaining only 5000 or so? Shouldn't the ea be smart enough to close out trades when trend reverses?
Well, when the trend reverses and starts losing, lets see, more than 10% of account?
Edit: very smooth and profitable equity curve until those two trades, and then no other trades were entered??? surely on a big trend down it would enter some trades.
Set EquityStop=true and MaxLossPercent at desired value.
Alright, so heres my long, vague rambling: statistical money management! Well, first i should say good job every1, this looks really neat. OK, so here i shall ramble:
Well, the equity protector setting helps alot! I was thinking, i think people are having trouble defining when the trend has turned without making the others unprofitable?
How about this for money management (in addition to current strategy) : you determine what the most negative floating equity percent loss was for that trade. Then you make a 'distribution' curve in which you have all the wining trades (look up 'normal distribution curves', it probably wouldn't look like that, but you get the idea, it would actually be half of a curve)
So, now you have this statistical curve which adapts over time which represents how many profitable trades had a max floating equity loss of whatever (or over market conditions, one per condition?). So the question is: if most of the trades do not have an floating negative equity of more than x percent, than why should we ever let a trade go over that? Certainly not out of a whimsical hope for it to win! Of course, the past doesn't mean the present,so maybe taking historical data for the curve over long periods wouldn't be too good, but something along this lines?
Now, lets consider this case. Lets say we place an equity 'protector' (a setting of the ea) at 15%, but too many trades are not working, but would have eventually been profitable if we had let it go above 15% and then back up! Well, now lets take that distribution curve and define a non-hedging area - say +-4 % equity (or whatever is optimal, based on curve). Anything beyond that we hedge, as we know that it statistically likely that either it will continue going down before it turns up or it will be a loser. Now we have protected even more of our equity at the cost of a small spread. Now, if our price pulls up, then thats great. If not, then youve protected yourself and can safely close the trade when the original trade would have gone 50% equity loss or whatever.
Now, i have an amusing (in my humble opinion
) equity curve for the ea with setting 15% equity protector, eur/usd, 1 hour chart from , 6/18/2009-4/18/10, one direction only = false, maximum risk = 2, micro account = false. Lol, i wonder if those straight lines correspond to good trends for the euro/usd. 
I haven't yet taken a look at why those big drops occur, that would probably be useful to know (anyone?, i haven't bothered reading 200 pages, too much!)
And then i have a screenshot with 50% equity protection. See where the above ideas could help out?? Many of those 15% losers could've actually been winners, we just need to optimize how we hedge based on history, how many times can we can hedge one trade, if it goes up , down , up , but then back down should we still hedge, etc. All based on the statistics our ea keeps.
Now this is just our 'base line.' If you think you have a tactical advantage to break the statistics based on a certain indicator, than the statistics should be ignored, and an exception noted/categoried in the ea's statistical history.
Remember, our goal is NOT to have an ea in which you 'optimize' for a certain period of time. The goal of ea optimization should NOT be to optimize it into an unstable setting which might create a 'bomb' moment later on. That is just an incorrect way of thinking. Our goal is too have one which adapts to all conditions (and if it can't, stop trading without too much loss), even the worst, and is consistent (well, easier said than done
). Therefore, for consistency, we should not expect our trades to be floating into an equity loss of more than x % (15?) Otherwise, we're just gambling with luck, no sense in winning all the time and then losing 50%, is there?
Now, i know this might all seem like rambling, but i know stock portfolio optimization uses math/statistics and future predicted volatilities vs returns, so maybe this is a viable idea. The idea of the normal-like curve comes from monte carlo simulation using historical data with a 'bias' inserted, btw, which is also used in stock portfolio optimization...
hmm.. i think my next ramble will be about how we can identify trends with our different equity curves with different equity protections. straight line = trend, volatile line = no trend, if on lower equity protection = better trend. and then theres the equity curve abnormal pattern recognizer = large spike, better watch out (base on those equity curves above!), change tactics to keep profit. haha, just joking...
EAs cannot work for such a long period of time. You NEED a manual system. its the only way.
i disagree, i think someone somewhere does have a good EA!
Alright, so heres my long, vague rambling: statistical money management! Well, first i should say good job every1, this looks really neat. OK, so here i shall ramble:
Well, the equity protector setting helps alot! I was thinking, i think people are having trouble defining when the trend has turned without making the others unprofitable?
How about this for money management (in addition to current strategy) : you determine what the most negative floating equity percent loss was for that trade. Then you make a 'distribution' curve in which you have all the wining trades (look up 'normal distribution curves', it probably wouldn't look like that, but you get the idea, it would actually be half of a curve)
So, now you have this statistical curve which adapts over time which represents how many profitable trades had a max floating equity loss of whatever (or over market conditions, one per condition?). So the question is: if most of the trades do not have an floating negative equity of more than x percent, than why should we ever let a trade go over that? Certainly not out of a whimsical hope for it to win! Of course, the past doesn't mean the present,so maybe taking historical data for the curve over long periods wouldn't be too good, but something along this lines?
Now, lets consider this case. Lets say we place an equity 'protector' (a setting of the ea) at 15%, but too many trades are not working, but would have eventually been profitable if we had let it go above 15% and then back up! Well, now lets take that distribution curve and define a non-hedging area - say +-4 % equity (or whatever is optimal, based on curve). Anything beyond that we hedge, as we know that it statistically likely that either it will continue going down before it turns up or it will be a loser. Now we have protected even more of our equity at the cost of a small spread. Now, if our price pulls up, then thats great. If not, then youve protected yourself and can safely close the trade when the original trade would have gone 50% equity loss or whatever.
Now, i have an amusing (in my humble opinion
) equity curve for the ea with setting 15% equity protector, eur/usd, 1 hour chart from , 6/18/2009-4/18/10, one direction only = false, maximum risk = 2, micro account = false. Lol, i wonder if those straight lines correspond to good trends for the euro/usd. 
I haven't yet taken a look at why those big drops occur, that would probably be useful to know (anyone?, i haven't bothered reading 200 pages, too much!)
And then i have a screenshot with 50% equity protection. See where the above ideas could help out?? Many of those 15% losers could've actually been winners, we just need to optimize how we hedge based on history, how many times can we can hedge one trade, if it goes up , down , up , but then back down should we still hedge, etc. All based on the statistics our ea keeps.
Now this is just our 'base line.' If you think you have a tactical advantage to break the statistics based on a certain indicator, than the statistics should be ignored, and an exception noted/categoried in the ea's statistical history.
Remember, our goal is NOT to have an ea in which you 'optimize' for a certain period of time. The goal of ea optimization should NOT be to optimize it into an unstable setting which might create a 'bomb' moment later on. That is just an incorrect way of thinking. Our goal is too have one which adapts to all conditions (and if it can't, stop trading without too much loss), even the worst, and is consistent (well, easier said than done
). Therefore, for consistency, we should not expect our trades to be floating into an equity loss of more than x % (15?) Otherwise, we're just gambling with luck, no sense in winning all the time and then losing 50%, is there?
Now, i know this might all seem like rambling, but i know stock portfolio optimization uses math/statistics and future predicted volatilities vs returns, so maybe this is a viable idea. The idea of the normal-like curve comes from monte carlo simulation using historical data with a 'bias' inserted, btw, which is also used in stock portfolio optimization...
hmm.. i think my next ramble will be about how we can identify trends with our different equity curves with different equity protections. straight line = trend, volatile line = no trend, if on lower equity protection = better trend. and then theres the equity curve abnormal pattern recognizer = large spike, better watch out (base on those equity curves above!), change tactics to keep profit. haha, just joking...IF you solve this , you can quit yor job
Those big drops occur when EA place an order at end of trend.
Indicators say is in trend, but cannot say is trend is ending, so EA place a order then start incremental grid (cost averaging, martingale or whatever you want to call it).
Better to limit number of trades and have enough distance between orders, or just allow one open trade.
Or use stop loss.
To make statistics it should write somewhere various parameters(file, database), is not possible to calculate maximum drawdown only from history.
lol X) "Adaptive optimal hedging based on probability of maximum floating equity loss per trade" Well it would be fun to try
Edit: well, it looks like this has been studied before. Efficient hedging: "Here we look for strategies which minimize the shortfall risk defined as the expectation of the shortfall weighted by some loss function. The resulting efficient hedges allow the investor to interpolate in a systematic way between the extremes of no hedge and a perfect (super-) hedge, depending on the accepted level of shortfall risk." All theory, of course...
Auto Trailing the Max loss percent
Set EquityStop=true and MaxLossPercent at desired value.
The fix MAxLoss Percent good when the market reverse, but little bad when the market tranding
Hallo Mr.Enforcer, How about to trailing the max loss percent, example
using TMA
1. start max loss percent = 50%
2.when the fast MA cross the MediumMA + 5pips, the max loss percent = 25%
(+ 5 pips, keep the false signal cross MA)
3. when fast MA cross slow MA, the max loss percent = 10%
Hello, first of all I want to thank you for that EA
it's looks really amazing
but I can't test it I have the EA in /experts and indicators in /experts/indicators but when I want to test it journal says
"2010.04.24 02:08:27 2010.04.23 11:34 Pipmaker_V17_3: the comment parameter for OrderSend function must be a string"
"2010.04.24 02:08:27 2010.04.23 11:34 Pipmaker_V17_3 EURUSD,M5: OrderSend error 4062"
someone can help ?
Thanks