learn how to earn money villagers [Episode 2] ! - page 28

 
I don't have any authority with Soros ;)
 
YOUNGA:
I don't have any authority with Soros ;)
That was the point: "Let's also look at them and take them into account... who's in the way?" :-)
 

Hello, villagers!!!! Are you still alive? More specifically your deposits!? :)))))


I have a present for you. I`ve been developing my Expert Advisor this winter, and then I gave up and abandoned it. But now I`m going to finish it again and at least made a version for public use.

The name of the Expert Advisor is "ThirdHand.Tractor" or simply Tractor. Based on the well-known Expert Advisor Buldozer, it was named! ;))) The Expert Advisor works based on the ATR_Bands indicator. But it doesn't require an indicator. I wrote the indicator myself. But there is its analogue in CodeBase. I found it much later! :))))))


Currency pair: any, timeframe: any.

I only tested it on EUR/USD. The best solution for the parameters implied in the Expert Advisor is the TF M5.

Parameters attached to the Expert Advisor are for four-digit symbols. If you want a five-digit number, multiply by 10 the parameters highlighted in bold below. There are only four of them.

The Expert Advisor has no permanent magic number. One magic number for one series of orders. This was done in order to calculate the profit/loss of any series irrespective of the fact whether all orders are in the market or some of them have already been closed. When a new series is opened, the Expert Advisor searches for the free magic number and opens the order with it. This feature is problematic when trading on the real account because after reinitialization the EA will not pick up current magic numbers of orders, unless they are written manually in the code! The orders will hang until you close them manually or enter the magic numbers in the code! There are a few more subtleties... But there's no point in writing about them. The Expert Advisor is not ready to trade on a real account!


Now let's talk about the parameters themselves, which are not few:

Setting MM:
- Risk if 0 - lot constant; Risk > 0 - lot dynamic;
- Lots, constant lot (if Risk = 0);
- LotsExp, by how much to multiply the lot when placing the next knee.

Slippage:
- Slippage.

Hold Mode:
- ModeHold, enables/disables hold mode, which tries to balance the drawdown from BUY and SELL orders.

Signal setting:
- SloserAtCrossWithMA, whether or not to close when it crosses the MA;
- FromWhatOrderClosed, from which knee the orders are to be closed when they intersect the MA (SloSeAtClosewhidewhMA should be True);
- HoldSignal, hold the signal until the action or cancellation;
- CloseOnlyForBands, to close the oredre only behind the bars of the indicator or not;
- CloseOnlySeries, close only the entire series (the Expert Advisor will become similar to Buldozer's) or profitable orders separately.

Settings of stops (virtual-conditional) and step:
- MinTP, minimum level of profit to be taken (in pips);
- MinSL, the minimum level of loss that will be taken (if 0, the loss level is disabled);
- MinPipStep, minimum distance from the last order.

Why virtual? TakeProfit and StopLoss are not set for an order. Why are they conditional? The Expert Advisor is waiting for the signal to close the order. If there is no signal but the order has passed the minimum profit or loss level, the Expert Advisor will not close the order and will continue to wait for the signal.

Settings for the indicator:
- ModeBands, St0 - ATR Bands; 0 - Bollinger Bands;
- Per, moving average period;
- Dev, Deviation multiplier;
- MinStep, minimum step (band width) beyond which we open an order.

Overlapping mode:
- Overlapping, enabling/disabling the mode of overlapping a profitable order with a losing one;
- LeadingOrder, from which knee the overlapping works.

SafeMode:
- ExtremeDrawdown, if drawdown > ExtremeDrawdown, SafeMode is enabled (try to close BUY+SELL series and return to the balance where the losing series started).

Other settings:
- Debug, allow/forbid displaying additional messages in "Experts" tab;
- Info, allow/prohibit displaying of information on the chart;
- DrawArrows, allow/forbid to draw arrows when prices intersect MA and bars;
- ColorArrowUp, colour of arrows up;
- ColorArrowDn, colour of arrows down.


And almost immediately the second version of the Expert Advisor was released.

What's new?

1). Added variable ModeLadder to disable ladder mode. The reason is that the first version of the Expert Advisor can only open BUY orders below the previous BUY orders of a series, and SELL orders only above the previous SELL orders of a series. Therefore, the Expert Advisor has been missing good signals. Now you can disable this "ladder" if you want!

2). Completely changed the logic of the Hold mode, which is activated using the ModeHold variable. In this mode trades are closed only with Overlapping. If overlapping is disabled, the Expert Advisor will not close trades.

I still cannot make this mode complete. I have all kinds of thoughts in my head. I want one of the series of orders to not be too heavy otherwise it will quickly reach the level of drawdown. To do this, we should not close the whole profitable series but only a part of it. I cannot find the best way to do this yet, except for covering the orders by overlapping. All my suggestions should be sent here.

3). Dynamic lot has been fixed. In the first version it could not work!


ATTENTION: Do not put the Expert Advisor on the real, as well as any averaging. Now it is not the best time for the Expert Advisor in the market, we need to finalize TS. Also I have already written that after reinitialization the EA does not pick up current magic numbers of orders, unless they are written manually in the code (this is what I did when I went to the real account)!


I attach two versions of the Tractor Expert Advisor and the indicator to this post. Please share your impressions! ;))) What a toy! :DDD

 

By the way, I lost my real account in May. I installed the Expert Advisor in late March. Increased my deposit almost 3 times. I was relying on a sluggish market. My hopes did not come true. You can say the Expert Advisor was developed for this sluggishness. Below is a test that didn't leave My terminal till May. That's what made me even more optimistic and hopeful! ;)


Test with aggressive Risk parameter settings (EUR/USD, test period: 02.01.2012-28.04.2012, deposit: 10,000, leverage: 1:100):

Strategy Tester Report
ThirdHand.Tractor
FBS-Demo (Build 432)

SymbolEURUSD (Euro vs US Dollar)
Period5 Minutes (M5) 2012.01.02 02:00 - 2012.04.27 22:55 (2012.01.02 - 2012.04.28)
ModelAll ticks (most accurate method based on all smallest available timeframes)
ParametersParamerts01="MM setting:"; Risk=6.5; Lots=0.1; LotsExp=1; Paramerts02="Slippage:"; Slippage=2; Paramerts03="Hold mode:"; ModeHold=false; Paramerts04="Signal settings:"; SloSedWithMa=false; FromWhatOrderClosed=2; HoldSignal=false; CloseOnlyForBands=true; CloseOnlySeries=false; Paramerts05="Stop and Pitch Settings:"; MinTP=10; MinSL=0; MinPipStep=10; Paramerts06="Settings for indicator:"; ModeBands=0; Per=45; Dev=3; MinStep=1; Paramerts07="Overlapping mode:"; Overlapping=false; LeadingOrder=2; Paramerts08="SafeMode:"; ExtremeDrawdown=100; Paramerts09="Other settings:"; Debug=false; Info=false; DrawArrows=false; ColorArrowUp=Blue; ColorArrowDn=Red;

Bars in history25161Modelled ticks1351017Simulation quality25.00%
Chart mismatch errors0




Initial deposit10000.00



Net profit153860.61Total profit219048.05Total loss-65187.44
Profitability3.36Expected payoff381.79

Absolute drawdown4345.24Maximum drawdown44149.16 (43.76%)Relative drawdown88.53% (43631.51)

Total trades403Short positions (% win)202 (78.22%)Long positions (% win)201 (83.08%)

Profitable trades (% of all)325 (80.65%)Loss trades (% of all)78 (19.35%)
Largestprofitable trade4526.00losing deal-7975.58
Averageprofitable deal673.99losing deal-835.74
Maximumcontinuous wins (profit)21 (13906.45)Continuous losses (loss)5 (-7478.67)
MaximumContinuous Profit (number of wins)18313.60 (9)Continuous loss (number of losses)-13352.52 (4)
Averagecontinuous winnings6Continuous loss1

It is not certain that other quotes will pass the test. But quite possible! :)))) If anything, reduce the Risk parameter. Better yet, stop messing around with the strategy tester! ;) This set is built into the Expert Advisor.


After some sit and twisting, I made Tractor EA trade in the style of famous Buldozer EA. Also there are a lot of orders in the market, short targets, tight averaging... Well, this kind of trading may make you go crazy! The set is called "BuldozerStyle".

It's not pouring in May. But the test results leave a lot to be desired! ;))

This is what the test looks like (EUR/USD, testing period: 02.01.2012-30.06.2012, deposit: 10 000, leverage: 1:500):

Strategy Tester Report
ThirdHand.Tractor
FBS-Demo (Build 432)

SymbolEURUSD (Euro vs US Dollar)
Period5 Minutes (M5) 2012.01.02 02:00 - 2012.06.29 23:00 (2012.01.02 - 2012.06.30)
ModelAll ticks (most accurate method based on all smallest available timeframes)
ParametersParamerts01="MM setting:"; Risk=0.16; Lots=0.1; LotsExp=1.6; Paramerts02="Slippage:"; Slippage=2; Paramerts03="Hold mode:"; ModeHold=false; Paramerts04="Signal settings:"; SloSedWithMa=false; FromWhatOrderClosed=2; HoldSignal=false; CloseOnlyForBands=true; CloseOnlySeries=false; Paramerts05="Stop and Pitch Settings:"; MinTP=10; MinSL=0; MinPipStep=10; ModeLadder=true; Paramerts06="Settings for indicator:"; ModeBands=0; Per=3; Dev=0.5; MinStep=0; Paramerts07="Overlapping mode:"; Overlapping=false; LeadingOrder=2; Paramerts08="SafeMode:"; ExtremeDrawdown=100; Paramerts09="Other settings:"; Debug=false; Info=false; DrawArrows=false; ColorArrowUp=Blue; ColorArrowDn=Red;

Bars in history37973Modelled ticks2035571Simulation quality25.00%
Chart mismatch errors0




Initial deposit5000.00



Net profit74016.46Total profit176496.16Total loss-102479.70
Profitability1.72Expected payoff18.34

Absolute drawdown2997.44Maximum drawdown23228.50 (54.54%)Relative drawdown82.51% (9448.71)

Total trades4036Short Positions (% Win)2003 (62.46%)Long positions (% win)2033 (60.65%)

Profitable trades (% of all)2484 (61.55%)Loss trades (% of all)1552 (38.45%)
Largestprofitable trade4700.00losing deal-1170.25
Averageprofitable deal71.05losing deal-66.03
Maximum numbercontinuous wins (profit)13 (854.83)Continuous losses (loss)10 (-3764.32)
MaximumContinuous Profit (number of wins)9635.44 (4)Continuous loss (number of losses)-5321.54 (7)
Averagecontinuous winnings3continuous loss2

What can I say about the test? The chart shows a lot of snot (drawdown of half of the depo), short targets are involved (which makes the test in the tester unimportant), low report figures...

Both tests were obtained with the second version of the Tractor EA.


Just for the sake of curiosity, I ran the Expert Advisor with the "BuldozerStyle" set on a demo.

EUR/USD pair, leverage: 1:500, deposit: 10 000, starting lot: 0.06.

I will also post here the demo account number and investment password:

Account number: 245551
Investor password: 0yh3shw
Server: FBS-Demo

The Expert Advisor has been tested since Monday (09.07.2012). Yesterday was a slow day. I did not see anything interesting! ;))))))


I attach the "BuldozerStyle" set to this post. But this set so far on the demo I am not satisfied. We need a long and persistent hours to optimize it!!! Which I do not want to do...

Files:
 

What are the adviser's innovations compared to other averagers I've seen here (although I haven't been here for a month or two)?

1). I tried to find a decent indicator for figuring out flat areas of the market... I think I've found something. I use this indicator in manual trading as well. But I still have not learned how to interpret its readings correctly! :)))


And not exactly an innovation, but still:

2). Virtual conditional stops.

MaxZ:

Why virtual? TakeProfit and StopLoss are not set for an order. Why are they conditional? This EA is waiting for the signal to close the order. If there is no signal but the order has passed the minimum profit or loss level, the Expert Advisor will not close the order and will continue to wait for the signal.

I am not satisfied with fixed ones. We exit at random. Usually, we enter (open a new series) just as we left it, again at random.

There is another test worth citing here as well. It will show how the indicator works on history. We leave profit level (MinTP) and include the loss level (MinSL) in the Expert Advisor. Testing results (EUR/USD, testing period: 02.01.2012-30.06.2012, deposit: 10 000, leverage: 1:100, trading at a fixed lot: 0.1):

Strategy Tester Report
ThirdHand.Tractor
FBS-Demo (Build 432)

SymbolEURUSD (Euro vs US Dollar)
Period5 Minutes (M5) 2012.01.02 02:00 - 2012.06.29 23:00 (2012.01.02 - 2012.06.30)
ModelAll ticks (most accurate method based on all smallest available timeframes)
ParametersRisk=0; Lots=0.1; LotsExp=1; Slippage=2; ModeHold=false; Slippage=false; FromWhatOrderClosed=2; HoldSignal=false; CloseOnlyForBands=true; CloseOnlySeries=false; MinTP=10; MinSL=10; MinPipStep=1; ModeBands=0; Per=45; Dev=3; MinStep=1; Overlapping=false; LeadingOrder=2; ExtremeDrawdown=100; Debug=false; Info=false; DrawArrows=false; ColorArrowUp=Blue; ColorArrowDn=Red;

Bars in history37973Modelled ticks2035571Simulation quality25.00%
Chart mismatch errors0




Initial deposit10000.00



Net profit1930.05Total profit15254.79Total loss-13324.74
Profitability1.14Expected payoff2.44

Absolute drawdown551.22Maximum drawdown1561.68 (13.32%)Relative drawdown13.32% (1561.68)

Total trades790Short positions (% win)391 (71.61%)Long positions (% win)399 (67.17%)

Profitable trades (% of all)548 (69.37%)Loss trades (% of all)242 (30.63%)
Largestprofitable trade188.99losing deal-159.02
Averageprofitable deal27.84losing trade-55.06
Maximum numbercontinuous wins (profit)17 (522.93)Continuous losses (loss)11 (-685.09)
Maximumcontinuous profits (number of wins)522.93 (17)Continuous loss (number of losses)-685.09 (11)
Averagecontinuous winnings5continuous loss2

The first long fall is the end of February, beginning of March. And the balance and funds flat in the last third of the chart is May and June.

I am attaching the set file to this post. You can pick up the MinTP and MinSL parameters as well as the indicator settings and get sweeter test results... Only why??? :DD


3). Also I am not satisfied with a fixed step for opening an order. Even if it's there with coefficients of some kind.

If the series goes in the red and a trend starts, the Expert Advisor does not open with a certain step up to the end, it just waits for the signal. But it will not help save us from the trend anyway.


I haven't been following Selian's branch, maybe something new has been invented? For example, some kind of trend-following two-way? Not an avalanche! ;) By the way, if we set this EA differently (by making variable MinTP very big and setting MinSL ~ 50-150), then it becomes like an avalanche... Only with stops! ;))

Files:
 

2MaxZ:

THIS IS THE BOMB!!! Pulls a thesis!

Try translating it to netting and posting it in the marketplace fives! :-)

Max, if you're familiar with the subject (code) of this owl "from" and "to" - it should be fine-tuned for the real world, to solve the issues with these magicians...

I have just prepared a platform for the real world!

There are several owls working there + more on the way...

The layouts on the owls and settings will be posted in a branch later, if there is a live real.

 
Roman.:

THIS IS THE BOMB!!! It's a thesis!

Try translating it into netting and posting it in the marketplace fives! :-)

Max, if you are familiar with the subject (code) of this owl "from" and "to" - it should be fine-tuned for the real world, to solve the issues with these magicians ...

What's your diploma's speciality? "Ilanivod"? :DD

And how do you translate it into netting? :))))))

Yes with magicians the question can be solved, add OrderComment() for example... Or you can use global variables. Or even easier, after re-initialization type in magicians by yourself.

But I don't want to use this leaky agent for real... Sooner or later it will sell, as well as any averaging. Of course, it can make money before that. And it's even possible to withdraw something. But all this is a lottery! ;)))


Roman.:

I've got a site ready for the reals!

There are several owls working there + more to come...

I will add the breakdown of the owls and the settings a little later, if I have a real one.

You better tell me, the villagers survived the month of May well? :DD

 
MaxZ:

1. What's your diploma speciality? "Elanologist"? :DD

2. and netting - how? :))))))

3. Yes with magicians the question can be solved, add OrderComment() for example... Or you can use global variables. Or even easier, after re-initialization to enter magicians by oneself, what I did.

4. But I don't want to put this plummer on the real... Sooner or later he will sell, like any averaging. Of course, it may work till then. And it's even possible to withdraw something. But all this is a lottery! ;)))


1. A lot of people here have been asking for "summer internships"... And there for a diploma! after such BOMBO-PRACTICE and DEPA! :-)

2. one tool - one pose in the market. This question (trading both in buy and sell simultaneously on one account for one instrument) was raised once on the fifth forum - something like virtual order bookkeeping should be organized there... I do not have the exact link at hand.

How does it work? I.e.: simply reinsert the values in the tab "external variables" in the terminal? If so, or something similar (where to enter?), then it's not a problem - I myself have Lavina (now it also starts again at me), when it was still a little crude (does not pick up the control number of coups Owls to calculate the next lot), entered in external variables, these current values coups and all.

4. I love lotteries! :-)

 
MaxZ:

...You better tell me, did the villagers survive the month of May OK?? :DD


I guess not. There's still last episode to watch. I got one of the owls bitten before Talogo on futobax and didn't take it out on a no-return... The deep was drained, which is not the case with the first two owls - I'll post later with the settings... Testing on history over this period - passing fine.

Poured the DEP, left the same settings - started up again...

 
Roman.:


1. Many people here have asked for "summer practice" themes for themselves... And then for a diploma! After SUCH BOMBO PRACTICE and DEPUTY! :-)

2. one instrument - one pose in the market. The issue (of both buy and sell trading simultaneously on one account for one instrument) was raised once on the fifth forum - something like virtual order bookkeeping should be organized there... I do not have the exact link at hand.

How does it work? I.e.: simply reinsert the values in the tab "external variables" in the terminal? If so, or something similar (where to enter?), then it's not a problem - I myself have Lavina (now it also starts again at me), when it was still a little crude (does not pick up the control number of coups Owls to calculate the next lot), entered in external variables, these current values coups and all.

4. I love lotteries! :-)

1. I read... And just like Some had deja vu! :DD

2. If you find it, send me the link... By the way, I'm thinking of converting this EA to mql5 for the Championship. I will have to create multicurrency EA only! :DD And I will have to learn how to balance the tools portfolio! ;))

3. These variables are not in the tab (they are not external), but you can bring them out easily if you want to. The variables are called MagicB and MagicS. You look at the orders that are now in the market and if the EA has to be reinitialized, you type the Magic of those orders into the code of the EA! :) But we can also encounter some problems with the indicator there. If we reinitialise it on a peak, one more order will be opened. The EA also has a variable for this purpose, which is DirMA. It should also be set. This EA can work on a real account, but I wouldn't advise it... That's why I'm writing this as a complicated settings so that nobody would think about using it for real trading. :)))))))))))

4. I'd play the 99 out of 100 lottery... :DDDDD


Roman.:


I guess not. There's still last episode to watch. I had one of the owls bitten to Talogo on a futobax and did not take out on a recoil... The deep was drained, which is not the case with the first two owls - I'll post later with the settings... Testing on history over this period - passing fine.

Poured the deep, left the owl settings the same - started again...

Sadness... I also lost my money in May... True, I lost pennies. I had lost pennies, but it was still a shame. I thought this idea would work for a long time.


This EA does not make money on increasing lots... I remember there was such a question that the bigger the drawdown of the two-sided EA, the larger the profit will be... It's all nonsense... The bigger the drawdown, the sooner the drawdown will come! :DD

You need an EA that does not multiply lots... And it holds approximately the same ratio of lots in both BUY and SELL series... I'm still trying to use ModeHold:

MaxZ:

2). Completely changed the logic of the hold mode which is activated using the ModeHold variable. In this mode, trades are closed only using Overlapping. If the overlapping is disabled, the Expert Advisor will not close trades.

I still can't get this mode right. I have all sorts of thoughts in my head. I want one of the series of orders to not be too heavy otherwise the drawdown rapidly occurs. To do this, we should not close the entire profitable series but only a part of it. I cannot find the best way to do this yet, except for covering the orders by overlapping. You can send all proposals here.

Otherwise, when in one series a lot on half of the deposit, and the initial one on the other, the cream is very close already! ;)))

And just for this holding mode you need at least some more or less working signals... Closing orders in random order is not a good idea!!! :)))))))

Reason: