PTT auto trading robot by Steve Hopwood: D1 IB strategy

SERIOUS WARNING

 

1. Most Forex traders lose all their money.
2. Using the robot posted here in trading Forex does not guarantee success.
3. Trading this robot could lead to serious financial loss.
4. Trading this robot without understanding its underlying trading strategies guarantees traders will lose their money.
5. This is not a set-and-forget ea; there is no such thing and anyone who tries to claim there is, is either stupid or lying. This ea requires some manual intervention.
6. At best, a trading robot is only 90% as good as the manual strategy it trades. At best. At worst, it can be much less effective. If the strategy is rubbish, so is the robot.
7. To trade this robot, you have to understand:
   1. the trading strategy posted at http://www.forexfactory.com/showthre...89#post4887789
      

Introduction

Eamonn pm'd me asking if I thought KelvinM's strategy could be automated; it is a prime candidate. Warm round of applause, Kelvin, for posting it.


A strategy rarely comes along that is more simple than this. It will consume little of your time and computer resources; run the platform a few minutes before the start of your next D1 candle and for about 15 minutes afterwards, then it can be shut down for the rest of the day (unless you want PTT to carry out step 8 below). There are three key posts in Kelvin's thread:

 

1. http://www.forexfactory.com/showthre...89#post4887789
2. http://www.forexfactory.com/showthre...68#post4898168
3. http://www.forexfactory.com/showthre...99#post4898199

So, my understanding of all this is:

 

1. Load a D1 chart with a 50 EMA.
2. Pull out as far as possible to see as many candles as possible. Ascertain the overall trend from this.
3. When I can see a visible trend, trade the bounce off the 50 EMA in the direction of the trend, as long as the market is the right side of EMA and EMA has a decent slope in the direction of the trend, so the requirements are:
   1. Buy:
      1. rising trend
      2. upward EMA slope
      3. market is above EMA
      4. market has bounced off the EMA
      5. place a buy_stop trade at the high of the IB

   2. Sell:
      1. falling trend
      2. downward EMA slope
      3. market is below EMA
      4. market has bounced off the EMA
      5. place a sell_stop trade at the low of the IB

4. Stop loss 30 pips.
5. No take profit.
6. Use the hi-lo of the previous candle as a jumping stop once the trade is in profit.
7. Rinse and repead steps 3 to 6 whenever a new IB presents.
8. If a filled trade hits its stop loss, replace the pending trade at the same price as the original. Disable this feature by turning ReplaceTradeAfterStopLossHit to 'false' if you do not require it.

It is only possible to code an approximation of Step 3. The more a strategy depends on eyeballing charts, the less accurate automation becomes. I am going to leave PTT as a semi-automated robot, so to use it:

 

1. Carry out steps 1 & 2 from the list above.
2. Once the bounce has occurred (step 3) drag the robot onto the chart:
   1. Select your lot size.
   2. Select your trend direction.

From here, PTT will do the rest. Check your charts once a day so you can decide whether to leave PTT running or remove it.

Unless there are any bloops in the code, this EA should run ootb apart from the lot size and trend selection inputs. There is a TP option, but using it is not part of the strategy; I include it only because I know some people like to have the option.

Swap costs
Watch the swap display. If you trade the high negative swap AUD, CAD & NZD pairs, the swap costs are shocking.


Avoiding overtrading
To help avoid overtrading, I have included the usual margin filters:

 

1. ScoobsMarginCheck: will stop trading if your account margin > your free margin divided by 100
2. ForexKiwi: will stop trading if your margin percent drops below your selected input in FkMinimumMarginPercent.

I also suggest that users pay heed to balanced trading. There are two filters to help with this:

 

1. UseZeljko: this ensures that a second trade including one of the currencies already being traded is 'balanced. The easiest way to explain this is with an example:
   1. Buy GBPxxx open, so only sell GBPxxx. So, if the buy is GBPUSD, then only sell GBPJPY, GBPCHF etc. The rationale is that forex markets are highly sensitive to big movements caused by international events. If something drastic hits the GBP and you are in two sell positions, then both are heading rapidly towards their stop loss. If one of the trades is a buy and the other a sell, then one is going down the tube but its balancing buy is heading for the stratosphere.

2. OnlyTradeCurrencyTwice: using the balanced pair example, the GBP is now involved in two balanced trades. Another trade involving the pound will unbalance the trades and leave the trader exposed to a horrible counter-movement. This filter prevents the GBP from being involved in any further trading of different pairs.

Two versions
There are two versions of the bot available:

 

1. The semi-automated version that works as described above.
2. A fully-automated version:
   1. this needs the All Averages indi residing in your indicators folder. You do not need it on your chart. Go to the Spider thread linked at the end of this post to read about it.
   2. you will find the full panoply of trend-detection. The one that approximates most closely to the manual method is the All Averages D1 50 - the default. Note that this does not detect a decent slope, only a rising/falling moving average.
   3. it is hard to code bounce detection. I have attempted to do this with the LookBackCandles input. When PTT finds an inside bar that is the correct side of the MA and is in accordance with the trend, it looks for a bounce. Taking a potential buy as an example:
      1. the Low of one of the previous LookBackCandles must have been at or below the moving average.
      2. if so, then one of the highs in the previous candles up to LookBackCandles x 2 must have been higher than the MA.

The potential fly in the ointment is the bounce-detection routine. The lower your LookBackCandles input, the closer the market is likely to be to the MA and so the trade will be safer, but the more bounces you are going to miss. The further back in time you go to spot a bounce, then the further from the MA the market could be, with the trade correspondingly more risky. I can always code ludo's max pips from MA filter if this proves to be a problem.

If the automated version misses a bounce, then run the semi-auto version for a while instead.

Hanover
The Hanover module allows us to trade in the direction indicated by hanover's Recent Strength indi. I consider this to be the most significant development in my time at FF. Details in the Hanover thread athttp://www.forexfactory.com/showthread.php?t=31970. Only the fully automated version is Hanovered.


In conclusion

Have fun.




Read this post:
http://www.forexfactory.com/showthre...11#post4923611 Newbie traders, read this for inspiration and warning. Those of you who have suffered losses, read it for inspiration.



EA coding
I receive may requests to code EA's. Traders, if I code your EA for free, then I will share it here at FF. If you want me to code your EA without sharing, then my fee is $100 payable into my paypal account. I explain why at http://www.forexfactory.com/showthre...94#post4431794, at the bottom of the post.



Useful utilities/EA's/scripts

 

1. MaryJane has posted a fabulous utility to allow us to resize EA input windows: http://www.forexfactory.com/showthread.php?t=299017
2. Multi purpose trade manager: http://www.forexfactory.com/showthread.php?t=89371
3. Account equity analyser: http://www.forexfactory.com/showthre...16#post4693716 Also at http://www.forexfactory.com/showthre...98#post4709098, which includes a template for both USD and the Euro.
4. Past-trades display indi by squalou; shows a graphical display of the success/failure of trades in your History tab. Breathtaking piece of work. http://www.forexfactory.com/showthre...10#post3830610

Most promising trading EA's

 

1. Spider: http://www.forexfactory.com/showthread.php?t=313680. I am trading this live.
2. BigBob: an implementation of Nanningbog#b V10. http://www.forexfactory.com/showthread.php?t=309980
3. Trendy: http://www.forexfactory.com/showthre...82#post4840182, but should be left behind by BigBob.
4. Reversi: http://www.forexfactory.com/showthread.php?t=294255 – I am trading this live.
5. Rip: http://www.forexfactory.com/showthre...98#post4777998
6. The Beast: http://www.forexfactory.com/showthre...=1#post4135863

He does it again!

Should we rename this forum, "Steve Hopwood's Forex EA Factory"?

Thanks for another EA based on brilliant concept. Something exciting is happening in FF. Whereas up to couple of years ago the threads were full of people having half baked ideas now days people are putting forward tried and tested systems based on sound principals backed by live testing.

I think you are aiming to be in the Guinness book of records for the production of the highest number of EAs in one year. Many thanks,

Kind regards
p.s. I suggest traders choose the pairs according to Kevin`s criteria and only load the EA on the qualifying charts.

Making sure we are current.
Thanks Sir Hopwood Et Al

Bless!

I coded the bot because I had a few minutes to spare and had nothing better to do. I had missed this post at http://www.forexfactory.com/showthre...68#post4898168 (just reading K's posts in his thread).

This requires some additions to the ea - starting with a slope detector. I shall add this today.

In light of my comments in the 'Subsequent edit' bit of post 1, the EA could do with some refinement. At the mo, it is only coded with reference to Kelvin's post no 10 - the first of the three links I have posted.

Until I get around to adding some refinement, I suggest users only drag the robot onto charts that meet the criteria in the 2nd and third posts I have linked - hardly an onerous task.

Following a re-read of Kelvin's thread and a bath, I have made some changes to the EA and completely re-vamped post 1.

GBPCHF just needs an IB today and should fulfil all the criteria for an excellent short to test the EA.

Lovely. Cheers.

Hello Steve! Thanx for this one.. Has this bot ability to ride again the trade whe SL was hit according to Kelvin rules?

"One important thing about the stop loss…if the stop loss is taken out we enter the original trade again (a sell stop or buy stop exactly as the original entry—at the same exact price). This is because if the market retraces back to our order entry price, after having taken out our original stop loss, the momentum will usually continue, and we absolutely do not want to miss this trade."

I think this is the sexiest part of this concept.. What do You think?

I red this few years ago somewhere in commercial part of FF...

I doesn't at the moment; I read this statement in Kelvin's posts but forgot about it.

Coming up as soon as I can work out the coding logic........

Nice one. Cheers.

Done this, or think I have. It seems to work in strategy tester. The relevant input is ReplaceTradeAfterStopLossHit.

Update in post 1.

First trade taken.

this famous thread from 2007 based its success on a similar re-entry strategy, although it lets price go 50 pips in the wrong direction first:

http://www.forexfactory.com/showthread.php?t=27286

Steve

The replace after SL hit seems only to work if the SL is hit during the day after the inside bar. Using ST, I have run this were the sl gets hit a couple of times, then the next bar opens (so previous bar is no longer an IB) and the pending trade is not placed. From reading the original thread, KelvinM states he can keep a pending trade in for up to a week before it gets triggered. Therefore I think we need the ea to keep placing the pending trade at the high/low of the last IB, which may not necessarily be the last bar, so that the trade gets taken when the trend resumes. Obviously if the trade doesn't resume then we can delete the pending trade manually.

Paul

There is nothing in the code to force a wait until the next day for a replace.

Steve,

Thanks for the great work.

Would it be possible to add a lot size filter based on risk percentage into the EA.

Thanks for this Mr H, I was reading the PTT thread and thinking this would be a great bot, but thought you'd be too busy to make another EA - Presto, you've already made it. I'll take a slap for my dopiness and go and add it to some charts, but it is a great manual system so hopefully it transfers to the EA well.

Cheers and fine wine to you

Have fun.

Hi, Steve,


Great start but I see two bugs.

1. EA applied the same pending order price for most pairs, and
2. It duplicated trades ie, audcad and usdjpy trades.

See attach picture.