Jollyroger: an inside bar auto-trader by SteveHopwood

Latest update in post 1:

 

1. fixed a bloop where I had forgotten to add tp to the first sell trade
2. added Matt's OR stuff to check that a trade has been recognised by the platform
3. a couple of other thingies that are highly significant safety features that were not available when I coded Ibuks' ea.

One thing you guys should be aware of is this: the utter, total, unmitigated crap that is represented by MT4 makes coding nearly impossible - presumably deliberately.

One of the most impossible things to achieve is sending out two trades quickly. It cannot be done. Therefore, there will be occasions where:

 

1. the open price of the second trade differs from that of the first by several pips
2. the second trade will not always send at all.

The only way to overcome this is to send pending trades rather than market orders. Before we decide to go down this route:

 

1. remove JR from your platform
2. start a new account
3. re-download the bot and start again

Let's see what happens for a few days before making radical changes.

Steve: Would it be easier to double the lot size in the bot on the initial order and code in the TP for half the initial order letting the other half run? This would accomplish the same thing as placing 2 initial orders.
Not being a coder this might be a stupid suggestion but....

Roundrock: Great observations in your Post 20. I concur with your observations particularly that the EA should have the option to take ALL IB trades occuring during the trading period. While this would undoubtedly lead in many instances to multiple trades on the same pair in the same direction it is an integral part of the system.

The whole basis of the IB system as I see it is to establish as many trades as compatable with good money management in the direction of the long term trend with minimal loss in the setup of same. THEN TO LET THE TRADES RUN AS THE LONG TERM TREND CONTINUES with appropriate adjustment of the stops either manually or using Steve's MPTM. Just as you allow produce in a garden to ripen over time, harvesting as it ripens and weeding out weaker plants to allow stronger ones room to grow and eventually reaping a bountiful harvest.

Steve, I believe we’re off to a very promising start with this EA but there are a couple of things that need to be tweaked.

 

1. The EA needs to open 2 trades at the trigger line and not at the beginning of the next candle. *
2. EA needs to open trades in the proper direction. i.e. “Only take an upside breakout if the breakout price is higher than the day’s open and downside breakouts of an hourly IB if the breakout price is lower than the days open”.
3. The EA should open trades at all qualifying IB’s that occur during set trading hours.**
4. Lines defining IB’s need to be shortend to 3 or 4 bars so as not to clutter chart.

At this time I haven’t paid much attention to the TP settings on the “free” trades. After we get the bot placing orders properly we can evaluate the TP’s on the bots “free trades”.

*If it is difficult to code two trades to open at the same time at the same price (A few points would make little difference) would it be easier to code 2 pending trades at the proper trigger point when the IB first forms?
** If a breakout occurs in the wrong direction and the price reverses and re-enters the IB and exits in the proper direction no trade should be taken. A breakout in the wrong direction should cancel any pending or trigger of an order to be placed in the proper direction in that IB.

Steve, Your time and talent is appreciated. I hope that this EA develops as well and as profitably as has TB. I am a little disappointed however that this thread has not generated more interest. Hopefully that will occur as the EA becomes more successful.

Roger

Done and uploaded to post 1.
This is how the bot is coded.
This is how the bot is coded. You just have to put up with the clutter.

Fantastic! Don't you ever sleep?

What are the inputs TakeProfit1 & TakeProfit2?

Would they be a fraction of the lot size?

Steve, Great work. One small tweak needed in the bot. When the 2 initial orders are opened at the trigger line both orders should have the same stop loss but only one with a take profit. Right now the bot is leaving one order with no SL.

I am getting very excited on the progress of this bot. Many of the "free" orders are beginning to accumulate positive pips. Correcting the SL problem will get rid of the "free" orders that are accumulating negative pips.

Jollyroger, how do you manage the free trades?
In post #1 I read somewhere: Do you manually place the trailing stop every candle (when the LWMA21 moves) or do you use one of the managementfunction of the bot? (like candlestick trailing stop / jumping stop / trailing stop.)

I have it on demo now, OOTB.
I'll share myfxbook link here, so you can follow it progress
(click https://www.myfxbook.com/portfolio/jollyroger/87226 to go to the myfxbook page)
http://www.myfxbook.com/widgets/87226/medium.jpg

Prior to the bot I was using a 21 day LWMA to set the SL on the free trades. Since I don't really understand the methods included in the bot I am letting them run to see what the bot does. (currently set on default SL) I am also reading several of Steve's manuals on other trading methods to understand the choices coded into this bot. Hope this helps

If any reader of this thread could post relative posts on the three SL methods coded in the bot it would be appreciated.

The code for both trades is the same, so the same stop is sent. This shows up in backtesting.

If stops are not being sent with your demo trades, the problem is at your end. Make sure you have the latest version.

You could use Steve's MPTM to follow up and place missing SLs. You could also use it to do jumping stops to track the "free" trade.

Filling in missing SL manually. Jumping stops is the default of the EA and am waiting to see how it works since I don't backtest. I put it on demo and let it roll.

Hi JollyRoger

What I meant is that Steve's MPTM has a setting by which you can have it automatically add SLs to trades that it detects does not have them.

This will mean you don't have to chase round behind it entering them manually.

Cheers

Invisible

Thanks Invisible.
I'll check it out. Most of the orders have a SL but occasionally one slips through for whatever reason.

(My first post after all this time reading and its in one of Steve's threads, I must have lost it!)

I'm experiencing the same:

http://www.freebits.nl/images/378insidebar.gif

Two missing SL and one quite ambitious TP.

I understand the principle of the system, but not quite yet how this EA does it. If I trade this manually the GBPUSD trade SL would have been moved to at BE by now.
I also had one instance where the EA decided to close a position but the price moved on. It kept on trying to close it 500 times until the price had moved back...

Having said all this I'm really glad to see a EA emerge for this system.

JR just took 2 trades on EUR/USD for an unknown reason. Does anyone have a clue why the bot took the trade. Experts file indicated that the bot generated the trades? The TP on the trades is also most optimistic and does not match any system criteria. See attached:

That TP is hilarious. What does the bot know that we don't?

Well, the TP was exactly double the entry price. (I have since changed it to a more realistic level) Since I don't read code I can"t even speculate on what happened. The trade itself does not conform to the system. Maybe the bot should be renamed "Aladdin's Magic Robot". So far I am very pleased with the bot. I feel it has considerable potential but I really would like to know why it took that trade and what caused a TP of exactly double the entry price.

Should anyone want to follow my EA:
http://www.myfxbook.com/members/fate...ntal-ea/116966

So far I've only seen the missing stoploss problem in eurusd.
The order send error (127) happend twice at eurgbp.

there is a serious bug in the EA, i mentioned it in prev posts. as per DIBS, current day open price is as of 6 AM GMT. Whereas the EA determines this as yesterday close. so if yr broker doesnt close the day at 6 AM GMT, then you may get wrong direction. EA may go long instead of short

i think the method has good potential to succeed, if we can get these small issues resolved in the EA we can do more tests. hopefully steve can find sometime soon