RoboPip systems implementation

[pioter]

Hello there,
I am planning to implement Robopip’s systems :

• Bollinger band
• Inside Bar
• SMA Crossover

I will create EA’s that are as close as possible to original rules, and run them on properly chosen pairs. Every week, I will try to compare them with Robopip’s results.

WARNING: This is not to prove whether Robopip is wrong. My time zone and broker is different, that will have an impact on results. However, after brief testing, I can see that methods are still profitable for my local setting.

Any ideas are appreciated, but please do not be upset if your post will be removed - I would like to keep thread clean.

May the pips be with you!

[ria_rose]

Good luck with this @pioter! I think there are a lot of great systems over here and I’m actually trying out the HLHB for myself! I’ll definitely watch out for this!

(I don’t mind if you want this post to be removed. Haha. I just want you to know that this is a great idea! )

[Ananais]

Cool idea! Sharing this one on BabyPips’s social media accounts! Let me know if you’d like us to tag you! Looking forward to your updates.

[pioter]

Thanks!
For sure more popularity will help, although I can not post source code right now - it is an early fast attempt to evaluate the strategies.
Once the mechanics will be tested, and result show some future, I will rewrite them in more OOP manner. Then I think we could collaborate on GitHub, to modify for better results.

@ria_rose If you create a HLHB thread, I will for sure follow that

[pioter]

Short-Term Bollinger Reversion Strategy

I am posting first results, to show that we are not only talking here

Week 02-09 August 2018
Transactions taken are the same as Robopip’s.

Week July 26 - August 02
No trades, the same as in original system.

—
I have also tested other weeks, and results were consistent.

Now comes the question, how is it performing then ?
Let’s see :

WARNING: My modeling quality is just above 50%, which is bad. However, system makes trading decision basing on complete candles of the same timeframe, so live tick emulation is not much important. It is still a factor, but for this overview, should be enough.
Used size: 0.4lot
Initial depo 10k USD

ibb.co/f6OuwU

USDCAD for year of trading is on breakeven, balancing around initial deposit. This is very good, leaves room for improvement. I have to remember, that EA can have glitches.
More detailed report will come in future.

EURCHF Test 06 06 2017 09 08 2018

This is just a nice chart of a pair that is providing profit.
I am planning to create configs for metatrader, to automatically test each strategy on all provided pairs. I will be able to find pairs which are best suited for given system. (but should not be “fitted”, that will be taken under consideration later )

SUMMARY:
Strategy is not behaving identically as original one, but definitely has potential! I will work on it.

Sorry mods for going around the rules

[pioter]

Inside Bar Momentum Strategy 2.0

This is a tricky one, as my timezone is not producing the same 4H candles. I am running this EA on 1H bars, and create 4H candles artificially, to match the ones from original strategy.
I am UTC+2 (Warsaw) right now, so my 4h candles starts when: hour % 4 == 2.

I do not know Robopip’s timezone, but I have aligned my charts with his to establish proper time offset.

Week 01-08 August 2018
I will explain all transactions in detail :

1

First transaction has been triggered, the same as in original thread.
Enter PRice: 112.011, StopLoss: 111.736
It coincides with original strategy’s entry, what is proofing correct time shift for 4H bars creation.

babypips-discourse-media-production.s3.amazonaws.com/original/3X/1/3/13e9c14dd60066a60546a2e6f03c47ef7d72d704.PNG

2

The second transaction has not been taken by my system, because High of 2nd bar is higher than High of 1st bar. This is very smal difference : 0.6 pip.
This situation of course will happen, as my broker has different source.

babypips-discourse-media-production.s3.amazonaws.com/original/3X/0/7/0718f347b3491654e24c94eedccf91923e168f3e.png

3

This trade has not been taken, because my quotation starts at 23:00, which is not taken by my system as start of new 4H candle. However, I will not fix that yet - I think combining candles from the end of week and beginning of the next does not have the same value as 2 consecutive 4H candles.

babypips-discourse-media-production.s3.amazonaws.com/original/3X/5/3/5387862fd5bb155a0503521c103dee1e7542a2dc.png

4

The same situation, Highs of 4H candles were bit different than in original system. (we can see that in original thread, the highs were very close too)

babypips-discourse-media-production.s3.amazonaws.com/original/3X/c/f/cf8670657e6fd2bcf804ebc1f339c503fa192d37.png

5

Now situation has changed 180 degrees. My system has taken the trade, but original didn’t. IT is the difference in quotations.

babypips-discourse-media-production.s3.amazonaws.com/original/3X/e/c/ec8fe4e4d7ef856a9b18328aba37255b43edc4dc.png

6

This is the same trade as in original system.

babypips-discourse-media-production.s3.amazonaws.com/original/3X/3/0/3076933a5ca5cf4fb6ae627df9d1d573c8e1380d.png

I think this shows that the Inside Candle EA can be treated as ready for forward testing.
Other weeks has also be tested, in a way described above.

What is the system’s performance?
0.4 lot, 5pip spread, since 06 June 2017 to 08 August 2018

USD/JPY

babypips-discourse-media-production.s3.amazonaws.com/original/3X/1/8/1895ab8f287bdac5b830a2558ec5857808e87612.png

GBP/JPY

babypips-discourse-media-production.s3.amazonaws.com/original/3X/7/6/76048ac5dbde30f53e736774d1303cee31a7c7d1.png

SUMMARY
Just having primitive EA, this system proved to be successful. It is definitely worth developing.

TO CONSIDER:
Is my broker changing summer/winter time ?

[ria_rose]

OMG! Hahaha. I’ve actually created the thread already.

I'm trying out the HLHB system!

Helloooo. :smiley: I've been demo trading for the past couple of months (thought I wasn't very active. :sweat_smile:) Back then, I usually just try to identify candlesticks or chart patterns. :sweat_smile: It was pretty much a guessing game for me. Huhu. :frowning: But then, going around the site again, I've decided to try out a system to hopefully set me on the right track. :confounded: I'm still trying to work my way through the system and see if I understand it. :slight_smile: I'll pro…

It would be great if you could also drop by every now and then. Looks like we’re gonna be part of each other’s trading journey! Haha.

[pioter]

SMA Crossover Pullback

Week August 7 - 14
Transactions taken:

babypips-discourse-media-production.s3.amazonaws.com/original/3X/a/d/ad272339b2258f5ca8aa1f56586ca4f9e1f0a6f7.png

babypips-discourse-media-production.s3.amazonaws.com/original/3X/4/6/46f7fa913e6407abafea0289a65e66d8489ce1d9.png

Transactions are the same as in original thread. Although EURUSD did not reach 300pips TP (it is 200 pips)

How the system performs?

Quite good, there is a future in this method .
babypips-discourse-media-production.s3.amazonaws.com/original/3X/a/d/adf3a0ad05c29c13dedbc619bb91661700b96e76.gif"

SUMMARY: I can consider my EA working.

I have developed 3 EA’s, where each of them follow rules. I have checked the results with RoboPip’s reports. I will run now each of them on 8 years history data, to assess the performance of each pair. For each method I will choose the best pair, and keep making weekly journal with quarterly summary.
During this time, I will be thinking on optimisation and creation of more robust .mq4 files.

@ria_rose thank you for invitation, I’ve made my first post on your thread. Do you know maybe how can I have granted right to post more pictues?

[pioter]

History testing:
I have ran the tests from June 2010 (some from 2013) and chosen the best pairs to trade.

• Inside candle strategy is not working for me, during forward testing we will see how it can be fixed

• Bollinger Reversion: EURUSD, AUDJPY, USDCHF, GBPUSD

• SmaCrossover: EURUSD, AUDJPY, GBPCHF, GBPCAD

LotSize chosen: 0.2
Initial depo: 10 000USD