Improve this grid EA

Hi George, thanks for your interest

As you said, there is a lot of noise and smoke in this world and avoiding it has to be a primary consideration for me. If as you say you have done this sort of thing before then you will know that looking for genuine collaborators is not the easiest of tasks. I can only go about it in a way that makes logical sense to me and doesn't waste large amounts of mine or other peoples time. One can only hope that those who are genuinely interested will make the effort to at least try and establish their credentials and be willing to consider some sort of collaborative relationship. As for my credentials, there is a fairly comprehensive CV with my profile and I have no problems providing further information about my development track record etc if and as required. I have already outlined in post #6 what this proposed project is about which is to find/develop an efficient risk management algorithm for grid trading. In my experience, attempting to debate ideas like this on this forum is a total waste of time.

Obviously I am not only a lot younger and less patient than you but also far less intelligent. While your concept may well be absolutely brilliant, unfortunately I haven't a clue what you are talking about. There is not enough information for someone with my limited capacity to figure it out.

Regards
Tim

After 15 years of membership in forum, I tend to believe the same ...
Thank you for your understanding, it was refreshing to see a young man with a sense of humor.

In the other thread I posted a little trading example from today.
It may not be entirely clear, but it can be useful.

Scientific method and trading

Scientific method, in simple language involves coming up with an idea that is called a hypotheses and then trying to prove that the idea is invalid. This sort of proof is known as a null hypothesis.

At first glance this may seem like a back to front way of doing things but there is a good reason for it. By trying to prove that an idea is not true or invalid we help prevent ourselves falling into the all too common trap of self delusion. If we can’t prove that an idea is invalid, then it may possibly be valid and therefore useful.

I would be the last person to claim the science and scientific methods can solve all human problems and issues but virtually everything from space vehicles to i-phones are the result of scientific method. I have also found it helpful for trading.

The idea or hypothesis is typically a written document with enough information in it for a knowledgeable independent person to understand the idea and be able to duplicate and test the results.

When I first started trying to apply scientific method to trading I found it very frustrating. Just about everything I tested that was based on conventional trading wisdom appeared to be invalid.

I am by no means the first person to discover this. Many academic researchers have found similar and come to the conclusion that trading Forex profitably is impossible. That of course isn’t true. A very small number of traders are consistently profitable.

Another challenge that befuddles many traders is that price movements in financial market have a large random or at least pseudo random component. Such characteristics are known as stochastic processes. Their presence tends to throw any “common sense” thinking about trading out the window. This is simply due to the way in which the human mind functions. If we ignore this problem, it tends to lead to deluded thinking.

There are of course ways of dealing with stochastic processes. Unfortunately this requires a a little bit of mathematics and the mere mention of the math word tends to send a good many traders rushing for the exit. This is rather unfortunate because with a little bit of study and effort, I believe most traders can master these subjects sufficiently well to be able to make practical use of them.

Learning to use scientific method and stochastic techniques together with the ability to think outside the square of conventional market wisdom will, in my experience, eventually lead to worthwhile results. Little by little we find things that do work and by carefully assembling these working parts, it is possible to create complete working systems.

In post #6, I outlined a current project that I am looking at which is to try and find or develop an efficient and practical risk management technique for grid trading. I invited anyone interested in the sort of approach that I have just outlined, an opportunity to become involved.

The shared reward will be knowledge and software. If a good solution to this problem is found, it could also lead to some very profitable trading.

Anyone interested is welcome to PM me.

Regards
Tim

And an old man with the same!

I have downloaded ducascopy tick data for the past 10-15 years data. And all my testing I do in my software with real ticks data. A lot of time I tested zigzag strategies, trying to guess when it ends, or how long does it continue. And it was with no luck. Zigzags are quite similar like grids. The main problem is long trend with no pullback. I thought I could prevent this with not being in market when high important news are released. But with some testing I found, that these long trends happen also without news. And it will kill account if you trade against this trend. To trade opposite is not an option too, because in the long term you lose spread in every trade. I've done hundreds of strategies based on price and everywhere the probability is ~ 50:50. And with that probability and rare long trends is hard to think some risk/money management if even it is possible.

Grid trading works well only in sideways market. Timing with distribution happening after accumulation has almost finished is what is required. Big Candles provide clues for sideways onset. If you have Excel skills to analyse previous bars we can work together. You have to code an Excel RTD dashboard for me in IB for analysing in real time futures and futures options data . An Excel novice, a friend developed the RTD client application to pull data from IB RTD server. The client which pulls data in real-time is quite an laborious work that is anyway done. Excel dash board to help lay people make analysis for the instruments is needed. In return, I will share my proft-making grid trading strategy with you. It has nothing to do with martingale. It is based on drip feeding for regular passive income generation.

Sorry for the bother, but reading your excellent contribution, I recall some quotes from Sherlock Holmes:
“… When you have eliminated the impossible, whatever remains, however improbable, must be the truth…”
"Eliminate all other factors, and the one which remains must be the truth."
"The world is full of obvious things which nobody by any chance ever observes."
“You see, but you don’t observe. The distinction is clear. ”
"I ought to know by this time that when a fact appears to be opposed to a long train of deductions it invariably proves to be capable of bearing some other interpretation."

There is a word in my native language:
"Don't spoil the barley on geese"
Perhaps one should consider what the discussion of scientific methods at FF has brought ...

Hi IenDzi,

I think of words like zigzags and grids as labels which can mean different things to different people. When people talk about zigzags I usually think about zig zag indicators which are a look ahead indicator. They therefore can't be used to predict future prices moves. This is one of many sorts of delusions that often trick and trap inexperienced traders.

As for grid trading, there are many different schemes that people use. The sort of grid trading that I am currently interested in works in a sideways market and is agnostic to direction. This strategy can therefore be thought of as being based on volatility rather than attempting to predict the direction of price moves which, as you say is typically going to be ~ 50:50 except when the market is trending. This of course is what trend trading is all about.

The problem with the style of grid trading that I am talking about starts when price breaks out of the grid and starts to trend.


There are various ways that traders try to deal with this problem. The simplest way is probably to just use a stop loss. A simple stop loss however is very inefficient and the cost is high in terms of lost potential profit. There are many other approaches. Some of them are unstable and when analysed demonstrate a high risk of ruin.

I have some ideas for a different approach which, as far as I am aware, does not exist anywhere in the public domain. This may well be a good starting point but when a small group of capable people start bouncing ideas off each other, who knows where things will finish up. This is the fun and challenge of working together in the type of collaborative arrangement that I am proposing. Sometimes the results from such a collaboration can far exceed what any one individual in the group would be likely to accomplish on their own.

The challenge is finding people with the right mix of technical skills and personality to participate in such a group.

Regards
Tim

Good one George
I'd like to send that to the editor of the newsletter for the little peer-to-peer mentoring group that I belong to. If you wish to PM me an email address, I will get him to send you a copy when it is published about mid Feb 2022



My guess is the discussion of scientific methods at FF has resulted in ~ FA but I am happy to be proven wrong, scientifically of course.

Regards
Tim

I did my version of Zigzag. It was minimum value of pips to start a ZigZag(for example zigzag with minimum leg of 10 pips) and then I looked how much it extends until new zigzag leg develops. And I wrote all zigzag legs and then done statistics with them. The maximum leg I got was 10 pips extended to over 200 pips.
I could team up with someone. I can code and test strategies on my own platform based on real ticks in c#, not metatrader. You must know some basic market stats: up candles = down candles, up pips = down pips, reversing losing strategy it doesn't become winning strategy, etc. I believe in CrucialPoint theories. Some theories I tested and it was true, others I didn't understand. I don't believe in indicators like stochastic, rsi, ma, etc.

The Grid Trading Development Project- An Outline

I thought it might be of interest to anyone who is considering becoming involved to see an outline of this project and how it is proposed to bring team members together and get the project underway.

The Environment

1. MetaTrader MT4 - A demo account is all that is required
2. R/RStudio - Current versions or close to it installed on a PC with reasonable capacity. Most PC's of around 5 years old or less should have no trouble running R and the sort of models I expect we will be looking at.
3. Discord - Will be used for most of the team communication. We may also use Google Meet on occasions (nothing to install)

All three of these items are free and easy to install on a windows PC.

Skills
While technical skills are important, the ability to think constructively and solve problems is essential. While not every team member needs to be an experienced programmer, it is essential that everyone is able to run scripts, programs etc in the above environments.

The Plan
Step 1. In order to ensure that we are all on the same page, we will start with an evaluation of a basic MT4 Grid EA designed for trading in sideways markets. Part of this evaluation will include an interactive tutorial that can be run on MetaTrader 4 using the strategy tester. This tutorial and subsequent evaluation exercises should illustrate the possibilities and potential pitfalls for this particular approach to grid trading.

Step 2. For the second stage we will evaluate another more sophisticated EA that uses what I regard as one of the more successful risk/money management techniques for grid trading that can be found in the public domain. This EA will also be run and evaluated using the MT4 strategy tester.
We will then introduce R/RStudio to create a model based on this technique which will then be used as the performance bench mark for future development work.

Step 3. The third stage will use R/RStudio to demonstrate an interesting Artificial Intelligence (A.I.) technique that could set the direction for future ongoing development work. While team members are not expected to have expertise in this area it is hoped that they will be interested enough to learn and gain an understanding of these concepts. This type of approach to trading systems can quickly become mathematically complex. I will do my best to present these ideas and develop our models and programs with the absolute minimum of math possible but there will inevitably be some.

The Next Steps
Where we go from here depends very much on the overall team response to the first 3 steps. The end objective of the project is to take the basic Grid EA that was evaluated in step 1 and expand it to include the best money/risk management system we can find or devise. This may or may not make use of A.I. techniques, depending on what we discover along the way.

Opportunities
There is still one place available on the team.
If interested please PM me.

Regards
Tim

Frequently Asked Questions

A number of people have contacted me privately with various questions. I have gone through all these and used them to create a FAQ list. There may be some duplication of information that has already been posted.

Q1. Who is running this project?
A1. The Melbourne Forex Trading Team (MFTT) is a small non commercial, peer-to-peer mentoring group based in Melbourne, Australia. The members of this group support and encourage each other in their trading and investment activities. Some members share their trading results with the rest of the group on a monthly basis. We also share trading strategies, custom software etc. The group was established in 2016. We currently have 9 members. I am one of the founding members.

Q2. Is it necessary to become a member of the MFTT to participate in the Grid Trading Development Project (GTDP)?
A2. No. This project is being run by but is independent of the MFTT.

Q3 How do I become an MFTT member
A3 MFTT Membership is normally through invitation by an existing member.

Q4. Are MFTT members involved in this project?
A4. Yes. For a start I am an MFTT member. There will be a mix of people from within and from outside the MFTT. All participants of this project group will have equal status within this group.

5. Do you normally recruit team members for these sorts of projects from outside the MFTT?Q
A5. No this is the first time this has been attempted.

Q6. Why don’t you just use your own members for this project?
A6. Not all MFTT members have suitable development skills. Some of those that do are currently engaged with other projects.

Q7. Have you had experience running projects like this before?
A7. Yes, since its inception the MFTT has run a number of projects. I have been involved in and led some of them.

Q8. How big is the team for the GTDP?
A8. Originally a team of 4 was envisaged. We have now met that target. There has however been more interest than expected and I am considering adding 1 or perhaps even 2 extra members to the team.

Q9. If I participate in the GTDP, am I required to sign a contract or a non disclosure agreement (NDA)?
A9. No. Apart from the fact that we would not be in a position to enforce any such agreement, we believe they have little if any real value. Obviously it makes no sense to go out and announce to the world at large what you are doing so hopefully team members will be discreet but concerns about protecting intellectual property, particularly from other retail traders are usually unfounded. There is so much misinformation and noise out there that it would be hard to get noticed, even if you really wanted to let everyone know what you were doing. On the other hand, if you start trading a system and consistently make a significant amounts of money you are going to be quickly noticed and investigated. The big commercial banks and their contingency of market makers and brokers etc have vast resources at their disposal and will eventually figure out what you are doing. It is their business to do so.

Q10. Aren’t you concerned that someone will join the project, not contribute and then steal all the intellectual property (IP) such as software etc?
A10 No. IMO this concern arises in some people who believe that trading success is all about finding the right strategy or technique. There is an old saying. “You can give an experienced trader any old system and they will make money. You can give an inexperienced trader the best strategy in the world and they will lose money”. With a project like this you tend to get out of it what you put in. In my experience, people with a short sighted “grab and run” mentality simply don’t “get it” and tend to do poorly at both trading and in life.

Q11. When will this project begin?
A11. It is planned to start this project on or about 1st Feb 2022.

Q12. Will people be able to join the GTDP project after this date?
A12. No. People joining the team after the start date could be disruptive to the project.

Q13. How long will this project last?
A13. This is a research and development (R&D) project and like any other R&D project it is impossible to determine its duration since the outcome is unknown. People do however set targets and dates for R&D projects but these are usually little better than educated guesses. With that said I am hopeful that we will see some worthwhile results within about 6 months from the commencement date.

Q14. What level of commitment in terms of time and effort is required?
A 14. There is no specified level of commitment. Each of us work in different ways. IMO it is impossible to regulate or control creativity in this way. The first part of this project involves three interactive tutorials that will hopefully enable us all to get up to speed with the technology, understand what we are attempting to accomplish and get to know each other a bit. It is estimated that it will require something between about 10 and 30 hours to complete these tutorials depending on existing levels of skill and experience. This would include some online discussion time with other team members.

Q15. What happens if for some reason I am unable to continue with the project?
A15. This can happen for various legitimate and unforeseeable reasons. Other times, some people just lose interest and drop out. Ideally you should notify the project leader if you are no longer able or wish to continue. If someone appears to no longer be active, and has not contacted us we would attempt to contact them before removing their access to the project.

Q16. Do you have a project plan?
A16. Yes It was posted on both the Forex Factory forum and the MQL5 community forum.
Forex Factory- Thread title - Improve This Grid EA-
Link - https://www.forexfactory.com/thread/...8#post13853818
MQL5 community – Thread title – Grid Trading Development Invitation
Link - https://www.mql5.com/en/forum/385682...mment_27012412

Q17- What skills and facilities are required to participate in this project?
A17- As a minimum you would need

1. The ability to think constructively and logically through problems.
2. A MetaTrader 4 demo account with the ability to run MT4 scripts and use the strategy tester. MT4 programming skills, while highly desirable are not mandatory. It is not recommended practice to carry out testing or development work using a live trading account.
3. R/RStudio installed and running on your computer with the ability to run R/RStudio scripts. R programming skills, while highly desirable are not mandatory.
4. Discord (https://discord.com). Installed on your computer for team communication.

Q18. Will all development work and software used in the GTDP project be made available to all active team members?
A18. Yes

Q19. Is the goal of this project to develop a fully automated Metatrader grid trading EA or robot?
A19. No. I have studied traders for a number of years and have an interest and reasonable background in Artificial Intelligence (AI). My conclusion is that successful traders have developed the ability to utilise and coordinate numerous different parts of their mind in quite complex ways that enable them to make better than average trading decisions.
While A.I. has made leaps and bounds in recent years, there are still a number of areas where it does not perform well. No doubt this will change over time and the whole trading game will then change accordingly.
There may well be fully automated systems that can perform as well or better than human traders but I have yet to personally see one. On the contrary, all the automated systems that I have looked at or created have performed poorly compared to what I know good traders can achieve.
For the time being, I consider it better to use automated computation and A.I. to augment the human mind rather than to try and replace it. IMO many traders miss good opportunities that a hybrid approach could offer them because their focus is totally on "set and forget" automated trading.

Q20. I wish to put together a project team for my own project. Can you help me?
A20. Putting a team like this together is not particularly difficult but does require quite a bit of work. As a start you could read through the threads on the Forex Factory forum and the MQL5 community forum. Then adapt this approach to your own needs. Essentially all I have done is try and explain to people what this project is about and invite interested and suitably experienced people to join me.

Regards
Tim Wilson

We finished up with a team of 7 good people which was a bit larger than I originally anticipated. We regrettably are not able to accept any further applications. I would like to thank everybody who participated on this conversation.