Im ashamed to admit but as a veteran trader, I still havent learned coding. I got involved with 2 neural network softwares, and neither made me any money although both showed me as much as 120% annual return 2 yrs backtesting on daily charts! My question is, I keep hearing from the MT4 marketing people how wonderful MT4 is in making your own EA's(expert assistants?) which I dont fully understand what that comprises of(funny everyone is too embarrased to ask. Am I the only one?) that will let me backtest and tweak my own system,add and tweak my own rules until I have a winner. So question #1 is:
1-I use MT4 everyday. Everytime I play with the drop down menu and hit anything to do with EA'S, It seems it is blackened out or non functioningor doesnt make sense to me?
2-If you could really program your own systems and win, why wouldnt everyone devote the time to learn this up and down and within 6 months everyone would have a winning system? IS IT WORTH MY TIME TO LEARN IT?
3-Is there something easier to understand that doesnt require coding I could use to try my hand at this?
4-It seems this is very similar to neural network software which is to me just letting you build your own black box without letting you see how they do it. Am I correct? And if so, why has Neural network software seem to fallen so out of favor. I used to see companys like Network Solutions and others promising the moon with their neural networks and now it seems those things have gone the way of black and white television. Why?
Id appreciate any feedback on this, At least tell me if MT4 is really capable of helping me create my own WINNING system? It seems Trade Station makes a big deal out of their EA'S and system development tools. Yet many squawk at learning their easy language. Anyone know of any success stories there ?
The following user says Thank You to Daytrader50 for this post:
I don't know much about MT as I use NT but winning systems can be programmed. That being said it is not the programming that makes it winning it is the method. So in a sense must develop a winning method that can then be programmed. You can borrow ideas from others but ultimately need develop something that is your own, fits your trading style and fits your observations about the market.
"The day I became a winning trader was the day it became boring. Daily losses no longer bother me and daily wins no longer excited me. Took years of pain and busting a few accounts before finally got my mind right. I survived the darkness within and now just chillax and let my black box do the work."
The following 3 users say Thank You to liquidcci for this post:
Based on my own experience I believe programming a winning strategy is possible. Many will say I'm naive to say that but that's fine with me. I also back test my strategies which many say is a waste of time for a variety of reasons. Without going into all those reasons I will say that my back testing results have played out the same way in forward testing in sim. Yes i know sim success doesn't mean real money success. However, a prudent person wouldn't turn a system loose with real money unless they have tested it first. At least I wouldn't. I have 3 strategies in forward testing now that are all making money. Been forward testing them for over 6 months. I had another one last year that I took live but made the age old mistake of being undercapitalized and was unable to withstand a record drawdown that came soon after starting. I didn't lose money but that's only because I stopped it before it could. I switched it to sim and it recovered after a few weeks. One of the 3 that I currently have is ready to go live in a supervised mode but I don't have the capital to get started. So, I'm saving for that.
That's my background so you have an understanding of where I am in this effort.
I use NT so can't really say anything about MT. I like NT just fine. Been using it for quite a while and understand how it works.
You can learn how to code which will be a very laborious effort or you could partner with someone who can if you can convince them that you have something they can benefit from.
As far as black boxes, I don't trust them which is why I embarked on the effort to build my own several years ago. Prior to doing that I had lots of experience building indicators.
Not sure if any of this helps but the bottom line is you are looking at an incredibly long and difficult journey to learn programming. I hope you're a young person.
I'm wondering why a veteran trader who I assume is making money consistently would want to build a strategy/EA.
The following 3 users say Thank You to MWinfrey for this post:
Just to add to what others have said. You're asking two completely separete questions which each have their own answers:
1 - Can a profitable system be programmed?
2 - Can MT be used to programme said system?
The answer to the first question is highly debated and you would be well served to search for the many threads on this forum discussing this.
The answer to the 2nd question is that MT is simply the tool. The hard part is creating the profitable strategy and that has more to do with your understanding and ability within the market than it does with the tool. There are many platforms that could be used to programme your strategy (eg: MT, NT, TradeStation etc). Personally I like NinjaTrader because of its feature rich C# coding possibilities. But keep in mind that with the best tool in the world, it is still up to YOU to develop the strategy,...which is why question 1 above is far more critical than question 2.
Last edited by DarkPoolTrading; September 12th, 2012 at 08:38 AM.
The following user says Thank You to DarkPoolTrading for this post:
There's been too many MT4 hyped EA systems out there over the past years mainly to profit from the forex craze as folks sought to get into the markets cheaply. A system may only work for a certain time before it stops being effective maybe because the market always changes as well as participation levels. On the few news reports of HFT firms they have "quants" constantly manning workstations observing the algos in process. They probably switch to a different algo or adjust if it loses effectiveness. re: Neural networks. The prof I had in my classes said AI still had a long long way to go. While they've got AI to help motorized functions such as those Japanese walking robots, AI probably still has a long way to go to think outside the limits of a chess board. Neural network EAs are supposed to simulate a human brain decision making process on trading, but most traders have a very hard time manually trading already. And the markets seem to be severely changed from just a few years ago with ever increasing participation % of big firm HFTs and algos. Maybe there are a few EA's on MT4 that do work, but maybe only for a period of time ,then they need to be adjusted based on the ATR or volume, volatility, etc. of the current markets.
Last edited by Cloudy; September 26th, 2012 at 01:50 AM.
Yes it is possible. MT4 is a good platform, but the brokers using it are not fair, manipulating price and entry price a lot.
I coded and tested hundred of EA, many of them work until...
But the goal is to survive, because one day the market do something strange and kills your account.
On my last EA I try to reduce risk and hedge positions. Of course profits are lower.
You can check on my blog MyTraderAdvisor , right column for real time/real account info and some backtests on some older posts.
I believe only in profits, not in prophets.
The following user says Thank You to MyTraderAdvisor for this post:
I believe so; I have an MT4 EA and an oscillator that I use that I wouldn't mind sharing as proof. I have a 5K account that is currently up 14% for the week. I have been using and modifying the code for many months now. I would like to communicate with / share with any interested persons who could assist me in further improving it or just testing it.