Can an automated trading system really work? | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Automated or semi automated trading? 3 replies

do automated trading systems work or... 21 replies

Why does Automated trading never seem to work? 64 replies

Trapping System: Way to Fully-Automated Trading System 47 replies

The Really Really Boring Forex System 97 replies

Can an automated trading system really work?

  • #1
  • First Post: Nov 12, 2007 3:28am Nov 12, 2007 3:28am
I have failed in forex so much. When I first started forex two and half years ago, I really thought I could do well and suceed. But I have faild so much, I lost a lot of money, I'm depressed and my life has been so bad. I just realized I can't really trade myself and I want to start to use automaed trading systems. So, here I am planning to study programming like C, C++, and Java. I will withdraw all the money from my trading account and just study programming until I master it. Because I don't know nothing about programming now, it will take a couple of months. And then I will come back to forex. But I want to know if automated systems can really work if they are really well programmed. I read a post that says, every automated system has some odds and none of them can work. If it is true, I don't really have to waste my time on studying programming. I was planning to go back to university next year but now I'm cancelling it just to study programming. But if automated trading system can't really work even if they are very well programmed, then I don't have to waste so much time and energy on it. Please gimme some comments. Thanks
When I buy it, it moves down. When I sell it, it moves up.
  • #2
  • Nov 12, 2007 4:41am Nov 12, 2007 4:41am
Your story sounds kinda like mine, 'cept that I have lost little money and it took me 1yr to get to your point. I'm now at the point where I give up on manually day-trading, and set my expectation for a yearly return to 30% tops. So now I focus on the large trends in currency pairs, and look for opportunities to hop aboard & get out. This costs me about 1-3hrs per day. The rest of the day I work for more capital, till the point that the 30% can support me for a living.

Now going on-topic; What I am doing right now is opening up an account with TradeStation (should have it funded within a week) and then play around with their Automation options. Their programming code is in Easylanguage and as the name suggests, it is pretty easy and feels very natural to write in. What I'm planning to do with this account next to regular longer-term trading is trying to develop a portfolio of automated systems that can make perhaps also 30% a year.
To be honest I think automated trading is just as hard as manual trading. It takes a lot of time to develop & test a system . And it takes a lot of tries to find a profitable system. And it takes a lot of profitable systems to make a portfolio that you can trust your capital to. And you think manual/discretionary trading can be frustrating? How about seeing your 'holy grail' system you worked for 3 months on failing miserably in it's first week of operation?

So don't fool yourself as I have fooled myself, and many others are fooling themselves & fooling others by their foolingness... eehhh right
Automated trading wont be a solution for your trading problems, and if you handle it with the wrong assumptions it can certainly add to the problems.

So get your goals realistic, have a safe plan of action. Dont rely on a single source of income, or only on automated trading. Be true to yourself, and certainly forgive yourself for failing in the past and losing money. It's a learning progress, and I'm sure if you keep on focusing on Learning and being the pessimistic optimist you (and I) will get there.

cheers
  • #3
  • Nov 12, 2007 4:56am Nov 12, 2007 4:56am
To be honest i've never heard of any completely automatic system that was profitable over more than a short period of time. A few months ago i saw one that won a contest with impressive results, but it was only for the duration of the competition so who knows how it would have worked out in the longer term.

Withdrawing your account it very sensible though. Why not try demoing with longer tf like the daily chart while you go to uni and teach yourself programming? Even if you code an automated trader it can only be as good as what it's programmed to do, so you'll still need to have a method anyway even if you don't want to implement it yourself.

The best of luck
  • #4
  • Nov 12, 2007 5:37am Nov 12, 2007 5:37am
A mechanic trading system works, but an automatic system dont work.

very strange, but it is true!
Trainee at UBS Investment Banking
  • #5
  • Nov 12, 2007 5:45am Nov 12, 2007 5:45am
Quoting FrankS
Your story sounds kinda like mine, 'cept that I have lost little money and it took me 1yr to get to your point. I'm now at the point where I give up on manually day-trading, and set my expectation for a yearly return to 30% tops. So now I focus on the large trends in currency pairs, and look for opportunities to hop aboard & get out. This costs me about 1-3hrs per day. The rest of the day I work for more capital, till the point that the 30% can support me for a living.

Now going on-topic; What I am doing right now is opening up an account with TradeStation (should have it funded within a week) and then play around with their Automation options. Their programming code is in Easylanguage and as the name suggests, it is pretty easy and feels very natural to write in. What I'm planning to do with this account next to regular longer-term trading is trying to develop a portfolio of automated systems that can make perhaps also 30% a year.
To be honest I think automated trading is just as hard as manual trading. It takes a lot of time to develop & test a system . And it takes a lot of tries to find a profitable system. And it takes a lot of profitable systems to make a portfolio that you can trust your capital to. And you think manual/discretionary trading can be frustrating? How about seeing your 'holy grail' system you worked for 3 months on failing miserably in it's first week of operation?

So don't fool yourself as I have fooled myself, and many others are fooling themselves & fooling others by their foolingness... eehhh right
Automated trading wont be a solution for your trading problems, and if you handle it with the wrong assumptions it can certainly add to the problems.

So get your goals realistic, have a safe plan of action. Dont rely on a single source of income, or only on automated trading. Be true to yourself, and certainly forgive yourself for failing in the past and losing money. It's a learning progress, and I'm sure if you keep on focusing on Learning and being the pessimistic optimist you (and I) will get there.

cheers
If you are so good at programming, I think you don't have to spend too much time to test. Why don't you just make a program that can automatically backtests three or four past years when you put the data in and analyizes the results?
When I buy it, it moves down. When I sell it, it moves up.
  • #6
  • Nov 12, 2007 5:56am Nov 12, 2007 5:56am
I don't want to disappoint you, neither I'm saying that automated trading doesn't work, but here goes my story:

In my early trading life, I was amazing with indicators and all those fancy lines, and I guessed how could I automate trading, since I'm (or I was) a programmer (Delphi and Java). For me, learn Metatrader language was relatively easy. After an endless quest to find a good Trading System, I realized that Technical Analysis DOESN'T HAVE to be SO complicated.

Presently I trade manually, using basic principles of TA, such trendlines and price patterns (flags and pennants). I look for good entries from weekly to 4h timeframes... and I'm doing pretty well.

Good luck anyway...
  • #7
  • Nov 12, 2007 5:56am Nov 12, 2007 5:56am
Quote
If you are so good at programming, I think you don't have to spend too much time to test. Why don't you just make a program that can automatically backtests three or four past years when you put the data in and analyizes the results?
Building a mechanic/automated system is all about testing, and this cannot be done by simply backtesting your rules over a 2-3 years period. Also there are already plenty of programs who can do this, like Tradestation. The only thing you need to know about programming is something like:
If LastBar_Close is >= 3BarAgo_close then Buy. One press on the button and you see all the occurances of this rule on your chart, and another press and you've got 5 years of data backtested.

The challenge lies in the reliability of your backtested data. If you test & keep optimizing your system for 1 set of data, you will get a system that is suitable for that set of data only. But finding a system that is suitable for multiple sets of data is a lot harder. Not to mention finding a system that is suitable/reliable for Unseen data.

Quote
Presently I trade manually, using basic principles of TA, such trendlines and price patterns (flags and pennants). I look for good entries from weekly to 4h timeframes... and I'm doing pretty well.
Exactly! that's the way I'm trading manually aswell. Easy & clearcut TA principles, only taking entries in the direction of the clear main trend. No clear main trend, no trade.
  • #8
  • Nov 12, 2007 5:59am Nov 12, 2007 5:59am
Yes it can, I am working on one right now.

The indicator worked excellently, I wish to see how it will do as an EA
  • #9
  • Nov 12, 2007 6:09am Nov 12, 2007 6:09am
Quoting pujolshomerun
I have failed in forex so much. When I first started forex two and half years ago, I really thought I could do well and suceed. But I have faild so much, I lost a lot of money, I'm depressed and my life has been so bad. I just realized I can't really trade myself and I want to start to use automaed trading systems. So, here I am planning to study programming like C, C++, and Java. I will withdraw all the money from my trading account and just study programming until I master it. Because I don't know nothing about programming now, it will take a couple of months. And then I will come back to forex. But I want to know if automated systems can really work if they are really well programmed. I read a post that says, every automated system has some odds and none of them can work. If it is true, I don't really have to waste my time on studying programming. I was planning to go back to university next year but now I'm cancelling it just to study programming. But if automated trading system can't really work even if they are very well programmed, then I don't have to waste so much time and energy on it. Please gimme some comments. Thanks
It depense on what your goal is; If you want to make a lot of money in a short time -greed-. Forget it...

I have a 100% hit EA: It works with crossing WMA's with different levels and 4H TF. It will give you (only) +/- 3 trades in a month with a profit of 1 or 2 pips per trade. Not worth it for me and not taking a risk to throw in multi-lots / orders at one time.

Or you can always follow the rainbow and scalp along the way...or just play poker -don't- :
  • #10
  • Nov 12, 2007 6:27am Nov 12, 2007 6:27am
Quoting FlashCall
It depense on what your goal is; If you want to make a lot of money in a short time -greed-. Forget it...

I have a 100% hit EA: It works with crossing WMA's with different levels and 4H TF. It will give you (only) +/- 3 trades in a month with a profit of 1 or 2 pips per trade. Not worth it for me and not taking a risk to throw in multi-lots / orders at one time.

Or you can always follow the rainbow and scalp along the way...or just play poker -don't- :
I don't expect to make a lot of money in a short time. I just want to build a system that can creat profits in a long period of time.
When I buy it, it moves down. When I sell it, it moves up.
  • #11
  • Nov 12, 2007 6:41am Nov 12, 2007 6:41am
Quoting pujolshomerun
I have failed in forex so much. When I first started forex two and half years ago, I really thought I could do well and suceed. But I have faild so much, I lost a lot of money, I'm depressed and my life has been so bad. I just realized I can't really trade myself and I want to start to use automaed trading systems. So, here I am planning to study programming like C, C++, and Java. Please gimme some comments. Thanks
Hi Pujolshomerun,

Could you please give some more specific details on the reasons you believe you have failed at trading. With more information perhaps I (and others) can help.

I really don't think attempting to trade an automated system is going to help you a great deal. To trade an automated system, you still need to be able to build a system. If you were able to do this you may have had at least some small success, from your post it sounds like you've had a very tough time.

If you have been able to build a system with a positive expectancy and good R:R, then maybe your problem is the psychological side of trading??? We really need some more info here..
  • #12
  • Nov 12, 2007 7:11am Nov 12, 2007 7:11am
I agree with Norman. Programming would only help if you have a proven system but are not able to trade it for whatever reason. Maybe it requires 24 hour monitoring, maybe it requires trading at a time when you must sleep. Or, maybe it is psychological.

I've been at this for 1 year full time and believe (again) that I am turning the corner. The thing I feel has gotten me to here is following bad and good advice enough times, and trying on my own, doing it right and wrong enough times, so that now I'm at the point where I can decide what is true for me and what isn't. I believe this is important.

http://www.forexfactory.com/showthread.php?t=40441

I've programmed many EAs of "proven systems" and none work. Over at forexTSD there are hundreds of people running EA's month after month, and tabulating results, which you can view. Look at the EA contests, check the results. A few people get lucky with the right settings for that time period, but almost all of them lose money.

(I do have an EA running on demo now that has been performing wonderfully for a month (50% return), however, I also know that under certain conditions, it will blow the account very fast. These conditions are happening right now, and the drawdown has eaten up that profit as I write.)

I've looked at hundreds of trading systems, free and purchased, and have seen none that can be traded mechanically. By mechanically, I mean that you take every signal the system gives, enter and exit by the rules, always. If there were such a system, it could be programmed and would make money.

I also know a professional trader who can trade any reasonable system profitably. How is this possible? Discretion.

All systems that I have seen require discretion. This is an art that is only learned through study and practice. You can't just follow the rules of the system and make money. It just doesn't work.

You've proven this to yourself, haven't you? If you can say "yes", then there is one thing you know is true. If you have one thing which you know is certain, you can build from there.
  • #13
  • Nov 12, 2007 7:16am Nov 12, 2007 7:16am
Automated systems do not work only for one reason: they are unable to adapt to changing market conditions which is why you can only use a "half-automated" system which trades on it's own but the market conditions are observed by the trader , once the market conditions start changing the trader interferes but if you just let it run you will lose for sure , it is only a question of time , an alternative would be to develop several systems for different market conditions and decide when to use which system
  • #14
  • Nov 12, 2007 7:29am Nov 12, 2007 7:29am
c4
Quoting fx_amateur
Automated systems do not work only for one reason: they are unable to adapt to changing market conditions which is why you can only use a "half-automated" system which trades on it's own but the market conditions are observed by the trader , once the market conditions start changing the trader interferes but if you just let it run you will lose for sure , it is only a question of time , an alternative would be to develop several systems for different market conditions and decide when to use which system
That's a good possibility.

I did read a post by someone, somewhere, once, who said he lost at forex for years and years and never made any money until he started using automated trading -- but that he closely monitored it and killed any obviously bad trades.
  • #15
  • Nov 12, 2007 7:52am Nov 12, 2007 7:52am
i want to tell u that automated is not reliable, u can easily make drastical loss because it is programmed by human being and u can easily know what u ve gained if you trade yourself.why can't u try technical analysis?So don't ever try automated trading system
  • #16
  • Nov 12, 2007 8:08am Nov 12, 2007 8:08am
nearly all completely automated systems fail in the long term. I'm not saying all, because maybe some do pass the test of time, but I have not seen one yet. There is no holy grail to success. It takes a lot of hard work and patience, you need to trade on demo and get a few profitable months under your belt before you even trade a live account even with a small amount.
  • #17
  • Nov 12, 2007 10:32am Nov 12, 2007 10:32am
There are a lot of good points in this thread. I have been trading for nearly 10 years and I have been doing the automated thing for about 4 years.

First off Pujolshomerun – The answer is yes, they do work for some. I do understand your want to learn computer programming. I did the same thing for different reasons. I had systems in place that I realized I could automate and make my life easier. You should first tinker around with meta trader or any other system and find something that works.

FrankS has a great point - If you are going to learn computer programming you should focus on a backtesting and walk forward testing program before learning to automate a strategy. This type of program will help you find good strategies.

Fx amateur also has a good point – market conditions change. Any moving avg trend following system has been doing really well the last few months, think of a strategy like CornFlower in the FF forums. The system has been doing really well the last year or so but when the market starts ranging it’s hard to see a good trade opportunity. So what you want to do is design 2 strategies in 1 – 1 for trending – 1 for ranging. You can use something like ADX – DMI on longer Time frames to turn one off and turn one on.

Blueruby recommended discretion – I use this for some of my strategies. I enter the market and I place stops with automation. I put my profit target on the high end and use my own discretion on when to get out, unless my really high profit target is hit.

I would recommend learning C#. The learning curve is pretty steep with any language but I was able to get a backtest and walk forward program up and running in 6 months with no prior programming experience. From my experience learning a language is a lot easier to learn when you know what you want to develop. I also use Trade Stations Easy Language to write all of my strategies and indicators. I do not recommend using them as a FX broker for other reasons but it is really good analytical software and Easy language is extremely Easy.

Mike
  • #18
  • Last Post: Nov 12, 2007 11:13am Nov 12, 2007 11:13am
Check out the MQL trading competition, those systems seem to be working.
There is no magic bullet in any type of trading, I could program like a f**kin Ninja when I started into Forex a couple of years ago, but I'm only just going live with a system now...it's all just hard work, but you also have to systematic & smart and nobody can teach that.
The breaking of a wave cannot explain the whole sea.
0 traders viewing now