one simple question, how can a system stop working?

¿how a sistem can stop working?

really, i cant understand...

a system works well for 2 years (for example) but then indicators begin to fail..

what is happening?

thanks!

Simple answer: the conditions that led to prior profitably have changed

Without knowing exactly what system you are talking about, as a simple example, consider a system that works well in a trending environment (maybe it uses moving averages). It will get chopped and screwed in a ranging environment, often times enough to deplete any previous profits.

thanks smittens.

A few days ago I discovered the Tlatomi method but at the end of the post, people say it does not work.

That's where I was puzzled.

thanks again!

ya. Easiest answer is probably that it was designed and traded best under the market conditions that were around at that time. Depending on how a system is constructed, it might only work for a certain amount of time.

That's why continuing education in this field is essential. Anything can change at any time.

Here's a simple example. For a very long time, the Asian market had fairly low volatility and tended to range. People built systems around these factors, such as buying/selling whenever price reached the range limits. Now, the Asian market is very active and not ranging like it used to. If you stuck to the ranging system, you'd have taken a beating over the last 6 months or so. Put up a Daily chart of the Eur with an ATR of 21. Zoom out so you can see as much as possible. Look at June 2007 and how low the volatility was...50 pips per day! Then look at October 2008...almost 300 pips per day. If that doesn't mess with a system, nothing will.

thank you all for clarifying my doubt!

Come on this of "system stop work" are myths, that's not true.

The market don't change, and the "system" always work as the same, whenever it is.

The thing is, you just didn't tested enough, or have bad-judge it, it's something that happen at subconscious level, traders always see good trades in the chart, past trades.. and the system is always nice, because human brain tend to give more importance to good trades then bad ones, but the system never stop to work..

my cent.

I know nothing of the strategy discussed, but just wanted to add my 2 pence worth!.

Often in todays market of conn artists and scammers selling robots that promise the world, you will find that the strategy was built and founded upon things which are not sound principles of trading - sometimes known as curve fitting. Essentially anyone can make a system that looks great, let me give you a quick example.

Say you notice that if you sell or buy on macd turns, sometimes you can get the turns nailed for a nice profit, unfortunately there are other trades that lose - if only you could filter them out! Well, how about I combine the stochastic and then the time of day and then things start to look a little better. Oh, and now I really fine tune things by backtesting for the absolute best results. I still have some losers which i am not happy with, so I examine these and notice that they would be winners if i opened my stop up to 700 points - ok its looking great now!!! This is just a quick example, scammers often include 10's or even 100's of variables which are almost selecting the trades that lost and taking them out of the results.

Well, let me just say, there is an extremely fine line between sound backtesting/tuning and curve fitting, and unless you are highly experienced in this you do run the risk of producing a strategy which will almost certainly fail in the future.

One test I always apply: Does this strategy work on other currencies? - if it doesn't I am immediately skeptical of it.

Does the strategy rely on huge stops? - beware of this, a common scammers way of sucking you in on strategies making 99% winning trades. Oh - these people make my blood boil.

Needless to say I am not a robot trader, however, I am a programmer and like to think that I know how to do things properly!

EA sellers (and systems sellers in general) are the worst ba***rd in this business, they just point in the direction of newbies traders (and there are tons), and steal their money without think about it, i would open a cause to let stop this kind of business, to let it become illegal!

It's a soo bad thing :|

Yes Tunera, it would be great if there was some kind of controlling body out there where these companies would all have to sign-up to and be scrutinised, but I'm also for cutting the red tape that hinders many businesses aswell, it's a very fine line. I'm sure there are some hidden gems out there somewhere, but its a jungle out there and for the majority of people they will get bitten by a snake, mauled by a tiger and stampeded by a herd of elephants, and if they are still standing they may eventually manage to actually find the gem.

Anyway, back to my post!

Many strategies out there will definitely become totally inefective with changes in market behaviour, it doesn't mean that they are poor systems, it just means that people need to be educated to use and adapt them with the market - and as a result are not useable by newbies as they won't have the experience to adapt.

Other strategies will not be effected by changes in the market dynamics, and those are perhaps the ones newbies should 1st learn and master before looking into other areas. Specifically here I am talking about the bread and butter of all traders, Support & Resistance and trend lines, learn this in your chosen market and you won't go far wrong - all charts throughout all history work with S&R. There are other offshoots, like Fibonacci, but i classify that as S&R too, just some guy gave it his own name!

So, you need to thoroughly understand the principles used in the 1st place to make the startegy work in order to ascertain how to use it in the future. If its based on principles that have always worked, then its probably a safer bet, however if its based on some crazy mix of lagging indicators, then you better be on your toes to adapt, or dump it.

First of all visit http://grailtrading.blogspot.com/ and read that tale of woe. The writer speaks of their thinking that six years was a sufficient period to backtest. Based on my testing experience I think not.

If you use anything that has parameters then you are exposed to the possibility of curve fitting. Back test, forward test it makes no difference. If you wish to avoid such situations then you need to make your parameters dynamic, thus self adapting. Such a system should work on almost any market. As one poster pointed out a system based on support and resistance should not be vulnerable to changing market conditions. Start from the premise that everything you read or hear is false until proven otherwise. This will save you a lot of wasted time and money.

My 2 cents worth.

my thoughts (i.e. noob trader thoughts) :

a system does not ever stop working. it simply never worked to begin with...

it just took time to reveal this fact to you.

Some years ago I tried very hard to make a living gambling, mostly on Horse Racing.

I used many systems over a couple of years and I found that most would work for a period, but all finally failed.

I could not understand why a system could be very profitable, for say six months, and then fail miserably.

Many people have tried to find an answer to this, and the concensus is, although you may not realise, subtle things do change that you may not even be aware of. New Trainers come into the game, tracks change slightly, start and finish lines are altered, unusual weather conditions, a change in the way that starting prices are calculated etc.etc.

This means that the ratio of winning bets to losing bets and the starting prices can change enough to make what was a winning system a losing one.

I believe that this applies to Forex Trading as well. There are many more subtle changes in the currency Markets than in racing, and this will mean that any winning system can fail at any time.

There is nothing that is predictable in Forex Trading, if there was we would all be rich.

Robert.

I totally agree with you, systems will and do fail, the best way to make money in this market is learn how to trade and get rid of the systems all together, you do not need indicators like the MACD and Stochastics, CCI ...etc, all you need to do is learn PA and once you have done that and have a good money Management Model you are good to go. But of course a good money management model needs to be inplemented even with a system.

consistent profits will only come by learning how to trade and the systems will never teach you how to trade.

--NM

Ask the banks trading derivitives. I doubt an automatic system wins long term. Recognize LONG TERM TREND and buy SHORT TERM WEAKNESSES "AGAINST THE TREND"

All systems will eventually stop working. This is because the market is always adapting and a trader must keep an arsenal of actionable ideas for any situation which the market presents.

While a system will eventually stop working, it will also eventually start working again. Complex, ain't it?

1. Create a plan.
2. Trade the plan.