I give up....... on technical trading

[Data_Miner]

You'll understand why indicators don't work very well if you step back and look at what they are and that is a way to reduce the complexity in the market so it can be graphed in a two dimensional matrix: X and Y coordinates. All are designed so anyone with a high school education can understand the concept in about 2 minutes. Now what's wrong with this? The market is a complex entity. If it were so simple that an indicator could capture its essence and dynamics then everyone would make a mint.... just add a few dozen more indicators than the next guy to your trade strategy and instant success.

So the concept is at fault: Complexity reduction, the reduction of information in the market so you can trade by looking at 6 cute colored lines on your computer screen. This is a sort of open secret.

The folks that sell you the nice software for calculating and graphing hundreds of pretty indicators won't tell you this, but why do you think they market trading software as opposed to using it themselves to trade the market. They know that their product doesn't make money in the market, so they sell it.
If it did they wouldn’t bother with marketing, tech support and customer service but just use it to trade and make money directly.

So what to do? What can handle complexity? The general term is soft computing. This includes Neural Networks, Genetic Computing, SOM, Fuzzy Logic and a host of other disciplines.

You might want to read the book called "The Predictors". It details how 6 grad students used such computing systems to trade. They started in the basement of a rented house and in a few years sold out to UBS for $500,000,000. They didn't use ADX, MACD or Bollinger Bands!!!

If you decide to go this route get industrial grade software for soft computing. Avoid the “toy” programs marketed to traders like NeuroShell. They are junk. Your software acquisition budget will be in the $20,000 to $50,000 range. That’s the route I went and haven’t had a weekly net loss in years nor do I use a single indicator.

 

[Grey1]

You'll understand why indicators don't work very well if you step back and look at what they are and that is a way to reduce the complexity in the market so it can be graphed in a two dimensional matrix: X and Y coordinates. All are designed so anyone with a high school education can understand the concept in about 2 minutes. Now what's wrong with this? The market is a complex entity. If it were so simple that an indicator could capture its essence and dynamics then everyone would make a mint.... just add a few dozen more indicators than the next guy to your trade strategy and instant success.

So the concept is at fault: Complexity reduction, the reduction of information in the market so you can trade by looking at 6 cute colored lines on your computer screen. This is a sort of open secret.

The folks that sell you the nice software for calculating and graphing hundreds of pretty indicators won't tell you this, but why do you think they market trading software as opposed to using it themselves to trade the market. They know that their product doesn't make money in the market, so they sell it.
If it did they wouldn’t bother with marketing, tech support and customer service but just use it to trade and make money directly.

So what to do? What can handle complexity? The general term is soft computing. This includes Neural Networks, Genetic Computing, SOM, Fuzzy Logic and a host of other disciplines.

You might want to read the book called "The Predictors". It details how 6 grad students used such computing systems to trade. They started in the basement of a rented house and in a few years sold out to UBS for $500,000,000. They didn't use ADX, MACD or Bollinger Bands!!!

If you decide to go this route get industrial grade software for soft computing. Avoid the “toy” programs marketed to traders like NeuroShell. They are junk. Your software acquisition budget will be in the $20,000 to $50,000 range. That’s the route I went and haven’t had a weekly net loss in years nor do I use a single indicator.

Do you use Fuzzy Logic in your analysis?

grey1

 

[Data_Miner]

Yes, mostly for trade decision logic.

To get to that point I use mostly nerual networks and SOM technology. It truns out that the same type of software that NSA uses to match voice prints of cell phone intercepts in the search for terrorists works very welll on the markets. Sifting throuch thousands of intercepts looking for subtle patterns in tone, inflection and such is a very complex task...much like the markets.

 

[Grey1]

Yes, mostly for trade decision logic.

To get to that point I use mostly nerual networks and SOM technology. It truns out that the same type of software that NSA uses to match voice prints of cell phone intercepts in the search for terrorists works very welll on the markets. Sifting throuch thousands of intercepts looking for subtle patterns in tone, inflection and such is a very complex task...much like the markets.

Thanks for the reply.

Could you kindly explain why you think fuzzy logic is more sutiable for time series modelling ?
Are you using a Knowledge based system with fuzzy logic incorporated in it or the market knowledge already coded?

Grey1

 

[Crap Buddist]

$500,000,000.

Anyone know what sort of return did they post to get that price tag ?

TYIA

 

[Market Wizard]

You turn your back for one day.......

Anyway... I must say I am sorry to have offended so many by my 6 months limit... but this is a personal conviction and I am putting it to the test....

like Mr.J-Arthur I have been risking very little amounts of money and occasionally making enough to recover...

I still consider six months to be enough.... but I will have to wait till that time has passed to find out...

As for my background... it is important...

Business and math can speed up your understanding of the technical indicators and what they mean... even though just using them as they were intended does not need education.

Programming is important when you want to use MQL4... which I picked up less than a week.... something that would have taken me more than a month if I had no programming background... and that is just to be able to write the simplest of indicators or EAs.

However... this is my take on things... and I am kind of "wet behind the ears" when it comes to trading so maybe I am being over-optimistic. I will find out at the end of the six months.

Time is not really the issue when it comes to trading. THe amount of time it takes you will depend entirely on your aproach to learning. Do you know exactly what needs to be learned? Can you seperate the worthwhile information from the rubbish? Do you know exactly where you have to be in understanding each topic?

These are some of the key questions that you have to answer. Analysis tself is very simple, the element that takes the most time is asking yourself which direction you should be heading in next. It is not as simple as it sounds because every trade has a completely unique perspective of the markets, the risk, the reward, the methods etc.

You have read on this very thread comments from people who will tell you that indicators will not work, some have said they will not work alone, some say just price. They are all correct. Because their understanding of the markets is built upwards from there and it works for them.

The best way to reduce the amount of time you take to learn is to have a plan. You need to have an idea of your trading plan so you can build your learning plan around it. If you can do this correctly, there is no reason why you cannot speed up your learning.

Something you should consider, though it may sound obvious is what you want out of the market. The hard part for many is not taking points out of the market, it is harder to leave some on the table.

But once you develop and choose your approach, you will discover that whether you use price, volume or other indicators, when used correctly all will give you an entry/exit around the same point. This approach may not be suitable for scalping, but daytrading, yes.

Good luck with your learning.

 

[jiggly]

I have a SIMPLE STRATEGY THAT HAS HISTORICALLY PROVEN TO WORK. over the long run this will make you money.

it is :

check first to trade in the direction of the longer term trend (the market tide), then look to ensure you are also trading in the direction of the medium term trend (the market wave), if these two are lined up then look to make an entry in to the market when the shorter term trend (the market ripples) are flowing in the same directions as these two larger trends. when all trends are in agreement look to enter a trade, use a technique such as breakouts pullbacks price cluster or what not to act as the trigger.

use support and resistance to measure potential risk / reward as well as having a plan to exit before you commit. exits give the trades 'potential' only, so if the market does not have the 'potential' to give you a risk / reward of at least 1:2, it is probably not worth taking.

use strict money management of about 1% risk per trade and place the distance of your stop loss in accordance with the structure of the market, or have a fixed pip risk amount set, this depends on how flexible you can be on trade size. experiement with what works best for you here. if the trade gets to your predetermined potential exit, you are gettin out.

a complete strategy that has worked an made a few individuals very wealthy in the past, only you can make this work for you.....

even still i dont think you will be profitable within 2 years either, i humbley submit this as someone who has been there, read and learnt everything i possibly could also, hell i read more material in a year on trading than i did for my whole degree spanning 4 years.

what you 'think' you know and what you actually do Know outside of trading, you will come to find they dont mean ****. you are learning to walk again my friend, and you dont know what you dont know.

good luck

 

[A Dashing Blade]

I have a SIMPLE STRATEGY THAT HAS HISTORICALLY PROVEN TO WORK. over the long run this will make you money.

it is :

. . . .


good luck

Good post and a succinct summary of Chick Goslin's book.

 

[jsc]

Being a good trader is rarely possible unless you have a good grounding in fundamental analysis, the ability to sort stocks through newsflow, results, P & L and cashflow.
then look at RSI, MACD, Stochastics, Chaikin Money Flow and Williams %R. Learn to plot trend lines. Learn they are not all pretty parrallel lines, plot the actual peaks and troughs. Learn all the major candlestick patterns.
That is enough knowledge to make money consistently. Use a 1 - 2% stop loss, trail winners and let bad trades stop out. Do not trade unless you can potentially make double the amount of your stop, i.e. 2-4%.
If in doubt, leave it out.

 

[Mohammad mirshekari]

Thanks for the reply.

Could you kindly explain why you think fuzzy logic is more sutiable for time series modelling ?
Are you using a Knowledge based system with fuzzy logic incorporated in it or the market knowledge already coded?

Grey1

Hi ostad

 

[ph_trader]

i see lots of very good advice and i hope traders take time to read them, its free knowledge and experience. thanks guys