Thread: Building a system based on price action

Hey everybody I have been wondering about here: I do not actually like using indiors, I discover they lag. What I am interested in is chart formations and cost action, but how can you integrate that into a system with rules and whatnot, since it is more optional than a mechanical system with indiors. Any ideas on this? Thanks.

Originally Posted by ;

Hey everyone, something I've been wondering: I don't actually like using indiors, I discover they lag. What I'm interested in is price action and chart formations, but just how can you integrate that into a system with rules and whatnot, as it is more discretionary than a mechanical system with indiors. Any thoughts on this? Thanks.

This is a superb question. I could probably write a few pages on this, but I'll try to be short as possible.

Computers now can perform just about everything your imagination can come up with. If you've got the time patients and ability (which can be heard) to inform it exactly what you desire.

There is actually no discretion in a price pattern. Their just how loosely you stick to the rules, which you'll be able to let the computer to perform.

An example of this would be tweezers tops, a favorite candlestick formation I could let the computer to follow specific rules. Or I could let it within a certain parameter considered it tweezer tops.

The exact same thing can be carried out with remember flags, triangles, wages and also a list can go on and on.

You can also let the computer to learn from its mistakes, or predict what is coming or to discover new price patterns.

And that is without utilizing neural networks, generic algorithms and fuzzy logic.

They can all be hardcoded to a system. Simply use your creativity. And see.
Read a lot.

Let's say you would like your system to market at immunity when a particular candlestick pattern happens.

First you'd define the immunity place, let us go with Fibonacci.

So price arrives at the 61.8 first you can not let it one special price. Very rarely will price hits the specific mark and turns. So what you need is 61.8 zone. I find the very best approach to create this zone is simply by expanding the area utilizing average true selection.

Now price has arrived in the 61.8 zone now you can let the computer to begin searching for a price pattern or hundreds of price patterns. Or perhaps tell the computer to think of a price pattern which works best.

I understand I have not been very specific, but should you have additional queries. All try to answer.

That I just don't wish to make this 20 pages long.

Originally Posted by ;

I don't actually like using signs, I find they lag. What I am interested in is price action and chart formations, but just how can you incorporate that into a system with rules and whatnot, since it is more optional than a mechanical system with signs.

Most price action-based systems like Vegas' or Jim16's are focused around discovering areas of important support/resistance, await the price to retrace to these amounts and then follow closely price action so as to choose whether to enter or not.

The question is what do you believe as a significant area of support/resistance. For example, Jim16 tracks 150 and 365 MAs, fibonacci clusters and trendlines. Vegas tracks bands that are fibonacci. Etc..

You could also use trendlines to filter cost action, in addition to the methods mentioned below.

Originally Posted by ;

Most cost action-based systems such as Vegas' or Jim16's are centered around discovering areas of significant support/resistance, await the cost to retrace to all those amounts and follow carefully cost action in order to choose whether to enter or not.

The issue is what do you think as a substantial area of support/resistance. For example, Jim16 tracks fibonacci clusters, 365 and 150 MAs and trendlines. Vegas tracks fibonacci bands. Etc..

Originally Posted by ;

Hey everybody, something I've been wondering about here: I actually don't actually like using indiors, I discover they lag. What I'm interested in is price actions and chart formations, but just how exactly can you integrate that into a system with rules and whatnot, as it is more optional than a mechanical system with signs. Any thoughts on this? Thanks.

You can make it easy, or complex. I like easy. One reason I developed my framework was because, as you say, more or less all indiors are historic. My framework provides you with a means to see just how far a currencies are likely to go in any particular direction. In addition, it shows you if the features of movements are changing, which is where many systems fall down in some stage or other.

Personally, I would not get stuck on growing too many specialized rules, because we're in a fluid situation, where the environment changes, therefore trading by too many fixed rules can reduce your profit potential because they can not work all the time. There is not any security in trading that may be provided by anybody else or their own systems. The best security is to build your own skills rather than rely on other people.

Originally Posted by ;

Hey everyone, something I've been wondering: I don't really like using indiors, I find they lag. What I'm interested in is cost action and chart formations, but how can you incorporate that into a system with rules and whatnot, as it is more optional than a mechanical system with indiors. Any thoughts on this? Thanks.

Check out MuddBudda's thread on Dow therory trading. It's the oldest, tried and true fashion I know of. He talks about MACD, But I don't believe the use of MACD is necessary, only a crutch. The cost holds all of the information there is to be gleened from the marketplace. After all, each indior out there's at it's core based on cost and volume, and as the FOREX doesn't use volume figures, all of the info is at the price.

At least, that's my thoughts on it. Thank you SMJ