Thanks learner2007 for the comment.
Can Price Action Trading Be Automated?
- Thread starter SimpleMeLike
- Start date
Thanks learner2007 for the comment.
No, by no means did I mean gut feeling/intuition. I meant an educated judgement based on the analysis of a situation regarding its probability. While a program could read a move from A to B, could it judge its strength based on the price movement and volume, its strength relative to relevant previous moves, its direction or importance according to the context etc.? Could a program read a move from A to B and determine if it is a move in the primary trend direction or a move in a correction to the primary trend? Or, can a program really read trends?
I think the OP hopes to be able to accomplish the above and more through a program. But as I said previously I cannot write programs and just rely on opinions such as your's. And while I can imagine that programs can analyse some offshoots of basic TA, given that something as simple as drawing a trend line apparently can't be done with accuracy, what the OP is looking to do seems out of reach. Am I wrong?
Thank you for your opinion.
learner2007,
Regarding the trend.
Honestly, I discretionary intraday trade now, and i define the trend by waking up and looking at the instrument chart I trade, whatever way price is moving is the trend of "that" moment. I just simply want to get on board of whatever way price is moving from 8:30am til about 1pm.
Sorry, based on the above I understood you to mean that you could not read the trend.
But can you determine if 'that' price movement is a move in the primary trend, a reaction, a correction, a resumption of the primary trend, or the beginning of a new trend etc.? If you can, no problem. If not, I would say that you'll have trouble with what you stated you are attempting to do.
Whatever, don't pay attention to me. I just wrote the above so that I could make my 1000th post!
Last edited:
learner2007,
lol haha no worries. Congrats on your 1000th post. We just discussing trading.
Here is a picture of the example how I define trend for myself.
5 minutes chart on /CL in real time just a few minutes.
Trend of the moment is downward. I want to get on board with price going down. I do not care about primary trend or anything at the moment, but getting on board with price going down, provided my rules are in place to go short. You see, I saw this with my eyes. Trying to program this is another story I am trying to figure out.
Attachments
Yes, to all of the above.
learner2007,
According to other comments on this thread, the answer is "yes".
As a discretionary/manual price action trader or whatever trader, I am learning that we can not ignore algorithm trading, especially when backtesting our ideas.
This is why i ask the question can price action be programmed because it will save me alot of time when I have these ideas that I think are profitable. I can not spend months and months backtesting or forward an idea by hand and eyes if a computer program can test that idea after a week or so of programming. Considering I have full time job and other family duties, it makes sense to try and program ideas.
Would you not agree in logic of my post now?
Thanks,
Of course a program could do all the above. You just have to spend more time developing the program, giving it rules, gathering the requisite raw data. It could easily be a full time job. The more experience with manual trading you have the greater the amount of nuances you have to take care of (things like you mentioned above). Each nuance probably multiplies the development time by 2 (just a figure from my head). So if you have only one nuance, may be 1 day of effort. By the time you have 10 nuances, that comes to 1,024 days and counting...
My wife keeps asking me when my program will be 'completed'. I laugh because I realize it will probably never be.
No, I'm sorry but I can't.
I was surprised by nonlinear5's reply above as it is opposite to the numerous explanations I've heard. But maybe he knows more than others. I can't say.
I don't know how you yourself go about back testing manually that takes months and months. But I've back tested more ideas over the last 40 years than you could shake a stick at. If I have a setup that I want to back test, I go over old charts and find 100 or more examples, and based on the fact if the setup worked out or not, I can determine its probability. If over xx%, I will use the setup in the future. Such back testing will take me 2 hours or so, not months.
Of course a program could do all the above. You just have to spend more time developing the program, giving it rules, gathering the requisite raw data. It could easily be a full time job. The more experience with manual trading you have the greater the amount of nuances you have to take care of (things like you mentioned above). Each nuance probably multiplies the development time by 2 (just a figure from my head). So if you have only one nuance, may be 1 day of effort. By the time you have 10 nuances, that comes to 1,024 days and counting...
My wife keeps asking me when my program will be 'completed'. I laugh because I realize it will probably never be.
Thank you. I can understand how the program would follow the rules you give it. But what I can't see is how it could judge the validity of a movement in price which is relative to other price movements, and to which the human would judge if this and that is too much or too little and that too much or too little will be different in each case. It would have to be able to judge if this price movement is too far out on the branch to be dependable. In other words it would have to 'think' and analyze as we do.
Last edited: