PAL EL system code for Multicharts

I recall some members here were Price Action Lab users. Has anyone used the EL system code generation with Multicharts? Does it work well? I am think of upgrading to v7.1 and your input will help.

It does what it is supposed to do by generating the simplest of trading systems built around the patterns. This saves a lot of time compared with how it used to work. They might not be systems you would want to trade live as is but they are a good starting point.

If you like how PAL works then this is a useful upgrade.

Alex

Thank for the reply. Those systems may be good starting point but I am looking for ways to build systems that are more advanced. I have noticed that when more patterns give a signal there is usually higher chance of a winning trade. I like the PAL concept but there must be ways to get more out of it. I arranged for a demo of the new version to help me decide.

I understand that the method of system code is based on the OR Boolean expression. Have you by the way considered or used more complex code based on both AND and OR?

My preference has been to use all the patterns. It all depends on how many patterns you can extract. I always make an indicator first which helps to see the patterns forming. It is quite possible to have multiple long or short patterns in sequence on successive bars and not just multiple patterns on a single bar. Adjust your position limits in the properties to see the effect of this. You may even find short and long patterns on the same bar.

If you have a lot of patterns then I could imagine that using "and" would be useful. I have been told that this does in fact improve the win rate. Unfortunately I can't say that I have been able to confirm this.

If your sample of patterns is small then excluding single patterns might not be ideal from the point of trading frequency. On the other hand, I have found it useful to use the genetic optimizer to sort these patterns. Sometimes using all the patterns does not produce the best systems. There is a lot of experimentation involved which is very time consuming.

I am very comfortable now using PAL but it took an enormous amount of time.

Your suggestions are of great help. Thank you!

Alex, if you are still around I hope, have you used the other workspace options of PAL and if you have what do you think about them? I am currently looking at the p-indicator and the scan while the program is working on a long search. I cannot figure out best way of using these other workspace results. Any input will be appreciated. Thanks in advance.

With the P-indicator, I export data going back to 01/01/2000 for 50 large caps and then look for directional bias on a daily basis. Here is an example of how Michael Harris does this on weekly data and you can extrapolate this idea to ETF's or other indexes. This is not the only way to use the data but this way is straight forward to understand and often useful.

http://www.priceactionlab.com/Blog/2016 ... r-results/

I suggest that you start saving results in some kind of data management system so that you can see over time what the scans predicted and what actually happened. I think this is the only way to really understand the capabilities and the limitations of this kind of analysis.

Since this is off topic a far as Multicharts is concerned feel free to contact me via Skype and I will try and help you further. My Skype name is hughesfleming.

Alex

I gave up trying to develop systems with PAL and another program about 3 years ago but after doing some research and reading the book by Michael Harris (Fooled by technical analysis) I realized some mistakes in my process. PAL works well and code for EL was updated recently. Certainly worth upgrading to v7.1 if you want to follow this route.