There is a factual mistake in the webinar. At 20:26 and 36:00 it mentions translating to C/C++. This is true. However, in order to do this, you must purchase MATLAB Coder which costs $6500. Matlab Coder only works with Matlab Professional which costs $2150, as mentioned in the webinar. Matlab Coder does not work with the $149 Home Edition.
I did find the webinar interesting, especially the Matlab part. The question is, do TradeStation users really need Matlab? Can't TS users simply use the tools they already have at their disposal? Unless you want to do some very complex calculations, I can't see the need for Matlab. Furthermore, accessing intraday pricing might be difficult with Matlab.
Yes and no. I do not want to get into a position of defending Matlab but it certainly has it merits over TS.
In the most simplistic terms, if you want more control over your backtesting and to see what's really going on during the process then Matlab is the way to go. TS certainly has a lots of bells and whistles programmed in, especially the optimization report; however, I do not backtest off those types of results as they can easily lead me to accidentally backfitting. What I prefer to do is create a system in Matlab and then once I am satisfied I will bring it into TS to test.
Matlab is just another tool, similar to a trading system or process. If it doesn't fit the users personality (not matter how great the tool is) it will be worthless to the individual. I recognize I have a personal bias to Matlab as I have used it for +20 years.
FYI - WFA is very simple in Matlab. Simply create a loop, iterate through each loop, record your results and then step forward using the new output. It's as simple as:
The following 2 users say Thank You to GoldenRatio for this post:
Though the point that you can call R functions from Python does have merit. I currently use the anaconda Python stack for data analysis. Spyder gives me pretty much everything rstudio does, yet I don't have to deal with R. (Or S, if we want to be pedantic...)
The following user says Thank You to skoa for this post: