Early Retirement Extreme Forums

Anyone have experience with tradestation and coding in easylanguage?

I trying to get a feel for how prevalent the language is among algorithmic traders. Subjective feedback on coding in the language would also be interesting.

I am principally interested in developing some backtesting skills and tradestation has the data at reasonable price (free - it seems at least for a flat rate equity account).

Never heard of either.

Algos are typically written in Java or C#. Research is usually done in R or Python. Backtesting is done in whatever you have. Ideally it should be done in the same language that the trading is done in to prevent rogue bugs during the translation.

These languages were near universal, at least back in 2015. Other languages (C or fortran) existed but mostly for idiosyncratic reasons.

As Jacob alluded to, the professional traders use Java or C# (because of the need for speed), research in R, Python, or combination of the two.

Tradestation's Easy Language is what is called a 4th generation language (derived from Pascal) and while it is essentially like scripting, there are a number of inherent limitations with the language. While it is probably still quite decent for fleshing out an idea, you really would be better served by learning Python or R (they aren't any more difficult) with the primary upside being flexibility from Tradestation's terrible broker/platform integration that makes you a slave to their system (with all it's flaws). I used Tradestation back in 2005-2008 when it was still the best of a bunch of bad choices for retail investors. Personally, I think it's time has passed.

If you are looking for free data there are probably better alternatives (depending on the data in question). Quandl in particular has a wide set of free (and reasonably priced) data.

Thank you both for the replies. I think the idea of backtesting and using the same platform was the attraction. Then I skip the possible bugs from going back testing to implementation. The free data is was also mesmerizing. No free lunch I guess.

That said, the old guys usually just had their system spit out the trades everyday and they entered them manually. I know I am still months/years from even that stage.

I hobby code in r and python. Realistically, I need to just get down the basics of the r quantstrat package on Quandl/yahoo data and once I can do that I'll start worrying about finding reasonably priced surviorship bias free data. I'll save that for another post way down the road.