Should You Absolutely Use Python for Quant Development?

2020-02-13

There was a great little tidbit on Twitter today that sparked some thoughts from yours truly. It stemmed from an observation posed by @MikeHarrisNY, and I loosely paraphrase: Python is a great language for quant development, but it isn't the only language for quant development.

This one hit close to home for me. You see, Python is not my first language. I think it may be a seventh or eighth language, actually. So it does beg the question, what's to gain?

First of all, a quick segue--a language itself is not difficult to pick up. With the exception of a few esoteric ones, we'll talk more about those in the future I promise, once you have learned one, you've learned them all. There are enough similarites to walk through them no problem.

Where the pain really comes in is the environment and ecosystem. There are nuances in every language. JavaScript, for example, requires learning the Node Package Manager and how to manage local dependencies versus global dependencies. In much the same way, there are quite a few different solutions to the same problem in Python, virtualenv, etc, and some are better than others. I really hate to say it, but there is NO quick way to understand any of this stuff. It takes time, heck sometimes years if you're not the sharpest scythe in the shack to pick up things.

And this brings it full circle to Mike's point: use the language you're most comfortable with because avoiding bugs is all that matters.

Do I fully agree with that? No, not fully... there are definitely benefits to using one over another.

But enough about that because the crazy fool below uses MATLAB, so get used to being flexible, friend.

https://www.indicativequant.com/offers/machine-trading-chan