Why you should learn to code

[Robert Morse]

Lol ... True. Coding for fun is not the same as coding to make a living. I guess same as trading in a way.

In conversation with a hedge fund manager (HF hedge fund all automated) with a blackjack past (Large teams playing in Vegas), I told him I like craps. He said do you ever make money and I said rarely, but my wife and I like to play as long as the losses are low, we look at it as paying for entertainment. He said, "I don't understand. Winning is fun, losing is not. I only play when I have a good count and the odds are in my favor. I won't make money every time, but overall, I win and winning is fun. I do not understand why you would play a game you can't have a positive edge."

Bob

 

[Turveyd]

Do you think I can still learn coding at the age of 30? I mean how long will I spend learning it.

It's not really age specific, just have a cheat sheet to remember commands as you find them, coding is more of a problem solving thing, constantly finding then solving problem till the code is written, if you haven't got a brain that easily and likes solving problems then maybe not.

 

[cafeole]

IMO getting good enough at programming where you put your money in the hands of a program takes YEARS. I personally won't do it. I program to backtest, write indicators, etc. But I will not turn over my hard earned money to it. Way too much can go wrong that you cannot anticipate and code for.

 

[cafeole]

Let me give you an example. Years ago the Globex went down for hours. If you were in a trade in an emini you were stuck unless you could hedge with a non-Globex equity. What does your program do now? Can it know the Globex is down?

 

[gaussian]

I agree that for day traders, automating their process can be very helpful. However, I speak to a fair number of programmers and math experts, looking to code an equity or futures strategy with little or no knowledge of market structure and little or no trading experience. They think they can make money because they can code.

I've been in the "tech" industry for 7 years now. This mentality is very pervasive especially among silicon valley hipster coders. I've reviewed a handful of systems (at the developer's request) and have yet to find one that is consistently profitable using standard metrics of risk. Majority of them are some combination of Tensorflow or Scikit-Learn with a smattering of "proprietary" (read: data mined) "indicators". Almost universally the data is overfit, and many of these people despite having an education in Computer Science lack the sufficient mathematical sophistication to realize the difficulties of applying machine learning to time series data.

I'm on the opposite end of the spectrum. I only program what is necessary. I'd work in Excel only if it was possible however I generally have to venture into other languages (typically C#, Python, R, or MATLAB depending on what I'm doing).

 

[d08]

Do you think I can still learn coding at the age of 30? I mean how long will I spend learning it.

Absolutely. To get things done and know the basics of a language takes 1-2 years. To get confident and understand some of the more complex structures well takes 5 or more years though. Or maybe I'm just slow.

 

[d08]

Let me give you an example. Years ago the Globex went down for hours. If you were in a trade in an emini you were stuck unless you could hedge with a non-Globex equity. What does your program do now? Can it know the Globex is down?

Your broker would know and if not then I'm sure there's a status page for the exchange. Putting on a hedge is not complicated, just another trade. Alternatively use a timeout, so if the order does not execute after some minutes, place the hedge assuming something is wrong.
This is all doable.

 

[d08]

IMO getting good enough at programming where you put your money in the hands of a program takes YEARS. I personally won't do it. I program to backtest, write indicators, etc. But I will not turn over my hard earned money to it. Way too much can go wrong that you cannot anticipate and code for.

It certainly takes effort but it's worth it, especially if you trade multiple markets. Automation helps you live a more "normal" life with flexible hours. I absolutely hated the fact that I had to be in front of a screen in certain hours.

 

[cafeole]

Not trying to discourage anyone, just stating what I see are the realities of starting from scratch. If you have the time and inclination then go for it. Most people I know think I am wasting my time learning to trade. But this is not only a dream, but a fun pursuit.

 

[zdave83]

...... I speak to a fair number of programmers and math experts, looking to code an equity or futures strategy with little or no knowledge of market structure and little or no trading experience. They think they can make money because they can code.

The strategy has to come first. Finding a process to make money must come first. I would suggest that too much time and effort learning coding takes time away from a process to make money which is vastly more important.

For some people that may be true ... however ... note that some strategies are nearly impossible to define, test and optimize without the data & processing capacity of a computer. Many traders code strategies that are far beyond just automating processes constrained by human ability. For example I'm sure that I could not have developed the strategy I trade today (baskets of options) without some form of coding to test endless variations of strategies.

So then for me coding/strategy looks kind of a chicken/egg thing Fortunately my first role in IT was software engineering, and that skill developed along side my personal interest in trading.