I’m a quant in power markets, focusing mostly on derivatives and exotic financial instruments. My background is operations research and finance; my peers come from math, financial engineering and physics. Now, with the emergence of AI and increasing acceptance of learning models, we’ve all become data scientist and analysts of sorts. But that’s not where the value is.
GPT can tell me the descriptive statistics, give me all the stationarity diagnostics and build me a Markov models in a day. I wouldn’t pay someone to do that. I would, however, pay someone who can connect the dots, so to speak, and come up with solutions that either make me money or save me money. That is, if you apply for a position in my team, econometrics, stochastics, calculus, Ito calculus, decision trees and neural networks, statistical inference and production grade Python are basic requirements. I need you to show me how you’ll make me money with all that knowledge. What trading strategy can you create with that knowledge? How can you improve our risk management? So my advice would be: learn Python and learn it well; look at Kaggle competitions for applications.
That’s off the top of my head; more than happy to provide tips or references. I just don’t want you to think that an analytics program is a magic wand. A STEM background remains more valuable.
Those are useless; I call them “asshole questions” and generally a red flag when I started my career. They say they’re meant to see how people reason; I call them schadenfreude.