Stats vs Applied Math Undergrad

[js_bsn]

Hi all, I'll be starting my undergrad (in Europe) this summer and I need help with deciding for my undergrad. I've done a bunch of reading and have always seen different opinions, so I need some more assurance. My ultimate goal is to break into QR, which I know is extremely rare from undergrad but I'd like to lay the foundations. It's also worth noting that I am also very interested in machine learning. So, my two options are:
1. Applied Math
- Courses: Analysis I - III, Linear Algebra I - II, Topology, Stochastics, Optimisation, Bayesian Statistics, Financial mathematics courses, functional analysis, probability theory, intro to ML

2. Statistics & Data Science
- Courses: Analysis I - II, Linear Algebra I, Statistical Inference I - II, descriptive statistics & probability theory I - II, multiple practical statistics courses and multiple statistical modeling courses, mathematical methods in statistics, intro to ML

My concerns are, some people say you should take as much math as possible, as it helps overall, but I've also read that most of quant finance involves a lot more statistics than some of the more calculus based courses (esp for buyside). But my concern is, is the statistics degree 'rigorous' enough per se? Any feedback would be greatly appreciated, i look forward to your replies !

 

[GWeitz]

Not sure by what you mean regarding calculus vs stats based courses, because calculus plays a fundamental component in stats. Regardless of the concentration you choose, a solid understanding of calculus and linear algebra is a must.
For QR, the stats track is probably going to be more useful. I would sprinkle in a course in optimization though.
Also, make sure you leetcode religiously at least 6 months before you start interviewing.

 

[js_bsn]

Not sure by what you mean regarding calculus vs stats based courses, because calculus plays a fundamental component in stats. Regardless of the concentration you choose, a solid understanding of calculus and linear algebra is a must.
For QR, the stats track is probably going to be more useful. I would sprinkle in a course in optimization though.
Also, make sure you leetcode religiously at least 6 months before you start interviewing.

Hi, I appreciate that you've responded. In terms of calculus vs stats based courses, I meant topology, optimisation, further analysis, and an extra linear algebra course (though not calculus). I'm aware that linear algebra plays a massive role, so does that one extra course justify choosing it? As for the recommendation for a course in optimisation, it's unfortunately not offered for the statistics degree, only the applied maths.

 

[GWeitz]

Hi, I appreciate that you've responded. In terms of calculus vs stats based courses, I meant topology, optimisation, further analysis, and an extra linear algebra course (though not calculus). I'm aware that linear algebra plays a massive role, so does that one extra course justify choosing it? As for the recommendation for a course in optimisation, it's unfortunately not offered for the statistics degree, only the applied maths.

If that’s the case I’d say to still stick it out with stats you’ll get more bang for your time

 

[Daniel Duffy]

Learning statistics after applied maths >> than the other way around.

 

[js_bsn]

If that’s the case I’d say to still stick it out with stats you’ll get more bang for your time

Thanks for the input then! Still have a few months before I ultimately decide but really do appreciate the input.

 

[js_bsn]

Learning statistics after applied maths >> than the other way around.

That is the general consensus i've collected after scrolling through various threads.. But why is it like that, if I may ask?

 

[GWeitz]

Thanks for the input then! Still have a few months before I ultimately decide but really do appreciate the input.

For sure.
Just as a disclaimer: I double majored in physics and applied math. I wasn't as constrained as you regarding my courses, and I ended up taking almost all my math electives in stats. Not sure what the other comments here mean by doing applied math before stats, it depends how you want to apply math?
It seems like both tracks cover your bases pretty well, so conditional on the fact you want to be QR, and from my experience interviewing with hedge funds, stats and coding are going to be most heavily emphasized. Even after doing an MFE, most of the technical questions I got were about my projects which involved statistical methods, regressions and other ML models.

 

[Daniel Duffy]

That is the general consensus i've collected after scrolling through various threads.. But why is it like that, if I may ask?

Statistics is an _application_ of mathematics to a certain domain.