Career prospects for a Math PhD student in pure math?

Question

I have scoured the internet and everyone seems to be saying different things. These range from

• "You will have no problem getting a job in quant/finance etc", to
• "You are in a hopeless situation and are completely unemployable"

Consider the following situation:

• I am at a PhD student at a mid-20s US institution studying pure math
• I have very limited programming skills
• I am an Australian national and want to live in Australia long term.

I am not overly optimistic about my future career prospects. Since I am near the beginning I thought I better think about this now rather than 5-6 years down the track.

Question: Is it likely that I'll be able to find a reasonably well-paid, stable job after graduation (even outside of academia)? Are there things I can do now to improve my chances (beyond "trying hard")? Would I be better off doing another masters in computer science or machine learning back in Australia instead of this PhD in pure math?

Answer 1

Given the current market for academic math researchers, I'd encourage everyone in math to make serious plans for an alternative career that you'd be happy with. Even for most people who end up with tenured positions, it's not clear until well after graduation whether that path is likely to work out, so even the strongest students should be considering other options. In math the main alternative tracks are data or coding focused industry jobs in finance or tech, or teaching-focused academic positions. In both cases, these job markets are competitive and they're not looking for people whose main qualification is that they didn't get a research position. If your alternative option is getting a job in industry, you should already be learning to code and trying to get relevant experience, while if your alternative option is getting a teaching-focused job, you should already be looking for teaching opportunities that will make you stand out as an attractive candidate at teaching-focused institutions. (In the US those are often small liberal arts colleges, but in Australia I'm not sure how many teaching-focused jobs exist.)

The good news is that there's lots of good jobs for math PhDs out there (much better than many fields), but you won't just stumble into them you need to develop relevant skills and experience.

Answer 2

It sounds like you have no idea what the career options for a Math PhD in Australia are like, which is understandable but not a good place to be. It's important to educate yourself about what the market is like, preferably as soon as possible. Since you're in the US, you won't be able to rely on your university's career center, so the burden of research is on you.

Go to an Australian jobs portal such as SEEK. Search for jobs requiring a PhD in math. If you know where you will settle in Australia already, then narrow only to jobs in that city. You'll get results like this one.

Our client is a leading provider of form and data software solutions for the global racing industry. They are an established and stable organisation who are perfect for those sick of the corporate life.

Working with some of the best mathematics and stats minds in the city in a small and focused team environment, you will work on challenging and rewarding modelling projects.

To be considered for this role you will have the following:

• A strong background in mathematical and statistical modelling
• Previous experience working as a Quant Analyst, Data Analyst, Data Scientist or Big Data Analyst - desirable
• Masters or PhD in maths / statistics / machine learning / econometrics essential.
• Experience using C++ or Python is highly regarded, although Matlab & R experience will also be considered.
• Strong algorithmic coding skill and solid knowledge of data structures
• Application clustering / Big Data experience using Apache Spark Experience with statistical forecasting essential.
• Understanding of machine / statistical learning techniques such as non-linear regression, kernel regression, support vector machines (SVM), neural networks, classification trees and similar techniques very beneficial.
• Experience working within the horse racing / wagering industry highly desirable

It's up to you to decide if you find this job attractive. If the answer is no, then search for another job until you find something attractive. Once you have an attractive job, do your best to pick up the necessary skills. For example to be competitive for this job, you really want C++ or Python proficiency, and preferably Matlab and R as well. You want modeling experience, you want to understand neural networks, classification trees, etc.

Your PhD is an opportunity to pick up these skills. As you say, right now you have very little programming experience which immediately rules you out of this job. If you still want to do it, then should you get the chance to direct the trajectory of your PhD - e.g. if your supervisor asks you "what do you want to do next" - this is your chance to learn those skills.

As for whether it's better to do a second Masters degree, well, you can apply the same methodology. Search SEEK for jobs requiring a Masters in computer science or machine learning, and see if you find those jobs attractive (relative to the ones that need a PhD in math). If the answer is yes, then it's probably better to do a second Masters.

Answer 3

Since a doctorate is geared toward research, the obvious place for you is somewhere that does a lot of research. Academia qualifies, of course, but most governments have/sponsor labs for research, especially econometric research for which a maths PhD is probably a good start. Insurance companies also hire a lot of mathematicians as actuaries and financial analysts.

There are other companies that, in pursuit of a business model, so a lot of model building of various things - populations, finances, products, etc. Mathematics is a good foundation in any model building scenario, though some domain specific knowledge is usually also needed. But most important is the ability to abstract away from the details, which is one of the most essential mathematical skills.

As to learning CS for use with mathematics, you don't really need to spend the time getting a MS for that. One can learn programming and data structures from books (and practice) quite well. That was my path, in fact. For the more esoteric studies of CS, a guide would be good, such as a degree program, but if you want to work as a mathematician, much of that wouldn't be essential, at least initially.