Finance Job Market for Astronomy Ph.D.?

[Physics2468]

I am currently applying to Astronomy, Astrophysics, and Physics Ph.D. programs. I am doing a Ph.D. because I can't wait to learn more about cosmology and to computationally model galaxies, but I understand that jobs as professors are rare. Looking at backup plans, I've heard that some academic jobs are less competitive and community colleges are an option. However, I was wondering how the field of quantitative finance is doing with respect to Physics Ph.D.s. Are Ph.D.s still being sought after with good salaries?

Also, if I get an Astronomy Ph.D. rather than Physics, will it be much harder to find a position as a quant? I already know C/C++ and some parallel programming.

This is just a backup plan if I can't find a postdoc etc., and I know it is hard to predict how things will be in 6 or so years. But how is the job market currently? There are a lot of posts on these forums but most I find are from 2010 or earlier.

 

[Stephenk53]

It seems physics jobs and astronomy jobs are going to grow according to this site https://www.bls.gov/ooh/life-physical-and-social-science/physicists-and-astronomers.htm. Here is another site that may help https://astro.berkeley.edu/astronomy-fans/aa/questions-about-careers-in-astronomy.

I will not offer personal advice as I do not have experience but I would suggest contacting a college admissions office or similar to get advice.

 

[Vanadium 50]

The word on the street is data analytics is the new quantitative finance. I can't speak to the accuracy of the word on the street, but I would argue that in any case you should build up your skill set for your target job beyond an astro PhD and knowing some C++.

 

[Physics2468]

Thank you Stephenk53. That is interesting information! But I am still curious about the job market in finance.

The word on the street is data analytics is the new quantitative finance. I can't speak to the accuracy of the word on the street, but I would argue that in any case you should build up your skill set for your target job beyond an astro PhD and knowing some C++.

Hmm. I will research more about relevant skills to learn; that is certainly imprortant to consider. I will probably be busy in grad school, and doing great research for a preferred career in astronomy will be my main emphasis. But I can very likely sneak in additional training along the way for my backup.

Still, I would love to hear fresh opinions from others working in the industry.

 

[Päällikkö]

The industry is very varied and nuanced whereas your question is very general. When one speaks about quants nowadays on the internet it often means "buy side statisticians" where it traditionally used to mean "sell side mathematicians". Both roles still do exist and are often fit for different kind of people (or for people with different backgrounds). Buy side as a rule is more geared towards machine learning and so on, whereas sell side is stochastic calculus and all that jazz.

Because the curriculum has become rather standardized: you can get a degree in machine learning or mathematical finance nowadays, the need and demand for the "all-versatile" PhD has too decreased. I think it will be very difficult to get a job without specifically training for the interview because the competition will, even though they may "only" have a Master's, have a degree directed to the industry's needs and have an idea what the job is about (often through internships). Just to clarify: What I am saying is not at all that it'd be impossible to get a job, but rather that it's unlikely to come by for free, without investing quite a bit of time to preparing for interviews, and so even though just a backup plan for you, you'd probably want to prepare at least for a couple of months (part time, while working on your PhD) and depending where you're applying, consider first going for internships rather than full time positions.

 

[CrysPhys]

Also, if I get an Astronomy Ph.D. rather than Physics, will it be much harder to find a position as a quant? I already know C/C++ and some parallel programming.

I don't have a direct answer to your question. But here's an issue for you to dig into further. I'm a PhD physicist. In the course of my career, in response to layoffs and industry meltdowns, I reinvented myself many times. I know other PhD physicists who have done the same ... including some who went into finance. APS News has frequent columns highlighting physicists working in alternative careers. There is a general perception that a PhD in physics is a difficult degree to obtain, that PhD physicists have valuable core skills and training (conceptual thinking, breadth of knowledge, attention to detail, problem solving, modelling, math, programming ...) applicable to a wide range of fields. So PhD physicists have an established track record of successfully switching fields.

I don't know whether PhD astronomers have a similar track record and what the general perception of a PhD in astronomy is. But if you are deciding between a PhD in astronomy and a PhD in physics, and planning to switch to an alternative career such as finance after your PhD, you should find out. Note that even if you have the same training whether your degree is in astronomy or physics, the perception that a hiring manager outside the field has plays a key role in whether you get your foot in the door or not.

 

[Physics2468]

The industry is very varied and nuanced whereas your question is very general. When one speaks about quants nowadays on the internet it often means "buy side statisticians" where it traditionally used to mean "sell side mathematicians". Both roles still do exist and are often fit for different kind of people (or for people with different backgrounds). Buy side as a rule is more geared towards machine learning and so on, whereas sell side is stochastic calculus and all that jazz.

Because the curriculum has become rather standardized: you can get a degree in machine learning or mathematical finance nowadays, the need and demand for the "all-versatile" PhD has too decreased. I think it will be very difficult to get a job without specifically training for the interview because the competition will, even though they may "only" have a Master's, have a degree directed to the industry's needs and have an idea what the job is about (often through internships). Just to clarify: What I am saying is not at all that it'd be impossible to get a job, but rather that it's unlikely to come by for free, without investing quite a bit of time to preparing for interviews, and so even though just a backup plan for you, you'd probably want to prepare at least for a couple of months (part time, while working on your PhD) and depending where you're applying, consider first going for internships rather than full time positions.

Ok! I am comfortable with taking some time to prepare, I will even try to build up more savings during my PhD in case I need a few months even after my defense. In fact, I think I would be less comfortable with going in knowing nothing about finance or specific maths involved that I don't know.
I understand that there will be non-PhDs with specific training; I want to learn about astronomy/astrophysics and would prefer to work in academia, and if I wanted to do finance I would follow them. I just wanted to check up with the industry. That being said, are there any other positions that maximize the value of the "all-versatile" PhD?
Thank you so, so much for your reply!

 

[Physics2468]

I don't have a direct answer to your question. But here's an issue for you to dig into further. I'm a PhD physicist. In the course of my career, in response to layoffs and industry meltdowns, I reinvented myself many times. I know other PhD physicists who have done the same ... including some who went into finance. APS News has frequent columns highlighting physicists working in alternative careers. There is a general perception that a PhD in physics is a difficult degree to obtain, that PhD physicists have valuable core skills and training (conceptual thinking, breadth of knowledge, attention to detail, problem solving, modelling, math, programming ...) applicable to a wide range of fields. So PhD physicists have an established track record of successfully switching fields.

I don't know whether PhD astronomers have a similar track record and what the general perception of a PhD in astronomy is. But if you are deciding between a PhD in astronomy and a PhD in physics, and planning to switch to an alternative career such as finance after your PhD, you should find out. Note that even if you have the same training whether your degree is in astronomy or physics, the perception that a hiring manager outside the field has plays a key role in whether you get your foot in the door or not.

I agree; this is my primary concern. Would it be okay for me to list my PhD as "Computational Astrophysics" on resumes, even if my degree is from an Astronomy Dept? If so, would that look better? I appreciate your help!

 

[Päällikkö]

Ok! I am comfortable with taking some time to prepare, I will even try to build up more savings during my PhD in case I need a few months even after my defense. In fact, I think I would be less comfortable with going in knowing nothing about finance or specific maths involved that I don't know.
I understand that there will be non-PhDs with specific training; I want to learn about astronomy/astrophysics and would prefer to work in academia, and if I wanted to do finance I would follow them. I just wanted to check up with the industry. That being said, are there any other positions that maximize the value of the "all-versatile" PhD?
Thank you so, so much for your reply!

Data science and machine learning are hot right now, though similarly to mathematical finance they are getting "saturated" by people with specialist degrees: Where some firms in their job ads used to write requirements down as "Master's/PhD in a quantitative field", which would have been the golden age for a PhD physicists to enter in some sense, they more and more often specifically want people with previous experience/degree in the techniques (and obviously the interview, too, will reflect this). It's not just the cool stuff of AlphaGo and self driving cars, but marketing companies, agriculture, finance and all kinds of firms have realized the need and the utility of collecting numbers and even just doing simple things like linear regression to guide them in asking the right questions.

Oil & Gas used to hire quite a few computational physicists, though the industry right now may be going through a slump (that said, to an extent at least it is just part of the normal business cycle).

There is also weather forecasting: big models, big data, big compute.

Management consulting is also a place where many ambitious physics PhDs have often found themselves, and in theory is in some sense the ultimate job for the generalist (though in practice a lot of it will be making powerpoints look nice and delivering solid pitches).

Those are some off the top of my head.

As for your CV, you can, and in fact should, list whatever you feel most appropriate.

 

[CrysPhys]

I agree; this is my primary concern. Would it be okay for me to list my PhD as "Computational Astrophysics" on resumes, even if my degree is from an Astronomy Dept? If so, would that look better? I appreciate your help!

As for your CV, you can, and in fact should, list whatever you feel most appropriate.

You need to be truthful; otherwise guaranteed to come back and bite you when the potential employer does a background check. In your resume, you put down the actual degree listed (or to be listed) on your transcript or diploma. If you got a PhD in Astronomy, you put down PhD in Astronomy. If you did your dissertation in Computational Astrophysics, in addition, put down concentration, specialization, or dissertation in Computational Astrophysics. But that may not get you past the HR filters if they specify PhD in Physics.

If you want to play it safe, get a PhD in Physics. If not, check carefully what doors a PhD in Astronomy will open.

 

[BWV]

ISTM the big wave of physicists into Wall Street in the 1990s and 2000s were employed in modelling derivatives, particularly exotic ones. Facility in stochastic calculus being the key skill. Everyone now has reasonable models for these instruments and there is little interest in pushing the envelope with new derivative models after the Financial Crisis and Dodd Frank. As Päällikkö said the emphasis today is on machine learning and big data. I am in finance on the buy side, just a lowly MBA, but from where I sit, I see the demand for people with advanced computer / data science degrees being the key area of demand. This is the area hedge funds are pushing into.

 

[Päällikkö]

ISTM the big wave of physicists into Wall Street in the 1990s and 2000s were employed in modelling derivatives, particularly exotic ones. Facility in stochastic calculus being the key skill. Everyone now has reasonable models for these instruments and there is little interest in pushing the envelope with new derivative models after the Financial Crisis and Dodd Frank. As Päällikkö said the emphasis today is on machine learning and big data. I am in finance on the buy side, just a lowly MBA, but from where I sit, I see the demand for people with advanced computer / data science degrees being the key area of demand. This is the area hedge funds are pushing into.

While what you say about modelling exotic derivatives is true, many other things on the sell side have changed in the past decade: the demand for quants is increasing like never before, and it is not being primarily driven by machine learning. Instead, much of it is regulatory pressure, for example banks employ teams to oppose the front office quants and to challenge their models (called model validation or something similar) and these teams can be very large and growing. This means that the front office teams, too, must grow to be able to answer the many regulatory and in-house queries.

There's also the relatively recent rise of market risk and XVA, which is obviously a huge topic (and FVA in particular is still hotly debated in academia) and an area where real big losses can be made if done wrong. The financial crisis, rather than stifling model development, actually caused fundamental changes to models, in particular forcing everyone to use multicurve bootstrapping. While all sell side institutions will surely have implemented models and operational procedures by now, put everything together and you could say that nobody still knows how to discount future cashflows.

Finally, many of the (even vanilla) derivatives have been priced with models that were known to be wrong (well, of course all models are always wrong, but some more fundamentally than others) and it is only recently that more proper solutions been suggested by both academics and practitioners. For buy side it may not matter as much (as they may be doing pure speculation), but for a market maker hedging with the wrong instruments does mean leaking profits and/or being hit with unnecessarily large valuation reserves.

 

[Physics2468]

Data science and machine learning are hot right now, though similarly to mathematical finance they are getting "saturated" by people with specialist degrees: Where some firms in their job ads used to write requirements down as "Master's/PhD in a quantitative field", which would have been the golden age for a PhD physicists to enter in some sense, they more and more often specifically want people with previous experience/degree in the techniques (and obviously the interview, too, will reflect this). It's not just the cool stuff of AlphaGo and self driving cars, but marketing companies, agriculture, finance and all kinds of firms have realized the need and the utility of collecting numbers and even just doing simple things like linear regression to guide them in asking the right questions.

Oil & Gas used to hire quite a few computational physicists, though the industry right now may be going through a slump (that said, to an extent at least it is just part of the normal business cycle).

There is also weather forecasting: big models, big data, big compute.

Management consulting is also a place where many ambitious physics PhDs have often found themselves, and in theory is in some sense the ultimate job for the generalist (though in practice a lot of it will be making powerpoints look nice and delivering solid pitches).

Those are some off the top of my head.

As for your CV, you can, and in fact should, list whatever you feel most appropriate.

Thank you SO much! I will do some more research about important skills etc. How important is your grad school reputation for these types of jobs and quants?

 

[Päällikkö]

Thank you SO much! I will do some more research about important skills etc. How important is your grad school reputation for these types of jobs and quants?

It's only important in passing the first filter; if you get as far as talking to a person, it no longer matters. Different companies have different hiring guidelines and get different pools of applicants, so any general statement would probably be wrong. I feel it's more important that your CV plays to your strengths and also shows that you understand what the job is about by listing concrete skills. Search for people from your school on LinkedIn and see where they've ended up at. Keep in mind though that what happened say more than 10 years ago will be irrelevant in today's job market. You can also ask recruiters, and still being in university, go see job fairs and ask the representatives.

 

[Ericpoke]

Also, if I get an Astronomy Ph.D. rather than Physics, will it be much harder to find a position as a quant? I already know C/C++ and some parallel programming.

These jobs are rare but not impossible to find. I believe what helped me was years on going articles that got published in several local a couple international magazines. However, unlike myself, your backup plan looks more enticing. Tons of solutions are required in the programming field today.