PhD in physics, first job as a quant

[Daniel Duffy]

I see. I put a bad wording by saying I can write anything. I definitely cannot. In fact I do not know if I know enough for a specific position or not. I meant I can write a program to find noise in a 1-milion line code, or I can sole a partial differential equation (or some simulations). I am definitely not an advanced programmer.

@ShowMeTheLight : I understand that you are frustrated that physicists who say things like that. I did not want to look like a very arrogant person who think he knows anything. I am pretty sure that most of the things that I learned in school is not going to be useful in finance.

@Daniel Duffy : I did not mean I am pretty advanced in C++. I will definitely be more careful in future in my statements. In fact, that is exactly why I want to go to industry, so I can get a better understanding of how to deal with a real problem.

@TraderJoe : thank you for your suggestions. I will read them before going to an interview.

In academia you have a solution to all problems. If it doesn't work you modify the problem to fit the solution.

Industry ... just solve it.


This can be a major shock initially.

 

[Xiangyu Gu]

Oh Feynman Jr. There are so many Physics phds in industry just doing average (aka sh*tty jobs). Please bare that in mind. Even if you are indeed better than most (outsmart them maybe?), there's still no guarantee you can do better at work. Let me put it this way, come up with some trading strategies and show it really works. If you can do that, everyone in the industry will agree you are good and offer you jobs.

No intention of discouragement here, but you might wanna lower your expectation a little bit.

 

[Daniel Duffy]

To be honest, I am trying to find the overlap between physics and finance post 2009. Markets and (software) technology have moved on.

My hunch is statistics, applied/numerical maths and programming if you want to have a shout.

my 2 Euro cents.

 

[rnavarro]

@feynmanjr There is no harm to take a postdoc to evaluate yourself deeply
whether academia or industry is for you.

Otherwise, an MFE route might be an alternative.

A PhD in Physics is no longer enough.
You need to differentiate yourself further why they would like to hire you.

If you subscribe to LinkedIn Premium, you will know how tough it is to break into
Quant Finance Industry. There might be somewhere in the range of 100 or so applicants.

A rule of thumb for Quant roles, there are around 20%-40% of the applicants have PhDs.
Let say, there are 100 applicants and 30 have PhDs ... you will be competing with 30 of them plus some exceptionally talented MS/MFE candidates.

Besides, with the high turnover in Quant finance industry playing tug-of-war with talent, you will be also competing with Quants with experience.

How many they will hire? 1 ... 2?

my 2 cents of advice ...

 

[feynmanjr]

I totally agree with all these comments. I already know that I may not be good in industry, but it is hard to solve an industry problem while you are in academia (specially theoretical physics). I think my advantage compare to most people is that I like to learn and I am willing to do so (I do not say that others do not have it though).

@Daniel Duffy : I really appreciate your advice. In academia (specially if you want to apply for a grant), you also need to solve the problem that is needed (you cannot modify the problem). However, I guess you do not get credit in industry, if you cannot find a solution for a given problem. Also, I am trying to be (relatively) good in all three subjects (math, programming, and statistics in that order), but I am not good in either of these compare to someone who only studied one of these.

@Xiangyu Gu : I say that it is pretty hard to come up with a strategy for trading, since all I did in my life was in theoretical physics. The point is that I need to expose to such an environment for a while to be able to come up with new strategies. I can try new things, but they may sound very unrealistic for someone who is already working in that area. I guess you mean that the best way to get a chance in industry is to learn it before leaving academia. What should be my expectations? I already know that there is a good chance that I cannot get a job in next few months. I am in this forum to find what I am facing (I have no expectation).

@rnavarro : I already know that a physics PhD is not enough, and as of most comments that I got from here, it is almost impossible to go to industry from a physics background. I am going to apply anyway, but I think I should stick to academia for a while (a postdoc or two), and try to learn more about this area. BTW, the 30 to 1 job is very fair compare to 500+ applicants for a tenure position in a normal university in US (for a position in my university there are 30+ on-site interview). Note that they are apply for each position that opens, and I do not expect to land a dream job as my first out-of academia position.

 

[Xiangyu Gu]

How about dab the water of MFE programs? Not sure if you are ready for another period of time in school though.

I do believe you could get somewhere without further education (like bloomberg if you try?) but it doesn't sound like what you want to do. I would say at least give yourself a couple of months to study before trying hardcore companies, and MFE is the shortcut, offering both knowledge boosting and degree that backs you up.

 

[feynmanjr]

@Xiangyu Gu : I would not mind getting another degree, but I am in US in visa and I can work for 30 months after PhD. Getting another degree will use it. Also, I do not think I will learn many new things in MFE, but the degree might help.

I would like to get a job in a place like Bloomberg (or somewhere similar), specially for my first job. Also, I am hopeful that I can learn more in a real job than getting another degree (I may be wrong though).

 

[MartinOvie]

@feynmanjr : I'm in a similar position. I will be graduating with a PhD in Mathematics in two weeks and I'm also pursuing a career in quantitative/computational finance. With only academic research in Math (numerical methods, computational fluid dynamics), Physics, Biochemistry under my belt, it is a bit difficult getting a quant role in the finance industry - especially as an international student. But I am very much determined to do so.

My experience so far: Matlab - which I used extensively for my dissertation - only goes so far. C++/Python are more appropriate, but require an attention to detail and creative thinking. Those books listed on the first page are a great help. I got them a couple weeks ago and plan to practice the coding problems every day during the summer. They are very good and contain a lot I didn't know. So even if you believe you are good enough, just get them and test yourself. At the very least, it might improve your coding skills if not your confidence level . As was previously mentioned, pick up at least one statistics/machine learning/big data stuff: R/SQL/SAS/Hadoop. I'm currently learning R now and if I've time, I will start learning SQL.

Getting a challenging quant role might probably take some time. So be prepared for that. If you need to get a 'short-term' (a year or less) gig in the mean time, do so. Applying to MFE programs is a good way to get into the industry, but with the deadlines past for the Fall 2016 session, the earliest time to get into a program is Fall 2017.

If you are interested, I would be happy to work with you this summer, on a computational finance project requiring C++ and/or R. Any tips on possible projects guys? @Daniel Duffy @Xiangyu Gu @APalley @rnavarro

 

[pingu]

If you are interested, I would be happy to work with you this summer, on a computational finance project requiring C++ and/or R. Any tips on possible projects guys?

Yeah, find a job. Don't waste time. Concentrate in job hunting.

 

[rnavarro]

Learn and Implment Algo Trading Strategies

Algorithmic Trading, Quantitative Trading, Trading Strategies, Backtesting and Implementation (their ebooks have Python/R implementaitons)
Quantopian

in between job interviews.

 

[pingu]

Learn and Implment Algo Trading Strategies

Algorithmic Trading, Quantitative Trading, Trading Strategies, Backtesting and Implementation (their ebooks have Python/R implementaitons)
Quantopian

in between job interviews.

why? he/she might not want to do algo trading. That's a very niche and narrow field.

 

[rnavarro]

The skills you will learn in Algo Trading is not just trading strategy nor programming.
It will open the doors of learning improtant Quant tools such as Time Series, Machine Learning as well as Statistical Modelling

 

[pingu]

The skills you will learn in Algo Trading is not just trading strategy nor programming.
It will open the doors of learning improtant Quant tools such as Time Series, Machine Learning as well as Statistical Modelling

you can learn all those things without learning algorithmic trading specifically

 

[Parallax]

you can learn all those things without learning algorithmic trading specifically

Exactly.

Comments like the previous one drive me insane. One of the biggest indicators, to me at least, of someone who has no idea how the industry works are incoherent discharges of massive amounts of "quant" terms when they either a) have minimal relatedness to the topic or b) are largely overlapping terms.

 

[pingu]

... One of the biggest indicators, to me at least, of someone who has no idea how the industry works are incoherent discharges of massive amounts of "quant" terms when they either a) have minimal relatedness to the topic or b) are largely overlapping terms.

There are plenty of those around here.

 

[MartinOvie]

So...focusing on the subject at hand . While job hunting, what would be the advised order for reading/practicing the material presented in the following unordered list of books?

1. Finite difference methods in Financial Engineering - Daniel Duffy
2. Intro to C++ for financial engineers - Daniel Duffy
3. C++ Design Patterns and Derivatives Pricing - Mark Joshi
4. Modeling Maximum Trading Profits with C++ - Valerii Salov
5. Cracking the coding interview - Gayle Laakmann
6. Quantitative Trading - Ernest P. Chan
7. 150 Most Frequently asked questions on Quant Interviews - Dan S., Rados R and Tai-Ho Wang
8. Elements of Programming Interviews - Adnan Aziz, T-H Lee, A. Prakash
9. Practical Guide to Quantitative Finance Interviews - Xinfeng Zhou
10. Cracking the Coding Interview: 150 Programming Questions and Solutions by Gayle Laakmann McDowell
11. Stochastic Calculus for Finance I - Steven Shreve
12. Learning Boost C++ Libraries - Arindam Mukherjee
13. A Primer for the Math of Financial Engr - Dan Stefania
14. Options, Futures and other derivatives - John Hull

I was thinking of reading 7, 9 every day and rotating between them bi-weekly, while reading the remaining books in this order

• 1, 13 - to cover basics in FDM and FE
• 2, 3 - to build solid C++ foundation
• 10, 5, 8 - to test that C++ foundation
• 14, 6 and then everything else.

 

[pingu]

I would avoid (1) and (2). Amazon reviews are awful.

I will concentrate in interview related books.

 

[MartinOvie]

I would avoid (1) and (2). Amazon reviews are awful.

I will concentrate in interview related books.

Yikes. Noted. I do believe the Quantnet/Baruch C++ Pre MFE is based on one or the other (I forget which). It serves as a good foundation in C++; I learned/am learning a lot. Only complaint would be that I'm currently at level 8 and haven't really done much finance. I know that changes at level 9 but I would have been happier if that change came about in...maybe level 4/5.

 

[pingu]

Yikes. Noted. I do believe the Quantnet/Baruch C++ Pre MFE is based on one or the other (I forget which). It serves as a good foundation in C++; I learned/am learning a lot. Only complaint would be that I'm currently at level 8 and haven't really done much finance. I know that changes at level 9 but I would have been happier if that change came about in...maybe level 4/5.

I tend to trust Amazon reviews overall (more if I'm going to be parting with $$$)