Preparing for a Career in Quantitative Finance: What to Do in 2 Years

[zpconn]

Me: an undergraduate math major, two years left, very well qualified, very disillusioned with the idea of an academic career, considering a switch to finance.

I'm wondering what would be involved in making this switch. This is all hypothetical, but if I were to do this what should I do in the next two years to best prepare for this sort of career? My mathematical background is at probably the second-year graduate level at a normal graduate school but my financial background is very limited--I have some knowledge about it but I haven't taken actual courses. So I'm asking in an ideal world what would be the absolute best way for me to spend the next two years if my goal is to go into quantitative finance? I have a good programming background so I wouldn't need to take any classes for that.

I understand that there may be no single answer to this question, depending on what area I want to go into, but I imagine that surely specialization like that wouldn't occur until being in the workplace or working on a master's degree/Ph.D. in the subject.

Internships are probably very important. I've already got an REU lined up for this summer, but what internships should I aim for the following summer? What are the coolest ones for stuff like this? I've heard good things about DE Shaw but don't know about any others really.

And finally, is there any decent introduction to this career? I've looked a good deal, actually, and all the information is incredibly fragmented with no clear target audience at all.

Thanks very much. I'm just trying to judge my options right now.

 

[armolinasf]

While I don't have any advice to give you, you might want to take a look at wallstreetoasis.com I'm sure you'll be able to find answers to anything finance related, even QF there.

 

[doodle_sack]

Academically, attending internships related to Finance will do a great deal of help with admissions. If you are aiming for PhD in Quantitative Finance/ Financial Engineering, then engage yourself in knowing about topics related to ODE/PDE/Probability/Stochastic Process/Analysis. See if you can do any research work related to that, anything at all will help.

For an introduction to this field, download the text files and see if it helps you.

If you need any information, don't hesitate to PM me.

 

[twofish-quant]

Me: an undergraduate math major, two years left, very well qualified, very disillusioned with the idea of an academic career, considering a switch to finance.

It depends. If you are disillusioned with the idea of an academic career, but you love academics, then it might be a good idea to just get a Ph.D.

I'm wondering what would be involved in making this switch. This is all hypothetical, but if I were to do this what should I do in the next two years to best prepare for this sort of career?

There are a dozen different careers in finance. If you want to do something similar to what I'm doing then either get a Ph.D. in math/physics/engineering/statistics or get a masters in computer science and work a few years for a software company like Google.

So I'm asking in an ideal world what would be the absolute best way for me to spend the next two years if my goal is to go into quantitative finance?

The good news is that there are about five different ways of getting in. A lot depends on what you want to do with your life. Personally, I'm a fan of the "get your Ph.D. in physics" route since that worked for me, but your preferences might be quite different.

One good question is answer the jobs "why do you want a job in quantitative finance?" For me, it's because it is "graduate school with money."

I understand that there may be no single answer to this question, depending on what area want to go into, but I imagine that surely specialization like that wouldn't occur until being in the workplace or working on a master's degree/Ph.D. in the subject.

Everyone is a specialist. You have to be really, really, really good at one or two things for someone to hire you. What that one or two things are are different.

And finally, is there any decent introduction to this career? I've looked a good deal, actually, and all the information is incredibly fragmented with no clear target audience at all.

Kuznetsov's "The complete guide to capital markets for quantitative professionals" is the best reference to the "so what do people do on Wall Street." However, you do need to realize that things get very, very quickly out of date. A lot of things have changed sense 2006, so the book is somewhat out of date, but not hopelessly so.

 

[twofish-quant]

One thing about Hull or Baxter&Rennie is that you have to realize that large parts of those books are pretty out of date.

 

[snipez90]

I don't know about Baxter&Rennie, but Hull is the standard text on derivatives pricing. Can you clarify the parts that you think are "pretty out of date" and why? Do you mean that for instance no one really uses the standard models of options pricing anymore? Well the answer seems to be yes, but it seems that you should still learn the basics to understand why models that were once considered useful are now considered failures.

 

[twofish-quant]

I don't know about Baxter&Rennie, but Hull is the standard text on derivatives pricing.

Yup. It's the standard. That doesn't mean that it isn't mostly wrong or irrelevant.

Hull is a very, very good textbook. It's an easy read, and it will teach you the language so that when someone tells you that gamma calculations from a local volatility model are wrong because they fail to account for counterparty risk, you know what they are saying.

I'd treat Hull, the same way I'd treat an astronomy textbook from 1975. Useful. But you just have to realize that a lot of the stuff is very, very out of date.

Can you clarify the parts that you think are "pretty out of date" and why?

It will be easier for me to start with the parts that *aren't* totally out of date. Probably about 70% of Hull includes models that no one uses, and contain assumptions that are wrong. Just to list problems with the interest rates models, they don't take into account counterparty risk, they don't tell you how to deal with correlations between interest rates and spot equity prices, no clue how to handle hybrid derivatives, short rate models are useless if the fed is holding short term interest rates near zero, log normal models seriously break if you have negative interest rates. That's just with one chapter...

The really good parts of Hull are the later chapters in which he talks about the history of derivatives and when things go wrong. Also, if you read and understand what a short rate model is, then when I tell you why it's totally wrong, you will understand what I'm saying.

Do you mean that for instance no one really uses the standard models of options pricing anymore?

No one uses the models in Hull straight out of the book.

Well the answer seems to be yes, but it seems that you should still learn the basics to understand why models that were once considered useful are now considered failures.

Sure. But that also explains why banks hire physicists and mathematicians rather than finance majors for some of the jobs. You can give a textbook to a finance major, and they may be able to do something useful with it, and that works for some jobs. However, if you tell someone that "this model is hopelessly broken for reasons A, B, and C, we want you to fix the problem" then most finance majors end up way, way over their heads.

The reason that Ph.D.'s in particular get hired is that Ph.D.'s know what to do when the textbook is wrong. What I like about quantitative finance is that after you've *finally* figured something out, the rules change. :-) :-) :-)

 

[zpconn]

Thanks very much for the replies and book recommendations.

Do fractals show up much in the field? My research project this summer has a lot to do with fractals, so if these show up in finance I could keep that in mind over the summer.

 

[ahsanxr]

For me, it's because it is "graduate school with money.

Could you please expand on this a bit.

 

[twofish-quant]

Do fractals show up much in the field?

Not really. The reason that they don't is that any sort of pattern in financial data, gets quickly destroyed. If you found out that there was a fractal pattern in stock prices, this would tell you when to buy and sell, and by buying low and selling high, you end up destroying any pattern that exists.