Career direction for Ivy League math major

[robbentheking]

Hey guys,

I'm interested in a career in quantitative finance, but I just don't know where to start.

I'm technically a senior at an Ivy League university where I studied math and german. I say technically because I decided to basically spend my last semester abroad which for various reasons doesn't allow me graduate until this december although im done my course load. I have a very good GPA (3.91), and care about math a lot (like the process, the formality etc.). I did some work for a small Asset Management firm last summer, but didn't particularly like it and have decided to stay in germany until december and intern at an electronics firm just because the position was easy to find, seemed like a good way to improve my mathematical maturity, and allowed me to get closer to fluency in german, which is something i also care about.

Anyway, I'm looking for direction post december. I've had an affinity for financial concepts for the last 1.5 years, but really haven't found my thing. Most of what I've done is just independent reading. I can't stand what i perceive as a lack of rigor in trading (I've read that much of the academic consensus is that short term price movements are essentially independent of previous movements, which makes charting, technical analysis etc. bullshit), equity research seems more rigorous but has no interesting math, and M&A just doesn't appeal to me at all. I'm definitely more attracted to the quant side of things, I don't know, like derivatives structuring or something. But here the problems really begin.

My programming experience amounts to matlab and a tiny bit of python, and I feel like many quant positions really are for PhDs. I don't intend to get a PhD to go into finance. I think that's bullshit and probably torturous. Plus, I really don't think I'm that good at math. I like/am good at probability, but although I really respect pure math, I don't think I could even pull of a decent PhD dissertation. I've been exposed to some very smart people at my school and don't think im on that level. I was thinking about maybe a masters in financial mathematics/ MFE just to give some legitimacy to the reading I've already done, but I've read that a lot of these programs aren't worth it and really don't get you to being a quant anyway.

So I guess my question is twofold. One, what industry should someone like me be looking into and two, what sort of steps should I take to get there?

Sorry for the long question.

 

[mhy]

I can't stand what i perceive as a lack of rigor in trading (I've read that much of the academic consensus is that short term price movements are essentially independent of previous movements, which makes charting, technical analysis etc. bullshit),

Many quants will tell you that technical analysis is 100% BS.

This is a quant forum so everyone's going to tell you to at least look into quant finance. Some MFE programs are truly outstanding and may help you get where you want to be. But you need to commit to finance first.

 

[tfors]

Side tracking here, i hope you don't mind, but finding this an interesting topic: "I can't stand what i perceive as a lack of rigor in trading (I've read that much of the academic consensus is that short term price movements are essentially independent of previous movements, which makes charting, technical analysis etc. bullshit)". Maybe academic consensus lacks rigor (or lacks touch with reality)? Does this look like independence to you http://revolution-computing.typepad.com/.a/6a010534b1db25970b014e8bb95c67970d-pi ? ps. would academics say these are outliers? well this (and other) kind of sh*t happens all the time, financial markets are very messy and almost impossible to model accurately. that's why warren buffet, Dalio, .. outperform 99.9% of math phd's in the markets (when working with large AUM). Even Renaissance Mgmt only do well with their smaller closed insider fund.

Charting and TA weren't serious contenders to start with anyway, no serious traders (that I know of ) work with that.

In response to your question, instead of a PhD you could indeed look into a master's to get a better idea of what you like, doesn't necessarily have to be a MFE or financial mathematics. (statistics, data science, computer science, .. all could work).

 

[tfors]

continueuing my previous post: at my university they also teach this advanced monte carlo stuff assuming independence running on expensive super computers and the lack of rigor here is that it gives false confidence, results are claimed to be approximately accurate while it is anything (it's better than nothing) but. When sh*t hits the fan, as in 2007, you can throw all these results and super expensive investments in the garbage can.

 

[robbentheking]

Side tracking here, i hope you don't mind, but finding this an interesting topic: "I can't stand what i perceive as a lack of rigor in trading (I've read that much of the academic consensus is that short term price movements are essentially independent of previous movements, which makes charting, technical analysis etc. bullshit)". Maybe academic consensus lacks rigor (or lacks touch with reality)? Does this look like independence to you http://revolution-computing.typepad.com/.a/6a010534b1db25970b014e8bb95c67970d-pi ? ps. would academics say these are outliers?

First, thanks for the reply. I appreciate it.

To answer your question, I haven't read the literature in a while, but the idea is that the short term movements of prices can be understood to a degree of accuracy that can all but assumed to be perfect by a string of independent random variables, right? You can never look at a chart and say "oh, that's independent" or "oh, that's dependent". The dependence of random variables is a tricky thing that can't be determined post hoc by looking at the result of a single experiment.

 

[tfors]

First, thanks for the reply. I appreciate it.

To answer your question, I haven't read the literature in a while, but the idea is that the short term movements of prices can be understood to a degree of accuracy that can all but assumed to be perfect by a string of independent random variables, right? You can never look at a chart and say "oh, that's independent" or "oh, that's dependent". The dependence of random variables is a tricky thing that can't be determined post hoc by looking at the result of a single experiment.

Sure, but if you assume independence and build models on that assumption, those models are not going to be much worth in reality if the underlying assumptions are incorrect. You have to decide beforehand, and it's empirically just incorrect to assume independence. Disclaimer: I am not a quantitative trader (yet), but a master's student (with skin in the markets), so i in a way i am also speculating here

 

[robbentheking]

Sure, but if you assume independence and build models on that assumption, those models are not going to be much worth in reality if the underlying assumptions are incorrect. You have to decide beforehand, and it's empirically just incorrect to assume independence. Disclaimer: I am not a quantitative trader (yet), but a master's student (with skin in the markets), so i in a way i am also speculating here

Of course you can't assume independence for no reason, but the actual research points strongly towards it. There's no strictly empirically correct approach. I see nothing wrong in building models assuming short term independence as long as you give the appropriate disclaimers

 

[BeepImaJeep]

I did math and finance at an ivy league for my undergrad. I would say the strongest applicants in my class with an affinity towards math and interest in finance largely went into trading. The best of those applicants went directly into quantitative funds or private equity.

I don't think you have a good understanding of what trading is. Only the least reputable firms on the street employ technical strategies. This isn't really because it's BS per say, it's just the edge that once existed in these strategies have long since been marginalized.

 

[robbentheking]

I don't think you have a good understanding of what trading is. Only the least reputable firms on the street employ technical strategies. This isn't really because it's BS per say, it's just the edge that once existed in these strategies have long since been marginalized.

I would never claim to have a good understanding of trading, but I fail to see how any technical strategy could have any solid theoretical basis.

Obviously so much is proprietary, but what would be your candidate for most mathematically rigorous class of strategies employed today?

 

[BeepImaJeep]

I would guess that the reason that technical strategies used to work is rooted in behavioral finance and inefficiencies in order execution. However, the edge in any such strategies is generally engulfed by high-frequency algo strategies today.

For the most rigorous strategy, I can't really give you a definitive answer. My background is in statistical arbitrage and equity options trading. The truth is neither really requires you to have a PHD or to prove anything rigorously (and I suspect there isn't really anything that does), even though most people in stat arb do have PHDs and that generally how it's portrayed. If you are interested in finding "the truth" or working with rigorous math, then perhaps you'd be better suited in academia. Sorry.

 

[robbentheking]

For the most rigorous strategy, I can't really give you a definitive answer. My background is in statistical arbitrage and equity options trading. The truth is neither really requires you to have a PHD or to prove anything rigorously (and I suspect there isn't really anything that does), even though most people in stat arb do have PHDs and that generally how it's portrayed. If you are interested in finding "the truth" or working with rigorous math, then perhaps you'd be better suited in academia. Sorry.

No, thank you. Obviously proofs aren't going to play a roll. I would just feel really weird basing a career off something I don't have a deeper understanding of, if that makes sense

 

[tfors]

I guess deeper understanding of financial markets points you to complex systems theory or the like, more econophysics than math.

 

[BeepImaJeep]

Edit.

 

[robbentheking]

I'll look into those for sure. Any with a particularly good reputation come to mind?

 

[BeepImaJeep]

The ones that recruit at your school which have the most people attend their information session.

 

[tfors]

Statistical finance - Wikipedia, the free encyclopedia just for reference, this theory assumes no independence. (not many firms use it though)

 

[tjk]

equity research seems more rigorous but has no interesting math

I don't know that this is necessarily true. Quant equities generally focuses on factor modeling, so it can usually include some fairly advanced multivariate statistics-- PCA, discriminant analysis, some machine learning, etc. The problem to watch out for, though (coming from someone currently doing it), is that as an analyst in quant equities, they'll generally have PhDs doing most of the heavy math and your job will generally just be to write the code to run the backtests (although you can still be a fly on the wall for the research, read the relevant academic papers, learn from the researchers, and then when one of them quits, if you've been impressing everyone with your understanding of what they're doing you can make a pitch to move into a more research-heavy job).

If what you consider "interesting math" is what's generally included in an MFE curriculum, though (PDE, stochastics, monte carlo etc), you won't see it in quant equities.