Career: Trader vs. Quant - Pros and Cons for Physics Graduates

[Patrick_Nth]

One other thing about masters degrees. If you have to choose between a masters in computer science and a computer science Ph.D., it's not obvious that the Ph.D. is going to be the better programmer, and curiously the odds are that the masters degree holder will program better. I've know a lot of people with Ph.D.'s in CS that are just totally incompetent programmers. Brilliant researchers and theorists. Totally incompetent programmers. I don't know anyone with a masters in CS that is an incompetent programmer. Some are better than others, but no one that is incompetent.

Would you be able to comment on how easy/difficult it would be to show some sort of competency in programming if one were applying with another technical Masters' degree (EE and Mech. E in particular)?

 

[twofish-quant]

Would you be able to comment on how easy/difficult it would be to show some sort of competency in programming if one were applying with another technical Masters' degree (EE and Mech. E in particular)?

No one really cares what you got your degree in at all. It's not that hard to see how good/bad a programmer is in an interview setting. It is a bit tricky in a resume to show that you have programming skills, but it can be done. Anyone can say "I know C++" but it's harder to say "I was a programmer on system X with Y lines of code and wrote Z."

 

[Clover_job]

One other thing about masters degrees. If you have to choose between a masters in computer science and a computer science Ph.D., it's not obvious that the Ph.D. is going to be the better programmer, and curiously the odds are that the masters degree holder will program better. I've know a lot of people with Ph.D.'s in CS that are just totally incompetent programmers. Brilliant researchers and theorists. Totally incompetent programmers. I don't know anyone with a masters in CS that is an incompetent programmer. Some are better than others, but no one that is incompetent.

In physics in the US, the masters degree is often seen as a consolation prize, but the situation is very different in CS, where the masters and the Ph.D. are just different degrees.

Also, I don't know any physics Ph.D.'s that end up as traders. I don't think that there is any rule that keeps physics Ph.D.'s from being traders, and I think the reason why there are so few Ph.D.-traders is the same reason there are few Ph.D.'s that are airplane pilots or movie producers. Having a physics Ph.D. simply does not help you be an airplane pilot, movie producer, or bond trader.

There is no offense to people in master in physics, however, based on everyone's reply, sounds like it's very difficult for master in physics to get a job, either as a quant or a trader, if my understanding is correct? BTW, what do you suggest a physics graduate student to do if he is not interested in physics at all and wants to be a trader eventually? Should he go like physics master -> apply for a trader job or physics Ph.D. -> apply for a trader job, or physics Ph.D. -> apply for a quant job (if trader is impossible right after phD)-> transfer to trader?

 

[twofish-quant]

There is no offense to people in master in physics, however, based on everyone's reply, sounds like it's very difficult for master in physics to get a job, either as a quant or a trader, if my understanding is correct?

It's difficult in general to get a job either as a quant or trader. If you don't have a Ph.D. in physics, then there is nothing to make it easier.

BTW, what do you suggest a physics graduate student to do if he is not interested in physics at all and wants to be a trader eventually?

Get an internship or something that will let you see what trading is really like. It's generally a seriously bad idea to make career decisions based on the movies or from brief exposure. If you hate it, you found that out early. If you love it, then you have some contacts in the business.

Should he go like physics master -> apply for a trader job or physics Ph.D. -> apply for a trader job, or physics Ph.D. -> apply for a quant job (if trader is impossible right after phD)-> transfer to trader?

If you hate physics, then it's a seriously bad idea to go for the Ph.D.

 

[xbomber88]

I have a question. I'm going into my last year of my undergraduate Physics degree and for a long time my goal was to become a tenured professor at a research university but lately I've been realizing how incredibly stressful and difficult that would be and how little my chance of success is. So now I've been thinking of alternative career paths and I've been thinking that working as a quant is something I might like to do. So now I'm trying to decide if I should still try to get a PhD in Physics even though I don't think I want to be a professor before or if I should end my physics education with my BS in physics. From what I understand it's not that difficult to get a job as a quant if you're a Physics PhD but is there a more direct route? Is it stupid to get a PhD in Physics knowing beforehand that I probably won't want to stay in Physics after I'm done? What type of Physics should I study in grad school if I think I might like to be a quant? Do quants tend to be theorists or experimentalists and do they come from certain fields of Physics more often than others such as high energy or condensed matter?

 

[Locrian]

From what I understand it's not that difficult to get a job as a quant if you're a Physics PhD but is there a more direct route?

I think the opposite has been stated pretty clearly in this thread by people who would know. I know lots of physics PhD's who would, with their current skill set, have 0 chance of getting a job as a quant - none of them have any experience with any programming languages (or any other skill useful in finance).

You need to change the way you're thinking. People in the private business world rarely care about your level of education - they care about whether you have skills that are valuable to them. Figure out what skills are necessary for the job you want to do (and that shouldn't be too hard, given this thread and the other 10,000 like it online) and get those skills.

 

[xbomber88]

I think the opposite has been stated pretty clearly in this thread by people who would know. I know lots of physics PhD's who would, with their current skill set, have 0 chance of getting a job as a quant - none of them have any experience with any programming languages (or any other skill useful in finance).

I may be wrong about this but I swear I've heard twofish-quant say in another thread that he doesn't know of anybody with a Physics PhD that wanted to get a job on wall street and was unable to do it and that people with PhD's in Physics have a much better chance of getting a job on wall street than people who studied finance. I don't expect to be able to do it without picking up any of the necessary skills in grad school such as computer programming but it's my understanding that if I do that and market myself well I should be able to get a job as a quant.

 

[twofish-quant]

So now I'm trying to decide if I should still try to get a PhD in Physics even though I don't think I want to be a professor before or if I should end my physics education with my BS in physics.

It's simple.

If you love physics, then do the Ph.D. If you don't love physics or are indifferent to it, don't do the Ph.D.

At some point you have to do things for the sake of doing them. Do the physics Ph.D. if you think it's worth doing the physics Ph.D. Doing a physics Ph.D. is great if you like doing physics.

From what I understand it's not that difficult to get a job as a quant if you're a Physics PhD but is there a more direct route?

If you want to do finance and are indifferent to physics, there are hundreds of easier ways of getting into finance than doing a physics Ph.D. However, if you are indifferent to physics, I don't see why you want to be a quant, since there are hundreds of easier ways of making money.

Also why do you want a more "direct route"? It's easier to get to the top of Everest with an airplane than climbing it, and it's easier to move 27 miles by driving a car than by running a marathon, but what's the point? To quote Kennedy, we do things not because they are easy but because they are hard.

Is it stupid to get a PhD in Physics knowing beforehand that I probably won't want to stay in Physics after I'm done?

If you love physics, then you spend six to eight years of your live doing physics which isn't stupid, and then you can spend the rest of your life doing physics-like things.

Also, I want to stay in physics. It's universities that I can't stand. That's something different.

What type of Physics should I study in grad school if I think I might like to be a quant?

Hard stuff with lots of math and maybe computers.

Something that needs emphasis, is that Wall Street banks hire physics Ph.D.'s because they are physics Ph.D.'s. Do whatever you need to do a quality dissertation and be a good physics Ph.D.

Don't try to study something that you think is "relevant" to finance, because you'll probably guess wrong, and no one really knows what will be relevant in five years. Do study something that you feel passionate about.

Do quants tend to be theorists or experimentalists and do they come from certain fields of Physics more often than others such as high energy or condensed matter?

Not really. People do tend to come from parts of physics that are heavily computational.

Let's suppose you study something like polymer physics. I have no clue what polymer physics has to do with finance, but you are the expert in polymer physics, you tell me.

 

[twofish-quant]

I know lots of physics PhD's who would, with their current skill set, have 0 chance of getting a job as a quant - none of them have any experience with any programming languages (or any other skill useful in finance).

On the other hand, it's more of an attitude thing. If you've mastered quantum field theory, then you have the intellectual skills to learn basic C++, write a resume, and learn to talk on the phone. Whether you have the desire and mental flexibility to do it is something else.

Figure out what skills are necessary for the job you want to do (and that shouldn't be too hard, given this thread and the other 10,000 like it online) and get those skills.

Also those skills may be something that you didn't think is that important. It's trivial for me to learn stochastic differential equations. Learning to talk on the telephone and realize that MBA's do something useful was quite a bit harder. Something that's *really* hard for Ph.D.'s to do is to shut up and take orders.

 

[xbomber88]

It's simple.

If you want to do finance and are indifferent to physics, there are hundreds of easier ways of getting into finance than doing a physics Ph.D. However, if you are indifferent to physics, I don't see why you want to be a quant, since there are hundreds of easier ways of making money.

I absolutely do love Physics it's just that I don't want to allow my life to become completely consumed by Physics like the lives of so many of the people I know who have been successful in Physics. I want to have a life outside of Physics and I'm not sure that you can do that and be successful in Physics at the same time unless you're a genius which I'm not. Also, money definitely is not my primary goal though I will admit that the money does have something to do with my interest in being a quant. Primarily what I'm looking for is a job that involves a lot of math and problem solving which isn't as stressful as trying to become a tenured professor of Physics and if I can make a lot of money while I'm at it then that's great too.

 

[twofish-quant]

I may be wrong about this but I swear I've heard twofish-quant say in another thread that he doesn't know of anybody with a Physics PhD that wanted to get a job on wall street and was unable to do it and that people with PhD's in Physics have a much better chance of getting a job on wall street than people who studied finance.

That's true. However, Lociran and I may know different people and different Ph.D.'s. Among the people I know personally, I don't know anyone that couldn't get a job on Wall Street if they wanted it.

However, I do know a lot of people that don't want it, once they see what it's really like.

I don't expect to be able to do it without picking up any of the necessary skills in grad school such as computer programming but it's my understanding that if I do that and market myself well I should be able to get a job as a quant.

There is a difference between "getting a job on wall street" and getting a job as a quant. There are a *lot* of different jobs on Wall Street. Some are clearly quants, some clearly aren't, and some maybe are and maybe aren't. If you are going on the market next year, then right now Wall Street is hiring lots of mathematically literate people to fill out the documentation that the regulators are requiring. That's not what people think of as "quants" but it pays the bills.

The jobs that most people associate with "quants" (i.e. derivatives pricing) has been stagnant for the last few years.

If you going on the market in three years, then everything will be different. Different good, or different bad I don't know.

 

[twofish-quant]

One other thing. Banks right now aren't hiring physics Ph.D.'s in large numbers as quants or traders. The big hiring right now is for a job that I don't think has a name yet.

 

[boseliquid]

You are going to be a lot, lot more marketable if you have some (even minimal) C++ ability. If you pick up a book on object-oriented programming and spend about a month reading it, you are going to be much more marketable.

The problem is that C++ is sometimes overkill for a simple numerical project, but in industrial programming, the monte carlo code is going to be a small part of a much larger system, and if you know even basic C++ you can interface your code with the production system. If you don't know *any* C++, then the company will have to hire someone else to interface your system.

The will do this, but then your monte carlo skills have to be outstanding. I do know of a few people that are that good, but they are professors in major research universities that have literally written books on the subject. If your mathematically skills are "merely good" then a little C++ knowledge will help you a lot.

The other thing is if people assume (usually correctly) that if you can handle C++, you can handle anything else. So if it turns out that your MC code has to work in a system that works in java, C#, MATLAB or perl, if you have C++ on the resume, then it's assumed that you can deal with those other languages.

Hi Twofish,

Thanks for your reply, which I truly appreciate. I still have one or two projects to complete before starting to write my thesis, so I still have some opportunity to do "real" C++ coding in my grad studies.

Thank you

 

[snowjoke]

Hi, I just registered on the site to say: thanks to twofish-quant for taking the time to write these very useful and frank replies.

I worked in management / IT consultancy in London for nearly 5 years after university and I didn't like it much, as it was totally non-stimulating intellectually, while oscillating between the very boring and the insanely stressful. So I'm looking at a career change into finance. I like the idea of being a fund manager in the long term. I've studied for (and passed) the CFA exams in my spare time, but I'm also thinking about applying to some master's courses in financial maths. This would clearly be the minumum for working as a quant, but my fear is that, without a bachelor's in maths or a PhD, I might end up being very mediocre in this field. I have no insecurities about my raw intellect; it's just that I will be surrounded by many people who have studied a lot more maths than me. Do you think the best mathematicians make the best quants? I might just be better off looking at non-quantitative analyst (equity research etc) roles, which might be interesting as well, although not in quite the same way. Any thoughts on this?

 

[Artus]

This post is awesome. Thank you guys for the information. I wrote a post about using math for Economics:

https://www.physicsforums.com/showthread.php?t=420301

And I really appreciate your opinion. From this and other threads I understand models in Finance are not that useful but I don't know if I'm correct about that.

 

[Onamor]

Thanks again to all repliers - all opinions, questions and discussion are valued.

Maybe this should be a new thread, but its slightly related so excuse me for sliding it in here.

It's quite a standard question, but I don't think you can ask it enough times:
Can you be a "good" trader?
Can you be right more than wrong? Are some strategies better than others? Why shouldn't I just flip a coin? Does it depend on how well you can trick or manipulate the other players in the game? Are people predictable?
This includes the speculators (Buffet, Soros etc), quants (whatever that means), the automated trading models or any other magician.

Reason I ask is this:
If people are ready to dedicate their careers to this banking thing, the answer better be yes. Because otherwise, surely this part of the finance sector is just a (very old) bubble.
Why would anyone want to study or work in a field where the strongest (or only) factor affecting success is luck?

Would be great to hear everyone's opinion.

 

[twofish-quant]

I've studied for (and passed) the CFA exams in my spare time, but I'm also thinking about applying to some master's courses in financial maths. This would clearly be the minumum for working as a quant,

It's more complicated than that. There are a lot of different parts of finance and each uses different types of math. It turns out that the math that people use for CFA (statistics and covariance) is just different from the math that is used in derivatives valuation (stochastic differential equations).

Also "quants" don't exist anymore. The name has been over used to the point it that's pretty much a meaningless label.

I have no insecurities about my raw intellect; it's just that I will be surrounded by many people who have studied a lot more maths than me. Do you think the best mathematicians make the best quants?

Don't know. I'm not sure what quants do.

I might just be better off looking at non-quantitative analyst (equity research etc) roles, which might be interesting as well, although not in quite the same way. Any thoughts on this?

Hard to say. A lot depends on what you really are interested in, who you work with, and a dozen other factors that can't easily be summarized.

 

[twofish-quant]

Are some strategies better than others? Why shouldn't I just flip a coin?

Maybe you should. However then flipping a coin to trade a large number of shares is quite non-trivial. If it turns out that making money depends on flipping a coin, then the challenge becomes flipping coins faster than the person across the street.

Maybe you should flip a coin. Maybe you shouldn't flip a coin. If you have a good trader that understands a market, then they'll tell you when it's random and when it's not.

If people are ready to dedicate their careers to this banking thing, the answer better be yes. Because otherwise, surely this part of the finance sector is just a (very old) bubble.

I don't follow the logic. Much of the markets are random, but characterizing how the type of randomness then becomes useful. Also there are dozens of jobs that don't involve trading.

And sometimes being a good trader means knowing when not to trade. If you know that things are random, then momentum and doubling strategies become very dangerous.

Why would anyone want to study or work in a field where the strongest (or only) factor affecting success is luck?

Because luck and randomness can be quantified and managed. Also, you end up further being a "failure" in finance than a "success" in academia.

 

[Onamor]

you end up further being a "failure" in finance than a "success" in academia.

Do you mean "richer" rather than "further" here? And this is my point; if "trading" is not an acquirable skill (I didn't say teachable), then why do people become traders? To be richer? Fun? Because it's the only (size-able) way to exchange assets/commodities etc in our current world society?
If this activity called trading is contrived and not really useful, surely one day (at least in a decent/honest/ideal society), it will cease to exist?

This was the logic behind the bubble analogy... But I suppose this is really a different question to whether or not traders can be "good." If no trader really knows what they are doing, there could still an argument for having traders - lack of a better system..?

Chances are more than likely that nobody knows whether or not the current market model is the "best" way (best meaning the most efficient and stable way to match supply and demand, and to ensure a free market? Question in itself). But for historical reasons (and possibly greed), we have the current state of affairs and procedures which are being studied in the mean-time.

If you have a good trader that understands a market, then they'll tell you when it's random and when it's not.

But if they truly understand, why aren't they always right? Personally I don't think you can be a better trader than someone else. Not unless you swindle someone directly, but this is a different topic. And even if you could manage risk better than someone else, you will still need to enter the market at some point to make money.
But I'm also not sure there is another possible way to have a market than what is current.

If it turns out that making money depends on flipping a coin, then the challenge becomes flipping coins faster than the person across the street.

I love this situation. Just because somewhere there could be a kid with a coin beating Goldman (Or a monkey and a typewriter...)
Anyways, I think if I don't aim on making more money than anyone else, just to make money, then it doesn't matter how often I flip my coin?

 

[snowjoke]

Onamor -

http://www.businessweek.com/magazine/content/03_29/b3842001_mz001.htm" to an interesting article I read recently about Steve Cohen, whose company makes millions of dollars per day from trading. It's a bit old and markets have probably changed, but it provides good insight into active trading.

Of course some traders are better than others, but it's not possible for everyone to be better than average.

 

[Onamor]

Hey, thanks for the article. Yes, these things are always interesting, and such success stories not even too uncommon. But I agree, it's very rare to get a decent look into what the hell it is that traders/hedge funds/banks do.

But I'd like to bring something up just for arguments sake,

Of course some traders are better than others

Why "of course"?!
Such stories like this are just a small (overly focused on) view of a much larger picture. For each Cohen, Soros and Buffet, I bet there are a million or so people who lost money. Or worse yet, made a lot of money, then lost that too. I bet there were times when each of these great and very wealthy speculators were very close to bankruptcy. If they had gone under, would you ever have heard of them?

I would argue that this is not the norm for an investor - to be "right" over and over again.

 

[snowjoke]

You're right that for every winner there is a loser - that's what I meant by saying that not everyone can be better than average - but it's not necessarily random who the winners are. Being a good trader can range from not making mistakes, to acting on information more quickly than others. Ask yourself why Goldman Sachs makes so much money from trading every year, while 80% of day traders sitting at home lose money.

 

[Onamor]

It's a good point, but are Goldman http://www.reuters.com/article/idUSTRE6781YC20100809"? Why not, if they know what they are doing? Scientific experiments are repeatable.
If you take the traders out of Goldman, sit them at home and have them trade like a day trader, I'm not sure they will do better. Else, why do they insist on trading from inside Goldman offices? Do they enjoy sharing their profits with the company?

I'm well out of my depth here, but I'd say that companies can have an edge from information and other external influences. Not a closed system I suppose.

 

[snowjoke]

Trading isn't a scientific experiment. Markets are largely stochastic rather than deterministic. You're sometimes right and you're sometimes wrong. You make money if you're right fractionally more often than you're wrong...

 

[Onamor]

Trading isn't a scientific experiment. Markets are largely stochastic rather than deterministic. You're sometimes right and you're sometimes wrong. You make money if you're right fractionally more often than you're wrong...

Completely agree. But are you saying there are strategies that are always right more often (even fractionally) than they are wrong? This is in essence the same thing, no?

 

[twofish-quant]

If this activity called trading is contrived and not really useful, surely one day (at least in a decent/honest/ideal society), it will cease to exist?

Let me tell you what a trader does.

I want to sell 100 shares of Exxon stock so I go to my computer and issue an order to sell 100 shares of Exxon stock. Now it is likely that somewhere in the world someone in the world wants to buy 100 shares of Exxon stock, but the odds are that they are not staring at the computer at the exact moment that I want to sell my stock. I could sit around at my computer screen for hours on end waiting for someone to buy my stock, but I have better things to do.

So you can pay someone to sit around for hours on end staring at a computer screen looking for someone wanting to sell 100 shares of Exxon stock and then wait for someone wanting to buy 100 shares of Exxon stock. That's a trader.

But wait... Now that we have computers, can you program a computer to automatically look for people wanting to sell stock and then match them with people that want to buy stock. Sure. That's algorithmic trading.

If no trader really knows what they are doing, there could still an argument for having traders - lack of a better system..?

I said that the price of stock is random, that's a different statement. If a trader has just bought 100 shares of Exxon stock, their job is to find someone that wants to buy 100 shares of Exxon stock as fast as possible so that they can make money off the difference. Traders do not buy and hold because the longer they hold on to stock, the higher the chances that "something bad" will happen. So someone that trades Exxon doesn't care whether Exxon rises or falls.

Now portfolio managers do, but that's a different job.

But if they truly understand, why aren't they always right? Personally I don't think you can be a better trader than someone else.

Part of true understanding involves knowing what you don't know. Traders in general try to get rid of their positions as quickly as possible. I buy Exxon stock, I sell Exxon stock. The reason that people want to do this is that if you have a position something bad could happen, and your job is to provide liquidity and not to gamble on the long term direction of Exxon stock. You don't care what Exxon is going to do in a month. If you are a human trader, you are buying and selling on the order of hours. If you are a machine trader, you can react in milliseconds.

I love this situation. Just because somewhere there could be a kid with a coin beating Goldman (Or a monkey and a typewriter...)

There almost certainly is. The thing about random processes is that someone somewhere will win the lottery. If you have enough monkeys typing random stuff in a typewriter, someone will make insane amounts of money. You can mathematically show this to be the case.

But the difference between luck and skill is that your luck will run out. If you have someone win a lottery, the chances that they will win the lottery again isn't high.

Anyways, I think if I don't aim on making more money than anyone else, just to make money, then it doesn't matter how often I flip my coin?

Yes. Transaction costs. Every time you flip the coin, you have to pay someone something.

 

[twofish-quant]

http://www.businessweek.com/magazine/content/03_29/b3842001_mz001.htm" to an interesting article I read recently about Steve Cohen, whose company makes millions of dollars per day from trading. It's a bit old and markets have probably changed, but it provides good insight into active trading.

It really doesn't. Part of the reason people have huge misconceptions about finance is because of articles like these. One problem is that it gives people extremely inaccurate ideas for what most people in Wall Street actually make. Yes, if you run your hedge fund, you can make huge amounts of money, but you can also make huge bucks running a technology start-up.

Cohen is a hedge fund manager. His company trades stocks, but he isn't a stock trader (i.e. a market maker).

 

[twofish-quant]

Hey, thanks for the article. Yes, these things are always interesting, and such success stories not even too uncommon. But I agree, it's very rare to get a decent look into what the hell it is that traders/hedge funds/banks do.

I wouldn't post as much as I do, if there were better sources of information. People get their information from the movies and from articles about mega-millionaires, but that gives you a *really* bad idea of what people in banks do, and that distorts decisions making.

For example, people think that everyone that works for an IB is a mega-millionaire, which makes people drop everything to work at an IB and forget about working at Google and Microsoft. If people got a better idea of what the real salary levels are like, then maybe Google doesn't look that bad.

Salaries in finance are "good". They aren't insane.

I would argue that this is not the norm for an investor - to be "right" over and over again.

Correct, and its a big, big warning sign if you are making too much money. One thing that you absolutely have to do is to know when you are investing and when you are gambling. Gambling can be OK, you just have to realize when you are going it, and make sure people whose money you are using are also OK with it.

Also, you have to think economically. I can imagine arranging the world economic system so that everyone in the world makes say $200K/year. I can't imagine setting things up so that everyone makes $20M/year.

So if someone offers a whole bunch of $200K jobs, that seems sustainable. If someone offers a whole bunch of $20M/year jobs, then I get suspicious since I don't see where the money is coming from.

 

[twofish-quant]

Ask yourself why Goldman Sachs makes so much money from trading every year, while 80% of day traders sitting at home lose money.

Because day traders have to go through GS in order to get to a counter-party.

Let's suppose you are a "day trader" (and I use the words in quotes since no one that I know of that day traders is actually trading). What do you do when you have to buy or sell a stock. You go through a broker.

***BZZZZTTTTT*** game over, you've just lost.

You make money from trading by *being* the broker. However being a broker requires lots of money. You have to have dozens of lawyers and then you have a lot of computers, and then you have to hire a ton of people to run the computers.

Which is where I come in...

Also trading as an economic activity generates value. It's a good thing that I can go online right now and sell 100 shares of Exxon and get cash for it.

 

[twofish-quant]

Markets are largely stochastic rather than deterministic. You're sometimes right and you're sometimes wrong. You make money if you're right fractionally more often than you're wrong...

That's one way to make money. There are others.