One less lawyer in the world, one more scientist

[twofish-quant]

Given the choice between spending my lots of my own money to own just one ferrari, or being paid a salary to play with a whole factory full of them - no contest

Some of the algorithms I write handle tens of billions of dollars each evening. It's more scary than fun because in the back of your mind you are always worried about getting something wrong and screwing up the world economy. One reason grunt work is relaxing is that while you are coding something, you stop consciously thinking about the possibility that you could screw up (or that you at that moment could be screwing up) the world economy,

However, one thing I really, really like about my job is that you get to ask some deep questions. You don't always get to ask deep questions *out loud*, but you get to ask some deep questions. What is space? What is time? What is gravity? What is money? The concept of money and the mathematical rules for how money works confuses me more than quantum mechanics. Some things about QM is that 1) the rules don't change from moment to moment or year to year 2) there are some basic fundamental principles that you can use as a foundation to figure out what is going on and 3) there are some basic fundamental logic rules that you can use to get from known statements to unknown ones.

None of that is true with money or for that matter for law, either.

 

[mal4mac]

Mihaly Csikszentmihalyi... interesting person

Yes. I'd recommend reading his book "Flow", my favourite "philosophy of life"/psychology book.

 

[mal4mac]

Some of the algorithms I write handle tens of billions of dollars each evening. It's more scary than fun because in the back of your mind you are always worried about getting something wrong and screwing up the world economy.

What's even more scary is that you might do it "right" and still screw the world economy!

It wasn't bad programming that caused the sub-prime problem.

Most of my programming involved creating educational software - that wasn't very scary. If the package crashes, just go back to chalk & talk, no real harm done...

 

[twofish-quant]

What's even more scary is that you might do it "right" and still screw the world economy!

By definition if you do it right, the world economy is fine.

It wasn't bad programming that caused the sub-prime problem.

It was bad mathematical modelling. One thing about working as a "financial scientist" is that you aren't a mere coder. You are responsible for developing mathematical models, and your responsibilities include non-technical aspects including thinking about the broader implications of what you are doing (which is why banks hire people that think about the universe in the first place).

There are a ton of jobs in finance which are "shut up and code what we tell you to". But those jobs are extremely unappealing to Ph.D.'s.

This incidentally means learning a fair amount of securities law, since the details of the law in fact can change the mathematical model in lots of ways (i.e. when modelling North Elbonian bonds we have to include this term to take into account the fact that North Elbonian handles default in a different way than South Elbonian,)

Now if you did spot a problem and you raise the issue and are told to go to hell, then at that point you could have resigned since you've done everything you could (i.e. theoretically you could go to the regulators but before the crash they would have ignored you) and then when it falls apart, you have clean hands. I know of people that were in that situation pre-crash. Fortunately, the situation is different now, and one thing that the regulators did was to create mechanisms so that you can raise critical issues before the world blows up. But that means that "they wouldn't listen to me" is no longer an excuse.

On the other hand, if you push the panic button for everything then that doesn't work either. If you push the panic button for everything then that can kill new modelling which can also blow up the world because it means that you are running models which are obsolete and wrong because you can't get the new ones in.

I suppose the one good thing about financial programming is that "it's only money" you can bankrupt someone and destroy the world economy, but you won't directly kill anyone. Now the people that I know that design hydrogen bombs on the other hand...

Most of my programming involved creating educational software - that wasn't very scary. If the package crashes, just go back to chalk & talk, no real harm done...

A lot of my programming ended up being "emergency room" programming. Production is down or failing, and it's like bringing in a patient into the ER. It's really awful, and a lot of programming involves making life easy for yourself when the system gets into the emergency room. A lot of it involves managing emotions, both your own and others. You want to focus at stabilizing the system, and calm down the half a dozen people that are frantically screaming at you so that you can get the information that helps you resolve the issue.

Getting back to the original poster. Before you bite into the apple you really need to know what you are getting into. The thing about scientists is that you are dealing with the elemental forces of the universe, and those can be quite dangerous. There is the initial thrill of discovery, but that passes very quickly, and a lot of what happens next depends on what's left afterwards.

 

[H2Bro]

Thank you all for your input. Some of what I am saying may come off as naive, or rubbish, but there are worse reasons for making a career decision than following what I find most appealing. If I end up regretting it I will say to myself "My god those physics forum lads knew it all along" and wallow in grief in my rental apartment and 15 year old Honda.

But, who knows, maybe I will find some satisfaction in it. Only one way to see.