It was invaluable having a mentor to talk through all the options.
I did my undergrad work at Caltech with a major in mathematics, and then went to UCLA to study theoretical physics. My dissertation was on applications of string theory to questions in black hole physics and hydrodynamics.
The most useful part of being fluent in mathematics is the ability to extract the simplest and most useful piece of information from what you are doing. Once you have small chunks that you can understand, it’s much easier to find effective solutions to difficult problems.
My favorite interview question was, "When you heat a sausage in the microwave, the tear is always lengthwise. Why is that?" I liked this problem because the real challenge was figuring out what equations to even begin working with. Once you got that part figured out, the math was actually pretty easy, but sitting there and looking at the blank page and trying to figure out how you were going to get the math to start solving the problem, that was the real challenge. That’s the challenge here at GS too. There are a lot of smart people that are great at solving equations once the problems are set up, but here the challenge is always in posing the right questions.
As someone with very little computer science experience, the question, "How could you set up a recursive function so that a really smart language / compiler could evaluate the function and never run out of memory?" was an interesting one. I certainly didn’t list theoretical computer science on my resume, but I guess they liked my answer.
There are two things I remember from my early days here. The first was the leadership and support I got from my supervisor and other team members. They were the ones that brought me up to speed quickly.
The other was the opportunity to make a contribution almost immediately. There was an expectation that every team member was there to deliver.
In order to execute transactions, we need to have confidence in our prices. Whether it’s in electronic markets quoting liquid products, to illiquidly traded derivatives, our models enable us to get the best possible prices for our clients.
In risk management, one of the key problems is correlation, and you can’t just say two products are correlated and expect that to be parameterized by just one number. Correlation arises from any number of different mechanisms in price dynamics, and I found working with and creating different models to capture these dynamics to be particularly interesting.
There have been two people who have influenced my path here. One was a colleague who was visiting from Bangalore when I first joined the team. He brought me up to speed, but more importantly, he showed me how flat and interconnected the firm was. While he was an analyst by title he was essentially given responsibility for the whole team involved in my first project.
The second would be my first manager, who was a clear culture carrier for the firm. He would always work hard to get everyone under him as much visibility as he could, but would also be as hands-on as needed to get a project done.
I have an informal mentor. I switched roles recently, and while I was figuring out where I wanted to move, it was invaluable having a mentor to talk through all the options. You need someone a little removed from your day-to-day work to talk to, and it helps when they’ve been around longer and have seen more of how the firm operates.
I’m in a great place with lots of opportunity. Electronic trading is becoming a bigger and bigger part of our business every year, and here in interest rates, we are right in the middle of that transition.