FinTech Female Fridays: Barbara Zhan, Vice President, Quantitative Research, Two Sigma
What made you interested in pursuing a career in quantitative research?
I have always been interested in the intersection between mathematics and the humanities. When I was looking for a job, I was particularly interested in quantitative finance because of the way it combines policy decisions, international relations, and human behavior in the markets with quantitative outcomes. Through the vast amounts of data in the market, qualitative events become measurable phenomena that can be figured out and predicted. I learn something new every day, because the world and the markets are constantly evolving. It is not an exact science, but the uncertainty is part of what makes it fun.
Can you please explain to our audience what your role entails?
Two Sigma takes a quantitative approach to investment management, and I work as a quantitative researcher. My role entails helping our firm make investment decisions based on statistical models that tell us how much we think an asset will be worth in the future, whether it is worth investing in, and how much of it to buy or sell. We use all sorts of data available in the world as inputs into our predictive models.
What is the typical day for you in quantative research?
There are many different moving parts and many other people around the firm that are necessary to get from an idea to a tangible trade. It all starts with an idea. Ideas can come from anywhere -- from reading academic papers, reading the news, attending conferences, or observing the market. I spend most of my day coding up and testing the ideas and analyzing the results. If the results look good, then I put those ideas into production. Whether the ideas work out or not, analyzing results can spark new ideas again, which creates an iterative cycle of discovery.
What skills does it take to be a quant researcher?
The foundational skills it takes to be a quant researcher are strong statistical knowledge, creative out-of-the-box thinking, and coding skills. Creative ideas are necessary to come up with the initial idea and statistical skills are necessary to evaluate whether that idea works despite noise in the results. Coding skills are the building blocks that power our idea testing, our trading, and the evaluation of the results. In terms of creative thinking, there are many types of predictive models that people can make at Two Sigma, and therefore many different kinds of ideas behind successful models. Some people specialize in specific algorithms that can find complex relationships amongst different datasets to predict future asset prices. Other people can have specialized expertise in certain domains that help them predict those industries better. Besides the types of models people can make, Two Sigma values cross-functional collaboration and people who can be good teammates.
What resources does Two Sigma have specifically for women in quantitative research?
Two Sigma is dedicated to making sure talented women are hired and have all the resources they need to succeed. For women up to a few years out of school, Two Sigma has a mentorship program designed to pair each woman with three senior mentors (of any gender), specifically aligned in goals and experience. Two Sigma also participates in the Grace Hopper Conference, a leading recruiting conference for women in tech, to hire full-time employees and interns in engineering and quantitative research roles. The culture is designed to be flat and meritocratic – there is nothing stopping people from getting well-deserved opportunities, as long as they rise to the challenge. Other recruiting opportunities, like on-campus events, sponsored conferences, and open roles are available on our website.
How do you create algorithms to trade? How did you learn to write algorithms?
Writing trading algorithms is a prediction problem. Our models predict future asset values, much like a weather model predicts weather. I learned to write prediction models during my undergrad at Princeton, as well as through recreational sports analytics -- leading to winning first place at the 2017 NBA Hackathon. Two Sigma also hosts modeling competitions through Kaggle.com for people to practice modeling. These competitions are a great way to understand the modeling process at Two Sigma, and to get recruiting opportunities.
Reach out to Barbara on LinkedIn.