The most recent Data Science For All meetup took place on Thursday August 8th, 2019 in Midtown Manhattan, and featured a discussion about Data Science within the field of Finance. Data Science for All (DS4A), the fastest growing data science Meetup in New York, was created to help both practicing and aspiring practitioners learn more about the evolution of the field and find new job opportunities at some of the world's top firms.
Interested in job opportunities at IEX? Take the Quant Researcher assessment or the Data Engineer assessment. High scores will be shared with hiring managers at IEX.
The August meetup featured a great conversation with extremely impressive panelists who shared insights into the role of data science within their firms and guidance on how to get hired. We also packed the room with nearly 200 attendees and broke a record for our most attended meetup to-date.
Our panel was headlined by:
Elaine Wah, Head of Quantitative Research at IEX
Alex Martins, Head of Data Science Solutions at Bloomberg
Tong Zhan, Head of Technical Product at Correlation One
Shane Wilson, our Director of Sales, and moderator extraordinaire drove the evening's discussion.
For those who missed the talk, here were some of the evening's best quotes:
Alex Martins (Bloomberg): "One thing I always look for in a candidate, in addition to core skills, is whether or not this candidate is indefatigable, meaning she won't give up. If you bite down on a problem, you will fail 100 times but you'll keep chewing away at that bone until you extract as much as you can from this problem and you get as close as you can to this answer."
Tong Zhan (Correlation One): "Quants need to be very good about how they separate signal from noise. The classic example is that if you know the average American male is 5'10, it's very, very unlikely you'll find someone who is over 8 feet tall. But in the markets, maybe on average if stocks move 1% a day, you're going to find many, many instances when stocks move 5% in a day. That sort of non-normal behavior is the result of all the players interacting in the market. This leads to a lot more noise as well as a lot more variability. It's probably the most difficult part of what quants do."
Elaine Wah (IEX), on the difference between being a Data Scientist or Quant on a small team versus a big team: "IEX is probably no longer really a startup. We're already over 100 people at this point. But I think working for a smaller company, data scientists, or just data science-type people kind of play multiple roles. So it's not like there's just one distribution per team. There's really only three quants total at IEX but there are a number of data engineers, there's a quantitative analyst, there's the business analytics team and there are also quants on some of the newer business lines as well. But we all wear many different hats. Sometimes one of the quants on my team might work with regulation on something related to a role filing or someone might be working with product about a new idea and trying to see what the market potential is, so it really varies."
Tong Zhan, on the application of Data Science in finance beyond computer driven firms: "Within the finance space, even under hedge funds and investment firms there are many different kinds of strategies that can be data science. Computer-driven hedge funds is a very small category. So if we branch out broader than that and we look at say hedge funds or investment firms that make decisions based on economic data or company information, sort of financial analysis, their data scientists are not the ones who are building final trading strategies that end up getting deployed in the market via computers. They actually work with fundamental company researchers in order to figure out the best companies to invest in and the best way to execute those strategies and build the entire portfolio. So for example, let's say a person who is a fundamental company analyst wants to look at an agriculture company, right? One thing that's popular these days is to look at alternative data. So they may look at satellite images of large farms or property areas. I can use that as a proxy to figure out how much supply there is of corn or wheat in the world. In order to process this data in an efficient manner, we can't just have the fundamental company analysts go through and take a look at all those pictures. You can use a data scientist on the team who's responsible for developing the models to do image processing and then from there quickly develop an estimate of what's going on to inform the final decision of whether to buy or not."
Elaine Wah, on what she looks for, skill-wise: "We are pretty language-agnostic when it comes to candidates so if you have strong programming skills in a language like python, that is something we definitely look for. I think strong communication skills for my team are key because my team is actually very client facing, so I spend actually probably 20% of my time with clients. So being able to communicate your research results and your methodologies in an easy to understand way is very, very important as well. I think another broader skill that we look for us is the ability to apply a very hypothesis-driven approach to solving problems because we are very research focused. I think one of the differences between my team and some of the other more 'data scientist' or 'quanty' teams is that we work very exploratory, open ended problems where sometimes it's not very clear what methodology or machine learning models should be employed. It takes a lot of patience to come up with a hypothesis, test it, see why it doesn't work, come up with a new hypothesis and then iterate from there."
As with all of our meetups, jobs and how to land them, was an extremely hot topic. Here were some of the open positions that came up in conversation with links for anyone interested in applying:
IEX Trading's featured jobs:
Bloomberg LP's featured jobs:
Data Scientist for Customer Analytics
Global Data Manager - Data Science (London)
After the main panel discussion, we got a handful of great questions from the audience that our panelists tackled during the 15-minute Q&A portion.
The panelists were also gracious enough to stick around to answer as many questions as possible from attendees.
Before and after the panel, attendees took full advantage of networking time to make new connections, talk about the job market, and discuss developments in the field.
All in all, Data Science For All's August meetup was a huge success!
We'll be announcing our September meetup very soon on our LinkedIn page (be sure to follow us there). Soon thereafter, we'll release event details on our meetup.com page.
We very much looking forward to seeing you at our next event! If you are interested in sponsoring a Data Science for All event, please get in touch.