While banks and hedge funds battle over the best quants, usually armed with a PhD in either maths or statistics, Desmond Lun, a professor at Rutgers University and the co-founder and chief investment officer of Taaffeite Capital Management, has a different theory - study hard science.
He says that the quantifiable data and mathematical models used to study sciences physics, chemistry, biology and genetics give you a better grounding for working as a quant than degrees in math, statistics, computer science or business administration
“In terms of quantitative trading in particular, get a PhD in a field of science with a strong quantitative component,” Lun says. “When we look to build out our team, to get hired, you have to have passion for the projects you’re working on, a good scientific background and someone who can think critically like a scientist.
“Practical science fields are empirical studies that are fundamentally about predicting the future, which are fundamentally like the problems you’re encountering in financial markets,” he says. “I would say science is even better than math or statistics, because science is about studying the real world, and financial markets are a real-world problem.”
The more abstract parts of mathematics provide you all technical tools you can use, but not the right thinking one needs to have in quantitative trading, Lun says. Computer science is a field that can be very empirical and scientific and on the other hand can be very abstract. Advantage: the hard sciences.
That assertion is all the more noteworthy given that Lun has studied and taught many of those subjects. He got his PhD in electrical engineering and computer science from MIT. He worked as a computational biologist at the Broad Institute of MIT and Harvard and studied advanced bacterial genetics at Cold Spring Harbor Laboratory before becoming a research fellow in genetics at Harvard Medical School.
In addition to running Taaffeite, at Rutgers Lun currently teaches applied probability and introduction to programming, which includes instruction in Python and computer science fundamentals – problem-solving using computers, as he describes it. The programming languages that Lun uses for development at Taaffeite include C#, Java, Matlab and some Python scripting.
“[Quantitative trading] was primarily an intellectual interest for me to begin with,” Lun says. “I was working on problems in computational biology, developing learning algorithms and extracting all kinds of information out of all of these very large data sets.
“The problems prevalent throughout modern biology, reconnecting them to a network and analyzing it to predict future behavior – there are clear parallels between that problem and the problems in financial markets,” he says. “Prices are constantly moving up and down, and they affect each other in some sort of complex interaction network – you have to figure it out through observation.”
Lun started developing trading algorithms and testing them with his own capital in 2006. In 2014, he launched Taaffeite together with his partner Howard Siow, who works on all the business aspects of the hedge fund while Lun handles the technical aspects.
They recently hired Ronald Raymond, who has previously worked at Blueshift Capital, Vermillion Asset Management (acquired by The Carlyle Group) and Dreman Value Management, as the COO and CCO of Taaffeite. The fund is up to $21m under management, and the co-founders have identified $100m in AUM as the point at which they would want to make a hiring push to aggressively build out their team.
Lun is currently supervising three PhD students. What he looks for when selecting the students who get the opportunity to work with him are equivalent to what he looks for in candidates for jobs at Taaffeite – the two key components to success are motivation and talent. He defines the latter as possessing the technical skills that are required.
“When I’m looking for talent, I’m primarily looking for mathematical ability, but in the field that I work in, not a mathematical ability that is purely abstract,” he says. “They must be willing and able to do something that is a little dirty in mathematical terms, which means working with real-world data."
Photo credit: aluxum/GettyImages