Hedge fund quant posting 21% return says biology is secret sauce
The computer scrutinizes so many variables that even Lun doesn’t entirely understand how it makes calls. But no matter. The returns at his $20 million firm have blown by the S&P 500 since 2014 while the average hedge fund has lagged far behind the benchmark, according to an investor report seen by Bloomberg News.
Source: Vincent Bielski
Hedge fund manager Desmond Lun’s 21% average return over the last four years springs from an unlikely source — a petri dish of algae.
Lun, 37, is a new kind of quant, combining AI wizardry with old-school biology to trade futures. Although his Taaffeite Capital Management is small, Lun makes a big claim: His research into one of the natural world’s most byzantine systems — the biological cell — has given him an edge in untangling the secrets of financial markets.
Computational biologists like Lun are late to the quant wave that’s upending hedge funds. Physicists and mathematicians were the first disruptors, who found that their statistical models, neural networks and machine-learning tools have had as many stumbles as triumphs. Now comes Lun with artificial intelligence derived from algorithms he developed to figure out the mysteries of cells. The idea sounds more like a lofty dissertation topic than an investment strategy, except for one thing: Lun’s AI is beating the market.
“Causal interactions are important for both biological and financial systems, it’s what makes them complex,” said hedge fund manager Desmond Lun. Bloomberg News
Lun, who has a Ph.D. from the Massachusetts Institute of Technology, spent a decade developing models that decipher how genes interact and influence each other — and published 18 academic papers related to predicting cellular behavior. Turns out, he says, that his models, tweaked for financial markets, are pretty good at figuring out how traded instruments like stocks and commodities interact and influence prices.
“Causal interactions are important for both biological and financial systems, it’s what makes them complex,” said Lun, a tenured professor at Rutgers University in New Jersey. “If you have a causal relationship, you can make predictions and be somewhat confident in them.”
Lun’s machine-learning system hunts for what’s pushing prices of global indexes up or down. It examines tens of thousands of prices (called nodes in quant-speak) from thousands of securities, commodities and indexes. That means the algorithms potentially consider billions of interactions between pairs of nodes before making wagers. Two of its biggest bets as of June 30: long the FTSE 100 Index and short the MSCI Emerging Markets Index.
“It’s not a black box. It’s not that you stick stuff in and have no idea what will come out, we do have parameters,” said Lun, whose computer holds trades for an average of six days. “But on any given day, I can’t explain all the factors that led to the system reaching that decision.”
Lun’s field of computational biology has taken off since its birth two decades ago, propelled by the landmark Human Genome Project that gave us our genetic blueprint in 2003. Top universities from UC Berkeley to Princeton offer Ph.D. programs. Hedge fund billionaire David Shaw also jumped in, devoting most of his time to D. E. Shaw Research, a computational biochemistry group he started in the early 2000s with the ultimate aim of advancing treatments for diseases like cancer and diabetes.
As the genome project produced reams of data, Lun saw an opportunity to break ground in computational biology and in 2006 joined the Broad Institute of MIT and Harvard, a crossroads for scientists and hedge fund managers. There Lun met senior computational biologist Nick Patterson, a former cryptographer who had spent a decade at Renaissance Technologies making mathematical models. Another Lun colleague, genomic researcher Jade Vinson, left Broad for the same pioneering quant hedge fund for 10 years.
Wall Street firms are now chasing these biologists who are on the cutting edge of the data science revolution. They not only work with huge data sets. Cloning gives biologists another advantage: They can repeat experiments again and again, providing a better view of causal interactions inside cells, said Richard Bonneau, a New York University professor who studies computational biology.
“We are kicking butt more than any other data science,” said Bonneau, the incoming director of NYU’s Center for Data Science. Financial firms have noticed. “They steal my Ph.D. students, which is annoying, but good for them. They get nice jobs,” he said.
GENES AND STOCKS
A cell and a market seem to have nothing in common: one is microscopic and natural; the other is global and made by humans. But Ilias Tagkopoulos, professor of computer science and a researcher at the Genome Center at UC Davis, says the two are somewhat analogous.
“A biologist tries to understand a cell’s complex organization and function by observing snapshots of gene expression, which can be chaotic,” said Tagkopoulos, a former quant at Credit Suisse Group AG. “This is very similar to the stock market. A quant tries to make sense of time-series price data that at first look chaotic because we don’t know the different parameters and their relations. So you try to piece that together, as one would do with biological data.”
Overvalued? Funds with the highest price-to-book ratios
If these ratios increase along with stock prices, how high is too high?
There are skeptics, too. Emanuel Derman, who was among the first physicists to work on Wall Street, doubts that biologists possess secret sauce for investing. Derman rose to lead the quant risk strategies group in his 17 years at Goldman Sachs. He found that as physicists applied their expertise of the laws of motion, atoms and mathematics to investing, their models didn’t work nearly as well as they did in a lab.
Newton’s law of gravity hasn’t changed for eons, Derman said, but human behavior in markets changes all the time, wreaking havoc on even the best models made by scientists.
“I’ve developed a lot of skepticism about anyone bringing their expertise from one field to another,” said Derman, author of the book “Models.Behaving.Badly” and a Columbia University professor of financial engineering. “They say stocks are like atoms, or like genes. But stocks are not atoms or genes. There is a resemblance, but ultimately they are very different.”
Lun has reason to believe in his machine, which made money in June 2016 after the surprise Brexit vote and again in November following the election of Donald Trump. His flagship fund was up 9.8% this year through June, after gaining 22.1% in 2016 and 33.8% the prior year, according to the report. The fund lost 2% in 2014. It’s only down year.
Lun, who was born in Hong Kong, splits his time between his firm in Pennsylvania and lab at Rutgers, where he’s undertaken an ambitious long-term project: creating computer models that predict how cells behave, using data from blue-green algae and other sources. The models allow Lun to re-engineer genes for useful purposes: he has modified E. coli for production of bio-fuel for transportation.
His other ambition: to raise assets under management to as much as $200 million in the next 6 to 12 months, mostly from family offices and hedge funds of funds. But he has encountered some resistance from investors to betting on AI because it’s new.
“It’s hard to explain simply why and how it works,” said Lun, who grew up in Melbourne and named his firm after the rare gemstone taaffeite found in Australia. “But we are on target with our fundraising.”