LONDON (Reuters) - Getting man and machine to work together is crucial for Robert Frey, professor in quantitative finance and former colleague of Renaissance Technologies star James Simons, as he builds his latest hedge fund firm.
The 57-year-old lecturer at Stony Brook university in New York has brought his experience developing algorithms that went into Renaissance’s top-performing Medallion fund to the battered funds of hedge funds sector with the launch of Frey Quantitative Strategies.
He uses his own super-complex computer models, programmed in C++ and Python, alongside the more traditional methods used by most of his competitors — meeting managers face-to-face, scouring CVs and working out how robust a manager’s firm is.
The goal is to work out how much of a fund’s returns are simply down to market moves and the like — the so-called beta that any investor can earn — and how much is down to alpha — or manager’s skill, the holy grail of fund management.
In a sector still reeling from 21.37 percent losses in 2008 and the failure of some big name firms to spot the Bernard Madoff fraud, Frey aims to bring a new angle in the quest to work out exactly what a hedge fund manager is investing in.
“It’s not quantitative versus qualitative. It’s figuring out what’s going on versus not having a clue,” he said from FQS’s London offices with their spectacular views over the city.
“It’s statistical analysis overlaid with traditional work,” he added. “We’re trying to classify sources of risk and return — how much is from alpha, how much from superior risk controls. We’re trying to decompose a manager’s return into components.”
It was an accident at the age of 14, in which Frey broke his hip and leg and suffered soft tissue damage, that changed the course of his life, ruling out the sports he enjoyed and making intellectual activity his great outlet.
However, it was a new direction that was to lead to success. Having invested in a restaurant, Billies 1890 Saloon, to fund a PhD, Frey worked on statistical arbitrage programmes trading pairs of stocks at Morgan Stanley, before starting his own fund.
It was whilst raising money that he met Simons. The secretive founder of Renaissance, who earned $2.5 billion (1.5 billion pounds) last year, bought Frey’s firm, rebranded it as Nova, and integrated it into Medallion, the flagship fund with astonishing average returns of 45 percent a year since 1988.
Frey left Renaissance in 2004, but the wealth he had accumulated, plus his friendships with the rich, helped shape his thinking to make wealth preservation a key aim for FQS.
“The clients are people who’ve made money and are worried what to do with it,” says Frey, whose firm raised $300 million, including money from staff and via wealth network Tiger 21.
“You don’t do anything stupid, which is very, very important — in the hedge fund space that can lose you a lot of money. The S&P never blows up the way a hedge fund can blow up.”
Making sure a hedge fund’s returns are not going to disappear when market conditions change is therefore critical.
“You’ll see two funds, which on the surface both have very good performance. One fund is exposed to things that have done very well recently. The other is not exposed to them, but has been generating significant returns and significant alpha.
“For long-bias long-short equity managers, if the market rises 30 percent then of course they will do well. For distressed credit or event-driven funds, it’s harder to see. We’re explaining 60-70 percent of the variance in these models.”
The Frey Multi-Strategy fund is so far down 0.88 percent, albeit from launch in March to the end of August. While only just over $100 million has been invested, the fund has substantial weightings in credit and distressed and global macro funds, among others.
Rather than chasing a strategy in vogue, Frey’s funds are based on a long-term asset allocation ‘anchor’, which changes only gradually over time. For the final portfolio the team will then tilt these allocations, based on their research.
Frey also believes he has an edge by analysing what he calls ‘regime shifts’ — using his own algorithms, rather than the simpler method of moving averages, to assess what sort of volatility environment exists and how that might change.
In an industry still struggling to emerge from the ravages of the credit crisis, identifying such changes can be key.
“You’d better understand what regime you’re in now. If you’re building a house, you want to know what winter weather looks like,” he says.
“In the period before 2008, volatility was around half normal levels. People got used to such a low volatility world, but when volatility spiked it was devastation.”