Quant Fund Gives Robots 364 Days Off to Best Currency Rivalsby
Systematic models make calls only once a year each January
PGI’s Global Time Diversified Strategy returns 13.8% in 2016
Like a lot of hedge funds, Principal Global Investors’ Macro Currency Group uses computer-driven models to devise trading strategies. What sets it apart is that its machines only make calls once a year.
Each January, the computers’ longer-term economic projections establish the firm’s key trading positions for the next 12 months. The bets don’t always seem intuitive at the time. At the start of this year, the models projected the yen would appreciate against Swiss franc, even as forecasters saw the pair little changed in 2016.
“Being very long yen at the beginning of the year wasn’t a very popular call,” said Ivan Petej, a money manager and head of quantitative strategy at PGI in London. Still, he said he was confident in the models. “What I was more worried about was when it was going to happen.”
It didn’t take a year for the trade to pan out. The yen has strengthened 14 percent against the franc in 2016 on haven demand and doubts that the Bank of Japan’s stimulus policies will stoke inflation. The $830 million Global Time Diversified Strategy’s 13.8 percent return this year is third among 30 funds tracked by Citigroup, and almost five times the 2.9 percent return for Hedge Fund Research’s HFRI Macro Currency Index through July.
The secret to the once-a-year call is the selection of economic gauges the robots track, like manufacturing data and other leading indicators, that drive long-term currency moves. The other key to the fund is that PGI’s computers aren’t doing it alone. Throughout the rest of the year, the firm relies on humans, led by money manager Mark Farrington, to trade around events that produce short-term fluctuations and adjust the portfolio for risk.
“If you trade this basket every year, you will get a proxy for global growth,” Petej said. “It trades once a year because the time horizon for different drivers -- valuation drivers, growth drivers, benchmark drivers -- is a yearly time horizon. Everything that happens intra-year, one week, six months, is dealt with a discretionary component.”
Currency strategies have been a bright spot this year as the $2.9 trillion hedge fund industry faces a growing backlash from investors for lackluster returns and high fees. The Bloomberg Global Aggregate Hedge Fund Index is up 1.2 percent in 2016, on pace for its third year of sub-2 percent returns.
Petej, 40, who has a Ph.D. in physics from Oxford, joined the firm in 2006 as a risk analyst. He and Farrington saw a need for a fund that could capture world economic growth, yet went beyond standard carry strategies -- borrowing in countries with low interest rates to buy higher-yielding currencies -- that traders had used for years.
“We thought, ‘wouldn’t it be nice to create a benchmark strategy in currency markets that targets global growth and has a long-term time horizon?’ At the time, there were no real benchmarks in foreign exchange.”
They identified indicators of growth by back-testing over 20 years of economic data. They put the strategy on as a paper trade in 2007 and it went live in 2011 with $1 million in assets. The fund has averaged a more-than-5 percent annual return since its inception, with gains about equally split between the computer-based calls and human trading, according to PGI. The London-based group also manages a separate fund based on a purely discretionary strategy.
Most "systematic funds are running these computers constantly and are often making a lot of trades throughout the day,” said Don Steinbrugge, managing partner of hedge-fund consulting firm Agecroft Partners in Richmond, Virginia. “It’s very unusual to put a trade on and not take it off for an entire year.”
Some of the robots’ annual predictions have been ill-timed. The fund placed a short call on the Swiss franc at the start of 2015, just two weeks before the country’s central bank removed the currency’s cap against the euro, sending it 21 percent higher against the dollar in one day. Still, PGI kept the short call on for the remainder of 2015. The currency had recouped its losses by December, and the fund maintained its position into this year.
“It was a big shock for a lot of people -- even though the trade was volatile, the rationale for the Swiss franc weakening was still there," Petej said. “The fact that we kept that Swiss franc trade on actually helped us.”
While the computers have called the Swiss franc and the yen right this year, it’s the humans of PGI who have salvaged the firm’s bullish long-term dollar bet. They’ve made money by trading around markets’ perception of the Federal Reserve even as the greenback has slumped about 4 percent in 2016.
“The human bias is to take profits, the human bias is to trade around events,” he said. “You need time for the systematic component to play out."