Quants: is it time to tweak the code?

Quant investment managers began 2018 in an optimistic mood. The much-hyped computer-driven hedge funds were among the fastest-growing parts of the investment industry and on course to hit $1tn of assets under management.

The 12 months that followed were a shock to the system. Algorithms that were meant to beat markets in all conditions short-circuited during two brutal bouts of turbulence. Investors pulled their money for the first time in almost a decade.

Now managers of some of the most popular hedge funds have to decide whether to recalibrate their computer programs or merely dismiss 2018 as a blip.

“When a year like last year happens, it certainly drives internal questions,” admits Anthony Lawler, co-head of GAM Systematic, the quant arm of the Swiss asset manager. “As you would expect, any time there is significant volatility we undertake research to try to improve.”

Poor performance and outflows aside, what spooked managers most was the fact that quant investing’s foundations were shown to be decidedly shaky.

“It was a challenging year . . . a number of factors performed slightly differently to expectations,” concedes Ian Ashment, head of systematic and index investments at UBS Asset Management, the $781bn fund group. “There were periods when a lot of the factors were underperforming all at the same time.”

Cliff Asness said AQR suffered a ‘crappy’ 2018 (Chris Goodney/Bloomberg)
Cliff Asness said AQR suffered a ‘crappy’ 2018 (Chris Goodney/Bloomberg)

There is no doubt that 2018 was unusual. Many quant funds were hit as both equities and bond markets turned south for only the second year in three decades. Other quant strategies — such as those that rely on cheaper stocks outperforming and rising assets continuing their trajectory — suffered too. The funds’ popularity also led to a suspicion of overcrowding, where too many computer programs are trained to exploit too few market anomalies.

The sector’s returns fell on average by 3.5 per cent; the last time the funds fared as badly was 2008, when they dropped 3.9 per cent.

Over the decade, quant funds produced an average annual return of 2 per cent, slightly below the 2.8 per cent of the hedge fund industry overall. However, quants’ returns had annualised volatility of 3.9 per cent compared with 6 per cent for all hedge funds.

For the first year since 2009, the funds suffered net outflows, with investors withdrawing more than $8bn. The sector shrank for the first time since HFR began tracking it.

This was not what managers expected at the beginning of the year. Over the previous 10 years quant funds, also known as systematic or scientific funds, had been one of the few bright spots in the global hedge fund industry.

While other hedge fund strategies stuttered and faith in individual stockpickers fell away, quants continued to grow at a steady clip.

Money poured in and more funds were launched. Between 2009 and 2017 assets in quant funds more than doubled to $962bn, in part because investors hoped that improved technology would give algorithms a further edge.

Anthony Lawler: ‘We undertake research to try to improve’
Anthony Lawler: ‘We undertake research to try to improve’

Then came the strong volatility of February and October, when quant funds lost 2.9 per cent and 2.3 per cent respectively. The February fall was the biggest monthly dip in performance for the funds in HFR’s records.

Choppy markets may become more frequent as central banks rein in stimulus measures and some quant sector commentators suggest that algorithms need a tweak to factor in changing patterns.

Mr Lawler, however, says volatility is not the whole story. “Last year was not as awful as it might have been,” he says. “Volatility levels in our funds were down about one standard deviation. That’s not as big an event as it might seem. Our funds in 2017 were generally up more than one standard deviation.”

He says that despite 2018’s market moves, “this has not resulted in us chucking out our models. We are always making decisions about types of data being more or less relevant”.

The leading light of the quant fund industry is AQR, which has more than $200bn of assets and 1,000 staff. Despite this, it suffered a “crappy” 2018, founder Cliff Asness told the Financial Times in January.

Several of its most popular funds suffered big falls, including its $7bn Managed Futures Fund, which lost 8.9 per cent, its $3bn Style Premia Alternative Fund, which dipped 11.9 per cent, and its $2bn Long-Short Equity Fund, which dropped 16.3 per cent. In January the company said it would trim its staff.

Hal Reynolds: ‘Each year is a learning opportunity’
Hal Reynolds: ‘Each year is a learning opportunity’

Elsewhere, quant funds run by Man Group, the UK-listed hedge fund manager, fell on average by 15.6 per cent, with some dropping as much as 17.9 per cent. Man said group performance was “heavily influenced by volatility across asset classes”. Assets in Man’s quant funds fell from $27.9bn at the end of September to $24.7bn three months later.

One category of quants that suffered more than most was alternative risk premia funds, one of the most popular types. The products are designed to benefit from the higher returns of investing in riskier assets but limit downside by diversifying holdings. They aim to provide systematic exposure to a variety of sources of excess return, including value, carry, momentum and risk-aversion strategies.

Ian Ashment: ‘A lot of the factors were underperfoming all at the same time’
Ian Ashment: ‘A lot of the factors were underperfoming all at the same time’

These funds have attracted up to $200bn globally, according to MJ Hudson Allenbridge, the consultancy, but last year lost 7.5 per cent on average and experienced 6.1 per cent volatility.

Fund managers fear such a performance could knock investor confidence. MJ Hudson polled alternative risk premia managers and close to 90 per cent saw last year’s experience as the biggest impediment to attracting new investors.

Hal Reynolds, chief investment officer at Los Angeles Capital, a $25.5bn group that has run quant funds for more than three decades, says: “Each year is a learning opportunity.”

He points to several tricky periods for quant investing since the 1970s and says these have tested managers and helped the algorithms develop. “Quant managers are always learning. If you believe that rules-based strategies that were fit for the past are fit for the future, you are destined for failure,” he adds.

Additional reporting by Siobhan Riding


Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.