Headlines knocking the quantitative investment industry have not told the whole truth, writes Gill Wadsworth.
‘Crappy’; this is the term quant manager AQR’s chief executive Cliff Asness used to describe his time investing in quants at the back end of 2018 and into early 2019.
Asness’s frank comments about the less than joyous experience shared by some quant managers came at a Morningstar conference in May this year, and were a reaction to 2018 in which the composite HFR Bank Systematic Risk-premia Multi-Asset Index lost -18%, in comparison with a loss of -4% on the S&P 500 total return index.
This 2018 horror story resulted in significant outflows for both quant and fundamental equity managers. Research from eVestment shows an exodus from long-only quant managers amounting to $14.7bn in the second quarter this year which perpetuates the outward trend that started in Q4 last year.
Quant underperformance was not restricted to equities. Multi-asset and some fixed income quant managers were not helped by the correlation between stocks and bonds towards the end of 2018, which saw yields move lower as equities fell.
More recent performance, however, suggests that the mass exodus from quant strategies could have been rather short-sighted.
In June, the HFRI Equity Hedge Quantitative Directional Index gained +3.9% for the month, meanwhile there has been a recovery for fixed income quants too.
By way of example, the Robeco QI Global Multi-Factor Credit fund leapt from -0.93% in the year August 2017 to August 2018, to 7.48% in the year August 2018 to July 2019. Meanwhile the AQR Core Plus Bond Fund rose from -0.18 in April 2018 – underperforming its benchmark by 0.22% – to 2.11% in May 2019 representing an outperformance of 0.19%.
Georg Elsäesser, portfolio manager in Invesco’s quantitative strategies team, says quant funds’ underperformance should not have resulted in investors pulling their money since the subsequent recovery was entirely predictable.
Elsäesser says: “In the long context this [underperformance] is just noise and represents a very short period of time. Value and momentum are negative for one of the few first times in 30 years then everyone is panicking.”
He adds: “Every winter it snows in Germany. If there are two winters where it doesn’t snow people think this is the way it will be forever; that it will never snow again. Clearly it will snow again, it always will in winter.”
Elsäesser says the factors that form the foundations of quant strategies were always going to return to form – and while it was not entirely obvious when that rebound would come – if investors understood why they were in the quant strategies in the first place, they should have been able to sit tight and wait for performance to improve.
“Some [quant strategies] have had a tough time over the last year to two years but now the picture has changed and they are delivering as we speak. That doesn’t mean investors should hold a fund blindly, but they should underwrite the evidence on which they invested and decide if it is still useful before disinvesting,” he says.
Sucking up ‘crappy’ performance in the belief that better times are around the corner is all well and good, so long as investors do not see asset managers offering similar strategies outperforming their own.
Man Group was one of the few quant managers to buck the underperformance trend, suggesting it is somewhat disingenuous to write off all negative quant performance as a function of the market itself.
Graham Robertson, head of client portfolio management at Man AHL, says: “It is a mistake to say the whole quant industry has struggled in the last 18 months. There is a significant dispersion in some intuitively similar strategies. If everyone is allegedly trading the same strategy, the dispersion [in performance] is remarkable. Manager choice is important.”
What drives success?
At its core, a quant strategy’s success lies in the algorithms used to determine the best investment choices and the people behind that technology. It is clear then that the winners in the quant race will be those that make the most of the huge amounts of data available and hire the right brains to make sense of it.
Jason Williams, portfolio manager at Lazard Asset Management, says: “Quantitative investment processes are capable of exploiting inefficiencies in the market in a systematic fashion. Academic studies have shown that certain market anomalies persist owing to how investors price certain stocks. The pursuit of generating consistent returns from investment anomalies entails significant research, investment experience, and risk management expertise.”
While quant managers will always struggle with maintaining performance during market anomalies and extreme volatility such as that experienced in the final quarter of 2018, Elsäesser argues systematic investors’ use of technology and data means they will have an edge in the long run.
“The big thing is data which offers a terrific chance for quant investors. The ones who are capable of navigating the jungle of data and who can distinguish between the good and bad stocks and bonds, have a real edge.”
However, the quant manager must continually evolve and be at the cutting edge of the latest tech if they are to exploit this advantage.
For example, Elsäesser says investee companies are already wise to the way in which quant managers use natural language processing to analyse reports.
He says: “These [reports] are scripted, companies know what words to include to look favourable when analysed by machines. Quant managers need to look at live presentations, and improve dictionaries and language processing to find the truth; how self-deprecating is the CEO or how confident; did they say, ‘um and er’, were they excited?”
Elsäesser goes a step further and notes that in the future, quant managers may use facial recognition software to analyse a CEO’s expression to ascertain the reality of a company’s financial health.
“This is the reality for quant managers, it is fascinating to see where technology can lead the way we invest,” he says.
More immediately, quant managers are using machine learning to improve execution. Man Group uses Adaptive Intelligent Routing to ensure order flow is allocated to the ‘optimum execution algorithm’ and allocates eligible flows to brokers in an automated and data-driven manner.
Chris Woolley, Man Group’s director of trading, says machines learn and remember how traders execute, which improves the trading process and frees brokers and managers to focus on the qualitative side of the job. “We have leveraged machine learning to help optimise where and how we trade. There is a huge amount of data to be captured and machine learning has allowed us to do that in a scalable way,” Woolley says.
Streamlining execution is of paramount importance to quant managers since any efficiencies made translates into return for investors.
“If you are paying less for execution that is pure alpha. There may be marginal differences between managers, so the speed and efficiency of execution can make all the difference,” Goodworth says.
Up to the early part of 2019, some quant performance has been ‘crappy’, but so were returns from some active managers. The markets were cruel at the end of 2018 and there were few places for anyone to hide. But quant investing is for the long term and for a rewarding experience, the focus should be on technology and the people operating it rather than short-term dips. ●