(Pictured: Michael Hunstad)
With the increasing popularity of smart beta strategies, discussion among investors has moved beyond the underlying ‘factors’ – whether they are sustainable, for instance – to product design and implementation. Managers such as Northern Trust are now refining ‘pure factor’ exposures to avoid unintended consequences.
Michael Hunstad, Northern Trust Investments’ head of quantitative research in global equity management, says that the dominant consideration in the selection of factors – such as low volatility, quality, size, momentum and dividend yield – is the investor’s time horizon.
Each factor has a different cycle. Investors should choose a factor or factors whose cycle length is less than their investment time horizon. “The rule of thumb is the intended holding or evaluation period should be twice the cycle length,” Hunstad says.
Northern’s research based on the US market shows that cycles range from an average of just 12 months for low-volatility stocks to 106 months for small-cap stocks. In between are 22 months for high dividend yield, 39 months for high momentum and 47 months for high value.
Investors are therefore increasingly using either more than one single-factor strategy or are building multi-factor strategies, with the evidence suggesting better long-term performance from a blend of factors. This, coupled with the search for ‘pure’ factor exposures, has led to increasing sophistication in smart beta strategies.
“A key consideration and a key risk is the unintended consequences,” Hunstad said on a visit to Australia from Northern’s Chicago head office last week.
“For example, low volatility tends to give you a negative value bias, a utilities bias and a number of other biases. Do I necessarily want these? What we’d like to see is more pure exposure to the factors.”
He said that if you look at performance histories of a range of products offering the same factor tilts you see a wide dispersion of returns. This is primarily because of the unintended exposures among some of the manager strategies.
Northern sought to quantify the unintended exposures and came up with the concept of a ‘factor efficiency ratio’, where the numerator is the factor and the denominator is the unintended consequence.
“We call our products ‘engineered equity’ [rather than smart beta],” Hunstad said. “We have much tighter bounds on unintended consequences. We tend to be very efficient.”
If more than one factor is to be blended in a portfolio, it is much better to choose the stocks which contain those factors rather than combining two separate groupings of stocks. Northern research on the period 1979 to 2013 shows a small-cap index delivering 13.7 per cent annual return and a high quality index delivering 16.8 per cent. If you simply blended the two indices you’d have a return of 15.3 per cent. But if you chose those stocks at the ‘intersection’ of the two factors, the annual return would have been 20.1 per cent – a massive difference over such a long timeframe. Similarly good results can be achieved by picking stocks at the intersections of low volatility and high quality, high dividend yield and high quality and high value and high quality. The simple 50:50 blend dilutes the benefit of the factor exposures.
Other trends in the types of discussion that investors are having about smart beta, Hunstad believes, include the integration of factors in asset allocation decisions and the incorporation of ESG considerations.
“ESG is a hot topic everywhere,” Hunstad said. “The two big questions investors ask are ‘what will happen to returns?’, with the implications from a fiduciary’s perspective, and ‘what will happen from a risk perspective?’. They are asking how we can use the lessons from smart beta.”
Research shows that a quality factor is a good complement for an ESG overlay to get the returns up, because they have complementary themes.
To resuscitate debate about factors, for those of us who haven’t yet moved beyond this stage or simply enjoy the arguments, Northern’s quants believe there are only six pure factors where it is clear an investment is compensated for the risk. They are: quality, value, volatility, momentum, size and dividend yield.
They put in an “uncompensated or yet-to-be-determined” category: leverage, currency, duration, ESG or SRI, low carbon, sector exposure, country exposure, industry exposure, commodities, idiosyncratic risk, macroeconomic exposure, beta and growth.
– Greg Bright