How to use smart beta

While smart beta is a much-talked about concept, implementation is slow. Part of the reluctance of investors is the risk of sustained underperformance, but that can be overcome by matching portfolio liquidity requirements with factor cycle duration. Amanda White speaks to Michael Hunstad, head of quantitative equity research, global equity management, at Northern Trust.

Sustained underperformance of a group of stocks according to a factor is a primary risk in factor investing, and further, the length of underperformance is not consistent across factors and so needs to be managed. But this provides an opportunity for investors to tilt their portfolios according to the time and liquidity needs.

According to Northern Trust research, cycles vary in length across common equity factors, from 12 months for the low volatility factor to 106 months for small size factor.

The research, which spanned data from 1979-2013, found the high value factor to have a cycle of 47 months, low volatility 12 months, high divided value 22 months, small size 106 months and high momentum 39 months.

In applying this research, Michael Hunstad, head of quantitative equity research, global equity management, at Northern Trust, says the combination of factor tilts depends on an investor’s time frame to liquidation.

Portfolios that are evaluated frequently or are near liquidation should opt for short-cycle factors like low volatility. Factors like size, value and momentum have long cycle lengths and should be held in long-term buy and hold portfolios.

Northern Trust says the optimal multi-period allocation of MSCI factor indices can be ordered according to time to liquidation.

Factors with relatively long cycles appear first in the portfolio to eventually be replaced by factors with shorter cycles.

As an industry, Hunstad says the discussion around smart betsa has moved past the initial stage of determining what compensation factors are and now investors are faced with the question of what factors to choose to tilt their portfolio.

He says the starting point for investors is to determine what their current portfolio factor allocations look like.

Northern Trust does this analysis for clients, strongly advocating that failing to base future investment decisions on a strong understanding of the current portfolio can lead to unintended bias or cancel out intended bias.

“For clients and potential clients we do mapping of their existing holdings and we see exposure to style biases exist, for example large cap. We’ve done it now for about 50-75 clients. The next step is to take corrective action so the portfolio can achieve their objectives, and then we discuss which strategies would work for them.”

Hunstad says in this way, even if a client has no intention of buying smart beta, it can be used as a way of thinking about risk.

Last year, Northern Trust conducted a quantitative survey of 139 global institutional investors to gain insight into how they were addressing strategic risk. It then looked at how “engineered equities” might better achieve the desired portfolio outcome.

The results outlined in the paper, Through the Looking Glass, found that only 18 per cent of the 139 investors surveyed felt they were certain of their overall equity portfolio’s actual factor exposures.

Further the survey found that regardless of the approach used to define the asset allocation –

asset liability management, core satellite, tactical or strategic – the portfolios didn’t always reflect the investors’ goals, objectives and intended exposures.

The paper says, to bring the portfolios in line with the investors’ expectations would require a substantial factor exposure – a deliberate and substantial commitment.

“One conclusion of the paper was that investors need to have to have a meaningful allocation to smart beta for it to be worthwhile, to actually move the dial on return expectations outside the noise of the portfolio,” Hunstad says.

In practice this is happening, and Hunstad says some clients are allocating 100 per cent of their equities portfolio in, for example, a quality, value blend or a quality, low volatility blend.

“You have to have a lot of conviction to do that,” he says. “More and more clients are realising it requires conviction. One client allocated 2.5 per cent of $10 billion but that doesn’t do anything.”

Hunstad says there are two variables in assessing factors – expense and whether it will be arbitraged away.

He firmly believes that while there is concern that, for example, low volatility will be expensive and the price is increasing, it doesn’t mean that low volatility will be unreliable or uncompensated.

This is due, mostly, to the total value of all smart beta being dwarfed by the total market – it’s less than 1 per cent, he says.

“It will never approach the volume you see in indexing, plus you see trades on the other side. It might be expensive or the wrong time for that factor, but it won’t be arbitraged away,” he says.