© Cat O Neil
Experimental feature

Listen to this article

00:00
00:00
Experimental feature
or

Smart beta, as a marketing concept, has taken the investing world by storm. Investment managers are funnelling money into funds that track indices based on factors other than companies’ market capitalisation. It has allowed those who missed out on lucrative business in passively managed exchange traded funds (ETFs) — dominated by the big three: BlackRock, Vanguard and State Street Global Advisors — to catch up. Some 103 companies now sponsor smart beta ETFs, according to BlackRock.

Some analysts hope that smart beta could rescue markets from themselves. As passive indexing has come to dominate flows into equity markets, concerns have grown that market cap-weighted funds distort asset prices and leave investors over-concentrated in larger companies.

Art Cashin, a UBS veteran on the floor of the New York Stock Exchange, says: “ETFs and passive investments, by their very nature, are beginning to build inefficiency in the market, and I don’t believe that will be resolved happily.”

Paul Woolley, who heads the London School of Economics’ centre for the study of market dysfunction, says moving from an underperforming active fund to a cap-weighted fund involves selling stocks that have been losing and buying winners.

That is the opposite of what investors should be doing, as it aids price momentum and makes the market prone to bubbles. Smart beta, by systematically leaning against the market, may correct that. However, the initial simplicity of the idea has degenerated into the kind of mess that used to typify old-fashioned active mutual funds.

The term smart beta is self-contradictory, and the ideas behind it are not new. William Sharpe, the Nobel laureate who originated the concept of beta, says the term makes him “definitionally sick” and doubts the strategies can work in the long term.

Cliff Asness, co-founder of AQR, one of the world’s largest quantitative hedge fund groups, dislikes the term, but has admitted defeat.

“Language is democratic,” Mr Asness says. “I like the strategies and not the term. It’s relabelling. It’s been a good marketing relabel, and it’s probably benefiting me too.”

So what has smart beta come to mean? Beta is drawn from Mr Sharpe’s capital asset pricing model, which holds that the return on any stock can be broken down into the return attributable to the market as a whole (beta) and the idiosyncratic risks of a given company (alpha).

Beta, by its definition, is dumb. All you need to capture it is to buy stocks in proportion to their index weight. Smart beta uses passive investing techniques — minimising expenses on salaries, research and trading — while trying to beat the index.

The term smart beta started with actuaries helping passive managers to cut their costs. If you could replicate the index while not buying every stock in it, you could save money.

With the advent of US groups Research Affiliates and WisdomTree more than a decade ago, it came to mean including all of the stocks in an index, but weighting them by a means other than market capitalisation. Even equal-weighting, where every stock in a 100-member index would have a 1 per cent weight no matter what their size, could count as smart.

Intech, a subsidiary of fund management group Janus Henderson, had been offering rigorously quantitative funds for years before, based on the ideas of Robert Fernholz, published in 1982, which showed that cap-weighted indexing was not efficient.

Intech targets a “rebalancing premium”. Weight an index by other means than market cap and periodic rebalancing becomes vital. This allows a manager, in Intech’s words, to “capture the volatility created as stock prices move up and down”.

Rebalancing will involve selling relatively successful stocks that have gained and buying losers. It is an automatic discipline to sell nearer to the top and buy nearer to the bottom.

However, asset manager Robeco’s head of selection research, David Blitz, says different investment factors “are widely regarded as distinct phenomena”, and adds: “The fact many factor investment strategies do not require much rebalancing to begin with also makes it unlikely that these modest amounts of rebalancing are the main driver of their return.”

Some indices are weighted in favour of factors that have been shown to outperform over time.

Such tilts are not new. They go back at least as far as the three-factor model, published by Eugene Fama and Kenneth French in 1992, which expanded the formula for investment returns to include value (or cheapness) and size. Their research found that cheap and small companies both tended to outperform in the long run.

Active quantitative funds try to exploit those factors and find others. The agreed list includes: value, size, momentum (winners tend to win and losers to lose), low volatility (stocks that do not move as much as the market do better over time) and quality (which refers to strong balance sheets and reliable profitability). All have been staples of quantitative hedge funds for years.

A factors arms race is under way. One academic paper in 2014 counted at least 300 putative factors that had been put forward. Funds now offer mixtures of investment factors, while advisers can offer to manage portfolios by timing which ones they buy.

Whether this is possible is a matter of fierce debate. Research Affiliates chief executive Rob Arnott argues that factors can be timed, based on how expensive they are, and the most popular factors can become overcrowded, such as momentum, which has performed very well this year. But Mr Asness believes timing is difficult and expensive, and valuations of popular factors are not extreme.

The evidence from practice is that factor timing is very difficult. Aneet Chachra, a portfolio manager at Janus Henderson, says evidence from ETFs suggests “the average investor is terrible at factor timing”.

Mr Chachra analysed the largest 20 smart beta ETFs listed in the US, which were all invested according to equal-weighted, low-volatility or high-dividend-yield factors from May 2014 to May 2017.

Comparing the days when dividend ETFs saw net buying (when units were created) and net selling (when units were redeemed), they undershot the S&P 500 by an average of 2.4 percentage points. The dividend ETF matched the S&P over the whole period. This suggests the ETFs were return-chasing — in other words buying whatever is doing well. The implication is that timing does not pay off.

For low-volatility ETFs, investors, on average, lagged behind the S&P by 2 percentage points even though the factor was about 3 percentage points ahead.

For equal-weighted ETFs, investors, on average, paid 1 per cent more than they sold them for. This would compound poor performance, and equal-weighted ETFs did indeed lag behind the S&P by 3.5 percentage points over the period.

Smart beta ETFs can offer sensible ways to try to beat the market, and lean against anomalies, at a low price. But their problems are growing clearer. The danger is they will reintroduce the problems of traditional active management — managers try strategies that only work momentarily, while investors lose money by trying to time factors.

Copyright The Financial Times Limited 2017. All rights reserved.