A basket of 500 individual stocks ought to be diversified. Indeed, that has been a key element of the marketing efforts behind many of the funds tracking the S&P 500 index.
The SPDR S&P 500 ETF, for example, describes the benchmark as ‘a diversified large-cap US index that holds companies across all 11 GICS [global industry classification standard] sectors.’
There are some obvious caveats. It’s not an internationally diversified index, notwithstanding the proportion of its constituents’ revenues derived overseas, and it offers narrower exposure than the likes of the Russell 3000.
Most buyers of the S&P 500 will doubtless be familiar with these limitations. However, they may not appreciate how poorly diversified it is by factor. New research suggests that the S&P 500 is typically driven by only two or three of the five primary factors.
High quality, and not in a good way
The study – by Ananth Madhavan, Aleksander Sobczyk and Andrew Ang, all at BlackRock but expressing their own views rather than the firm’s – focuses on long-only value, size, quality, momentum and minimum volatility, as defined by the respective MSCI indices. It uses these indices to estimate the factor loadings of the S&P 500’s underlying names.
These loadings could then be used to determine both the concentration of factor exposures in the S&P 500, and to measure how much of the variability of its performance could be explained by specific factors.
Madhavan, Sobczyk and Ang found that between 2002 and 2016, the S&P 500 had an average factor breadth of 2.9 factors, so at any one time it was likely that the index would have minimal or no exposure to two of the factors.
For instance, as of their most recent data point, March 31 2017, the S&P 500 had effective weightings of 46.6% to quality, 27.2% to value and 26.2% to momentum. The absence of the size factor is probably no surprise given it is a large-cap index, but the lack of low volatility exposure may be less expected.
Indeed, since 2002, the minimum volatility factor has often been marginal, and it was almost invisible in the S&P 500 between 2009 and 2013.
‘In practical terms, this means that the effective number of factor bets has been small, and there was much less diversification than might be suggested by a 500-stock portfolio,’ Madhavan, Sobczyk and Ang commented.
One implication for them is that ‘a passive investor who blends a minimum volatility portfolio with the S&P 500 index can increase diversification by improving the diversity of factor exposures.’
The inverse is also true. Quality, for example, has long been a dominant factor in the S&P 500: it has accounted for more than 50% of the index several times since 2002. Even at its lowest decomposition in December 2008, quality represented 28% of the index. Any investor complementing a core S&P 500 ETF with a quality-biased satellite – active or passive – is heavily reliant on this single factor in their portfolio as a result.
Madhavan, Sobczyk and Ang extended their investigation to the Russell index family too. As would be hoped, the far broader Russell 3000 index did offer greater factor exposure, with an average factor breadth of 4.2 since 2002. However, as with the S&P 500, the minimum volatility factor had a zero weighting in the Russell 3000 as of the end of March this year.
They noted, though, that the Russell 3000 is not significantly more diversified in factor terms than the Russell 1000. ‘Even though there are around 2,000 stocks that are different between the Russell 3000 and the Russell 1000, the addition of 2,000 smaller-cap names to the Russell 1000 index does not significantly expand factor diversification,’ the authors explained.
All this matters because factors appear to be driving an ever-larger share of the market’s returns. ‘We find that the proportion of the index movements explained by factors has materially increased in recent years, which is consistent with a more top-down, macro-driven environment or the increasing importance of economy-wide risks for financial markets,’ Madhavan, Sobczyk and Ang reported. ‘This makes factor analysis and allocation – a deliberate and thoughtful top-down approach to understanding the factors present in portfolios – increasingly important for investors seeking maximum performance, reduced risk or superior diversification.’
With fund buyers focusing on the factor exposures of active managers, it is worth remembering that market-cap index trackers are not necessarily factor agnostic. In the case of minimum volatility in particular, passive investors may well need to supplement their core positions.
‘The lesson here is that passive investors are accidental factor investors and need to be aware of their benchmark factor exposures from a diversification and return viewpoint,’ Madhavan, Sobczyk and Ang concluded.