A recent article in the Journal of Indexes (Blitz, [March/April 2013]) provides an assessment of smart-beta indexes. The author’s main argument is that smart-beta indexes provide access to factor-risk premia, but capturing such premia may be done in a more efficient way. This column aims to point out a number of shortcomings in David Blitz’s article, which provides his views without providing much substantiation. It has the potential to lead to confusion about smart-beta indexes.
Firstly, the article takes issue with the constraints on sector exposures used in defensive smart-beta indexes such as minimum-volatility and low-volatility indexes. It states: “In our view, both approaches are too extreme. The MSCI Minimum Volatility index is overly constrained, while the S&P 500 Low Volatility Index is overly concentrated. Our assessment is that the optimum lies somewhere between these two approaches.” While the reader does not learn what the factual basis—if any—for these claims is, it seems surprising to blame index providers for providing both too few and too many constraints. Moreover, while there is extensive literature on the use of constraints in portfolio construction,1 the article includes nothing on alternative ways of defining constraints. More generally, the question raised by such statements is, Who should define the appropriate level of constraints? Perhaps, rather than claiming that there is some optimal level of constraints without further disclosing what this level is, one should recognize that, depending on an investor’s context and beliefs, he or she may want to use a strategy with no sector constraints, with tight sector constraints or with a medium level of sector constraints.
Secondly, the article discusses the properties of maximum diversification indexes and the efficient maximum Sharpe ratio (MSR) indexes. The article argues that “Although the way in which expected risk and return are defined is not identical, the differences are relatively small.” Such a statement is inconsistent with the fact that considerable differences exist between these two approaches, both in terms of methodology and in terms of results.
In terms of methodology, these two smart-beta approaches pursue different objectives. The maximum diversification approach aims at maximizing a diversification ratio (see Choueifaty and Coignard ), while the efficient maximum Sharpe ratio approach constructs a proxy for the maximum Sharpe ratio portfolio of modern portfolio theory. The efficient maximum Sharpe ratio approach does not have any objective in terms of a diversification ratio.
In terms of risk and return properties, an analysis of the index returns of the corresponding U.S. indexes2 reveals that the two methodologies lead to pronounced differences. The annual return difference over the year 2010 was 5.6 percent, and the correlation between the relative returns over the S&P 500 Index of the two indexes was 0.5. Moreover, the market exposure varies from a value of 0.99 for the efficient MSR index to a value of 0.79 for the maximum diversification index. Furthermore, the article argues that efficient MSR indexes are exposed to higher volatility than cap-weighted indexes over the long term, even though the article cited by Blitz to justify his argument (Clarke, de Silva and Thorley ) does not specifically refer to efficient MSR strategies. Besides, this statement is inconsistent with an article analyzing the long-term performances and risks of this strategy, which finds that the volatility of efficient MSR is lower than that of a cap-weighted index containing the same stocks (see Amenc, Goltz, Martellini and Retkowsky ).