There's nothing smart about 'Smart Beta'
‘Smart Beta’ presents a beguiling prospect to investors: a set-and-forget investment approach that can regularly outperform market capitalization based indices, but we believe appearances can be deceptive. Our analysis suggests there’s nothing special about smart beta, certainly nothing that can’t be replicated by using an investment process based on something as arbitrary as the length of a company’s name.
The concept of ‘Smart Beta’ is far from revolutionary. At its core, it repackages style investing as a strategy with the superfi cial characteristics of an index. ‘Smart Beta’ is the marketing term used to refer to this new type of indexing where stock weights deviate from traditional capitalization weights in a way that attempts to beat the market. Most of these strategies appear to have impressive back tested performance.
However, its recent popularity has probably more to do with the growing awareness of the shortcomings of traditional benchmarks than a resurgence of interest in style based investing. These shortcomings include an anti-value bias as market capitalization weighting has a tendency to favor expensive stocks over cheap stocks and excessive concentration caused by a momentum bias pushing investors into themes that have outperformed in the recent past. The result is that larger stocks are favored over smaller ones, leading to an inefficient use of the investment universe.
In our opinion ‘Smart Beta’ strategies tend to reduce the anti-value bias by breaking the link between market capitalization and weight. The problem is that many weighting schemes that do not rely on market capitalization appear to work just as well. The weighting mechanism can be anything that is quantifi able that helps remove the anti-value bias and return drag from mega stocks. The key is that it should involve some degree of rebalancing back to a value not determined by price movements. Obviously the simplest form of non capitalization weighted index is one that applies equal weights to each stock (although, in reality, equally weighted indices tend not to be very investible). In our opinion, it is the act of rebalancing itself, rather than the choice of weighting scheme, that is important here as it is this that helps investors avoid chasing the most overvalued companies.
To illustrate the point we took the MSCI World Index over the last 25 years and reweighted its constituents using two schemes, fi rst using an equal weighted version and second, an arbitrary scheme deliberately divorced from the investment fundamentals of the underlying stocks. In this case, we used the length of a company’s name, so that the stocks with the longest names were allocated the largest weight in the index. The choice of weighting scheme may seem virtually pointless, but it meant that the weighting scheme didn’t suffer from performance bias (e.g., by leaning towards cheaper or lower volatility stocks).