Smart beta has become the topic du jour. With some $140 to $150 billion invested in these strategies, one could argue that its time in the sun is indeed overdue. But the question this blog will address is what is it that makes beta smart? After all, one commentator pointed out correctly, it is just old wine in new bottles. Towers Watson coined this apparently controversial name. Many object to it, alternatively preferring to call it “strategic” or “structural” beta.
But what was Towers Watson thinking when it conjured up the moniker “smart”? That is not immediately clear and as one digs deeper into the topic, the reasons getting muddier.
Initially smart beta consisted of factor or style portfolios. Academics pointed out that while beta represented the overall risk in the market, all markets consisted of a number of embedded factors which added up to the market portfolio. These factors could be considered to be sub-betas each with an associated risk premia. Soon early adopters were creating value versus growth, small versus large, low volatility versus high volatility, quality versus junk and momentum versus drift portfolios to harness these factors. They had quite different risk/reward characteristics than the market portfolio. They could also be created following a set of rules making them qualify as a passive construct.
Given this, these factors could have a number of valuable uses to an investor seeking to amplify or diminish bets in its portfolios, thereby changing their risk/reward characteristics.
It wasn’t until Research Affiliates (RA) introduced its Fundamental Index that the process took on a new dimension because it was cogently pointed out that the market portfolio as represented by cap-weighted indexes was based on inefficient constructs. Its analysis showed that the cap-weighted feature caused at least a 2% per annum performance loss versus some other weighting system. RA used a set of economic variables as measures of size to weight index constituents as opposed to the number of freely trading shares multiplied by price. RA noted the obvious: that stocks which had better relative price action gained in weight in an index, a process akin to buying winners and selling losers. RA’s process weighted the fundamental measures of size which, it argued, is a more efficient index creation.
This then begged the question, “where was the improved performance coming from?” There was no longer an explicit factor bet, though RA admits that there is some value and smallness bias in its Fundamental Index. Its own research answered the question. The improved performance arises from periodic rebalancing that occurs in the portfolio. Cap-weighted indexes self-rebalance as prices change. The Fundamental Index must explicitly rebalance through security transactions to bring the weightings in line with changing economic variables.
RA has undertaken some extensive research to show that if you created a portfolio with the inverse of fundamental weighting better performance still occurs versus the cap weighted index. Indeed, it repeated this study for all factor portfolios and found exactly the same feature in every one.
Going back to the early days with the factor portfolios and the myriad of academic studies which support the returns to the embedded risk premia, it is necessary to appreciate how these studies were conducted. As it turns out, the portfolios are implicitly rebalanced at each measurement point because the predetermined factor loading had drifted from its initial setting. Practitioners creating these portfolios in real time had to deal with the transaction cost implications of this, but the fact remains, they were regularly being rebalanced. Academics had never attempted to assign a value to the rebalancing, assuming the return was arising from the risk premia.
So now what is smart beta? It does not appear to be solely related to factor or style configurations but likely it is not solely related to rebalancing. (Though the RA research strongly suggests it is). It appears to be a reasonably sophisticated arbitrage of the cap weighting process.
Whatever it is that determines the performance changes in both risk and reward dimensions, the intelligence in smart beta is it can be harnessed and intelligently utilized in a portfolio context. Therein lies its value.