When discussing smart beta strategies with some investors, I am sometimes struck by their visceral negative reaction to the topic. As much as I appreciate the clever moniker given to these strategies, I suspect it can strike others as a little too clever. Perhaps they interpret smart beta as implying that their beta is dumb – for serious investment professionals, akin to calling their baby ugly. For those impacted individuals, please feel free to substitute scientific, factor, advanced or systematic beta for smart beta, but I suspect it is in your best interest to take note of the smart beta phenomenon.
Despite any sort of negative reaction from some, smart beta strategies have become popular with many investors. According to a study by Cogent Research, 29% of the U.S. ETF net inflows went into smart beta ETFs in 2013. In addition, they found that 53% of institutional investors expect to increase their use of smart beta ETFs. The study may actually understate the adoption rate, since this study only focused on the ETF implementation of smart beta. Potential performance might also have something to do with the trend toward smart beta with a study by Cass Consulting finding that between 1968 and 2011 various fundamentally weighted indexes would have outperformed traditional market capitalization weighted indexes.
What is Smart Beta?
Interestingly, one of the complications of smart beta is that there is no standard industry definition. Simply put, one could define smart beta as a strategy that differs from the traditional market capitalization weighted method. Most, including me, would argue that smart beta should also be systematic and based on methodology that has been shown to provide excess returns historically.
Theoretically, if one doubts that the stock market is always efficient then there is an opportunity to form portfolios with better risk/reward characteristics. I like to think of smart beta as attempting to systematically harvest returns utilizing some strategies that have historically been employed by some active managers. For example, a smart beta strategy might weight a portfolio based on valuation criteria (stocks with lower valuation measures are weighted higher in the portfolio) rather than by market capitalization. Consider that market capitalization weighted indexes automatically add to their allocation of stocks that are increasing in price or decrease the allocation to those falling in price, while smart beta indexes can base the decisions on other factors. It seems very possible that designed correctly, these smart beta strategies might be able to avoid at least some of the curse of over-allocation to expensive stocks that befalls market capitalization weighted indexes.
Separating Good from Bad Smart Beta
While so far I have painted a rosy picture of smart beta, things are a little more complicated than they may seem on the surface. The beauty of market capitalization weighted indexes is that they give exposure to the market portfolio in a very straightforward manner.
While an entire article could be devoted to the attributes that I believe are needed to develop a robust smart beta strategy, we’ll focus on just a couple illustrative issues here:
- Factor persistence; and
- Liquidity/Cost considerations.
If one were to invest in a smart beta strategy, I would argue that the most important thing was that the strategy was rooted in a market anomaly that is likely to persist. In other words assuming that data is analyzed and the strategy has been shown to produce good results in the past, you now really want to know if it will work while your money is invested in it in the future. Accomplishing this is really no mean feat. Among the criteria one should consider in my opinion would be:
- Does the strategy work outside the U.S. as well?;
- Is the strategy relatively stable – or do small changes in inputs drastically change the results?;
- Has the strategy withstood the test of time even after it was discovered?; and
- Is there a logical reason why this strategy should work?
Even dealing with the above listed complications skipped a possible problem in the first place which is to find an attractive strategy. A well-known issue for those trying to backtest these strategies is that much financial data is revised over time. Consider that these strategies must make intelligent decisions in real-time in the future once implemented, but the backtest could have relied on revised data that would not have been available at the time of the actual trade. As you might imagine, this can certainly make a difference between an effective and ineffective investment strategy.
So once we assume that all the other hurdles have been cleared, these strategies should have significant liquidity so as not to impact the results too drastically. While earlier in the article I noted that giving larger weights to higher market capitalization stocks in an index was a possible drawback, it can be considered a positive for liquidity. Since an investor is putting more assets to work in the larger companies, it should make possible liquidity concerns less of an issue. Extreme liquidity issues could again turn an effective smart beta strategy on paper into an ineffective or unimplementable one. It should also be noted that many smart beta strategies have more turnover than traditional passive strategies, so the cost of transactions especially in areas of limited liquidity could impact the efficacy of the strategy.
The Complexity of Real World Examples
I believe price momentum has one of the strongest cases for being a legitimate smart beta factor. In fact, I was involved in the creation of a smart beta strategy based on price momentum and consider it one of the most compelling factors known. Momentum is simply the tendency of securities that have performed well in the past to continue to perform well over the relatively short term.
Even if you have never read about it before, momentum is a well-documented phenomenon in academic literature. Further when considering some of the questions I posed earlier regarding possible factors, momentum has very strong answers to these questions. Momentum is fairly unique in that it has been shown in studies to outperform across U.S. and international stocks, equity indexes, bonds, currencies and commodities over long periods of time. One study I read went back to 1927 for U.S. stocks. Our studies showed that price momentum using between 3 to 15 months all performed quite well over time. In other words the momentum component is fairly stable even with changes in the inputs. It is my belief that momentum should continue to work despite being a well-known factor for two reasons. First, it has withstood the test of time even in very liquid markets. Second, it benefits from two core human emotions of fear and greed which are not likely to change anytime soon.
Lest anyone believe that momentum is a smooth path to continuous gains and outperformance, the style certainly has periods of underperformance. For example despite shorter periods of outperformance, momentum has generally underperformed since 2008. These fairly long periods of underperformance are not necessarily uncommon. In addition, our studies have shown these momentum strategies can be more volatile than the benchmark index. Certainly it can take a bit of an iron stomach combined with a stubborn streak to possibly profit from this smart beta factor over the long-term.
One other well-known factor that others might point to as a possible smart beta input is the small capitalization bias. This is the belief that smaller stocks as measured by market capitalization outperform larger stocks over the long-term. While initial data seemed to indicate that the small cap effect exists, more recent data has added some questions as to its existence or at a minimum the persistence of this effect. At least one study argues that some errors in the securities database used for most of these small capitalization studies incorrectly accounted for delisted stocks and thereby overstated the historical returns to small cap stocks. Even if one takes the simple approach of looking at more recent small and large cap index data for the last 20 years, the small cap outperformance ceases to exist and the Russell 1000 index has actually outperformed the Russell 2000. While I don’t deny that small-cap stocks are likely a good addition to most investment portfolios, the case to systematically overweight them in a smart beta strategy due to a long-term outperformance effect is not an open and shut case by any means.
Closing Thoughts
While I am loath to predict the future, I do believe that smart beta will continue to grow and become a very significant part of the investment landscape. Why do I say that? As illustrated, I do think there are factors that can be systematically mined to produce long-term outperformance. Performance success is certainly likely to cause assets to find these strategies. A recent article in the Journal of Indexes looked at a subset of smart beta indexes and found that all of them outperformed their comparable market cap weighted index over the last 1, 3 and 5 years as well as since inception. Also, these strategies can be argued to find the sweet spot between traditional passive and active investing. Since these strategies are systematic, they tend to cost less to own than active strategies. Investors are likely willing to pay the additional cost over traditional capitalization-weighted indexing for at least a portion of their portfolio if they judge the probabilities of outperformance and the level of risk to be acceptable. At the same time, some portion of assets currently invested in active managers should be drawn to the lower costs of smart beta while these strategies use and harvest returns from many of the similar strategies of some active managers.
What could derail this success? Like any successful investment story, success might be its own undoing or at least provide a significant setback. Investors will need to be wary of too many new products looking to slake the thirst of those looking for smart beta. In their haste to market new solutions, firms could rest these products on more suspect factors or those not tested thoroughly enough. Despite the phenomenal performance of some of the products thus far, even well-researched smart beta factors will work or not work at different times. As illustrated in the earlier discussion, these strategies can have significant periods of underperformance despite strong long-term track records which may test the patience of some investors.
While I believe the future of smart beta is bright in the traditional investment space of stocks, the concept is already moving into the alternative investment or hedge fund area which only adds to the opportunities. Consider that some investment strategies traditionally used by hedge fund managers are ripe for being systematically implemented. Many call this systematic replication of hedge fund strategies, alternative beta to distinguish it from smart beta but it shares many of the same underlying concepts. While most alternatives don’t have the passive investible indexes that stocks do, the fees for active management tend to be even higher and thus provide incentive to harvest returns from some of these hedge fund strategies at a lower cost. This alternative beta will certainly be a topic for a future article, since it may provide some interesting opportunities for investors.
The views expressed in this article are those of Bill Stone individually and should not be construed to be the position of PNC Bank, National Association or any of its affiliates.
