- Investors seeking to add smart beta and factor strategies to their portfolios should consider the four craftsmanship elements of product design: 1) universe coverage and weighting mechanism, 2) signal definition, 3) measurement period, and 4) rebalancing frequency.
- Design decisions guided by these craftsmanship elements can help preserve a portfolio’s performance potential by reducing implementation costs.
In the previous article in this series, our colleagues John West and Trevor Schuesler explained the challenges of seeking positive alpha through manager selection. In this article, we highlight how financial advisors may add value with due diligence efforts on, perhaps, a more reliable source of excess return—product craftsmanship.
Technology today pushes us to move faster in almost every area of our life. We are encouraged to consume our news in 140-character bites, race through podcasts on 1.5x speed, and make every effort to “10x” our efficiency, condensing what we used to accomplish in 40 hours into a 4-hour workweek. Overawed with big data, we tend to be impressed by the glitter of complexity, and let slip from sight an appreciation for the simple beauty of the straightforward. When we rush at warp speed, much can be lost. In the investment industry, the application of nuanced, masterful strokes of design and development are rapidly becoming a fading art.
Investors face a dizzying array of decisions as they navigate the hordes of “latest and greatest” factor and smart beta strategies and new product launches. Many clients are seeking specific long-term factors as robust drivers of returns, without recognizing that the world is constantly changing, that mean reversion can flip historical alphas on their heads, and that equilibrium relationships are themselves not static. We suggest shifting due diligence efforts to a critical, and often underappreciated, awareness of what we and others call craftsmanship1—the product design and implementation elements that are tangible, measurable, and impactful.
We begin by explaining what qualifies as a robust factor and then what we believe are the essential dimensions of quality craftsmanship in smart beta strategies. Whereas we focus on one of the most commonly accepted robust factors—value, as represented by the Fundamental Index™ approach—the underlying lessons and considerations drawn from this example are broadly applicable to all smart beta strategies.
Our colleagues recently proposed a framework for how to determine which of roughly 300 factor strategies are robust (Beck et al., 2016). First, a factor should be grounded in a long and deep academic literature. This academic scrutiny should identify a theoretical basis for the factor return premium, such as an investor psychological bias (behavioral explanation) or an undiversifiable risk exposure. Ideally, this theoretical foundation should be determined before looking at the data, so that we do not use data to build the theory. Second, a factor should be robust across definitions, meaning that small changes in the measurement of a factor should not destroy its demonstrated performance. Third, a factor should be robust across geographies to show out-of-sample performance, and fourth, a factor should be implementable without incurring large trading costs that erode the factor’s return premium. On a separate note, as our Alice in Factorland series of papers observes, we have also learned that in a backtest much of past performance is tied to revaluation (i.e., a factor becoming more expensive). Without controlling for the valuation level of the corresponding factor, past success likely presages future disappointment.
Once an investor identifies the desired robust sources of return premium they wish to capture in their portfolio, the logical next step is to select the product(s) that can deliver these premiums. An investor new to the smart beta arena can easily be overwhelmed with the immensity of product offerings.
To address this challenge, we believe investors should be aware of the elements that constitute quality craftsmanship in the design of smart beta strategies, and make their product selection decisions accordingly. These elements can be grouped along the following dimensions: 1) universe coverage and weighting mechanism, 2) signal definition, 3) measurement period, and 4) rebalancing frequency.
A commonality underlying all of these dimensions arises from an overriding lesson we have gleaned from our combined decades of designing investment strategies: thoughtful product design work includes striking a balance between simplicity and effectiveness. Albert Einstein aptly captured this notion: “Everything should be made as simple as possible, but not simpler.” We lean toward simplicity because it tends to lead to more predictable results and easier governance (Brightman, Kalesnik, and Kose, 2015), and we are highly conscientious in our design of preserving the effectiveness of the strategy.
As we explore each of the four craftsmanship dimensions, we will use the example of a strategy—the RAFI™ Fundamental Index—that relies on fundamental measures of company size to systematically rebalance against the market's constantly shifting expectations, and thereby harnessing a value premium (Arnott, 2006). Investors can apply the following framework to evaluate the craftsmanship elements of just about any smart beta strategy.
Dimension #1: Universe coverage and weighting mechanism
Decisions surrounding the universe coverage and weighting mechanisms of strategies can meaningfully impact investor portfolios because they influence the amount of liquidity available, which in turn affects transaction costs. For instance, if we invest in a strategy that allocates equal weights to all stocks with no consideration of the size of a company (an equal-weighted scheme), we inherently have a higher exposure to smaller companies, which tend to be less liquid and more expensive to trade. Unfortunately, most investors ignore liquidity issues because of the difficulty in observing their tangible impact on portfolio returns.
This dimension partially informs why we chose to avoid an equal-weighted approach, even though it is the simplest non-price-weighted weighting mechanism of all! Consider the comparative analysis conducted by Aked et al. (2014). The starting universes of the portfolios in the analysis include all stocks within the largest 85th percentile rank, either by cumulative market capitalization (for the cap-weighted and equal-weighted strategies) or by cumulative fundamental weight (for the fundamental strategy).2 Both essentially cover a fixed portion of the market and are consistent with the practices of the most popular index providers.
A couple of observations: First, both an equal-weighted approach and a fundamental-weighted approach exploit a weakness in the cap-weighted approach—the tendency to overweight overpriced stocks and underweight underpriced stocks. So, as expected, both strategies meaningfully outperform the cap-weighted benchmark over the long run. Second, and more surprising, we observe a substantial difference in the performance of the fundamental approach and of the equal-weighted approach across the G7 countries: Australia, Canada, France, Germany, Italy, Japan, United Kingdom, and United States.