Comparing Smart Beta Options
For more than three decades, Dimensional has invested using a systematic approach designed to outperform conventional market cap-weighted indices. Our equity strategies seek to increase expected returns by pursuing size, relative price, and profitability premiums while sharing many of the benefits of conventional indices, including low fees, low turnover, and broad diversification. In efforts to outperform conventional indices, we focus on all aspects of implementation, from research into the dimensions of expected returns to intelligent portfolio structure to efficient portfolio management and trading.
In more recent years, investors have been presented with many new indices that seek to outperform conventional market cap-weighted indices by breaking the link between a stock’s desired weight and its market cap and instead deriving desired weights from characteristics such as book value, earnings, or recent performance. They go by many monikers, but smart beta has become the most commonly used. Smart beta index-based approaches have been garnering both interest and assets. Exhibit 1 shows that smart beta approaches have grown to more than $500 billion over the last 15 years. While still a small percent of the overall mutual fund and ETF market, their growth in net assets implies they are becoming a more popular choice for investors.
Data shows that smart beta indices sometimes provide investors with exposure (sometimes inadvertently) to size, value, and profitability premiums but may do so inefficiently and may subject investors to unnecessary risks. Given Dimensional’s experience pursuing such premiums, many of our clients have asked what could happen if smart beta indices become even more popular. Would these premiums go away? Like all questions dealing with complex subjects, the answer is “it depends.”
To begin, it is not clear whether smart beta indices are attracting additional demand to certain types of securities or if the inflows represent a transfer of assets from managers who target a similar set of securities using a more traditional active approach. For example, some smart beta indices focus on securities with a low price relative to fundamentals such as book value or earnings in an effort to increase expected return. Traditional value managers also tend to focus on securities with low prices relative to fundamentals in an effort to improve expected returns. If the inflows to value-based smart beta indices come at the expense of traditional value managers, the total demand for low relative price stocks may not greatly change. A similar argument applies to the size or profitability premiums. Without analyzing aggregate demand at the security level, it is not clear if inflows to smart beta will greatly change the demand for certain types of securities.
While there is no compelling evidence that demand for securities held by smart beta strategies has increased, one might ask; what could keep the expected premiums from diminishing? Consider the equity premium—the higher expected return of stocks over Treasury bills. Although realized equity premiums can be negative, there is little evidence that expected equity premiums are ever negative. This is sensible because investors generally demand compensation to bear the greater uncertainty around the value of their assets when investing in equities versus Treasury bills. Not everyone chooses to invest only in equities (even with the expectation of higher returns) because investors differ in their risk preferences, needs, and goals. We also know that it is not feasible for all investors to hold only stocks because the entire investible market, including stocks and bonds, must be held collectively by all investors. For everyone who wants to overweight equities, there has to be someone who wants to overweight bonds.
Within equities, how probable is it that all stocks have the same expected return? Just as it is reasonable to expect equities to have higher expected returns than bonds, it is reasonable to expect different securities to have different expected returns. There are many reasons for this. For example, Robert Merton’s2 multifactor approach links expected returns to state variables representing non diversifiable hedging needs against future risks or changes in opportunity sets. Under this general framework, the possibility of all stocks having one unique expected return is close to zero. Different stocks can provide different hedging needs and risks. This may be investor dependent if some investors are the natural holders of certain risks and others are not. Alternatively, in behavioral finance, tastes and preferences drive expected returns without associated “real” risks or hedging preferences. So under this framework, there is also a very low possibility of all stocks having the same expected return.
To identify information that can be used to determine differences in expected returns between securities it is useful to begin with the valuation equation. The valuation equation links expectations about a firm’s future profits to its current price through a discount rate. Algebraically, the valuation equation says for a set level of expected future profits, the lower the price, the higher the discount rate. It also says, for a set price, the higher the expected future profits, the higher the discount rate. The valuation equation implies the expected return of a stock is driven by the price paid and what is expected to be received. So, regardless of whether differences in expected returns between stocks are due to risks, investor tastes and preferences, or a combination of both, market prices and profitability contain information about these differences in expected returns. A low relative price is one indication that the market has discounted a company’s expected future profits more heavily. Applying the same valuation logic, companies with similar price characteristics but different levels of expected profitability should have different market discount rates.
What happens if everyone becomes a value investor? The answer: It is not possible. Collectively, all investors must hold the entire equity market. For every investor who wants to overweight stocks with low relative price or high profitability, there have to be investors who want to overweight stocks with high relative price and lower profitability. As long as there are differences in the discount rates market participants apply to different stocks (a very reasonable assumption), a strategy that uses a combination of current market prices and up-to-date firm characteristics can identify those differences today, tomorrow, and out into the future. Using current prices is the key to identifying the differences in discount rates the market has applied now.
While we expect positive size, value, and profitability premiums, not all strategies that pursue those premiums are created equally. Investors always have the option to invest in a plain vanilla broad market index fund. This is a decent option as these funds are transparent, low cost, low turnover, and well diversified. The success of conventional market cap-weighted indices can be explained in part because they have delivered what they set out to deliver: market rates of return. This success is likely one driver of the recent interest in smart beta strategies. Investors should be careful, however, when extrapolating the success of conventional indices to infer smart beta indices will be able to deliver on their goals of outperforming the market. Unfortunately, good back-tested research is not nearly enough information to make this assessment. The implementation details matter.
For example, if a smart beta index ignores market prices, it is difficult to infer if it will have a higher expected return than the market going forward. In back-tested research, the index may have provided inadvertent exposure to stocks with low relative prices and high profitability and outperformed the market. If current market prices are ignored in index construction, however, this implies the index is not directly managing that exposure and may not outperform in the future. Another detail to understand is the risk of a smart beta index generating excessive demand for less liquid stocks and creating unnecessary turnover. As the assets attached to that index increase, there may be a drag on returns due to an index reconstitution effect.
When designing strategies at Dimensional, we begin with research into how markets work. A sensible story can boost one’s confidence that a premium is positive in theory, while a low-cost approach improves the chances of capturing premiums in practice. No premium is a sure thing. Costs, on the other hand, surely lower investors’ net returns. Investors should consider whether a strategy would be a good investment even if the premiums are smaller in the future or do not appear at all. Some questions to consider when evaluating a strategy include:
- Does the strategy use current prices when choosing securities? The expected return of an investment is driven by the price investors are willing to pay and what they expect to receive. If a strategy ignores current prices, a vital component of what drives differences in expected returns is omitted.
- Is there unnecessary turnover? Security weights that are not tied to market cap weights, such as equal weighting or rank weighting, may incur excessive turnover that increases implementation costs without increasing expected returns relative to a market cap based approach.
- How is the strategy rebalanced? Trading that demands a lot of liquidity from the market can be costly, even if turnover is low.
- Are there avoidable risks? Is the portfolio well diversified given its mandate? Investors should be cautious of an approach that allows for extreme positions in a few securities. This approach can yield good back-tested results but can also result in significant company-specific risk in a portfolio.
It is important to have a solution that will be at least as good as the market portfolio in most scenarios, including if the premiums do not show up. A strategy with high implementation costs will have lower expected returns than the market portfolio if the targeted premiums do not exceed the costs. Keeping opportunity costs low helps a strategy maintain market-like expected returns even if premiums do not appear in the future.
The underlying driver behind the question “What happens to premiums if more investors adopt a smart beta approach?” is an old one. It is: “Will the dimensions of expected return be there in the future?” While we cannot say what the returns of size, value, and profitability premiums will be, it is far from certain the aggregate demand for the stocks that drive these premiums has increased such that the premiums are expected to be smaller going forward. Setting this aside, the details around how to pursue those premiums are the most important details to focus on. A well-designed strategy that seeks to capture size, value, and profitability premiums should minimize its opportunity cost relative to a broad market index. Using current prices in every part of the investment process, remaining well diversified, pursuing premiums that lead to low strategy turnover and eliminate unnecessary turnover, and having a flexible approach to portfolio management and trading that balances competing premiums and considers explicit and implicit trading costs are among the key tools that can be used to minimize opportunity costs. Dimensional has been making these considerations for our clients for more than three decades.
1. Total net assets in open-ended mutual funds and ETFs identified by Morningstar as strategic beta.
2. Robert Merton provides consulting services to Dimensional Fund Advisors LP.
All materials presented are compiled from sources believed to be reliable and current, but accuracy cannot be guaranteed. This article is distributed for educational purposes, and it is not to be construed as an offer, solicitation, recommendation, or endorsement of any particular security, product, or service.
Performance data shown represents past performance. Past performance is no guarantee of future results and current performance may be higher or lower than the performance shown. Consider the investment objectives, risks, and expenses carefully before investing.