How aggressively a fund pursues value is an important consideration.
A version of this article was published in the October 2016 issue of Morningstar ETFInvestor. Download a complimentary copy of ETFInvestor here.
Value investing isn't new. Benjamin Graham and David Dodd first advocated the approach in the early 1930s, and many investors have applied it with great success. More than 80 years since that time, countless academic studies have also provided strong support that the value premium exists in nearly every market and asset class. The idea is simple, in theory. Purchasing stocks that are trading below their intrinsic value should lead to higher expected returns. However, there are many different ways to implement a value strategy.
There is no shortage of options for investors interested in trying to capitalize on the value premium. There are currently more than 40 single-factor exchange-traded funds focused on value, and an additional 150 multifactor ETFs that include value as one of their factors. Adding rules-based or indexlike actively managed funds to the list, such as those offered by DFA and AQR, brings the final tally to well over 200.
All value-oriented index funds and ETFs construct their portfolios based on observable measures of value, but they differ in approach. Exhibit 1 shows the value measurements that three of the most popular index providers, FTSE Russell, S&P Dow Jones, and CRSP, use to construct their value index portfolios. DFA is also included because it utilizes a simple, rules-based approach to construct its value portfolios. Each provider measures value in a unique way, as shown in Exhibit 1, but many of these differences are not as important as they might appear. The success of index value strategies is less dependent on how they measure value than on how aggressively they pursue it.
How Is Value Measured?
For nearly a century, active managers have tried to identify value using a wide variety of measurements, such as those listed in Exhibit 1 and countless others. In 1993, Eugene Fama and Kenneth French popularized the now commonly accepted, simplified, academic definition of value, which is based on stocks' price/book ratios, as part of their classic three-factor model (1). They showed that portfolios with a high average price/book ratio outperformed portfolios with a low average price/book ratio. However, there is no theoretical reason why a stock's price/book ratio is a better measure of value than other valuation ratios, such as price/earnings or price/cash flow. In fact, many follow-up studies reported a similar value premium, but they used a multitude of different value measures, including some not based on fundamentals but other measures such as prior performance.
In 2015, AQR Capital Management published a study entitled "Fact, Fiction, and Value Investing"(2) in which it tested portfolios constructed based on various definitions of value to determine the optimal measurement. AQR tested price/book, price/earnings, price/cash flow, dividend yield, and negative trailing five-year returns over a 63-year period. The results showed that all the portfolios constructed based on different value measures produced positive returns and were highly correlated to one another. It also showed that the effectiveness of each value measure varied drastically during each decade of the 63-year period. The price/earnings portfolio resulted in the highest risk-adjusted returns during the entire time period, but the authors conclude that a composite measure that equally weights all five value measures was the superior measurement because it reduced variation in risk-adjusted returns over time.
Based on these findings, it's not surprising that many value indexes have similar historical performance despite using different value measurements. Exhibit 2 plots the total returns and the standard deviation of monthly returns for four U.S. large-cap value indexes and DFA US Large Cap Value DFLVX over the trailing 15 years through October 2016 (CRSP index performance begins in 2001). The high correlation between different value measurements explains why the S&P 500 Value, Russell 1000 Value, and CRSP U.S. Large Value indexes all show a similar risk/reward profile. More intriguing is the significant difference between the risk/reward profiles of the S&P 500 Value and S&P 500 Pure Value indexes, because they use identical value measurements. What separates the S&P 500 Pure Value Index and the DFA US Large Cap Value fund from the other three portfolios in Exhibit 2 is that they are constructed in a way that more aggressively targets value stocks.