- While value generally underperformed the broad U.S. equity market over the past decade, it did so at levels that were less than half those of the late 1990s, a period followed by 10 years of outperformance.
- In the U.S., underperformance in value was driven by the earnings yield factor and the technology sector. Value drove positive returns in several other regions and certain U.S. sectors.
- Within technology, the high-performing – but expensive and thus underweight – software and services industry has led the decline in value strategies.
In 1934, Benjamin Graham and David Dodd identified value as an equity investment style.1 Their then-revolutionary idea was that investors could buy stocks that were undervalued, based on their intrinsic value, and then wait for the market to recognize its “correct” price. A glance at value’s recent history, however, raises the question: Are value’s glory days over?
As shown below, value’s underperformance relative to the broad U.S. market over the past decade has hurt both index-based and active strategies. The MSCI USA Enhanced Value Index — a proxy for index-based value strategies — has limped through the past decade ending Sept. 30, 2019, when compared to the broad market MSCI USA Index. It underperformed the broad market index by 7 percentage points on a rolling 10-year basis. The median domestic value mutual fund faced similar headwinds. Meanwhile, a relative factor neophyte — momentum (represented by the MSCI USA Momentum Index) — sprinted to a 30% cumulative outperformance since 2010.
Here we diagnose the past two decades of the value factor through the lens of MSCI equity risk models around the world. Model factor returns can be particularly helpful as they are designed to be “pure” and aim to neutralize exposure to all non-target factors. As we’ll see, the value premium was down but not out, and in fact persisted in most regions outside the U.S. Even within the U.S., performance varied by sector.
Value returns sagged in the US while momentum soared
Returns for the MSCI USA Enhanced Value and Momentum indexes are gross in USD. Returns for U.S. value managers are sourced from eVestment. Data was used only for those managers classified as mutual funds and with continuous monthly returns from 2010 to 2019.
Value around the globe
First, we explored whether value has declined in lockstep around the world and found that the performance of value has been very nuanced. The exhibit below shows factor returns across several regions for different definitions of value, since 2000.2 Compared to the decade from 2000 to 2009, most of the fall was in earnings yield — the ratio between a company’s earnings per share for the last 12 months and its current stock price — as opposed to the book-to-price factor – the ratio between a company’s book value per share and its current stock price.
Only in the U.S. did the return for earnings yield turn negative. In China, the earnings yield return improved over the past decade, and dividend payers in the emerging markets generally weathered the 2010s well.
Why have there been such wide differences between book-to-price and earnings yield? Book values are backward-looking and driven by a firm’s return on equity. Earnings yield, as defined in MSCI risk models, contains both historical and forward earnings expectations. The latter were high only for growth companies, as investors chased recent winners and bid up growth-oriented companies from specific industries, overlooking their valuations or earnings expectations. This focus on growth has largely left the earnings yield factor ignored.
Value performance varied across different definitions and regions
Factor returns are annualized. Earnings yield is part of the book-to-price factor in the Australian equity risk model. Dividend yield is part of the book-to-price factor in the Japanese equity risk model.
The decline in the U.S. was the outlier among the other regions. The next exhibit compares the value factor’s behavior in the U.S. across sectors, using MSCI’s U.S. sector risk models.3
Again, the fall in value from the 2000s to the 2010s was not consistent across market segments. During that time, value’s return flipped from positive to negative in consumer staples, technology and health care. Despite the falloff in the largest sectors, we found positive performance in materials, utilities and energy stocks.
Value performance differed across US sectors
U.S. equity sector models correspond to the Global Industry Classification Standard (GICS®) sectors, jointly developed by MSCI and Standard & Poor’s. Each sector model is estimated only from the constituents in that sector.
Software might not be eating the world, but it certainly snacked on value
The drop-off of value in the U.S. technology sector was stark compared to other sectors. Among all industries within technology, the high-performing — but expensive and thus underweight — software and services industry was the largest detractor to performance in the MSCI USA Enhanced Value Index during the study period.
Marc Andreessen noted in 2011 that “Software is eating the world.”4 While we can’t judge software’s dietary habits, there is ample evidence that software was a principal reason for the decline in value in the U.S. We trace the software and services industry’s active exposures to two value factors (book-to-price and earnings yield), along with the growth, profitability and momentum factors over the past decade.5
The book-to-price and earnings yield exposures indicate the industry looked consistently over-priced, and became more expensive on an earnings basis as the period went on.
Software and services industry’s active exposures
Then again, software and services was significantly, and consistently, more profitable than the rest of the broad market MSCI USA Index. In addition, its positive growth exposure — a measure of historical and future earnings growth — was nearly as consistent.
Despite their elevated valuations, the software and services industry outpaced the broad market by a cumulative 150% over the past decade and the momentum exposure increased commensurately. Given those exposures, many value strategies may have underweighted this industry, which in turn would have hurt their active returns.
Been here for years
While value’s recent underperformance appears stark, we also note that factor investing has historically been cyclical. The gap between the broad market MSCI USA Index and the MSCI USA Enhanced Value Index over the last decade was large, but it was less than half the 25% rolling 10-year shortfall in late 1999 during the technology bubble. And that underperformance was followed by a decade of above-market returns during the 2000s. (Of course, past performance is no guarantee of future results.)
It’s also important to note that value’s decline over the last 10 years was mirrored by momentum’s rise, and that factor behavior was not monolithic across region, sector or definition.
1Graham, B., and Dodd., T. 1934. Security Analysis. McGraw-Hill, New York. They first identified value investing as a style in 1928.
2Using multiple definitions provide a multi-dimensional assessment of a factor and improve model robustness and reduce estimation errors (Asness, C.S., Ilmanen, A., Israel, R., and Moskowitz, T.J. 2015. “Investing with Style.” Journal of Investment Management.)
3The value factor in the U.S. sector equity models comprises book-to-price, sales-to-price, cash-flow-to-price, dividend yield and trailing and forward earnings-to-price ratios.
4Andreessen, M. “Why software is eating the world.” Wall Street Journal, Aug. 20, 2011.
5We use the most recent MSCI’s Barra® Global Total Market Equity Model’s factor definitions.