The simple explanation for value’s underperformance

By LARRY SWEDROE

In a recent post for The Evidence-Based Investor, Value Premium RIP? Don’t You Believe It, I discussed value’s underperformance: the fact that the relative performance of value stocks in the U.S. has been so poor over the past 12 years that many investors have jumped to the conclusion that the value premium is dead. While it has not been the deepest drawdown for the value premium (that occurred during the Great Depression), at 12 years it has been by far the longest (the second longest was just about four years, and that occurred during the tech bubble in the late 1990s).

One explanation offered for the “death of value” is that the publication of the research led to overcrowding. The evidence presented showed that this is not the case, as valuation spreads have dramatically widened, the opposite of what you’d expect if there was overcrowding. The article provided evidence on why value isn’t dead, just dormant, showing that we have lived through such periods before, and thus the most recent performance was not unusual. It highlighted the fact that all risky assets go through long periods of underperformance. It also highlighted that when underperformance occurs due to relative valuations becoming cheaper, future performance is typically much stronger because spreads in valuation provide information on future returns.

Robert Arnott, Campbell Harvey, Vitali Kalesnik and Juhani Linnainmaa examined the cause of value’s recent underperformance in their December 2019 study, Reports of Value’s Death May Be Greatly Exaggerated. They began by noting: “There are many reasons why an investment style can suffer a period of underperformance. First, the style – or factor – may have been a product of data mining in the first place (i.e., it only worked initially because of overfitting). Second, there could be some structural changes in the market that renders the factor newly irrelevant. Third, the trade can get crowded leading to distorted prices and low or negative expected returns. Finally, the recent performance of a factor might simply be a result of bad luck and/or the factor plumbing new lows in relative valuation. If the first three reasons (among others) would imply that the style has stopped working and is not likely to benefit investors in the future, the last reason has no such implications.”

The evidence on the value premium, as presented in Your Complete Guide to Factor-Based Investing, is that it has been persistent over the long term, pervasive around the globe and across asset classes, is implementable (survives transactions costs) and is robust to various definitions (there’s a value premium whether the metric is price-to-earnings, price-to-book, price-to-cash flow, price-to-sales, dividend yield and so on). It is unlikely, therefore, that the premium is the result of data mining, overfitting or a random outcome.

To determine the cause of value’s recent underperformance, the authors decomposed the value-growth performance into three components:

1. A valuation spread: If growth stocks get relatively more expensive than value stocks, the process of value becoming cheaper relative to growth means that value will underperform growth. This is no different than what happens when stock valuations rise, impacting the equity risk premium. Over the period 1980 through 1999, valuations rose dramatically, increasing the historical (ex-post) equity risk premium but lowering the ex-ante (expected future) premium.

2. The migration of securities: Migration occurs when value stocks become more valuable, trading at higher price-to-book ratios, and migrate from the value portfolio to either neutral or growth. Migration also occurs when growth or neutral stocks become cheaper and drop into the value portfolio. Migration is a large and reliable contributor to the relative performance of value to growth.

3. Profitability: Growth stocks are more profitable than most value stocks. There could be a regime shift in relative profitability.

Arnott, Harvey, Kalesnik and Linnainmaa analysed the pre- and post-2007 period and found:

There is no significant difference between the migration of stocks (value to neutral or growth and growth to neutral to value) in the two periods, nor is there a significant difference in profitability.

Over 100 percent of the underperformance of value since 2007 is due to it becoming considerably cheaper relative to growth. Value cheapened by more than 4 percent per year, creating a performance shortfall of just over 3 percent per year. Note that in 2007, the valuation spread was narrow (22nd percentile); by July 2019, it had widened to the 97th percentile.

There is little difference in overall profitability of growth and value over the two periods. The average return on equity difference between value and growth is almost the same before and after 2007: -11 percent versus -12 percent.

The migration component easily overcomes the profitability advantage of growth stocks.

The bottom line, the authors wrote, is that “increasing valuations created a headwind that accounted for all of value’s losses over the past 12 years.”

Their findings led them to conclude: “Unless we choose to assume that the valuation spread between value and growth stocks will continue to widen indefinitely, value seems poised to outperform growth, perhaps by a startling margin (as has generally happened when it got as cheap as it is today).”

They added: “Even with a small mean reversion in relative valuations between growth and value, from 97th to the 95th percentile, value should outperform growth by 9% over the next year.” They also observed: “Even if valuations [were] to stay at the current levels the model would expect to see positive 5.1% premium.”

The authors did raise another important issue regarding how value is defined. Since we have moved from a manufacturing to a service economy, a simple measure of value (such as price-to-book) can be misleading. They explained: “For example, intangible investments (e.g. research and development, patents, intellectual property, and so forth) are presumably undertaken because they are expected to add to shareholder value. Yet these investments are treated as expenses and deducted from book value. This leads to many stocks being classified as growth stocks because they have tiny book values due to large investments in intangibles. Many of these stocks would be classified as value stocks if the expected capitalised value of the intangible investments added to the book value.”

This is why many have argued for the use of other metrics such as price-to-earnings, price-to-cash flow and EBITDA (earnings before interest, taxes, depreciation and amortisation)-to-enterprise value or multiple metrics (because some metrics work better in some industries than others, and some perform better in different regimes).

The authors added: “It makes sense to capitalise the investments in intangibles, to construct improved measures of company capital. Our empirical work shows that starting with the 1990s such improved measure increases average annual return of the standard HML factor [the value premium] by about 1.5%.”

For these reasons, many fund families have adopted the use of multiple value metrics and/or combining value metrics with such measures as profitability and quality.

Summary

Critics of value investing have declared the death of value. However, the empirical analysis suggests it is unlikely because the size of the current drawdown, while painful, is not an outlier, and the pre- and post-2007 data suggest there have been no statistically significant changes in the two drivers of the value premium — migration and profitability.

The simple explanation for value’s underperformance is the relative increase in the valuations of growth stocks.