Is Value Dead (Again)?

Is value dead? We believe the answer is no. There are certainly scenarios where we feel value could regain its mojo: namely, a rising inflationary and interest-rate environment that might reduce the attractiveness of longer duration (i.e., growth) assets. But whether or not that occurs, we believe value investors need to do a better job appreciating and valuing growth. The evidence we present in this paper does not suggest a once-in-a-generation opportunity in value, nor does it suggest that value is entirely broken and it is time to retreat.

27 SEPTEMBER 2018

Authors

Introduction

Value is, and always has been, an important part of our investment process. While we place significant emphasis on Information Flow1, intermittent periods of poor value performance generally make it more difficult to generate the level of alpha we hope to achieve. The internet and mega-cap bubble of the late 1990s, the Global Financial Crisis (GFC) in 2007 and 2008 and the broad struggle of value, particularly in the US, over the last few years are examples of such periods. This raises a question, indeed one that has been asked before: Is value dead?

Who better to ask than a firm specializing in systematic value investing?! In all seriousness, while we may be biased, we hope to take a more objective, empirical look into value. Why has value generally produced positive results historically? Why has value struggled lately? Is value still a useful investment approach? And on a related note, how should investment managers, asset owners and strategists respond to the recent challenges of value? Stay the course? Double down? Retreat?

Why Has Value Outperformed Historically?

A value strategy attempts to profit by going long undervalued securities and potentially hedging market exposure with short positions in overvalued securities. While there are many approaches to identifying cheap and expensive stocks (both quantitative and fundamental), most rely on one or more valuation ratios. A common premise across value managers, for better or worse, is that the market overreacts to information and growth prospects. Thus, lower-valued securities may offer more attractive returns than higher-valued securities.

Historical data seems to support this. For example, an internal Man Numeric model decomposes market returns into book value growth, changes in multiple on book value and dividend yields. From January 1964 to June 2018, the broad US market returned 11.3% annually, driven by 7.9% growth, 0.5% multiple expansion and a 2.9% dividend yield. However, if we look at the experience of growth and value stocks, we find very divergent experiences (Figure 1)2.

Over the entire period, value stocks outperformed growth stocks, and the margin was more pronounced in small and midcap stocks. The more interesting point is that growth stocks significantly outgrew value stocks, but were unable to overcome the effects of multiple contraction. In other words, changes in valuation multiples dominated differences in realized growth. While price-to-book is a very simplistic approach to valuation, this analysis suggests that one cause of concern for value investors would be changes in the relative magnitude of realized growth differentials versus multiple expansion or contraction.

Figure 1. 1964 - 2018 Value Attribution
    Ann Ret Book Growth ∆Multiple Div Yield
All stocks   11.3%  7.9%  0.5%  2.9%
Small Cap Growth  10.7%  29.5%  (20.4%)  1.5%
Neutral 15.2%  5.5%  7.5%  2.2%
Value  17.5%  (4.9%)  20.1%  2.2%
Mid Cap Growth  11.1%  29.6%  (20.4%)  1.8%
Neutral 13.7%  9.0%  2.0%  2.7%
Value  15.5%  (1.4%)  14.0%  3.0%
Large Cap Growth  10.8%  19.8%  (11.2%)  2.2%
Neutral  10.9%  9.8%  (2.0%)  3.1%
Value  12.1%  2.0%  6.3%  3.7%

Source: Center for Research in Security Prices (CRSP)3, Compustat.

When Has Value Outperformed? When has it Struggled?

Because value has historically tended to outperform in equities, let’s discuss environments where it has struggled. Broadly speaking, value has struggled during weakening economic environments (going into a recession) and growth-driven markets. Examples of weakening economic environments include the US recession of 1990 and the GFC of 2007 to 2008. Examples of growth environments include the Nifty Fifty of the early 1970s, the megacap and technology bubble of the late 1990s, and, dare we say, the recent market led by the FAANG stocks. The (generally) poor performance of value during weak economic environments is probably related to increased risk aversion; similarly, the performance of value during growth environments could be driven by over-extrapolation of firms’ prospects. It is pretty clear4 that market participants over-extrapolated the growth potential of the Nifty Fifty in the early 1970s and the Technology sector circa early 2000; it is less clear that this is the case today, with several of the FAANG stocks trading at, or sub, market multiples.

Using data from the Center for Research in Security Prices (CRSP) going back to the 1960s in the US and focusing on the largest 1,000 securities, we see difficult periods for value overlapping with the weakening economic and growth environments. Figure 2 shows the average monthly return on a 3-year rolling basis for raw and industry-adjusted earnings-to-price (E/P) and book-to-price (B/P) signals. After the drawdown related to the GFC, value staged a brief recovery, but has struggled over much of the last decade. In some ways, the last decade appears similar to the 1960s and early 1970s with low interest rates, low inflation and an increasingly concentrated market.

Figure 2. 1965 – 2018 US Top 1,000 Rolling 3 Year Spreads

Source: CRSP, Compustat, NBER.

If we look globally at an industry-adjusted E/P model (Figure 3), a similar pattern emerges. Value struggled during the technology bubble, followed by a strong resurgence until the GFC and has been fairly weak over the last decade. The only region where value has been consistently positive is emerging markets (EM). Value was productive in Europe up until the recent period, while Japan has arguably shown the biggest change over time. Japan used to be thought of as a market where value was consistently rewarded, and while we see that holding up until the GFC, it has been the least effective market for value over the last decade.

Figure 3. 1999 – July 2018 Industry Adjusted E/P Rolling 3 Year

Source: Man Numeric.

What Has Been Going On Lately With Value?

Outside of the short rebound following the GFC, value has produced subpar results over most of the last decade with a few brief respites, most notably in 4Q 2016. Interestingly, this difficulty occurred during a relatively positive period for global economic growth. Here, we examine several different possible explanations for value’s underperformance.

A) Has the relative magnitude of growth versus multiple expansion / contraction shifted?
Figure 4. 2010 – 2018 Value Attribution
    Ann Ret Book Growth ∆Multiple Div Yield
Small Cap Growth  11.9%  24.4%  (13.6%)  1.2%
Neutral  14.5%  2.4%  10.6%  1.5%
Value  14.4%  (8.9%)  21.9%  1.4%
Mid Cap Growth  14.5%  26.0%  (12.8%)  1.3%
Neutral 14.3%  6.2%  6.6%  1.5%
Value  13.6%  (3.1%)  14.9%  1.7%
Large Cap Growth  15.6%  22.5%  (8.8%)  1.9%
Neutral 14.1%  8.8%  3.2%  2.1%
Value  14.9%  2.7%  9.9%  2.2%

Source: CRSP, Compustat.

Focusing on the more recent period5 (Figure 4), we can see that there has been a subtle shift in the relative amount of growth and multiple changes relative to the longer period (Figure 1). In small-cap stocks, the spread in book growth was ~34% in favor of growth stocks over both the long and more recent history, while the spread in multiple expansion favored value stocks by 40% over the full period, but only 35% recently, producing a significantly narrower advantage for value stocks. In large-cap stocks, the value advantage entirely disappeared due to the diminishing of a substantial yield advantage and an increase in the spread of book growth. The change in relative multiple expansion was most pronounced in midcaps, where value stocks enjoyed a 34% advantage over the entire period, but only a 28% advantage over the recent period. While these seem like subtle changes in magnitude, the consistent nature by which growth and value stocks generate growth and multiple expansion, respectively, make these changes significant.

B) Are value companies undergoing more negative earnings drift than they have historically?

We calculate annual earnings drift as the difference between expectations at the beginning of a fiscal year versus reality at the end of the fiscal year. In the US, the experience of cheap and expensive stocks has varied over time (Figure 5)6. As the technology bubble imploded in the early 2000s, it coincided with significant negative earnings drift for expensive stocks, and this was a strong period for value. In fact, over the first half of this period, it was not uncommon for the annual drift of cheaper stocks to exceed that of growth stocks. However, over the last 10 years, the negative earnings drift of cheaper stocks has generally been on par with expensive stocks, and far worse in 2008 and 2009.

Figure 5. 1996 – 2017 Annual Earnings Drift – US

Source: Man Numeric.

These results appear to hold more broadly, as the US does not stand out in this regard. Globally, the annual earnings drift spread of cheaper versus more expensive stocks was less attractive over the recent decade, compared with the prior one (Figure 6).

Figure 6. 1996 – 2018 Annual Earnings Drift Cheap Minus Expensive

Source: Man Numeric.

C) Are value stocks producing poorer earnings growth outcomes relative to growth stocks than normal?

Historically, cheaper stocks have produced worse prospective earnings growth than expensive stocks (Figure 7). In fact, historically cheaper stocks actually had a difficult time producing any earnings growth at all, while more expensive stocks in some ways are justifiably so because they have shown the potential to produce meaningful earnings growth. Interestingly, the spread in earnings growth between more expensive and cheaper stocks widened significantly in the mid-1980s and has been relatively wide over the last 30 years, compared with the prior 20 years. This raises an interesting question which is difficult to answer: how much of the increase in spread is due to the market more efficiently pricing growth versus the increase in technological innovation, globalization and access to cheaper capital over the last 30 years? For value investors, the former is more worrisome, while the latter provides some comfort that value could prosper if some of those conditions abate.

Figure 7. August 1962 – December 2013, 3 Year Prospective Earnings Growth

Source: CRSP, Compustat.

Globally, outside the GFC, the last few years do not stand out as being particularly unique in terms of the spread in earnings growth (Figure 87). In the US, the spread in 3-year earnings expectations growth has been fairly stable, and is currently at low levels. The one region that stands out is Japan, where expensive stocks have materially outgrown cheaper stocks over most periods.

Figure 8. January 1996 – August 2015 3 Year Prospective Earnings Growth Spread Cheap – Expensive

Source: Man Numeric.

If we look at a 3-month window, we see a similar pattern to the longer-term outcomes (Figure 9).

Figure 9. January 1996 – April 2018, 3 Month Prospective Earnings Growth Spread Cheap – Expensive

Source: Man Numeric.

D) Are cheap stocks staying cheaper for longer? Is the increase in passive investing related to the decay of value investing?
Figure 10. July 1985 – January 2016 US Top 1,000 Industry Adjusted E/P Probability of Remaining Cheap or Expensive in One Year

Source: CRSP, Compustat..

With the exception of the GFC, it does appear that cheap stocks are staying cheaper for longer, while more expensive stocks are staying expensive for longer. Figure 10 shows the probability of remaining in the cheap (red) or expensive (blue) cohort from one year to the next. For the first 20 years, the cheapest stocks had 40% to 45% odds of remaining in that cohort the following year. But since the early 2000s, that probability has increased from 45% to 55%.

Figure 11. February 1996 – July 2018 US Top 1,000 Industry Adjusted E/P Probability of Remaining Cheap or Expensive in One Month

Source: Man Numeric.

Figure 11 produces the same conclusion using a higher quality dataset8 at a monthly frequency. The probability of remaining in the same cheap or expensive cohort from month to month rises gradually from ~85% prior to the GFC to ~90% over the last decade. Is it possible that the rise in passive investing in the US has altered the behavior of stocks? If the marginal investor putting money to work in the US is buying a passive broad market fund and not delineating on valuation, we might expect a higher propensity for stocks to remain cheap or expensive.

In EM, where a significant move to passive has yet to occur, we see no change in the autocorrelative behavior of value (Figure 12).

Figure 12. February 1997 – July 2018 EM Top 1,000 Industry Adjusted E/P Probability of Remaining Cheap or Expensive in One Month

Source: Man Numeric.

Lies, Damn Lies, And Valuation Spreads

While the performance of value is dependent on how accurately (or not) the market prices future growth expectations, it is also dependent on the dispersion of valuations. A wide dispersion of valuations may indicate a greater prospective opportunity for value. There are multiple approaches that can be taken to measure valuation spreads, but we will focus on E/P and price-to-earnings (P/E) spreads, which can give different signals depending on the overall level of market valuation. Higher market valuations lead to less dispersion in the E/P space, but more dispersion in the P/E space9.

Figure 13. 1962 – 2016
(A) US E/P Spreads Percentile vs Average

Source: CRSP, Compustat.

(B) US P/E Spreads Percentile vs Average

Source: CRSP, Compustat.

Figure 13 looks at long-term valuation spreads in the US using trailing net income from CRSP (and ending in 2016). We either compare the 80th percentile to the 20th percentile in valuation space, or the average of the best 20% versus the worst 20%. (A) shows E/P dispersion; both the percentile and average within buckets methodology suggest we are at low levels of dispersion (the latest observations are the 22nd and 26th percentile relative to history, respectively). But, one of the reasons E/P spreads are low today is because the market is near all-time highs on a valuation perspective; note in (A) spreads do not appear particularly wide during the technology bubble either. (B) shows P/E dispersion, which is admittedly more noisy because P/E is not a very well-behaved variable10. Putting that aside, we see that P/E spreads have generally been rising throughout this period and were relatively wide at the end of 2016 (81st and 80th percentile relative to history, respectively).

Today, E/P valuation spreads globally are somewhere between fairly wide in the US to normal or slightly narrow in Europe (Figure 14). In our traditional view of E/P valuation spreads (A and B), EM spreads continue to be wider than developed market (DM) spreads, although the spread was far narrower over the last 15 years than previously. In the US today, the 80th percentile earnings yield is 8.1%, while the 20th percentile earnings yield is 3.6%, leading to a 4.5% earnings yield spread. Industry-adjusted (A), that narrows to 3.5%, which is in the 72nd percentile of historical observations (wide, but not extreme). Similarly, if we take the average of the best and worst 20% earnings yields (B), we see a 6.9% spread in the US, which is in the 75th percentile of historical observations.

Figure 14. January 1996 – August 2018
(A) Industry Adj E/P 80pctile – 20pctile Spreads

(B) Industry Adj E/P Best 20% Avg – Worst 20% Avg

(C) Industry Adj P/E 80pctile – 20pctile

(D) Industry Adj P/E Worst 20% Avg – Best 20% Avg

Source: Man Numeric.

However, in P/E space, the valuation story in the US is a little more attractive; (C) and (D) show that valuation spreads today are as wide as they have been since the technology bubble, and relative to history reside in the 86th and 84th percentiles, respectively. The 20th and 80th percentiles for P/E in the US are 12.4 and 27.5 as of August, 31, 2018, respectively, and the average within the cheapest and most expensive buckets are 9.9 and 63.3, respectively. At long last, we have found a methodology that produces a positive view for value looking forward11.

Outside the US, the story for valuation is not quite as attractive. Japan has relatively wide spreads on an E/P basis, but average to below average on a P/E basis. Similarly, the view for Europe, EM and Global depends on what you look at. The value opportunity in Global is wider than the historical norm, but driven mostly by the US. EM valuation spreads are at historically average levels, but still significantly wider than DM spreads on an E/P basis. Note that they are in-line with DM on a P/E basis − the difference here is driven by the significantly lower valuation of EM relative to DM.

Where Does Value Go From Here?

Value investors, including ourselves, have been disappointed by value’s performance over the last decade. However, value has historically performed better after periods of poor performance: the mid-1970s, the mid-1980s, the early- to mid-2000s, and to a lesser degree after the GFC. We tried to determine why value has underperformed and how strong its prospects are going forward. As much as we would like to conclude that the future has never been brighter for value, the evidence does not unequivocally support this. At a high level, our conclusions include:

  • The market does a good job of delineating between value and growth stocks, but historically, changes in multiples overwhelmed realized growth;
  • The opportunity today for value appears better than the historical average, particularly in the US;;
  • Value has struggled over the last decade because the market has more accurately anticipated growth;
  • It is possible that the move to passive has reduced the efficacy of value, at least temporarily; and
  • Growth may have benefited disproportionately from technological innovation, globalization and access to cheap capital. A change in any of those conditions could potentially create a more robust environment for value.

So, is value dead? Shall value investors stay the course? Double down? Retreat? Every investor and allocator needs to come to their own conclusion. We believe that value is not dead, but we acknowledge that times may have changed. There are certainly scenarios where we feel value could regain its mojo: namely, a rising inflationary and interest rate environment that might reduce the attractiveness of longer duration (i.e., growth) assets. But whether or not that occurs, we believe value investors need to do a better job appreciating and valuing growth. The evidence presented here does not suggest a once-in-a-generation opportunity in value, nor does it suggest that value is entirely broken and it is time to retreat. Value will remain an important part of our investment philosophy and our potential to deliver alpha in the future, but we cannot stand pat with hubris. We must continue to learn, enhance our value and information flow signals, and strike the right balance in our process:

Appendix

Linking Growth and Profitability
  • Simplistically, the return of a firm can be decomposed into Growth (G), Multiple Expansion (M), and Dividend Yield (D)
  • Steady-state growth is governed by a firm’s dividend payout policy, financing choices and profitability

 

1. Our founder, Lang Wheeler, called Value and Information Flow the ying and yang of investing.
2. Analysis performed on the top 80% of the Center for Research in Security Prices (CRSP®) universe (virtually every stock listed on the broad American market) by market cap each month, covering January 1964 to June 2018 and assumes a one-year holding period. In each month, the universe is split into three equal groups based on P/B multiples to create Growth/Neutral/Value portfolios, respectively. Subsequent returns are based on market-cap weighted portfolios. Dividends are not compounded.
3. Calculated (or Derived) based on data from the Center for Research in Security Prices (CRSP®), The University of Chicago Booth School of Business.
4. Possibly with the benefit of hindsight, but countless market observers expressed skepticism in real-time.
5. From January 2010 through June 2018.
6. Within the US top 1,000 stocks. Cheap and expensive are the worst and best 25% of E/P, respectively.
7. Note Figures 7 and 8 suggest different magnitudes of the difference in the US but is likely due to Figure 7’s focus on backwards looking income while Figure 8 focuses on changes in earnings expectations.
8.Here we use forward earnings estimates which tend to be cleaner and more comparable than trailing net income.
9. Consider a market with a multiple of 10 and a cheap (expensive) stock trades at a 20% discount (premium). The P/E spread would be 12 – 8, or 4, while the E/P spread would be (1/8) – (1/12), or 0.042. Imagine the market, and these stocks now double in valuation. The P/E spread is now 8 (24 – 16), or double, but the E/P spread is now only 0.021 (1/16 – 1/24), or half.
10. For this purpose we are truncating P/Es at 100, and assigning negative earning companies to 100 as well.
11. Recall the title of this section is Lies, damn lies, and valuation spreads.

Download full article

Latest Research

Liquid PE alternatives allow investors to potentially achieve PE-like returns and risk exposures for lower fees.

John Lidington, CFA, Gregory Bond, CFA

In the context of the risks around China's A-Share market, active managers may be well-placed to add value.

Mickael Nouvellon, CFA, Ori Ben-Akiva

Important information

Opinions expressed are those of the author and may not be shared by all personnel of Man Group plc (‘Man’). These opinions are subject to change without notice, are for information purposes only and do not constitute an offer or invitation to make an investment in any financial instrument or in any product to which the Company and/or its affiliates provides investment advisory or any other financial services. Any organisations, financial instrument or products described in this material are mentioned for reference purposes only which should not be considered a recommendation for their purchase or sale. Neither the Company nor the authors shall be liable to any person for any action taken on the basis of the information provided. Some statements contained in this material concerning goals, strategies, outlook or other non-historical matters may be forward-looking statements and are based on current indicators and expectations. These forward-looking statements speak only as of the date on which they are made, and the Company undertakes no obligation to update or revise any forward-looking statements. These forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from those contained in the statements. The Company and/or its affiliates may or may not have a position in any financial instrument mentioned and may or may not be actively trading in any such securities. This material is proprietary information of the Company and its affiliates and may not be reproduced or otherwise disseminated in whole or in part without prior written consent from the Company. The Company believes the content to be accurate. However accuracy is not warranted or guaranteed. The Company does not assume any liability in the case of incorrectly reported or incomplete information. Unless stated otherwise all information is provided by the Company. Past performance is not indicative of future results.

1538896/US/GL/I/W

Please update your browser

Unfortunately we no longer support Internet Explorer 8, 7 and older for security reasons.

Please update your browser to a later version and try to access our site again.

Many thanks.