After a good run, value looks anything but exhausted

The 12 months to September 30 were the best for value relative to growth performance since the fall of 2001. It was a huge relief for those value investors who managed to survive the long period of growth ascendance since about 2007. Like many periods of value outperformance, however, it didn’t quite live up to the way that value investors had imagined such a year might be for a couple of reasons. 

First, while value stocks did outperform growth by a double-digit margin (MSCI ACWI Value beat ACWI Growth by 12.4%), value stocks didn’t actually make any money in absolute terms, falling by 4.4% in U.S. dollar terms (those of you who measure your returns in different currencies may feel a little better about value’s absolute performance). And second, it seems as if growth investing still got all the attention, even if in this case the attention was focused on the spectacular flameouts of a number of high-profile growth stocks and a few high-profile growth investors. 

Value’s strong performance seemed to be primarily about avoidance of really bad returns rather than any particularly notable successes. 

While long/short value strategies were able to generate double-digit positive returns, a quick look at the attribution shows that most of the returns came from the very negative performance of high-priced growth stocks our strategy was short.

Given value’s significant outperformance of growth this year, we have been asking ourselves several questions that we know are also on the minds of our clients. Is there still money to be made in the value/growth trade? Are there better and worse parts of the world in which to be investing with a value mindset today? How tightly do you need to define “value” to get to an interesting group of stocks? And what would happen to value if these stocks don’t revert to historically normal value spreads but stay at today’s levels indefinitely?

Value is still priced for significant outperformance everywhere

As for the question of whether the value opportunity is over, it seems pretty clear to us that the answer is no. Exhibit 1 is an updated version of the valuation of value chart I’ve been showing for the last couple of years.

Exhibit 1: Where are we on relative valuations?

This chart shows the valuation of the cheap half of the U.S. market against the expensive half, and it’s been renormalized so that the average valuation gap is 1.0. As of the end of September, value was trading at 0.72, which is the 11th percentile versus history. It’s certainly up from where it was a year ago, when we were at the 4th percentile versus history, but far from all the way back to normal. And the U.S. is certainly not an outlier compared to the rest of the world. Table 1 shows the percentile ranking for value versus growth in a variety of regions, sliced in a variety of ways.

Table 1: Percentile ranking of valuation spreads

While Europe is a notable outlier as the most extreme value spread in the world, we see a wider than normal spread of value in every region and in every different way we know how to slice value versus growth. The willingness to pay extreme prices for [supposedly] long-duration growth assets that benefited from low rates led to a massive growth bubble. Given the valuation extremes reached across regions and sectors, value’s recent outperformance corrects only part of the dislocation. For this reason, long value/short growth remains our highest conviction position.

Within the U.S., “deep value” (cheapest 20%) is the truly dislocated market segment

But while the halves of the market are a natural way to break things up from a style perspective, they can hide some important distinctions. In our Equity Dislocation Strategy we differ from the global style indices in a number of ways. We are much closer to equal weighting than capitalization weighting, we limit the size of our sector and industry bets between longs and shorts, and we focus much more heavily on the most extremely cheap and overvalued segments of the market rather than simply looking at the cheap and expensive 50%. Over 80% of the long and short sides of our strategy are in the cheapest and most expensive 20% of the universe, so it is natural for us to want to look at the quintiles of valuation rather than the halves. Exhibit 2 shows quintiles of valuation in the U.S.

Exhibit 2: Valuations ALUATIONS WITH A FINER-TOOTHED COMB

Quintiles of valuation in top 1000 U.S. stocks

While the five different lines of this chart make it a little busy, we can immediately see several interesting features of the U.S. market today. On the expensive end (noted by the dark blue line), the valuation of the very most expensive stocks has indeed come down relative to the rest of the market. They peaked out trading around 4 times the valuation of the average company and are now trading around 3 times. That is still one of the more extreme valuations of the last 40 years, scoring as the 88th percentile valuation versus history. The rest of the growth universe is less expensive relative to the market but, if anything, a bit more expensive relative to its own history, at the 90th percentile. 

On the value end of things, however, the story isn’t quite so simple. The cheapest 20% of the market (in red) is very cheap indeed, still trading at the 4th percentile versus history. But the rest of the value universe is much less attractively positioned, with the next quintile at the 70th percentile and the “middle” quintile of stocks that are neither particularly value-y or growth-y at just about its most expensive valuation in history. For the purposes of our strategy, nothing in this chart is particularly concerning. 

While the opportunity in the most expensive tier of stocks is less extreme than it was, it is still the group that seems worth focusing our short book on, whereas the long side gives us every reason to be focusing almost exclusively on this cheapest quintile of stocks. 

But the pattern within value is an intriguing one that does have some meaningful implications for long-only portfolios. Looked at in this way, it seems as if a value strategy in the U.S. should be avoiding the “shallow value” stocks that are mildly cheap relative to the market and focusing solely on the “deep value” quintile. But the version of this that is most relevant for a long/short that is close to equal weighting within the large cap universe is not necessarily the version you’d be most interested in for a long-only portfolio. 

But what we see across different weighting schemes is that the pattern is pretty similar whether we are using the weighting scheme most relevant to a long/short (at least the way we tend to run our long/shorts) or ones that are more relevant to long-only portfolios. The percentile ranking of the cheapest 20% of the market against the percentile ranking of the next 30% of the market is shown in Table 2.

Table 2: Valuation percentiles of deep and shallow value

Your portfolio may have minimal exposure to U.S. deep value today

The upshot to us is pretty obvious. If you want to take advantage of the attractive pricing of value stocks in the U.S., you need to be focusing on the very cheapest tier of stocks. The rest of value is actually pretty unattractively positioned. This does pose a bit of a problem for a lot of value-biased managers given the way they build their portfolios. Most of the U.S. value managers who managed to survive the long dark winter for value since 2007 did so by diversifying their portfolios beyond deep value.

Deep value stocks were both the worst performing cohort from 2007-2020 and the group that didn’t have particularly obvious compelling attractions beyond their cheapness. 

Your value portfolio would both have performed better and been an easier sell to clients if you blended value with other factors – value with a quality bias, value stocks that had some positive sentiment, or relatively inexpensive growth stocks (if allowed). Value-biased managers who had always built their portfolios in such a way would have outperformed “purer” value managers, as those other factors did significantly better than value in the 2007-2020 period. And value managers interested in not getting fired by clients annoyed with their performance would have been well served by moving to such a “multifactor” brand of value investing. At GMO we have long used a multifactor approach, albeit with a larger weight to value than many other managers. And we still believe that such an approach is the right way to build an “all-weather” value-biased equity portfolio that has a chance of keeping up even in an environment in which value is struggling. 

But today, I believe there is a good argument for putting together a more targeted portfolio that is not designed to do OK in all environments, but is instead tailored for the environment we find ourselves in where deep value is far better positioned than the rest of the value universe. 

Such a strategy would have done dreadfully in the value winter of 2007-2020, and I’m quite grateful that we weren’t running one. But now, when the forecast for value looks to be an extended period of clear skies, we have built a new U.S. equity strategy focused solely on the cheapest tier of stocks. (More detail on what the cheapest tier of U.S. stocks looks and acts like can be found in Part 2, below.)

Value is broadly attractive outside the U.S.

Interestingly, the pattern in the U.S. is not what we see in the rest of the world. Exhibit 3 shows the quintile breakdown of valuations in the MSCI EAFE and MSCI Emerging universes, where the attractiveness of value is much more uniform between “deep” and “shallow” value.

Exhibit 3: Brief tour of rest of world

In both of these regions, deep value stocks are very well positioned, but the next quintile of value is also quite attractively cheap. While the value opportunity is generally at least as good for these universes as it is in the U.S., the need to focus on deep value to the exclusion of other value-tilted stocks is much less obvious.

Value is positioned to win, even if it remains cheap

All this analysis of the attractiveness of value shares a strong underlying assumption: value is most attractive when it is trading at a larger than normal discount to growth stocks and least attractive when that discount is small. For anyone who believes that these valuation discounts tend to revert to historical averages, such an assumption seems obviously appealing. But what happens if such reversion does not occur? What if today’s abnormally wide discount is permanent? Can value outperform? We believe the answer is a resounding yes, driven by a factor that is often overlooked by investors when they think of value (or indeed growth) as an investing style. That factor is rebalancing, and it can help explain not only why value is such an attractive investing style today but also why it was such a dangerous parameter to focus too heavily on from 2005-2015. Exhibit 4 shows a breakdown of the components of return for value relative to growth in the U.S. in both the 1983-2006 period when value performed quite well and in the 2007-2021 period when value lost out to growth. 2 While the performance of value versus growth was quite different in the two periods, most of the components of return showed strong continuation.

Exhibit 4: Total returns for value versus growth

The single biggest difference between the first period and the second was the valuation component. From 1983-2006, value stocks on average got more expensive relative to growth, whereas from 2007-2021 they got much cheaper. If valuation spreads were to remain constant, that factor would definitionally be zero, and the other three components would determine whether value would win or lose. The first two factors are the obvious ones we think of when we contemplate value and growth investing – growth and income. Everyone knows that growth stocks grow faster than value stocks, and everyone knows that value stocks spin off more income than growth stocks. In this case, what everyone knows is absolutely correct. But it is also incomplete. 

Growth stocks do indeed grow faster than value stocks, to the tune of 7.5-8% in an average year, and value stocks offer investors more income – whether in the form of dividends or net stock buybacks – to the tune of 2.5-3% per year. 

The problem for value is immediately apparent, because 7.5-8% is a lot bigger than 2.5-3%. The extra income from value stocks does not come particularly close to making up for the extra growth that comes from growth stocks. Happily for value, that is not the end of the story, because there is an additional component of returns that is invariably positive for value relative to growth – rebalancing. Rebalancing is a bit of a catch-all term for all of the ways that the constituents of the growth and value universes change over time. The single largest piece of rebalancing is the swapping of stocks between the value and growth universes. This naturally happens whenever a growth stock drops in valuation enough to fall out of the growth universe or a value stock rises in valuation enough to graduate to the growth universe. While this type of turnover isn’t immensely large in the average year, it is almost invariably material. Exhibit 5 shows the percent of turnover between value and growth in the U.S.

Exhibit 5: How many exits/entries are there?

Style migration probabilities (%)

This series does bounce around a bit from year to year, but the average rate of migration is actually quite stable. I tried to engineer as large a gap as I could between the earlier period of value outperformance and the recent period of growth outperformance, and the largest gap I could come up with was a drop from 11.1% on average from 1983-2010 to 10.3% from 2011-2022. Even in the lowest years on record, this migration was in the 8-9% range.

This migration provides a large boost for value because the process of moving from value to growth is a positive one, whereas the process of moving from growth to value is invariably painful for investors. The size of the benefit for value and pain for growth is driven by the gap between the valuation of the exits from the value or growth universe and the valuation of the entries. Exhibit 6 shows a scatterplot of the size of this gap against the ratio of the valuation of value to growth stocks that we used in Exhibit 1.

EXHIBIT 6: VALUATION SPREAD DRIVES RETURN FROM REBALANCING

As the dashed regression line shows, there is a pronounced tendency for this exit/entry valuation gap to be larger when the spread of valuation between value and growth is larger. This relationship makes plenty of intuitive sense. If value stocks are trading far cheaper than growth, a growth stock falling in the value universe is likely to have fallen quite far in valuation, whereas a value stock that moves into the growth universe will have gotten a big valuation bump. 

The size of the expected gap varies widely with the spread of value. We can use the regression line to see the impact of different discounts for value in Exhibit 7.

Exhibit 7: Entry/exit gap at today's valuations

At a normal valuation spread of 1.0, the expected exit/entry gap is 46%. At today’s spread, that gap is much larger at 89%, and in the tight spreads in the middle of the first two decades of the 2000s, it should only have been around 31%. Translating from the size of the gap to the size of the expected rebalancing effect, we can estimate that rebalancing at an average valuation should be 7-8%, at a tighter spread of 1.1 it should be only 5-6%, and at today’s spread it should be 11-13%. 

To put it slightly differently, if valuation spreads are constant and value is trading at 1.1 times its average valuation, you would expect value to perhaps break even versus growth or even underperform slightly. At an average spread of 1.0, value would be expected to beat growth by 1-3% at a constant spread. And if spreads were to stay stable at September’s valuation of 0.72 times average, the expected return from value relative to growth should be in the range of 6-9%. While value will make money more quickly if valuations revert higher – as they indeed did over the past year – at a constant wide discount, a typical value portfolio could hope to beat the market by 3-4% per year, and a value/growth long/short portfolio could deliver a lovely 6-9% return on top of cash forever. Call me greedy, but I would love to see a world in which value stays at today’s spreads for at least the rest of my career. I’d have less to write about but managing money would at least be easy.  

A sliver of companies offer excellent opportunities 

Part II of this two part series, which looks at the stocks offering compelling value, will be released on Tuesday morning. To be the first to read it, hit the follow button below..