Value Does Just Fine in Recessions

As a result, we have no single answer to the question, “How does value do in a recession?” But honestly, the nuances of different value models don’t matter much in answering that question, since pretty much every way of defining value has done just fine across most recessions. ³ The only recession in which value models did notably badly across the board was the Covid recession of 2020, and that event was so utterly unique that I’d argue it was a different sort of event entirely.

Certainly, the next recession – whenever it comes and however it transpires – is much more likely to look like one or other of the prior recessions of the past 50 years than the COVID crisis, a rare event that completely shut the economy for on-premises businesses (think value) while simultaneously launching online companies and enablers to the moon (think growth). So I believe it is more instructive to look at the performance of value across the range of recessions we’ve seen over the last half century.

Historical Performance of Value in Recession

Exhibit 1 shows the relative performance of the cheap half of the market in every U.S. recession since 1969 on four different valuation models: price/book, price/earnings, Composite Value, and the blend of value models that we use in our Opportunistic Value strategies.

We would not recommend building value portfolios on the basis of standard price/book or price/earnings ratios. But even if an investor insisted on doing so, on average they'd have done decently well across recessions of the last 55 years. In fact, every version of value besides price/book performed better in an average recession month (even including Covid) than in a non-recession month since the end of 1969!

That analysis is for broad value and today we have been arguing that among U.S. equities it is only deep value stocks that are really interestingly priced. One could imagine that the cheapest tail of the market dials more heavily into recession risk, but that simply isn’t the case, as we can see in Exhibit 2.

Once again, apart from Covid, there isn’t a particular pattern of trouble across recessions for deep value. Let me emphasize this point: despite the common narrative about value in economic downturns, there is simply no empirical evidence to suggest value stocks consistently underperform during recessions. The cheapest price/book stocks did have quite a rough time of it in the Global Financial Crisis, but for the lowest quality way to define value one can readily imagine, ⁴ one notably bad outcome across seven recessions is substantially better than one would have naively expected. For versions of value that pay more attention to quality considerations, there were simply no noteworthy negative performances for value in U.S. recessions other than during Covid.

For those of you less uniquely oriented towards the U.S. stock market and economy, the pattern is not particularly different in the rest of the world. Exhibit 3 shows the performance of the cheap half of the market in recessions for the four largest developed economies outside the U.S.

While the very impressive performance of value in Japan is a bit of a standout, even the worst of these country/parameter pairs – cheap Composite Value in Germany – only underperformed by 2.6% in the average recession.

Moving into the cheapest 20% of securities, which we find quite appealing today in all these markets, is actually even more reassuring, as we can see in Exhibit 4. Even the 2.6% average loss for cheap Composite Value in Germany has turned into a 2% gain instead.

I’ll admit to finding it slightly odd that the larger quality bias in the blend of models we use in Opportunistic Value was a negative influence within European markets, but in no market and in no model was the performance of value in recessions particularly scary, and across the average market all the value models outperformed in recessions.

Why doesn’t value underperform?

So why doesn’t value struggle in recessions when it seems intuitive that it should do poorly? We think there are a couple of places where investors are misunderstanding value and growth that are relevant here. First, I think there is an assumption that value stocks are uniquely sensitive to the business cycle and that they are hence much more likely to disappoint investors in a recession than growth stocks are. This turns out not to be the case, as we will see looking at value traps and their counterparts in the growth universe, which I am attempting to get the investing world to recognize as “growth traps.” ⁵

Everyone has a good idea of what a value trap is, and investors are used to grilling their value managers on how they go about trying to avoid them. We came up with a definition of value trap that allowed us to measure just how problematic the issue is. Since a value trap is generally thought of as a company that looked cheap but proved to be less cheap than the investor thought because its fundamentals deteriorated, we define a value trap as a company that disappoints on revenues in a given year and also sees its future revenue growth expectations fall. We choose to use revenues rather than earnings for this purpose since it can be hard to know exactly what version of earnings analysts are forecasting, but the basic idea is simple. We are looking for companies whose fundamentals wind up disappointing, and where that disappointment is not merely a short-term problem but reflects a downgrading of the future prospects of the company that is at least as severe as the near-term headwinds. On average, about 25% of the value half of the market ⁶ is a trap in a given year. But as we see in Exhibit 5, ⁷ the percent of the value universe that winds up being a trap is not constant over time and generally rises in recessions (shaded in gray). ⁸ Exhibit 5 also shows the analogous group in the growth universe. Value stocks are not the only stocks that disappoint their investors. As we can see from the orange line on the chart, growth stocks on average are just as likely to meet our definition of traps as value stocks are. Furthermore, the shaded sections of the chart show that growth stocks see their odds of becoming a trap rise in a similar fashion in recessions.

We see spikes in value traps and growth traps in all three recessions. The last year has been a really interesting case as well. The U.S. does not seem to be in a recession at the moment, but the vast majority of growth stocks have both disappointed on revenues in the past 12 months and seen future revenue growth estimates come down. Each of the top 10 growth names in the U.S. were officially traps in the last year, although the scale of the issue varies by company. Apple disappointed by 4% on their revenues and saw their future growth expectations fall by 4% as well, whereas Tesla disappointed about 9% and saw its future growth forecast come down by a massive 34%. The huge prevalence of growth traps last year does call into question the common narrative about the 2022 growth carnage. Rather than being a case of growth stock valuation premia coming down due to rising interest rates, I’d argue that an awful lot of the pain was from growth stock valuation premia coming down because the growth stocks proved less growthy than investors had hoped.

At the risk of banging the same drum I have been on for some time: though value traps seem to be a much bigger concern for investors than growth traps, growth traps are every bit as prevalent (actually, they are slightly more so). And what’s more, they turn out to be much more damaging to investors when they occur. Exhibit 6 shows the performance of value traps and growth traps relative to the rest of the value and growth universes.

I am not claiming that value traps are not painful. They are extremely painful, and we see in Exhibit 6 that they underperform the rest of the value universe by more than 15% in an average year. They are a group well worth avoiding, and feel free to continue to pester your value managers about how they go about trying to avoid them. But growth traps are a whole lot worse, and you can see in Exhibit 6 that they underperform the rest of the growth universe by a stunning 22.9% per year. At some level, this does make sense. While value companies are capable of disappointing investors, and any company that disappoints will generally underperform, when a growth company disappoints, it means not only that their fundamentals are worse than expected, but that they are less worthy of being considered a growth stock in the first place. Since growth stocks trade at a premium to the market because investors believe they will grow strongly in the future, an event which calls that future growth into question is likely to be particularly painful since their valuation premium deserves to fall. For our purposes here, the basic point is a simple one. Value stocks and growth stocks are basically equally likely to disappoint in a recession and when growth stocks disappoint, they are generally going to be punished a lot more.

I’ll admit that the trap percentages might not tell the whole story about the vulnerability of value stocks in recessions. While similar percentages of value and growth stocks might disappoint in recessions, and we know that on average when growth stocks disappoint it hurts more, it might be the case that when specifically looking at recessionary periods, things are different. Since I can only take the trap data back to 1996, I’ve only got three recessions to look at. In the 2001 and 2008-2009 recessions, value traps and growth traps underperformed the rest of their cohorts by 19.8% and 24.1%, respectively. Once again, the Covid recession of 2020 proved to be the outlier, with value traps underperforming by 18.5% versus 12.2% for growth traps.

Value profitability and cyclicality is in line with the market

Still, I hear some of you saying, “But surely value stocks are more cyclical than growth stocks, and that must wind up hurting them in recessions somehow.” Here we need to be a little more careful about how we define our value stocks. If we define value on price/book (even after correcting book value for its various distortions with a measure we call GMO’s price/economic book), value is a lower-profitability group than the market on average, and it is noticeably more cyclical, as you can see in Exhibit 7. ⁹

On average, the cheap half of the market on either measure has a return on capital ¹⁰ about 1.5% worse than the market as a whole, and the gap grows to between 2.0% and 2.5% in recessions. In the GFC of 2007-2009, the gap was as much as 3.6%, which probably goes some way toward explaining why cheap price/book did as badly as it did in that event (i.e., the most severe recession in the U.S. since the Great Depression). But not all definitions of value are as junky as price/book, which is by a decent margin the worst of the lot.

Exhibit 8 shows the same data for the Composite Value model and the blend we use for U.S. Opportunistic Value

That substantial profitability gap we saw for price/book is entirely absent here. Composite Value has an average profitability decrement of -0.3% relative to the market, and the blended model we use in our Opportunistic Value portfolios almost precisely matches the market. And measured at the profitability low of each recession in our data, the difference is a profitability decrement of -0.4% for Composite Value and again zero on the blend. In other words, these value measures have almost exactly the same amount of cyclicality as the market and very similar levels of profitability in the first place.

When we tighten our definition to the best 20% of value on these measures, there is a little more cyclicality to value, but it’s only a couple of tenths of a percent of return on capital, as you can see in Exhibit 9.

If you squint very hard, you could make the case that deep value is slightly more cyclical than the market, but the benefit of low expectations for these companies seems to more than compensate for any mildly greater deterioration in profitability during the average recession.

Conclusion

The common assumption that value is destined to underperform in a recession simply does not fit the historical data. I’ll freely admit that value did strikingly poorly in the Covid crisis period, but when we widen our view to look at other more typical recessions, we see that the Covid event was very much the exception rather than the rule. Even in the GFC, only the junkiest definition of value substantially underperformed, with other definitions holding up decently well and a couple notably outperforming the market. The reasons for this are fairly straightforward. Most thoughtful metrics for defining value have similar profitability characteristics to the broad market. While plenty of value stocks will end up disappointing investors when a recession comes, historically they have not been any more likely to disappoint than their growth counterparts. It is the disappointing growth stocks that are the biggest headache for investors, since their disappointments call into question their premium valuations. Value stocks, by contrast, have the charm of low expectations. No one is expecting all that much from them, so they have less to lose in an economic environment in which companies of all stripes wind up having a tough time. In a market that seems fraught with risk, deep value stocks are underheld by most investors, despite their extraordinary valuations. As we reference above, we believe those holding back due to recession fears should take some comfort in history and have a second look.

¹ Equity Dislocation is long deep value and short the most expensive growth stocks, globally.

² U.S. Opportunistic Value captures the deep value (cheapest 20%) cohort within the U.S, and International Opportunistic Value does the same across non-U.S. developed markets.

³ Note the word “across.” Equity owners abhor the negative economic and earnings growth associated with recessions. This leads to price declines in and around them. But the timing of declines and recoveries varies across formally defined recessionary periods. As we will see below, however, value relative to the market has on average modestly outperformed across recessions measured beginning to end. There is also a reasonable question as to how value would perform in a true “Depression” as opposed to a standard recession. We don’t have great data back far enough to look at the performance of value stocks in the Great Depression, but on the Fama French data on Kenneth French’s website the cheap 30% of the market on book/price broke even with the broad market in the August 1929-March 1933 period that comprises the Great Depression. Given that that means it fell 80%, it is questionable whether anyone should take a lot of solace in the fact, even if it is a reasonably good showing for that group relative to the market.

⁴ “Lowest quality” in this case means the way of defining the value group that leads to a group of stocks with the lowest quality ranking versus the broad market. Our quality ranking looks at average return on capital across the economic cycle, stability of return on capital, and level of leverage to determine high- and low-quality companies.

⁵ I’ve been banging on about growth traps a fair bit over the last couple of years (notably in Dispelling Myths in the Value vs. Growth Debate, and Growth Traps Snap Shut) but I haven’t been able to get the term into the general lexicon. We define a growth trap analogously to a value trap in that it is a growth stock that winds up disappointing on current results and marking down future expected growth.

⁶ For those keeping score at home, I’m using the GMO Composite Value model to define my value (and growth) universe. The data would look substantially similar on other valuation models.

⁷ This is a slightly revised definition of traps from what we have used previously. The difference from the previous version is in the year-two revenue estimate. We are now using a slightly tougher hurdle to be considered a trap, where a company must both disappoint on revenues in a 12-month period and also see its future revenue growth expectations rate fall for the following year. This reduces the average trap percentages to 25% for the value universe and 26% for the growth universe.

⁸ Those recession dates are shifted by a year from the actual recession dating since a company is only a trap when it winds up disappointing on sales over the next 12 months. A recession period trap would therefore be a trap in which the reporting period includes the recession.

⁹ If you look at the two profitability series for the price/book portfolios you might wonder why we bother correcting price/book in the first place, as they seem to be more or less on top of each other. Despite that similarity, however, a price/economic book portfolio has done substantially better over the past 40 years. More to the point, however, price/economic book helps to both avoid screwy return-on-equity estimates and helps distinguish expensive companies substantially better than unadjusted book. Both groups, however, are pretty junky when you get to the cheap end, as one might expect.

¹⁰ Our return on capital measure is a blended profitability measure on several different proxies for estimated economic capital. The series would not look meaningfully different if we were looking purely at ROE, even though ROE is pretty screwed up by this point for a lot of companies.

Disclaimer: The views expressed are the views of Ben Inker through the period ending June 2023, and are subject to change at any time based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security and should not be construed as such. References to specific securities and issuers are for illustrative purposes only and are not intended to be, and should not be interpreted as, recommendations to purchase or sell such securities.

Copyright © 2023 by GMO LLC. All rights reserved.

A message from Advisor Perspectives and VettaFi: To learn more about this or other topics, please check out our most recent white papers.

© GMO

Read more commentaries by GMO