Good morning. I spent much of last week in Venice, for the FT’s Business of Luxury summit. I am happy to report that only one attendee fell into a canal (he was fine), that the luxury industry remains in rude health (even if the wild post-pandemic boom is ending), and that Tintoretto has moved up smartly in my personal rankings of the greatest painters (Velázquez and El Greco still secure in the top two spots). Send me your top 5: robert.armstrong@ft.com.
The valuation mystery
Something happened to the valuation of US stocks about thirty years ago. Take, just for one example, Robert Shiller’s cyclically adjusted PE ratio (the S&P 500’s price divided by its 10-year average earnings). Through 1995 it looked like a well-behaved mean reverting series, where the mean was about 15. Since then it has looked like a somewhat more volatile but still mostly mean reverting series where the mean is about 28:
If you think valuations are an important if imprecise indicator of long-term returns (as I do) this is important, and vexing. Even if you are prepared to accept that something about valuations changed thirty years ago (as many value investors struggle to do) the fact that it changed demands an explanation. If you don’t understand this, how can you know that another change is not approaching, or is not already here? Granting there is a statistical relationship between valuations and long-term returns (some think there isn’t), if you don’t understand why the relationship changes through time, it would be foolhardy to use it for decision making.
I have been thinking about this problem without coming to any particularly firm or useful conclusion for an embarrassingly long time.
My best guess used to be that what changed was inflation. When valuations were lower, inflation was generally higher and, what’s more important, less stable. Equities are often thought of as an inflation hedge, but this is only partly true. High and unstable inflation (and all high inflation is unstable) makes life difficult for companies, dampens real growth over time, and can contribute to recessions. It makes sense that persistently higher inflation should be associated with lower equity valuations. Unfortunately for this theory, we’ve just had the biggest inflation spike in 40 years and valuations have only gone up. That does not mean that we need to dump the idea altogether. Inflation expectations have remained generally under control over the past few years, and for valuations it is almost certainly expectations that matter, because stocks are such a long-duration asset. Still, 10-year inflation expectations on the Cleveland Fed’s model were around 2 per cent pre-pandemic and they are around 2.5 per cent now, and valuations don’t care a bit.
Another possibility is that the last several decades have just been really strange. They have featured two epic bubbles, a once-in-a-lifetime financial crisis, a revolution in monetary policy and a global pandemic. Maybe it would be unrealistic to expect any market regularities to hold in such a whirlwind, and old patterns will reassert themselves eventually. There could be something to this, but looking at a 30-year period and declaring it anomalous is a little more than the social scientist in me can bear.
Shiller himself was bothered that for many years his Cape ratio has been very high while stocks rose ever higher. His solution was to adjust for interest rates. His “excess Cape yield” (ECY) is the reciprocal of the Cape multiple, expressed as a percentage, less the real return on 10-year Treasuries. As it is a yield rather than a multiple, a higher ECY means stocks are cheaper. And historically, the ECY does seem to do a pretty good job of signalling good and bad times to buy equities. That said, while at the time Shiller first introduced the ECY in 2020, it did make the S&P 500 look more reasonably priced than the classic Cape ratio, it no longer does, because rates have risen sharply and stock prices have risen too. Just like with the Cape, there are quite distinct pre- and post-90s ECY regimes. Before, the average ECY was about five per cent; now it is half that level (and right now, the ECY is at a revolting 1.2 per cent).
In a recent post the blogger and former equity strategist Jim Paulsen offered five interesting explanations for the shift in the Cape ratio:
-
Economic cycles have become more gentle: “From 1854 to WWII, the US was in recession 42 per cent of the time. Since 1980, it has had recessions only 11 per cent of the time. Arguably, when one of the biggest risks for stock investors, recessions, occur less frequently, valuations can and should expand.”
-
The composition of the market has changed, with faster-growing technology companies making up a bigger part of it: “Since the 1980s, the US has enjoyed a rapid pace of modernisation including fiber optics, mobile phones, the PC, the internet, social media, streaming services, and now AI . . . innovations have always received higher multiples as much of their value is based on future business growth . . . growth stocks have always received higher PE multiples compared to cyclical stocks.”
-
The market is more liquid now: “Not only has individual [market] participation improved substantially, but international investors have also expanded. Technology advances have also improved liquidity as electronic trading now makes investing so much easier. Similar to any individual stock, when liquidity improves, volatility diminishes, and valuations rise.”
-
“Profit productivity,” or real profit per worker, has risen. “Since 1940, there has been a close relationship — a correlation of +0.69 — between the stock market’s PE multiple and profit productivity. What is most striking about this relationship is when the stock market was in its old stable valuation range, profit productivity was also in a stable range. When profit productivity broke this range in 1990 and surged higher ever since, the valuation of the stock market also broke its old valuation range and remained higher.”
I think the first two ideas are compelling and interesting, and the third and fourth are not. Starting with the bits I disagree with, I can’t see how the difference in liquidity between now and 30 years ago is enough to explain a significant move in average valuations. Stocks are more liquid now, but you did have daily liquidity back then, and on very bad days — the days that really matter — liquidity is still a problem today. Also, it is worth noting that low-liquidity assets — private equity and debt — do not trade at much of a discount in today’s market.
On productivity, I agree with Paulson that productivity is the most important thing, in the sense that it is the most important driver of growth, for both economies and businesses. From an investor’s point of view, however, what matters is profit levels and profit growth, not their drivers. At best, an explanation of valuations based on higher productivity is reducible to an explanation based on higher growth. If productivity growth is good for valuations, it is because it means higher growth rates.
This leads us to Paulsen’s second point, that the market contains more growth stocks now. This explanation makes broad sense, but it makes me somewhat nervous, because I think growth rates are very hard to predict. Higher future growth in profits must, all else equal, be reflected in a higher valuation, if asset prices are the present value of future cash flows. And if we look at the current crop of huge, dominant growth companies and their powerful market positions it is easy to think that their growth rates will persist. History teaches us, however, that dominant positions in markets, including tech markets, can disappear rather quickly (General Electric, IBM, Nokia, et al). It may be that today’s tech companies enjoy semi-permanent “increasing return to scale” dynamics that will help their growth persist for longer than past companies’. But I’m just not sure.
This leaves Paulsen’s final point, that the economy is more cyclical now. I agree with Paulsen’s instinct that the very worst economic moments are the ones that matter most for investors. It is in recessions when companies are most likely to fail, when margin calls tend to happen, and when panic selling wipes out portfolios. But, again, have the last 30 years, with its bubbles and busts, really been so calm — calm enough to justify a doubling of valuations? Perhaps modern central banking has taken the hardest edges off the economic cycle, but again, I’m just not sure.
Valuation matters. The idea that what price you pay for a financial asset does not matter to the future returns from that asset is a non-starter. But how to best measure valuations, and how to best deploy those measurements in decision making, remains a very tricky question.
One good read
Rent control does not solve housing crises.
FT Unhedged podcast
Can’t get enough of Unhedged? Listen to our new podcast, for a 15-minute dive into the latest markets news and financial headlines, twice a week. Catch up on past editions of the newsletter here.
