Make America Cheap Again
Is justified optimism baked into high US stock prices, or just elevated risk?
"Stock prices have reached what looks like a permanently high plateau."
Irving Fisher, Yale economics professor, quoted in the New York Times nine days before the start of 1929's Wall Street Crash.
"The concentration of value in the largest stocks has not been at these levels since 1929, prior to the Great Crash."
Charlie Morris, ByteTree Research, 5 January 2025
It should be no surprise to you that a lot of people think that US stocks are too expensive. But a lot of others believe that the high prices are justified by future prospects for earnings, particularly of the so-called "Magnificent Seven" (Mag 7) big stocks that sit atop the S&P 500 index.
To understand the situation, we don't just have to rely on subjective or biased opinions. There's a lot of data out there, and picking it apart can be highly instructive. That's my aim today, with a view to drawing some conclusions about the state of the US stock market.
Of course, I could just lay out a load of valuation multiples, showing that they're all at very high levels. But that would leave any conclusions open to derision by the "this time it's different" crowd, and claims that superior future prospects for profit growth justify the current "high plateau".
As you'll see, digging further into the detail indicates quite the opposite. Certainly in terms of trying to justify how the market got to where it is today.
The story of richly-priced US stocks extends much further than just the Mag 7. There's some really dull stuff with eyewatering valuations as well.
What's more, in aggregate, the data suggests that corporations within the S&P 500 have actually performed pretty poorly in the recent past (I analyse what's happened over 5, 15 and 18 years). This makes justifying the currently elevated index valuation multiples a bit of a head-scratcher.
In any case, when stock valuation multiples reach extreme levels, the result is that investor risk becomes highly elevated as well. If things are priced for perfection, then anything less than perfection will be a tragedy.
That risk can manifest itself in two main ways:
1. Something provokes a substantial market fall or crash, re-setting prices to cheaper - and thus less risky - levels. Depending on the size of the pre-crash bubble, and the size of the crash itself, it can take many years for investors to get back to even. (For example, it took about 15 years for investors in the NASDAQ 100 index to regain their losses after the tech bubble burst in year 2000. This tough lesson appears to have been completely forgotten by most market participants.)
2. The market muddles along, but long-term returns are extremely low. This means that scarce investor capital ends up being poorly deployed compared with the alternatives available.
In reality, the first scenario is more likely. That's because every bubble is prone to finding a pin, meaning anything that pops optimistic market sentiment. Examples include the severe downdraughts from 1929 to 1932, 1937 to 1942, 1972 to 1974, 2000 to 2002, 2007 to 2009, and 2022. If we look at things in inflation-adjusted terms, we should also include 1946 to 1949 and 1968 to 1982.
S&P 500 index since January 1928 (inflation-adjusted, log scale)
Source: Macrotrends
Valuation multiples are very high relative to market history
It terms of working out where things stand, the first place to look is valuation multiples. These are a short-hand way to express the value that the market places, at current prices, on all future earnings (or some other financial metric). That's when those earnings / metrics are expressed in today's money, taking account of the perceived returns that are required for the risks involved.
At the index level, I believe one of the most informative valuation ratios is the Cyclically-Adjusted Price-to-Earnings ratio, or CAPE. It's also referred to as the Shiller P/E and sometimes the PE10.
A standard P/E ratio just divides a company's stock price by its earnings-per-share (EPS). This gives the same result as dividing the whole company's market capitalisation at the current stock price by whole company earnings (net profits after tax).
The problem with the P/E is that it only looks at one year of earnings, and earnings can fluctuate significantly from year to year. In fact, in a market crash scenario when stock prices plummet, earnings sometimes collapse even more (or even turn into losses), even at the index level. In other words, a P/E can increase substantially even though a price has fallen sharply. The stock or index is a lot cheaper, but the P/E appears to show that it's more expensive.
CAPE gets over this by using earnings across multiple years. Specifically, it uses EPS from each of the past ten years, adjusts them upwards for inflation to the present day, and takes an average. This gives the "E" part of the P/E equation.
The idea is that it gives a sort of proxy for average corporate earnings power through one or two business cycles. Spread across a diversified index such as the S&P 500, the CAPE thus gives a pretty useful indicator of how high or low prices are in relation to the aggregate, smoothed earnings power of all the companies.
Here is a chart of the S&P 500's CAPE ratio since 1871.
Source: multpl
At 37.8, the CAPE is at one of its highest levels ever. It's above the elevated level of 32.6 just before the 1929 Wall Street Crash. It's well above the level of 27.4 in May 2007, when the market was riding high ahead of the Global Financial Crisis.
Also, it's only marginally lower than the brief October 2021 peak of 38.6, during the speculative silliness that occurred during the Covid pandemic. And, incidentally, when interest rates and bond yields were far lower than today. Ordinarily, we should expect valuation multiples to fall when interest rates rise, all other things being equal.
In fact, the only time the CAPE has been exceeded for any amount of time was between October 1998 and November 2000. That was during the speculative technology bubble at the time, which everyone definitely accepts was a bubble. The CAPE reached an all-time high of 44.2 in November 1999. The S&P 500 index entered a major, two-year bear market (prolonged crash, really) just nine months later.
The median CAPE ratio for the full time series is 16. Today's level is 2.36 times that (136% higher). Meanwhile, since 1971, the median CAPE has been about 20. Today's level is still 1.89 times that (89% higher). The period since 1971 is perhaps more relevant, since the US dollar was no longer pegged to gold after that year.
The upshot is that the S&P 500 index would have to fall by 47% just to get back to the post-1971 median level. The last time the CAPE was that low was back in late 2012. Such a fall would be very painful for investors, even assuming the market didn't overshoot to the downside.
Other index valuation multiples have also risen substantially in recent years, including the price-to-sales (P/S) ratio and price-to-book ratio (P/B). What's more, at 1.24% a year, the index's dividend yield is back around the extremely low level reached during the late '90s tech bubble.
I won't show all the detail of these ratios, in the interests of space. But they essentially tell the same story as the CAPE. On the face of it, US stocks look very pricey, at levels that are consistent with or exceed historical pre-crash conditions (1929, 2000, 2007, 2021).
In fact, by my reckoning, the current CAPE is in the 95th percentile of all monthly readings since 1971. That means it has only been higher in 5% of all the months. Looking back to 1881, it's in the 98th percentile, and has only been higher 2% of the time.
That implies a hell of a lot of optimism amongst the market mob.
Is the optimism justified?
I read a report the other day from a research outfit where they bragged that their 2025 EPS forecast for the S&P 500 indicates profit growth of +19%. There was a chart of EPS forecasts from 21 different outfits, including 10 big-name investment banks.
Most of the forecasts fell within the range of +11% to +18%. That compares with past growth of 9% a year over the past 15 years (more on that below).
In other words, the forecasts look pretty optimistic. However, in my experience, they're almost always too optimistic. That's because they leave out the inevitable negative surprises that are bound to happen at a fair chunk of companies within an index with over 500 constituents. Pretty much every year, there are enough negative developments to suppress over profit growth at the index level.
Of course, a lot of the optimism is centred on the Mag 7. So let's have a look at those, before taking a glimpse at the elevated prices of other mega-cap stocks.