What makes a bubble a bubble

Good morning. From the time we sat down to write this letter to the time we were half way done, martial law was announced and then called off in South Korea. The Korean won fell at first, but promptly recovered most of its losses. Another case of politics not mattering much to markets? Or are investors in South Korea especially prepared for turmoil? Email us: [email protected] and [email protected].

The America bubble, part II

Yesterday I wrote that US stocks were in a bubble, but that the bubble is more likely to expand further than to pop in the foreseeable future. Today, I’ll argue that this is a specifically American bubble, not a bubble in tech stocks that happen to be American. And then I’ll argue that saying “it’s a bubble that is not about to pop” is actually saying something, not just emitting a journalistic noise that has no actionable implications.

On Tuesday, we pointed out that there is a historically large valuation gap between US and European shares, even when the Magnificent 7 big tech stocks are excluded (I’m using Europe as a simple proxy for “cheap global stocks” here; the argument should be transferable). But it is still possible that the value gap between the S&P without the Mag 7 and the S&P Europe 350 is due to different sector weights. The S&P 493 is, for example, 17 per cent tech by market cap; the Europe 350 is just 7.5 per cent tech. This must be part of the explanation of the value gap. Of course if Europe had more large growing tech companies, its stock markets would be worth more. But it’s not all of the explanation. Below are the P/E ratios of the various industrial sectors of the US and European indices. I have treated the Mag 7 as a separate sector to make the comparisons cleaner:

Once you take the Mag 7 out, European tech trades at roughly the same valuation as US tech. But elsewhere in the index, especially where Europe is heavily weighted, US valuations are much higher. There are big gaps in consumer goods, energy, financials, industrials, materials and utilities (I’d ignore those high US real estate multiples, which seem to be an artefact of depressed earnings). There is more to the valuation gap than technology. 

The something more is mostly growth. Below is expected annual earnings per share growth by sector over the next two years (note that these growth rates, derived from “bottom up” analyst estimates for individual companies, are almost certainly a bit inflated, as company analysts have an incentive to keep their year-ahead estimates high):

In almost all the sectors with big valuation gaps, there is a gap of at least a few percentage points in expected near-term growth. Sometimes the gap is much bigger; European utilities and financials are not expected to grow earnings at all in the next few years.

The question, as we argued yesterday, is whether that growth gap, of perhaps 2-3 percentage points overall, justifies a historical valuation discount of 40 per cent (for markets outside of Europe the growth and valuation gaps will be somewhat different). If the US actually delivers 2-3 percentage points of growth over Europe forever, a 40 per cent premium is likely justified. But the US might not deliver that in 2025 and 2026 — that is, analyst estimates may be too high — and global growth rates seem likely to converge over time.

But just saying the valuation gap is too large in relative terms does not make a bubble. Bubbles are also characterised by high absolute valuations, and they have to be accompanied by very strong sentiment.

The absolute valuation box is checked. Robert Shiller’s cyclically adjusted price/earnings ratios, both adjusted for interest rates and unadjusted, are at the very top end of their historical ranges (though not quite as extreme as in 2000). The S&P 500 equity risk premium, as calculated by the valuation maven Aswath Damodaran of New York University, is at 3.85 per cent. When I interviewed Damodaran earlier this year, he said: “My red zone, usually when I start to get worried, is when [the ERP] hits 4 per cent and starts going below that.”

On sentiment, we may have a little way to go, but not much. Citigroup’s Levkovich sentiment index has moved into euphoria territory, but was actually briefly higher in 2021. The Bank of America asset managers survey shows that after Donald Trump’s election victory, the portion of managers who were overweight US stocks moved to an 11-year high. The proportion of respondents to the Absolute Strategy Research investor survey who expect US equities to outperform the rest of the world lept by 11 per cent, to 63 per cent, after the election. That is the highest level in the decade-long history of the survey. The AAII retail investor survey shows sentiment steadily improving over the past two years, and now sitting at the top of its cyclical range.

I posted a link to yesterday’s letter on Bluesky (this is called “bleeting,” I’m told). My friend George Magnus, who knows a thing or two about markets, scolded me as follows:

[The] best time to call time on a bubble is when it’s about to pop. Otherwise, you’re just a Cassandra w/o bearings. Surely people over-own and over-invest [in] the US partly out of lack of alternatives. Partly because US productivity and profits are on a roll.

It is a fair criticism. Saying there is a bubble that is not about to pop is almost like saying nothing at all. To mean something, the argument has to propose an action. Well, here is the proposal. Most investors, especially professional investors, cannot afford to significantly underweight the US right now, especially with the new administration set to turn every possible policy setting to near-term growth. Fair enough. There are good reasons to think the US bubble inflates for months or years to come. But surprises happen. I would argue that investors should have high cash holdings right now, perhaps double their usual levels.

This is not journalistic shadowboxing. In my tax-protected accounts (my retirement and my educational savings account), I have shifted heavily to cash, so that 15 per cent of my liquid wealth is in cash, on top of another 17 per cent that is in short-term fixed income. That’s a lot for a 53-year-old who needs high returns to retire. Real returns on cash of 2 per cent or so help only a little. That, and another 15 per cent in international stocks, does leave me with a lot of exposure to a US stock market I think is seriously overvalued from a long-term perspective (though a good part of my US exposure is in cheaper small- and mid-caps). I’m scared about this, and if US prices take another leg up, of say 15 per cent, I will sell the rest of my tax-protected US shares. This is not — and I can’t emphasise this enough — investment advice. But in response to George’s very fair point, I wanted to make it clear that I am making a substantive argument and living by it. 

One good read

The end of a ruse.

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.