Making US stocks cheap

Good morning. Ethan here; Rob is away. Another sleepy day on the street. Inconclusive debt ceiling talks seem not to be moving the market one way or another. With the S&P 500 still up 10 per cent this year, in the past few weeks I’ve noticed an uptick in bulls saying that maybe, just maybe, things will be fine. Are you among them? Email me: [email protected].

Can we make US stocks look cheap?

Last week I wrote this about the US stock market:

Standard p/e’s are slightly above historical averages, the Shiller cyclically adjusted p/e ratio is very high and the equity risk premium is near a decade low. If there is a valuation metric that makes the market look cheap, we haven’t seen it.

Several readers wrote in to suggest other places to look, and there’s a useful exercise here in understanding where valuations stand. That is, how long must we tickle the data before the US stock market appears cheap?

Start by comparing different S&P 500 valuation multiples, including price to cash flow, which a couple readers suggested is the superior historical measure. The chart below shows five ratios as a percentage of their three-decade historical average. Greater than zero per cent suggests that US stocks are more expensive than recent history, while below zero means cheaper. Some, like basic p/e’s, do look historically average, rather than expensive. But none looks cheap:

Measures based on enterprise value, like EV over ebitda, look much the same. Bloomberg’s data only runs through the 1990s, but if someone has a longer-running analysis, do email me.

The equity risk premium, the extra dollop of compensation over risk-free bonds that stock investors demand, offers another lens on valuations. The long-term chart below, courtesy of BCA Research’s Irene Tunkel, stirs together various methodologies for calculating the ERP. It confirms that the ERP is near a decade low but, as Tunkel pointed out to me, it looks bang on the historical average:

Might cheapness lurk below at the sector level? Mostly, no. Ian Harnett of Absolute Strategy Research sends over this slide of the S&P 500; the upper-rightmost chart breaks out p/b ratios by sector. With the exception of media and maybe banks, all sectors are trading above or around their 10-year average:

The next place to look is beyond the S&P 500. The Russell 3000, which captures almost all listed US stocks, looks a lot like the large-cap S&P:

Perhaps large caps are the problem. It’s been widely noted that the top slice of US stocks is dominating the market, both in terms of size and uncommonly strong performance. One much-discussed data point: Apple is now bigger than the entire small-cap Russell 2000. And indeed, unlike its large-cap cousin, the S&P 600 small-cap index does look cheaper. Across valuation measures, the index sits below the historical average:

Here, at least, is one significant pocket of cheapness, on top of media and banks, though these alone hardly make the broader market seem cheap.

But what if we’re going about this all wrong? One reader, Martim, sent over this recent post by Yves Bonzon, chief investment officer at Julius Baer, arguing that S&P 500 valuations are close to fair value. His point is that comparing S&P 500 valuations to history misses the fact that, overall, large-cap US companies have gotten better at making money and, crucially, handing it to shareholders:

Over the past few decades, S&P 500 index companies, in aggregate, massively increased their ability to generate free cash flow. The ability to return capital to shareholders therefore also increased: the combined amount of dividends and share buybacks as a percentage of revenues climbed from 4.7 per cent, on average, during the 2000s to as high as 8 per cent towards the end of the decade.

Moreover, investors pointing to an overvaluation of the S&P 500 are implicitly discounting a mean reversion of the secular surge in the free-cash-flow metrics of the index. This is unlikely, as it would require both a complete reversal of corporate tax rates back to 40 per cent and a fully-fledged reversal of globalisation.

In stark contrast to large caps, the profitability metrics of smaller companies have deteriorated dramatically over the past 20 years. Moreover, the financing structure of listed small caps is much weaker. According to a study by Empirical Research Partners, almost 40 per cent of their total outstanding debt is rate-sensitive, compared with only around 20 per cent for large caps.

Put another way, a stock’s value is just its discounted future cash flows. So if holding large-cap stocks simply earns you more cash now, and will continue to for the foreseeable future, then it’s rational to apply a higher multiple. A historically expensive multiple starts looking fair, and an average one starts looking cheap.

While Bonzon puts his argument in terms of cash flow generation, it’s closely related to Unhedged’s recent discussion of margins (see, eg, here). Our problem with this thinking is that the “secular surge in free cash flow” really might not last. As we wrote about in April, the economist Andrew Smithers points out in his recent book that pre-tax corporate profit margins as a share of output are both mean-reverting across history and particularly high now:

If I’ve missed any important valuation measures, I’m interested to hear from readers. But the punchline is that data torture alone can’t make stocks look cheap. You need to believe that something fundamental about corporate profitability has changed, and for good.

One good read

China’s debt bubble may not end with a pop, but a slog.