A scary, but noisy, inflation report

Good morning. Birkenstock’s shares kept on falling yesterday, and we are starting to wonder if the shoe business is all it’s cracked up to be (see Dr Martens, Crocs and Nike’s recent performance). If you have a theory about footwear-specific investment risks, email us: robert.armstrong@ft.com and ethan.wu@ft.com. 

September’s CPI: looks bad, could be noise

On the face of it, yesterday’s consumer price index release was bad. Monthly core inflation moved up for the second month running, this time fuelled by a re-acceleration in rent inflation. Supercore inflation, or ex-housing core services, shot up from 0 per cent in August to 0.6 per cent in September. Markets didn’t like it. Yields surged, led by the long end, and stocks sold off. Presumably because small caps hate high rates, the Russell 2000 fell 2 per cent.

The three-month trend in core inflation has turned up:

The rent inflation uptick, from a monthly 0.3 per cent to 0.6 per cent, feels unsettling because moderation in this category has been such an important part of the disinflation narrative. Weren’t the CPI shelter measures supposed to be following the private-market rent indices like Zillow and Apartment List down by now? The idea was that as new leases pass into the CPI, which also includes existing leases, you’d see convergence.

We put that question to Omair Sharif of Inflation Insights, who replied that the Zillow and Apartment List measures “tell you directionality, but don’t take them as gospel for magnitude”. This is because the underlying lease data is different. The CPI runs a recurring survey of a consistent sample of leases, spread across neighbourhoods both shabby and swanky. The private-market measures, in contrast, tell you what is happening to the new leases that happen to show up in online listings, which tend to be higher-end. 

Sharif’s preferred leading indicator of CPI shelter is the Bureau of Labor Statistics’ recently released new tenant repeat rent index (NTRR). Like Zillow or Apartment List, this measure focuses on new leases, but unlike them, the NTRR uses the same lease data as the CPI. (One downside is that the NTRR is subject to big revisions.) The NTRR leads CPI shelter by about a year and judging by it CPI shelter looks roughly on track:

And given that rent inflation in August came in weaker than usual, it’s possible that the September surge is just a reversal from the prior month. The rise in supercore, too, looked noisy. It was driven by the volatile hotels category as well as an 8 per cent monthly rise in sports tickets. 

Still, though, it is hard to ignore how hot economic growth appears. The Atlanta Fed’s GDPNow measure is tracking third-quarter growth at a staggering 5.1 per cent, of which half is consumer spending. After months of everyone (us included) dismissing this measure as overreacting to early data, there are just two weeks to go until its final third-quarter reading. The average historical error at this stage is 1.2 percentage points. It would take a whopper error for third-quarter growth to be anything but strong.

That strength may be showing up in some pockets of inflation, such as hospital services (up 1.5 per cent in September) and apparel (0.7 per cent). Car insurance inflation remains strong, and used car prices, a big recent source of disinflationary impulse, seem to be rising again in wholesale markets. That could show up in CPI soon. Yesterday’s report was not a five-alarm fire, but a bit of market nerviness seems justifiable. It’s too soon to tell if the Fed is really done. (Ethan Wu)

The curious case of consumer staples

Over the past five months, the worst performing sector of the S&P 500 is consumer staples:

Utilities have performed almost as badly; no other sector is even nearly as bad. This is a little odd. The appeal of the staples group, which features food and drink (Kraft Heinz, Coca-Cola), basics retailers (Walmart, Kroger, Dollar Tree) and household goods (Colgate, Kimberly-Clark), is its stability. A little underperformance in an expansion, fine. But this is ugly.

There is a broad explanation that is of some help here. In 2022, when interest rates were ripping upward, the yield curve was inverting hard, and everyone was betting on recession, staples outperformed — a classic flight to safety. In 2023, as economic growth exceeded expectations and the soft landing narrative took hold, staples have simply given all that outperformance back. Here is a chart of the performance of staples relative to the market going back to the start of the pandemic:

This is not quite satisfactory, however. Yes, growth is strong, but it is hard to argue that we are in a typical early-cycle expansion where one would expect staples to underperform. Indeed, as the first chart shows, most of the damage has been done to staples when stocks overall are falling. And if the falling staples are about shunning defensive assets, why hasn’t healthcare suffered more?

Several Wall Street analysts have explained the downturn by noting that staples are rate-sensitive. The idea is that, especially among lower-income consumers, the rising cost of credit card and auto loans pinch the household budget. But wouldn’t this affect discretionary purchases more than toothpaste, toilet paper and groceries? There might be more switching to cheaper store brands. This might help explain why many of the weakest performers among staples have been branded food companies, from General Mills to Smucker’s. This is consistent with comments from the CEO of ConAgra Brands (Duncan Hines cake mix, Hunt’s tomato sauce, Birds Eye frozen vegetables, etc). On the company’s earnings call last week, he said that

After three years of unprecedented inflation, along with other macro dynamics, consumers have felt increased financial pressure . . . This resulted in a near-term reprioritization . . . with the notable exception of summer travel, discretionary purchases have been down almost across the board . . . Within food, convenience-oriented items, typically a top consumer priority have lagged as shoppers have turned to more hands-on food prep to get additional bang for their buck . . . a reduction in wasted food and an increase in the use of leftovers

That sounds like trading down. But then why are the dollar stores — which would presumably benefit from trading down — the worst staples performers of all? And again, why isn’t a pinched consumer showing up in consumer discretionary stocks such as, say, Domino’s, Amazon or TJX? And how do we make the thrift story fit with good economic growth overall, and a tight jobs market?

The food companies’ poor performance has also been put down to widespread use of the new GLP-1 diet drugs; a Walmart executive attributed lighter shopping baskets to them. But do Wegovy injections cause lower consumption of toothpaste, toilet paper and bleach?

If you can make better sense of what’s going on here than I can, by all means, send an email. 

Geopolitics and market, revisited 

A few days ago we argued that, while markets were hardly responding to the outbreak of war between Israel and Palestine, that wasn’t informative, because markets are bad at pricing geopolitical risk. Well, the market continues to be calm: oil and gold, for example, are moving sideways. 

My argument, simplified, is that markets can’t estimate the risks, so they ignore them. Other market observers argue that the market’s indifference is actually a well-calibrated response because, simplifying again, geopolitical crises usually blow over. The average crisis ends up not mattering much to most investors. And under most definitions of what constitutes a crisis, this is true. Here, for example, is a clear and comprehensive table from George Smith, a strategist at LPL Financial, of several dozen major conflagrations going back 80 years. Note the title: 

The “on average, things turn out OK” argument is as true as far as it goes. The point, however, is that markets are lousy at estimating which crises are going to metastasise, which is precisely what one would want to know. Note that several of the crises that the market “took in stride” turned out to be very painful for investors, as the fourth column above shows. 

The second problem with averaging results of past crises is that it smells slightly of the turkey farmer metaphor. In every past instance, says the turkey, when the farmer shows up I have gotten some corn, so why not expect corn this time? A perfectly good inference, until Thanksgiving day. 

There is another reason — I would speculate — that markets might not bother carefully weighing geopolitical risk. The associated professional risks are asymmetrical. Say I manage money, and reduce risk exposure because I believe a regional conflict creates big risks to my portfolio. If the conflict blows over and the portfolio underperforms, that’s clearly my fault. On the other hand, say I keep my risk exposure high and the conflict turns into a war. Well, in that case, lots of people lose money in wars, and whose fault is it, really? 

One good read

FTX’s seven alternative balance sheets, The New York Times test and Sam Bankman-Fried’s “very valuable” hair.