The Fed’s job is getting harder

Good morning. Yesterday a movie reference Caddyshack caused many letters of approval. Does that mean most of our readers grew up in the 1980s, or do they just have great taste? Send other market-relevant movie quotes to: robert.armstrong and ethan.wu.

The time for compromise has arrived

We have been complaining for the past few months about confusing, contradictory or equivocal economic data coming out of the US economy. But recent reports, in a refreshing change, have mostly sung from the same anthem.

Wednesday’s CPI numbers were, overall, encouraging. A second consecutive month of falling housing inflation was particularly welcome. Yesterday’s producer price index was also good. Year-over-year, the core PPI plunged 10% to 3% in 13 months. “The peak inflation data continues,” as Strategas’ Don Rissmiller puts it. At the same time, an increase in jobless claims this year adds evidence that the tight labor market is weakening. Together, we have a slowing and deflating economy.

The graph below illustrates this. It shows smoothed data for unemployment insurance claims (dark blue), core PPI (pink), and trimmed average CPI (light blue), a measure that excludes the hottest 8% and 8 Coldest % of CPI sub-components. Unemployment insurance claims jump, PPI falls rapidly and CPI cut follows, but slowly:

For a year and more, the Fed has struggled to make live monetary policy decisions using historical data. But there was something interesting: the economy was obviously too hot. The two aspects of its dual mandate (price stability and employment) went in the direction of a tightening. Today, inevitably, the dual mandate comes back into conflict. We wrote this in July 2022:

In a sense, the Fed’s job is now easy. Inflation is very high and unemployment is very low. What he needs to do – raise rates, quickly – is clear. But imagine a scenario in which inflation is still far too high, say 5%, and falling. At the same time, imagine that unemployment is higher, say it is again approaching 5% and rising. What does the Fed do then?

Things are better now than we imagined then. Headline inflation is indeed 5% and falling, but the unemployment rate is still only 3.4%. It’s entirely possible that the tightening done so far – 500 basis points of rate hikes, $400 billion in asset write-offs and a few bank failures, to boot – will be enough to control inflation without increasing much unemployment. The soft landing dream, to our surprise, remains alive.

But even in an optimistic scenario, inflation will take several months to reach something close to 2%. Meanwhile, the economy, although relatively strong, is unbalanced. It depends almost entirely on the American consumer. If the labor market continues to weaken, growth could fall rapidly. And at the peak of a tightening cycle, we should expect more things to break, making the economy worse.

Tough choices lie ahead for the Fed. A break in rates in June seems reasonable, but the real question is how long to wait before a cut. Mr. Futures Market is betting it won’t be long at all: a 25bp drop in September and three by the end of the year. Here, Unhedged is still split. Rob is inclined to say that the Fed will “raise us longer” straight into a recession. Ethan thinks the economic deceleration later this year will force cuts. In a few months, our divide could well be reflected at the Fed. (Ethan Wu)

Uber versus Airbnb

Uber and Airbnb are companies with almost identical business models, built around different assets. Both run global marketplaces where owner/operators can lease an asset to a customer. In one case, the asset is a car; in the other, a house. Both companies make money by charging the seller network usage fees.

The similarity of the two companies is well evidenced by the fact that their values ​​are quite similar and evolve together. Here is the enterprise value of the two:

However, there is something profoundly different between Uber and Airbnb. One of them makes money and the other doesn’t.

Regular readers will recall that we recently argued that the correct way to think about Uber’s profitability is in terms of what we call true free cash flow: operating cash flow minus capital expenditures. and stock-based compensation (the failure to include stock in any measure of profitability is, everywhere and always, a dirty trick). Here is the true quarterly free cash flow of the two companies:

Airbnb has generated $2.8 billion in free cash over the past four quarters, a revenue margin of more than 30%. It is a very profitable business! During the same period, Uber burned over $900 million in cash. Given the similarity of business models, what explains this? I don’t know for sure, but I have a tentative theory that has to do with the basic differences between a car and a house.

As companies use a market model (“network”, “platform”), it is traditional to link profitability to scale. There is a phase of network construction where the company is loss-making, but when it reaches a certain size, costs stabilize and cash begins to flow. But given that Uber’s revenue, at $33 billion, is four times that of Airbnb, it’s hard to argue that scale explains the difference here.

A closely related, somewhat more compelling theory is that Uber, regardless of size, decided to invest more aggressively than Airbnb, sacrificing profits now for profits later. In other words, Uber could be profitable at its current scale but chooses not to be. It is true that Uber is growing its revenue faster than Airbnb (58% in the last 12 months, compared to 32%). It is also true that Uber is adding new services (deliveries, freight). There might be some truth to that theory, but I doubt that’s the whole story.

What makes me doubt is the big difference between the gross profit margins of the two companies:

In both cases, revenues are the fees charged to asset owners/operators. Here’s what Uber says in cost of revenue (i.e. costs subtracted from revenue to arrive at gross profit):

Revenue cost, excluding depreciation and amortization, primarily includes certain insurance costs related to our mobility and delivery offerings, credit card processing fees, banking fees, data center and network, mobile device and service costs, costs incurred with carriers for Uber Freight Services, amounts related to fare chargebacks and other credit card losses.

And here is Airbnb:

Revenue cost includes payment processing costs, including merchant fees and chargebacks, costs associated with third-party data centers used to host our platform, and amortization of internally developed software and acquired technology .

One difference here is that Airbnb includes technology amortization in the revenue cost line, while Uber puts it in a different line further down the income statement. But that should make its gross margins higher than Airbnb’s, not lower.

Another, perhaps more important, difference is that Uber flags insurance costs up front. This brings up a crucial point: you can’t pitch a house, or at least not without serious effort. I am guessing that insurance costs for each unit of revenue from Uber’s business are much higher than those of Airbnb, and that these higher costs are almost entirely variable, i.e., they increase with revenues (whether these insurance costs are borne directly by Uber or by the asset owner/operators, who then have to be compensated for them, does not matter).

I still have a lot of work to do on these two fascinating ventures; insurance could be a red herring. What is clear is that Uber’s cost of goods includes significant variable costs that Airbnb’s does not. I look forward to hearing from readers who know more. (amstrong)

A good read

Steven Kelly pushes back on technology and social media theory of bank runs.