A second month of sticky inflation

Stay informed with free updates

Simply sign up to the Global Economy myFT Digest — delivered directly to your inbox.

Good morning. Yesterday’s hot-ish consumer price inflation data didn’t keep the S&P 500 from another all-time high, or artificial intelligence hype from charging on. Nvidia, a serious AI bet, and Super Micro Computer, a half-serious one, both rose some 7 per cent. Microsoft, Amazon, Alphabet and Meta are set to spend a combined $180bn on capex in 2024, a 27 per cent increase, according to Bank of America. At this rate, we’ll have to revise our “it’s not a bubble” view to “well, it wasn’t a bubble until you all went nuts”. Email us: robert.armstrong@ft.com and ethan.wu@ft.com. 

CPI, part II

After last month’s consumer price index report, which was hotter than expected, Unhedged wrote a piece called “Everyone calm down about CPI”, which concluded:

If February is similarly hot, that will be a different story. For now, reserve judgment. 

Well, a month has passed, and CPI has come in hot again in February, though at a slimmer margin. How worried should we be? Last time, we argued that four pieces of context made the January inflation report seem less scary. In the interest of consistency, we thought we’d revisit them:

• “Inflation expectations, which the Federal Reserve thinks are the most important single input to realised inflation, are falling.”

  Still broadly true, but less convincing than a month ago. Three- and five-year consumer inflation expectations have ticked up in the past month, and so have break-evens, the market’s best guess at future inflation. The most dramatic move came in two-year break-evens (though the underlying market is thin and noisy) while longer-term break-evens still look calm:

• “January inflation reports tend to be volatile.”

  This looks right. Analysts suspected the “January effect”, created by annual employment and supplier contract updates that are hard to seasonally adjust away, had boosted certain labour-intensive price categories. Many of these reverted in February, including medical services, personal care and restaurants. A startling jump in owners’ equivalent rent — one so sharp that the Bureau of Labor Statistics held a public webinar to explain how it happened — reverted, too.

• “Recent news is much better for personal consumption expenditure inflation, which is what the Fed targets.”

  Not so much anymore. Last month, we used Goldman Sachs’ core PCE forecast to visualise the up-until-then good PCE news:

  But an above-expectations reading in January and limited relief expected in February paint a darker picture. In the updated chart below, we use Goldman’s latest February core PCE forecast (a 0.27 per cent monthly increase, or 3.3 per cent annualised). The six-month average is flirting with 3 per cent:

• “Shelter inflation should still cool further, though the timing is anyone’s guess.”

  No clarity. Overall shelter inflation slowed in February to 0.4 per cent, from 0.6 per cent in January. But that was just a return to the previous rate. We’re still waiting for the much-anticipated convergence between CPI shelter and new-lease rental data. The chart below from Pantheon Macro gives one optimistic story. But you could equally argue that CPI shelter has been worryingly high and stable for nearly a year:

So far, not encouraging. A few other February data points brighten the inflation outlook, however. January’s 0.9 per cent explosion in supercore inflation (core services ex-shelter) proved a blip. Some of the increases in February look fleeting, such as a 3.6 per cent jump in airfares (possibly a one-time response to higher jet fuel costs) and a 0.5 per cent rise in used-car prices (wholesale auctions have seen falling prices lately). 

Remember, too, that wage growth is slowing, likely helping disinflation along. A first-quarter inflation resurgence that fades in the second quarter is plausible. But the case that inflation is stabilising closer to 3 per cent than 2 per cent is growing slowly stronger. (Ethan Wu)

Why JPMorgan is so dominant, part II: structure over skill

On Monday I wrote about how JPMorgan absolutely towers over the US banking industry. Its dominance is somewhat mysterious. One expects to see the big and strong get continuously bigger and stronger in technology, but not so much in banking. Banking does have some features that create scale effects — fixed administrative and regulatory costs lead to operating leverage for larger players, for example. But banking’s core product, capital, is a commodity. And unlike some commodities, it is hard for a company to become the undisputed low-cost producer of capital. So one would not expect to see a handful of banks take most of a market, and for one bank to dominate that handful. How has JPMorgan done it?

Readers wrote in, most of them arguing that JPMorgan’s dominance comes down to superior management. Several pointed out that it has avoided the whopping big mistakes of rivals such as Wells Fargo and Citigroup (they all noted that the London Whale was not a good look). Others noted that banking is basically just risk management, and JPMorgan is just better at knowing which risks to take, and how much of them to take. A famous recent example is JPMorgan’s decision to hold low-yielding cash when rates were at their lows, rather than picking up a bit more return by investing in long-term government securities. Rival Bank of America took the opposite approach. Rates have duly risen, BofA has big unrealised losses on their long bonds, and JPMorgan is free to invest at a higher rate.

Many other readers said that Jamie Dimon is a very good chief executive, both on the level of operational details and strategic vision (a rare combination). Indeed, it is hard to find people who deny this.

But let me announce my bias: while I do believe that great management can make companies much better, large companies that consistently produce excellent results and dominate industries almost always have intrinsic structural advantages. So it is best to look for those structural features before thinking about management excellence.

Clearly, one of JPMorgan’s advantages is its diversified business model. Especially in the post-2008 regulatory regime, where high leverage and short-term wholesale funding is frowned upon and asset risk weightings matter immensely, having a large retail banking operation to match with investment banking and trading businesses is an advantage. The retail operation provides both deposit funding and assets ripe for securitisation, such as mortgages and credit card loans. And in a world where investors like steady fee businesses, wealth management and payments processing support a bank’s valuation and decrease its cost of capital. And JPMorgan has all of these things, at scale 20 years ago, before Dimon became CEO. He inherited an ideal platform for growth in the post-crisis world. 

And once a bank has the right growth platform, and starts growing, that growth will tend to compound. Recent studies have found significant and persistent returns to scale in banking. For example, David Wheelock and Paul Wilson, professors of economics at the Federal Reserve and Clemson University, respectively, argued in a 2017 paper that the conventional wisdom about returns to scale in banking was wrong. That wisdom held that returns to scale were cost-driven and were exhausted once a bank reached just a few hundred million dollars in assets. Looking at the pre-crisis year 2006 and post-crisis 2015, they found that there was a significant relationship between growth and increasing returns, even for the very largest banks, and at some of those very large banks, the returns came from revenue and profit, not just costs. Getting bigger helps increase returns, and those higher returns can be reinvested in growth.

General results such as Wheelock and Wilson’s don’t fully explain what JPMorgan has achieved, though. The bank’s growth rates over the past 20 years are remarkable. In its investment bank, commercial bank and retail bank, assets have grown at a compound rate of about 6 per cent a year (it is slightly hard to come up with precise numbers because JPMorgan has changed its unit reporting structure over the years). If that doesn’t sound like a high growth rate, remember there were $750bn in assets at the outset. Client assets in wealth management have compounded at 10 per cent. The other divisions have been almost equally impressive. Remarkably, return on equity in the bank’s key business (investment and retail banking) has increased even while leverage — the amount of borrowed money used to generate those returns — has fallen dramatically.  

There is more to JPMorgan’s dominance than a corporate structure that fits the times perfectly and generic increasing returns to scale in banking. More soon on what that might be.

One good read

Martin Wolf on China’s world view.