Where’s the credit cycle?

Receive free US banks updates

We’ll send you a myFT Daily Digest email rounding up the latest US banks news every morning.

Good morning. The role of artificial intelligence in filmmaking is a key point of friction in Hollywood’s big strike. Scriptwriters and actors are worried that AI-generated extensions of their own words and images will undercut their earnings power. But if they are worried, shouldn’t lawyers, doctors, journalists and other white-collar types — many of whom have weaker union representation than film and TV workers — be more worried still? Hollywood could be a test case for a much larger conflict between workers and capital. Curious to hear your thoughts. Send them to robert.armstrong@ft.com. 

Banks results and the credit cycle

JPMorgan Chase, Citigroup, and Wells Fargo all reported earnings last Friday. The results were not too far from Wall Street’s expectations. Given the rapid changes in the interest rate environment, analysts’ attention was mostly focused on whether asset yields are keeping pace with funding costs. As it turns out, that was mostly a wash in the second quarter, leaving interest margins stable.

Trends in credit quality are of more interest to investors generally, however. Unhedged and everyone else is looking for evidence of stress at the economic periphery. Are consumers and businesses falling behind or defaulting on their loans at a greater rate?

Results from the three banks suggest that the answer is no. 

Executives at all three banks used the word “normalisation”: while writedowns of bad loans are rising, they are rising gradually and remain below pre-pandemic levels. JPMorgan CEO Jamie Dimon set the tone, talking about consumer loans, and pointing out a credit-card charge-off rate of just 2.4 per cent. He said that the bank has been “over-earning” in its credit results, and he expected things to return to a normal trend before long — which would be a 3.5 per cent charge off rate. Revolving balances per card account are still below pre-pandemic levels at JPMorgan, too. Things are still unusually good on the credit front.

The theme sounded by Dimon was repeated across all three banks and across most lines of business. Non-performing loans are stable to down, as well.

This is consistent with the aggregate charge-off data from the Federal Reserve, which shows that as of the first quarter writedowns at US banks remained low:

One exception to the normalisation trend is Wells Fargo’s $33bn portfolio of office real estate loans. Nonaccrual (more than 90 days delinquent) office loans at the bank more than doubled to $1.5bn between the first and second quarters, and the bank has made allowances for credit losses equivalent to 6.6 per cent of the office portfolio. A trend worth watching, and one that will probably worsen, but not a catastrophic one as of yet. 

It is, in short, hard to see a traditional credit cycle of any sort in the bank numbers. What the numbers show, as of now, is a big hangover from the pandemic. We may see more of interest in the regional bank reports. Unhedged will be following closely. 

More on the excess savings puzzle

On Friday, we wrote about the simplistic way in which we and other people have been thinking about post-pandemic excess savings. On a naive view, households spend the money received from stimulus programs, and then it the money is gone, and aggregate spending slows. But this is wrong, because one household’s spending is another’s income. That does not mean that the stimulative effect of excess savings lasts forever, of course. Instead, it means that the stimulus cash moves through the economy until it ends up with a household (or company or government agency) that doesn’t spend it. We wrote:

Eventually some amount of [the excess savings] will get to someone who will simply hold on to it, or use it to pay off debt. 

If the money simply sits in a deposit account, and the bank with the deposit does not make a new loan on the back of that deposit, the money falls into a sort of coma. On the other hand, if the money pays off a debt, it is destroyed — the reverse of the process by which lending creates money. The “person” who pays off the debt and destroys the money might be the government, if they use a tax payment to settle a debt that is not replaced with a new one. 

Innes McFee, managing director at Oxford Economics, emailed over the weekend with some more perspective on this. He emphasised that tax is a particularly important part of the picture. In a note last year, he estimated that tax payments on transactions, dividends and so on took a $745bn bite out of the total stock of excess savings (so perhaps a third of the total). He notes, however, that most of these taxes were likely paid at the high end of the income distribution. But at the high end of the income spectrum, marginal propensity to spend is low, so the savings paid in taxes may not have been very economically potent, anyway (more on this dynamic below).

He also notes that there is evidence that households are indeed using stimulus to pay down debt, destroying excess savings. He points to this chart from the Fed (made with data from Standard & Poor’s), showing that credit card payments, measured as a proportion of outstanding balances, has been well above the trend:

Finally, he makes the point that, in an inflationary environment, excess savings’ impact on the economy could constrain themselves: the extra spending they encourage drives inflation up, bringing real incomes down and discouraging spending. Excess savings create an unstable equilibrium.

Several readers sent along a paper called The Trickling Up of Excess Savings, published early this year by Adrien Auclert, Matthew Rognlie, and Ludwig Straub. It attempts to model the way excess savings “trickle up” through an economy, finding their way to rich households with a low propensity to spend, driving their economic impact steadily down. 

The paper argues that, if you assume that the excess savings are initially widely distributed across households but that poorer households will spend more of the money, excess savings must move slowly up the income scale, steadily diminishing their economic impact. All the same, the model acknowledges that one household’s spending is another’s income, so despite the trickling up effect the stimulus effect still lasts longer than suggested by naive models in which the savings disappear when spent. The authors suggest that there could be a notable impact (perhaps half a percentage point of GDP) on consumption five years after excess savings’ peak, though tighter monetary policy could shorten the timeframe. 

The takeaway for investors of all this? The pro-growth impact of excess savings may be with us for a while. A sharp drop off in consumption in the third or fourth quarter, driven by the exhaustion of the excess savings stock, is unlikely. Instead, taxes, monetary policy, trickling up, and possibly sustained inflation will cause the impact to diminish slowly but steadily over several more years. 

One good read

Music festivals are bad.