Before the recent banking-related sell-off, US stocks were not pricing in a recession. Investors probably overinterpreted a few encouraging bits of data from the start of the year, which may have been flattered by exceptionally warm winter weather — with New York experiencing its longest winter period without snow in half a century — and problems with the adjustments government statisticians make to account for seasonal consumption patterns, which have changed since the pandemic.
The bigger picture is that the US economy has already slowed significantly since its boom in 2021. While GDP growth has been volatile due to swings in inventories and international trade patterns, measures of growth in underlying domestic demand had already slowed to a crawl by the end of last year.
At the same time, our quantitative modelling — which incorporates the signal from a range of variables with a track record of leading the economic cycle — suggests that recession risk increased significantly from mid-2022 and remains high.
The economy would have slowed even more last year had consumers not spent some of the extra savings they built up during the pandemic lockdowns — but there’s evidence to show that support from this source is fading. Loan delinquency rates began to rise again late last year, a sign that at least some households have already burned through their cash buffers.
Other households may not plan to spend any extra savings they accumulated during the pandemic. They may have used their stimulus cheques to build a precautionary cash buffer they always wished they held. In aggregate, we can see that so-called excess savings are now providing less support.
Delayed is not denied
Even more significantly, interest rates rose at the fastest rate since the early 1980s in the second half of last year. The key point here is that monetary tightening tends to hit the economy with a lag. Most evidence suggests that it’s typically more than a year before the peak impact of rate hikes is felt.
There are lots of reasons why it takes time for higher rates to bite. For example, loans aren’t all refinanced at once, so they roll onto higher rates over time. Meanwhile, layoffs in the most rate-sensitive sectors like construction typically happen only months after activity in those sectors weakens.
The delayed impact on the financial system, evident in the recent failure of three regional banks, should also be seen in this context. Rapid rate increases and the associated sharp moves in bond yields often create strains in the financial system that are not evident immediately.
We shouldn’t take this final point too far. There are good reasons to be optimistic that a systemic crisis in the banking system in the style of 2008, and the very deep recession that followed, will be avoided. The US Federal Reserve has acted quickly to quell the acute stress that followed the failure of Silicon Valley Bank (plus its smaller regional-bank peers Signature Bank and Silvergate Bank).
It has established a new lending facility (the Bank Term Funding Program) to provide liquidity to the domestic banking system. In addition, it has enhanced its swap lines with other major central banks — which help provide dollar liquidity to banks elsewhere in the world.
The banking system today also looks different than it did on the eve of the global financial crisis. Large banks around the world are much better capitalised, meaning they have much greater capacity to absorb losses. We don’t see the same evidence that they’ve engaged in widespread lending to risky borrowers, or fuelled an economy-wide credit bubble, as they had done then.
Yet none of this is cause for complacency. Banks had already begun to tighten their lending standards at the end of last year. Even if the immediate turmoil that has followed the regional bank failures subsides, that shot across the bows means that tightening is likely to continue in the coming months, with lenders turning more cautious once again.
Firms will find it harder to borrow for investment, and households will find it harder to obtain credit for big-ticket purchases. That will add to the chances of recession, regardless of whether any more banks fail.
Recession? What recession?
Given this economic backdrop, analyst forecasts for the earnings of US companies in aggregate still look much too optimistic. While their forecasts have been falling gradually since mid-2022, in our view they haven’t adjusted nearly enough yet. They remain consistent with earnings slightly higher this year than last, followed by strong growth (of about 12%) in 2024.
Those aggregate earnings forecasts look even more optimistic if we strip out the energy and materials sectors, where it’s widely acknowledged that the earnings are likely to be much lower this year with commodity prices now well below their 2022 highs. In both cases, analysts expect earnings to fall by more than 15% this year. Outside those sectors, the consensus among analysts is for solid growth.
Current forecasts clearly contrast with the typical experience in recessions. Nearly every recession is associated with a double-digit percentage decline in earnings. We sometimes hear the argument that earnings are nominal, and therefore might be cushioned by still-high inflation. Yet earnings fell even in the wake of the 1970s recessions, despite even higher inflation than we have now. Our top-down quantitative modelling also continues to point to significant further falls in earnings forecasts.
One particular challenge this time around is that profit margins are already very high by past standards. They jumped to all-time highs during the exceptional circumstances of the pandemic and have yet to normalise fully again. In other words, it made sense to expect some downward pressure on margins even without a recession.
Beware of falling expectations
When earnings expectations fall, the market usually struggles although there are occasional exceptions, which tend to happen when the valuation of the market (the multiple that investors are willing to pay for a given stream of future earnings) rises by enough to offset the decline in earnings expectations.
Unfortunately circumstances don’t seem aligned for that to happen this time around. In the past, these earnings multiples have reacted to changes in government bond yields. The higher the returns available from government bonds, which have minimal credit risk, the less appealing the prospect of paying a big multiple for riskier equity earnings.
The key problem this time is that equity multiples in the US have yet to adjust by as much as you would expect given the scale of the surge in real (inflation adjusted) yields last year. While the same is arguably true across major equity markets, it’s the US market that really stands out on this count.