
Bloomberg has an article with the following title and subhead:
How Torsten Slok Solved the ‘Sherlock Holmes Mystery’ of the Economy
When others thought a recession was inevitable, Apollo’s chief economist correctly predicted more growth. He did it by looking at the data.
Sløk seems to be one of the few economic pundits that is able to avoid the temptation to reason from a price change:
Adding to the difficulty, the 54-year-old has lately concluded that much of the economics he learned at school is broken. After the punishing series of rate hikes that began in 2022, the field’s models concluded with near certainty that higher interest rates would tip the US into recession. Instead, the economy powered ahead with barely any sign of a slowdown. Bad forecasts aren’t just an occupational hazard of Slok’s chosen discipline—they seem to be the default state. “The economics profession was totally wrong. And why were they wrong?” he asks rhetorically. “They were too wedded to the textbook. They basically said, ‘Oh, when the Fed raises rates, it always goes bad.’”
You might quibble that good textbooks don’t say that higher rates always lead to recessions, but in a broader sense Sløk is correct that most economists put far too much weight on changes in market interest rates. Long-time readers may recall that I was also skeptical of the claim that higher interest rates in 2022 constituted “tight money”.
The past year showed that high interest rates alone can’t trigger a downturn. That left Slok with a “little Sherlock Holmes mystery” to solve, he says. Why exactly did rates stop being that one metric a forecaster could depend on? Slok has a guess. On the one hand, the economy in the US is now much less sensitive to rates than it used to be, with homeowners and companies alike locking in low-rate debt during the pandemic. On the other hand, the country has economic “tailwinds” the rest of the world doesn’t, namely an artificial intelligence boom as well as fiscal stimulus from the Inflation Reduction Act and other federal legislation signed by President Joe Biden.
Higher rates are only a problem for the economy when the Fed pushes its policy rate above the natural rate of interest. As Sløk implies, both the AI boom and fiscal stimulus somewhat raised the natural interest rate in real terms, and of course higher inflation expectations in 2022 were a factor pushing up equilibrium nominal interest rates.
It’s also worth revisiting the issue of inverted yield curves. In the past, I’ve argued that while inverted yield curves are often correlated with subsequent recessions, the correlation is far from perfect. The fact that a recession did not occur in 2023 or 2024 is a black mark for the supposedly “infallible” yield curve prediction model. I suspect that if not for Covid the model would have also failed in 2020.
READER COMMENTS
Mark Cancellieri
Mar 21 2025 at 7:15am
“The past year showed that high interest rates alone can’t trigger a downturn.”
I have a problem with this claim by the Bloomberg author. What constitutes a high interest rate?
Scott Sumner
Mar 21 2025 at 11:47am
I think he’s saying the sort of rate increase we saw in 2022-23 doesn’t necessarily trigger a recession.
David Seltzer
Mar 21 2025 at 9:40am
Scott: Not reasoning from a price change. Nominal rates can rise with an increase in anticipated inflation. In a strong economy with increasing demand for money, wouldn’t rates also increase?
Scott Sumner
Mar 21 2025 at 11:47am
Yes, economic growth can also push up interest rates.
Thomas L Hutcheson
Mar 21 2025 at 9:49pm
“Competent” economist do not make predictions, they make conditional predictions. “If the Fed doesn’t keep inflation on track, we’ll have a recession” is a sensible precision (it could be wrong, but it’s sensible).
If there are n economists unwisely making predictions, what is the likelihood that the one whose prediction was right will believe it was becasue his model of the economy was better than the others?
Jon Murphy
Mar 22 2025 at 8:23am
I’m not sure I understand your point. All predictions are conditional predictions. The conditional may or may not be precisely stated (or even understood) by the person making it, but all predictions are necessarily conditional.
Thomas L Hutcheson
Mar 23 2025 at 4:31pm
Indeed, that is what is mean that they should spell out their conditions, especially their policy conditions. Maybe one should presume that the listener shares the same expectation of non-policy conditionals, but the predictor’s policy conditionals are their own.
Comments are closed.