The new week begins with longer-dated bonds recovering significantly better than short-term notes after Friday’s big selling spree. In fact, the shortest-dated Treasuries aren’t recovering at all with 2s and 3s trading higher in yield as of 11am ET. This only adds to the rapid move toward a fully inverted yield curve (2yr yields higher than 10s)–something that’s now within striking distance perhaps as early as this week. Much has been made of curve inversion, but it doesn’t necessarily mean what everyone thinks it means.
What does everyone think an inverted curve means? In a word: recession. Rarely do market concepts get as much instant traction and evade thoughtful critique as one that can be claimed to have “successfully predicted every major recession since the 1950s” (or whatever it is that everyone seems to be saying).
Inversions don’t predict crap.
Inversions are quite simply a historic comorbidity of mature economic cycles- -a symptom that coincides with other symptoms, but not one that necessarily occurs for the same reasons and never one that acts as a catalyst. The lag time between inversion and recession is also problematic considering it can be construed to be as short as a few months or as long as several years.
There also seems to be no correlation between the size of the inversion and the size of the ensuing economic contraction. For instance, the curve was MUCH more deeply inverted in 2000 before the relatively mild 2001 recession, but barely inverted at all before the massive Financial Crisis.
And how about the inverted curve in 1982 that preceded once of the biggest and certainly the most stable periods of economic growth in U.S. history? Or was that inversion a good thing because it was less of an inversion by 1982 than it had been in 1980?
Let’s also consider that the curve was within 7bps of inversion in 1994 and remained within 60bps of inversion for the next 6 years of economic expansion. The curve inverted once in 1998 and then again in 2000, so which of those two inversions gets credit for the 2001 recession?
In fact, which curve inversion gets credit for ANY recession? The answer: none of them! To reiterate, the curve inverts at the end of mature economic cycles and often because the Fed has decided to remove accommodation by hiking the Fed Funds Rate. Indeed, note the white vertical lines in the following chart at the intersections of inversions and the end of Fed rate hike campaigns. (NOTE: leading indicators data is used for recessions here as it is more granular than quarterly GDP data, and matches up just fine).
A few notes on the chart:
First off, we must understand the psychology of the market and the Fed following the early 80s. Inflation had finally fallen. The economy was expanding very nicely, but mortal fear of inflation remained. As such, at the first sign that inflation was no longer gently descending the Fed began hiking rates, slowly at first, but then abruptly (even though the economy was far from overheating). The combination of overly restrictive Fed policy and the lingering effects from the S&L crisis tipped the economy into a shallow, short recession. The curve inverted as a result of the Fed’s rate hike campaign.
Inflation fell steadily and, some would say, mysteriously throughout the 90s. The x-factors of globalization, automation, and e-commerce remain underappreciated as disinflationary drivers during this time. Under the mistaken belief that solid economic growth MUST result in inflation, and with Greenspan afraid of the “irrational exuberance” that would make the post-QE era look like a Daytona Beach spring break party , the Fed kept rates high enough to keep the curve within 60bps of inversion (and within 7 bps in 1994, as stated). The curve inverted in 1998, but the economy bounced back stronger as the dot-com bubble inflated. The Fed hiked again, further inverting the curve, but by that time, the dot-com bubble was bursting, and it deserves true credit for the 2001 recession.
Fed policymakers had another shot to combat irrational exuberance in the mid 2000s and this time, the phrase was far more appropriate. Once again, the curve inverted in response to the rate hike campaign, but the Fed was bringing way too little, way too late to a party that no one was really going to understand until years later. In fact, the curve only managed to invert very modestly before the Fed ended the rate hike campaign and inexplicably froze in its tracks for more than year while those of us working in the mortgage industry panicked about what was about to happen (the Fed hadn’t cut rates as of July 2007! If you were in the mortgage industry much before that, do you remember July 2007? I sure do…).
All that to say, the curve inversion had absolutely nothing to do with the Great Recession. It was going to happen regardless of rates because the exuberant homebuying and lending environment existed across multiple rate levels from 2003-2006.
And ever since then, we’ve had QE and ZIRP (zero interest rate policy). ZIRP is a great way to ensure a steep curve for as long as the Fed wanted it. When ZIRP ended, it was a given that the curve would begin flattening. It did so in a more gradual and predictable way than at any other time in history. It gently touched inverted levels in late 2019 and bounce out of inverted territory a few months BEFORE the inking of the first phase of the US/China trade deal (widely thought to be a critical step in staving off a trade-war-driven recession in 2020). Instead, we had a covid-driven recession–one that would have existed regardless of the status of the yield curve in the preceding months.
Bottom line : 2018/2019 was the Fed’s first really great shot at engineering a soft landing for the yield curve without causing recession. It’s entirely possible if not probable they would have succeeded without covid.
The current situation is unique in the sense that the curve is inverting VERY early in the rate hike cycle. That fact alone means it’s a terrible comparison to past instances of curve inversions, and that the dynamics are completely different. No one can say we won’t experience a recession at some point in the next 2 years, but anyone can say that the reason for that recession will be something other than the shape of the yield curve.