The Analyst’s Course · Macro Regimes
Inversions — Evidence, False Positives, and the Right Spread
The evidence
Since 1968, every US recession has been preceded by a 2s10s inversion — seven of eight cycles by the classic count (the 1966 inversion missed the brief 1970 recession's start but "hit" its neighborhood). Lead times run 6 to 24 months. The mechanism: inversions compress bank net interest margins (they borrow short, lend long), credit contraction follows, and the Fed usually confirms by cutting. This is the best-attested recession signal in conventional macro — and the panel tracks both its current state and how long the current configuration has run.
The false positives
1966: inverted, no recession followed (a near-miss often credited to a timely Fed pivot). 1998: the curve briefly inverted at the belly (2s10s touched zero during LTCM) with no recession — the Fed's intermeeting cut reset the curve. Both fakes share features: strong private balance sheets, Fed willingness to act fast, and shallow inversions. The 2019 inversion was more instructive still: a recession did start within its historical lead window — it just required a pandemic to manifest, muddying the scorecard in a way no model anticipated.
Why the Fed watches 3M-10Y
The Fed's research (and the NY Fed's recession-probability model) uses the 3M-10Y spread, for two reasons: the 3M bill sits closest to actual policy control (uncontaminated by the belly's supply/technical factors), and empirically it produces fewer false positives than 2s10s in the post-war sample. The panel computes both spreads side by side for exactly this reason — when they disagree (as in 2022's famous 2s10s inversion with positive 3M-10Y for months), the disagreement itself is information about how much of the signal is Fed-policy mechanics versus genuine macro repricing.
2s10s = 10Y − 2Y; 3M10Y = 10Y − 3M
The two signals — Panel shows both. When they conflict, policy mechanics vs macro repricing is the argument to have.
2022–2024: the deepest inversion since 1981
The 2s10s went negative in July 2022 and stayed inverted for a record ~two years, reaching depths not seen since Volcker. The 3M-10Y followed in October 2022. The "signal" then produced... a soft landing through 2023–24, with unemployment staying under 4.5%. Post-mortems focus on three unusual factors: pandemic-era household balance sheets, the Fed's balance sheet still holding rates down at the long end, and supply-driven inflation letting cuts happen without recession. The honest scorecard: the signal's sample is small (eight cycles), the world it was fitted to keeps changing, and it remains a risk-management flag — not a recession oracle.
What you'll practise
Historically, 2s10s inversions precede recessions by…
3 graded checkpoints · certification exam at the end of the track
Sources
Shiller, Irrational Exuberance (3rd ed.); Estrella & Mishkin (1996); multpl.com; US Treasury
Learn content is for education only — not individualized financial advice, a recommendation, or a solicitation to buy or sell any security. Options involve substantial risk. Examples are simplified and historical patterns never guarantee future results.