VIX vs 10Y-2Y Yield Curve

VIXCLS is computed from a strip of S&P 500 SPX options expiring 23 to 37 days out and is published by CBOE every 15 seconds during trading hours. T10Y2Y is the constant-maturity 10-year Treasury yield minus the constant-maturity 2-year, both sourced from the U.S. Treasury H.15 release and posted daily on FRED. The two are observed in different markets: VIX comes from equity-option dealers, T10Y2Y comes from primary-dealer Treasury auctions and the secondary cash market. They share one underlying input, the path of the federal funds rate, but encode it differently. VIX prices the variance of equity returns over the next 30 days; the 2s10s prices the difference between the average expected funds rate over the next two years and the next ten years.

The Federal Reserve Bank of New York maintains a recession-probability model built on the 10y-3m spread, but private research desks at Goldman Sachs and BofA Global Research have published since 2019 on the joint VIX/2s10s system because the curve alone misfired in 2022 to 2024. Estrella and Mishkin (1998, FRBNY) is the original term-spread reference; the joint indicator framework appears in Cieslak and Vissing-Jorgensen (2021) on Fed put dynamics.

T10Y2Y first inverted on April 1, 2022, briefly reverted, then re-inverted on July 5, 2022 and stayed below zero until August 27, 2024, a 25-month run. The trough hit -108bp on March 8, 2023 during the Silicon Valley Bank failure. VIX during that 25-month window averaged 18.9 and only closed above 30 on five trading days, all of them during the SVB and First Republic episode in March 2023. Real GDP grew 2.9% in 2023 and printed 3.0% annualized in Q2 and Q3 2024 per BEA, so the recession the curve was warning about did not arrive on the historical schedule.

Research at the Cleveland Fed (Haubrich, Bauer 2024) flagged that the 2022 to 2024 inversion was driven by term-premium compression rather than the usual policy-tightening expectations, which broke the historical link between curve depth and recession probability. The joint VIX/2s10s reading captured this: in every prior post-1976 inversion that preceded a recession, VIX averaged at least 22 during the inversion window. The 2022 to 2024 inversion was the first where VIX averaged below 20.

Three breaks reshape the historical pair series. First, the 1979 Volcker regime change moved the curve dynamics: the 1980 to 1982 episode produced a 2s10s trough of -240bp with VIX-equivalent realized volatility above 35, the most extreme joint reading on record. Second, the 1994 Greenspan tightening produced a yield-curve flattening to roughly 0bp without any sustained VIX spike, the cleanest example of a curve signal disagreeing with equity-vol calm; the 1995 to 1996 economy did not enter recession. Third, post-2008 quantitative easing structurally compressed term premium by an estimated 80 to 120bp at peak per Federal Reserve Board working paper estimates (Ihrig et al. 2018), shifting the inversion threshold downward.

The 2020 COVID episode is the cleanest cross-asset shock in the joint history: 2s10s steepened from 18bp to 50bp inside three weeks while VIX rallied from 17 to 82.69 on March 16, 2020. That bull-steepening with VIX panic is the regime-change signal: bond markets pricing aggressive Fed easing while equities priced disorderly liquidation. The April 2020 PEPP and Fed dollar-swap-line announcements collapsed VIX to 28 by month-end while the curve held its steepening.

Convex Risk Appetite Index (CRAI) blends VIX, the high-yield OAS (ICE BofA HY OAS series BAMLH0A0HYM2), and equity put-call ratios into a single risk-appetite composite. When CRAI prints in its lower quartile (risk-off), VIX is doing most of the work and the 2s10s reading is best treated as a slow-moving conditioning variable rather than an actionable signal. When CRAI sits in its upper quartile (risk-on) and 2s10s is inverted, the joint pair flags the configuration that produced the 2022 to 2024 anomaly: equities ignoring a term-structure warning.

Convex Net Liquidity Impulse (CNLI) tracks the change in Federal Reserve H.4.1 liabilities (reserves plus reverse repo plus Treasury General Account) and tends to lead curve moves by 4 to 8 weeks. When CNLI inflects negative and 2s10s subsequently steepens with VIX rising, the joint pair confirms that liquidity withdrawal is feeding into both legs. The June 2022 to October 2022 episode produced exactly this sequence: CNLI turned negative in June, 2s10s deepened to -55bp by September, VIX averaged 27.2 in October, and the S&P 500 hit its bear-market low at 3,577 on October 12, 2022.

T10Y2Y has been positive since August 27, 2024 and the curve sits comfortably steeper than zero in April 2026 per FRED. VIX has averaged in the high teens to low 20s over the trailing six months. This is the post-disinflation steady-state configuration: the curve has un-inverted, equity vol has normalized, and the joint pair carries less recession-warning content than at any point since 2021. Historically, the 12 months following 2s10s un-inversion have produced the bulk of recession risk because the un-inversion itself often coincides with the start of Fed cuts that reflect deteriorating activity.

The Bauer-Mertens (2018) FRBSF Economic Letter is the cleanest reference on this lag: of the seven recessions following sustained 2s10s inversions since 1955, the recession started 6 to 24 months after the curve re-steepened, with median 11 months. From August 2024 plus 11 months puts the median historical analog window in mid-2025, which has now passed without recession. The current configuration (positive curve, VIX low) maps most closely to the 1995 to 1996 mid-cycle correction analog rather than to the 2007 to 2008 pre-recession setup.

The pair is most actionable at three configurations. Configuration one: deeply inverted 2s10s plus low VIX (the 2022 to 2024 setup). The trade is to underweight long-duration equity multiples and prepare for either a curve-snap-steepener catalyst or a term-premium re-rating, both of which have historically arrived within 18 months. Configuration two: bull steepener with VIX panic (the March 2020 setup). The trade is to size into the cross-asset reflation scenario after the central bank intervention has been announced; the joint pair confirms when the disorderly phase has cleared. Configuration three: rising VIX into a flattening curve from positive (the 2018 Q4 setup). The trade is to reduce equity beta because both legs are confirming the same Fed-tightening-overshoot story.

Position sizing on this pair requires recognizing that the two legs have different cadences. VIX moves intraday on equity-option flow; 2s10s moves on the daily H.15 close. Allocators size against the slower leg, treating intra-day VIX prints as confirmation rather than as standalone triggers. The Federal Reserve's dot-plot release windows (March, June, September, December FOMC) are the cleanest catalysts for joint pair re-pricing because both VIX and 2s10s respond to the same dot-plot input.