For education only. Not financial advice. Figures come from public government sources and are checked, but may contain errors. Check the original source before relying on it.
Spreads rose above 800bp around each of the three recessions since 2000 (the 2002 peak came after the 2001 recession ended), and the larger credit scares (2002, 2011, 2016, late 2018, 2022, March 2023) took them above about 500bp. Smaller scares did not: August 2024 peaked at 393bp, October 2023 at 453bp and April 2025 at 461bp.
Basis points, ICE BofA US High Yield OAS. Episode values come from secondary compilations of FRED history (eco3min.fr, Convex, Trading Economics, CFA Institute) and are approximate where marked ~. The 2002 and 2011 figures are weekly closes. March 2023 is 522bp (Mar 24, 2023; single secondary source). The median is roughly 440–450bp depending on the period (New York Life Investments: 438bp over the 20 years to mid-2024).
Spreads stayed below 400bp for almost three years in 2004–07 before the crisis (New York Life Investments). Tight spreads mean investors get little pay for risk, not that trouble is imminent.
In 2007, spreads began widening in June, as Federal Reserve staff noted at the Aug 7, 2007 FOMC meeting, while the S&P 500 went on to a record close on Oct 9, 2007. Credit moved first that time. It does not do so reliably: in April 2025 spreads widened in the same days that stocks fell (3.42% on Apr 2 to 4.61% on Apr 7), and Reuters reported in November 2018 that high yield had lagged that sell-off.
Energy was about 11.5% of the ICE BofA US High Yield Index on Oct 8, 2026 (State Street's SPHY fund page), and has been higher: Muzinich put it above 14% by mid-2013 (single source). High oil prices support those issuers; a sharp fall in oil hurts them, as it did in 2016.
Forward earnings yield minus the 10-year Treasury yield. On this chart it was negative at year-end 1999 to 2001, positive at every year-end from 2002 to 2024, and above +3 pts at every year-end from 2008 to 2021. It was also negative before 1999: the Fed noted in July 1997 that the 10-year yield exceeded the forward earnings yield by the most since 1991.
A wide gap in 2008 did not mean stocks were safe; it was largely a product of the crash itself. The S&P 500 fell 56.8% from its record close of 1,565 on Oct 9, 2007 to a closing low of 677 on Mar 9, 2009, which pushed the earnings yield up, while the 10Y fell from about 4.7% to about 2.1% in December 2008.
The gap uses analysts' forward estimates, and those lagged the collapse. Actual 2008 operating EPS came in at $49.51 and reported EPS at $14.88 after write-downs (S&P figures, via Cambridge Associates). Part of the 6-point "cushion" was measured against earnings that never arrived.
At the end of 2007 the gap was about +2.8 pts on this chart's reconstruction, and stocks still fell by more than half. Credit gave the warning that time: HY spreads went from a record-tight 241bp in June 2007 to 2,182bp in December 2008.
Over years, not months. The high readings of early 2009 and 2011–12 came ahead of multi-year gains in stocks, even when the first months hurt. Two episodes are not a rule. A gap near zero does not predict a crash; it means little margin for error if a shock hits.
Nominal 10Y = real (TIPS) yield + breakeven. A rising real line with a flat breakeven points to a growth, Fed-path or term-premium story, not an inflation scare.
Mainly the market's price of default risk. This page treats ~3.5% as the line between watch and stress.
The 30-year fixed rate moves with the 10-year Treasury yield, not the 30-year, plus a spread that varies over time.
A lagging indicator. The Sahm rule signals the start of a recession when the 3-month average unemployment rate is 0.5 pts or more above its lowest 3-month average of the previous 12 months. It is not infallible: it crossed 0.5 from July to September 2024 and no NBER-dated recession followed.
Tightening typically reaches housing first, then business investment, then hiring. Rules of thumb put the peak effect of a rate change 12 to 24 months out, though Fed governor Waller argued in July 2023 that the lags may now be shorter (Federal Reserve).
Bar spans −3 to +3 pts around zero. Earnings yield is forward EPS ÷ price; it is a valuation yardstick, not a return forecast. On this page's reconstructed year-end values the gap was above +3 pts at every year-end from 2010 to 2019.
One row per headline number. "Checked" is the last date the update script read the value from its source; an older date means it was not re-read. All dates are 2026 unless stated.
Explanations audited Oct 18, 2026. Next audit Nov 18, 2026.