Why do credit spreads widen in recessions?

The short answer

A credit spread is the extra yield investors demand to hold a corporate or sovereign bond instead of a comparable government bond. When the economy weakens, the probability of corporate defaults rises, so investors price more risk and the spread widens. That much is intuitive.

The less intuitive part is the timing and decomposition. Spreads start widening before recessions are visible in GDP data, sometimes by 6 to 18 months. They tend to peak around the entry into recession, not at the bottom of the cycle. And much of the movement is not about default expectations at all.

Studies decomposing credit spreads find that expected default losses explain only about one-third of the variation; the rest is illiquidity premium, risk aversion, and a residual that captures market sentiment. The 2008-2009 episode is the extreme illustration of that decomposition, dissected in our anatomy of the crisis-era spread widening. This is why spreads can be useful as leading indicators even when default forecasts are stable.

→ New to credit dynamics? Financial education framework

What the data shows

The empirical regularities are well documented (FRED, Moody’s, ICE BofA, NBER, 1981-2024):

• HY spread peaks: 21.82% in December 2008, around 11% in March 2020 (FRED BAMLH0A0HYM2)
• IG spread peaks: roughly 6% in December 2008, around 4% in March 2020
• Lead time: HY spreads typically begin widening 6-18 months before NBER-dated recession starts
• HY spread average: roughly 5% across the full series since 1996
• Default rates lag: defaults peak 12-24 months after spreads peak in most cycles
• Decomposition (Elton, Gruber, Agrawal, Mann 2001 and later replications): expected default loss accounts for roughly 30-40% of spread; the rest is liquidity, risk premium, and tax effects

The exception that nuances the picture: in the 2014-2016 energy-sector spread blowout, HY spreads widened to over 8% without a US recession. The widening was concentrated in energy issuers facing the oil price collapse, and it eventually receded as oil stabilized. This shows spreads can widen for sector-specific reasons that do not generalize.

→ Dataset: HY credit spread leading indicator · Credit spreads and recession risk

Why it happens — the macro mechanism

Three transmission channels combine to explain the widening pattern.

The default expectations channel. When economic indicators deteriorate — PMI declines, jobless claims rise, earnings revisions turn negative — analysts and investors update their expectations of corporate default probabilities upward. Bonds with higher expected losses must yield more. This is the textbook channel and it accounts for some of the move, but not most.

The risk aversion and liquidity channel. This is where the angle most explanations miss appears. When uncertainty rises, investors demand a higher risk premium even for an unchanged level of expected losses. Simultaneously, dealers reduce their balance sheet capacity to warehouse bonds, so bid-ask widens and liquidity premia rise. Empirical work by financial conditions researchers finds that this combined risk-and-liquidity component explains the majority of spread movement in stress periods. The spread is therefore as much about investor balance sheets and behavior as about issuer fundamentals.

The forward-looking channel. Markets price the future, not the present. Spreads widen when investors anticipate worsening conditions, even before the data confirms it. This is why HY spreads have been a robust leading indicator: they incorporate forward expectations from market participants who lose money if they wait for confirmation. Gilchrist and Zakrajsek’s excess bond premium work formalized this with a measure that strips out default expectations and captures the forward-looking risk premium directly.

Synthesis by regime: at the cycle peak, with strong growth and tight spreads, the default channel dominates and spreads track fundamentals. In the entry-into-recession phase, the risk aversion and liquidity channels accelerate, pushing spreads wider faster than fundamentals justify; this is when the leading-indicator property is strongest. In late recession and recovery, default rates peak (lagging) but spreads have already started to compress because forward expectations have improved; the regime pivot is the change in expected future fundamentals, not current conditions.

Most of a credit spread is not a forecast of default — it is the price of holding risk in markets where liquidity and confidence have already started to evaporate.

→ Framework: Yield curve and credit channel

What it means for different economic actors

Savers. Bond fund NAVs reflect spread movements directly. A fund holding HY at the start of a spread blowout can lose 10-20% over a few months, even with no defaults in the portfolio. Understanding that spread widening can be a price-of-risk movement, not a fundamentals movement, helps explain why such losses can reverse quickly when sentiment improves.

Investors. Spreads are watched as a regime-state variable. When HY spreads cross thresholds historically associated with stress (5%, then 7%, then 10%), the probability distribution of equity outcomes also shifts. Many institutional risk frameworks use spread levels as inputs to position sizing.

Corporates. Widening spreads raise refinancing costs, which can force capex cuts, layoffs, or distressed restructurings. Companies with bond maturities falling during a spread spike face the worst funding conditions; this is why the maturity wall calendar is watched closely going into a recession.

A common error is to read spread widening as exclusively a default forecast. Most of the movement, especially in stress, is the price of risk-bearing capacity — and that capacity recovers when central banks act, not when defaults peak. This is treated at length in the false assumptions investors make about credit and spreads.

Practical observation

Go deeper

Related questions

• What is a CDS and how does it measure credit risk?
• What is the difference between investment grade and high yield debt?
• How does a credit cycle differ from a business cycle?
• How does bank supervision affect credit conditions?

Frequently asked questions

Is the spread widening always followed by a recession?

Not always. The 2014-2016 energy-driven HY spread blowout produced no US recession because it was sector-specific and reversed. False signals do occur, particularly when widening is concentrated in one industry or region. The strongest signal is when the widening is broad-based across sectors and accompanied by other indicators turning, such as a yield curve inversion or tightening bank lending standards.

Why do spread movements have a larger non-default component than expected?

Because credit markets are not frictionless. Bonds are illiquid relative to public equities, dealer balance sheets are constrained, and many holders are forced sellers under certain conditions. All these create a price of risk-bearing that varies with sentiment and balance sheet capacity, independently of fundamental default probabilities. This is the structural reason for the empirical decomposition results.

How fast can spreads tighten after peaking?

Very fast in liquidity-driven stress. After the Fed announced corporate bond purchase facilities in March 2020, IG spreads compressed from 4% to under 2% in roughly two months. After the European Central Bank’s 2012 backstop announcement, peripheral sovereign spreads compressed dramatically over several months. The speed of recovery often exceeds the speed of the original widening, because the risk premium component reverses faster than the default premium.

Last updated — 14 June 2026