High-Yield Default Rates, Recovery, and When Credit Cycles Turn
Corporate Financial Distress, Restructuring, and Bankruptcy by Edward I. Altman, Edith Hotchkiss, and Wei Wang (Wiley, ISBN 978-1-119-48180-5)
Previous: High-Yield Bond Market Overview | Next: Altman Z-Score: Fifty Years of Credit Risk Models
The first half of Chapter 9 described how big the junk bond market got. This second half answers the question every investor actually cares about: what do you lose when things go wrong, and what do you actually earn after defaults eat into those juicy yield spreads?
Altman and his co-authors built the rigorous default rate framework that the industry eventually adopted. Back in the mid-1980s, the average annual default rate was just 1.4%, and some market players argued that was too high. Over the full 1971-2017 period, the dollar-weighted average climbed to about 3.4% per year. Bad years can push toward 10%. The market has seen five years at or above that level since 1978.
Defaults Lead Recessions (Sometimes)
One pattern jumps out from the data: corporate default rates often start rising two to three years before a recession begins. The 2008-09 crisis was the exception, when mortgage defaults led and corporate defaults moved in sync with the economic drop. By 2017, the U.S. was in its eighth year of below-average default rates, the longest benign stretch on record.
Altman also tracks nonfinancial corporate debt as a percentage of GDP. In the last three stressed cycles, debt-to-GDP peaked within about a year of the peak in high-yield default rates. In 2017, that ratio looked like it was peaking again. The market did not seem to care.
Recovery Rates Move in the Opposite Direction
Here is the relationship that still shapes how professionals think about credit losses. When default rates spike (1990-91, 2001-02, 2009), recovery rates on defaulted bonds fall to 25-35%. When defaults are low, recoveries run well above the historical average of about 45%. Altman, Brady, Resti, and Sironi showed that default rates alone can explain 58-62% of recovery rate variation.
Put it together and high-yield investors have historically lost about 2.4% per year from default losses (including lost coupons). Average promised yield spreads from 1978 to 2016 were about 5.2%. Subtract expected losses and you get an expected return spread of roughly 2.8%. Actual return spreads came in around 2.4-2.8%. The math works. The spread is compensation for risk, not free money.
What Makes a Benign Credit Cycle?
Altman defines a benign cycle through four variables:
- Low default rates
- High recovery rates on defaults
- Low yields and spreads over Treasuries
- Easy market liquidity (measured partly by CCC issuance share)
Historically, periods between stressed cycles lasted 4-7 years, averaging 5.5 years. By 2018, the post-2009 stretch had run nine years. Spreads in late 2017 sat around 3.9%, well below the long-term average of 5.2%. CCC-rated new issuance stayed elevated. Easy money was still the story.
Altman flagged bubble characteristics similar to 2007, but also noted mitigating factors. The four benign-cycle indicators all pointed to continuation in the short term. His concern sat in the longer-term risks: unprecedented cycle duration, high corporate debt-to-GDP, heavy CCC issuance, and LBO leverage.
LBOs: The Supply Engine
Private equity buyouts have been a constant source of high-yield and leveraged loan issuance. When competition for deals heats up, purchase price multiples rise. The 2016 average public-to-private LBO multiple of 11.2x was the highest Altman had ever recorded, beating even the late-1980s binge that fed the 1990-92 default wave.
Skeptics point to lower interest rates and more equity in modern deals. Fair enough. But debt-to-EBITDA ratios still sat just below 6x at the median, meaning half of all LBOs carried even more leverage. And floating-rate loans mean rising rates could hit hard.
Private equity firms do default less than comparable standalone firms, controlling for credit risk. Their access to alternative financing helps. But the multiples still make me nervous.
Junk Bonds and Stocks: More Correlated Than You Think
High-yield bonds and the S&P 500 showed a 59% return correlation from 1987-2017. In stressed periods like 2008-09 and 2010-17, that correlation jumped to 72-73%. The old idea that companies can easily switch from debt to equity financing breaks down when both markets tank together. Waiting until the bond market shows stress to raise equity often means doing it at terrible prices, or not at all.
Mortality Rates vs. Rating Agency Defaults
Altman’s “mortality rate” approach tracks bonds from their original issuance rating, weighted by dollar amounts. Rating agencies track issuer-weighted cumulative defaults across all bonds with a given rating at a point in time, regardless of age.
The difference is huge for new bonds. A single-B rated issue has a 2.86% first-year mortality rate in Altman’s data versus about 5% for Moody’s. By year four or five, the methods converge. For a bank making a new loan or an investor buying a fresh issue, the mortality approach is the right benchmark. For a mature portfolio, agency cumulative rates make more sense.
My Take
This section is Altman doing what he does best: taking a messy, politicized market and putting hard numbers on it. The 2.4% annual loss figure is the one I keep coming back to. Junk bonds pay more because you will lose money sometimes. The spread is not a gift.
The 2017 warnings read differently today, after COVID and another long stretch of low defaults. But the framework still holds. Watch default rates, recoveries, spreads, and CCC issuance. Watch debt-to-GDP. And do not assume that because the market survived the last crisis, the next one will look the same.
Previous: High-Yield Bond Market Overview | Next: Altman Z-Score: Fifty Years of Credit Risk Models