Skip to content
Menu
Finabase Learn

Bonds · Credit risk

Bond default and recovery: what bondholders get back, and how often

What counts as a bond default, how bankruptcy and distressed exchanges work, how often each rating has defaulted since 1981, and what spread covers the loss.

On this page
  1. What counts as default
  2. What happens next: Chapter 11 and Chapter 7
  3. Recovery rate and loss given default
  4. How often each rating has defaulted
  5. The spread that only pays for expected loss
  6. Pitfalls
  7. How Finabase does it
  8. Frequently asked questions
  9. Summary
  10. Check your understanding

A bond is a promise to pay interest on set dates and the principal at the end. Default is the promise broken. For a Treasury the possibility is treated as remote; for a company rated B it is a real part of the bargain. Over the 44 years of S&P Global Ratings' records, about one B-rated company in five defaulted within ten years.

Default is not the same as losing everything. Bondholders usually get part of their money back, through a bankruptcy plan, a sale of assets or a new set of securities. How much they recover matters as much as whether default happens. This lesson defines default as the rating agencies count it, walks through what happens next, reads the agencies' default statistics, and turns them into the spread a bond would need just to break even.

What counts as default

S&P counts an issuer as defaulted on the earliest of four events: its rating cut to 'D' or 'SD', a missed debt payment, the announcement of a distressed exchange offer, or a bankruptcy filing. A missed payment counts unless S&P expects it to be made within five business days when there is no grace period, or within the stated grace period or 30 calendar days, whichever is earlier.

  • Missed payment. Many bond contracts give a short grace period for a late interest payment. A payment made inside it is late, not a default.
  • Bankruptcy. A filing under the US Bankruptcy Code, or its equivalent elsewhere.
  • Distressed exchange. A company in distress offers bondholders something worth materially less than the original promise, such as new bonds with a lower coupon, a later maturity or a smaller face value. S&P treats such an exchange as tantamount to default. No payment is missed, but holders take a loss.

In S&P's 2024 study the mix was lopsided: distressed exchanges made up 59.3% of all defaults, bankruptcies of various types 22.1% and missed interest or principal payments 18.6%. S&P counted more distressed exchanges than in any year since 2008, and linked the rise partly to private equity owners trying to cut debt while keeping control of their companies, and partly to persistently high interest rates.

Two contract features spread a default. Cross-default clauses make a default on one debt an event of default on others. Acceleration lets holders, usually through the bond trustee, declare the whole principal due at once rather than on the original schedule. Together they explain why a company rarely defaults on just one bond. The exact terms are in each bond's indenture; Bond seniority and covenants: who gets paid first in a default covers them.

What happens next: Chapter 11 and Chapter 7

In the United States, a company that files for bankruptcy usually picks one of two chapters.

Chapter 11 Chapter 7
Purpose Reorganization: keep the business alive Liquidation: sell the assets
Who runs it Usually the company, as debtor in possession A trustee
How creditors are paid Under a plan, often with new debt or shares From the sale proceeds, in the order the law sets
Creditors' role Classes whose claims are cut vote on the plan File claims and receive distributions

The US Courts describe Chapter 11 as generally providing for reorganization, with a plan that creditors whose rights are affected may vote on and the court may confirm. Chapter 7 is liquidation: a trustee gathers and sells the debtor's nonexempt assets and pays creditors from the proceeds. In Chapter 11 the company normally carries on as "debtor in possession" until a plan is confirmed. A business worth more running than sold in pieces has a reason to prefer it.

Either way, bondholders become claimants. What they receive depends on where their bonds rank against other creditors, on the value of the business, and on time: cases can take months or years, and recoveries are often paid partly in new securities rather than cash.

Recovery rate and loss given default

The recovery rate is the share of a claim that holders get back, usually measured as the defaulted bond's trading price shortly after default or as the value of what is eventually distributed. Loss given default (LGD) is the rest: 1 − recovery rate.

Example: $100,000 of a bond with a 3% probability of default in the coming year and an assumed 40% recovery. LGD is 60%, and the expected loss is 0.03 × 0.60 × $100,000 = $1,800.

Recovery varies widely with seniority, collateral and industry. Secured debt tends to recover more than unsecured debt, and subordinated debt less, because of the order in which claims are paid. The rating agencies publish recovery studies, but this lesson could not read a current one from the agencies themselves, so the 40% used here is an assumption for illustration, not a statistic. The method matters more than the number: the calculations below show how the result moves when the assumption changes.

How often each rating has defaulted

S&P's annual study follows every company it rated, grouped by rating at the start of each year, and records how many defaulted within one, two, up to fifteen years. The averages over 1981–2024:

Rating Within 1 year Within 5 years Within 10 years
AA 0.02% 0.28% 0.65%
A 0.05% 0.39% 1.03%
BBB 0.14% 1.36% 2.86%
BB 0.56% 5.75% 10.44%
B 2.93% 15.60% 22.02%
CCC/C 26.12% 46.53% 50.43%
Investment grade 0.08% 0.77% 1.69%
Speculative grade 3.54% 13.64% 19.15%

The averages hide large swings from year to year. The speculative-grade default rate was 9.70% in 2001, 9.89% in 2009, 5.54% in 2020 and 1.68% in 2021; in 2024 it was 3.94%, with 145 defaults worldwide, and 5.1% in the US. Defaults cluster in recessions, which is when holders can least afford them. Credit ratings and credit spreads: how bond risk is priced explains the rating scale itself.

From an annual rate to a cumulative one

If a bond had the same chance p of defaulting every year, its chance of defaulting within n years would be:

Apply it to each rating's one-year rate and compare with what actually happened:

Ten-year cumulative default rate: S&P history vs the 1-year rate repeated S&P average, 1981–2024 1-year rate repeated 10 years 2.9% 1.4% BBB 10.4% 5.5% BB 22.0% 25.7% B 1.7% 0.8% Inv. grade 19.1% 30.3% Spec. grade Percent of issuers. Source: S&P Global Ratings 2024 default study, Table 24. Repeated: 1 − (1 − p)^10.
Ten-year cumulative default rates, S&P Global Ratings 2024 study, Table 24 (averages 1981–2024), against 1 − (1 − one-year rate)^10. Computed by Finabase.

The shortcut fails in both directions. For BBB and BB it understates the ten-year risk by about half, because companies that start at BBB are often downgraded later and then default at a lower rating's rate. For B and the speculative grade as a whole it overstates it, because the weakest companies default early and the survivors are, on average, stronger. A constant annual rate is a useful approximation over one or two years, not over ten.

The spread that only pays for expected loss

A lender who wants to end the year no worse off, on average, than with a Treasury needs a higher coupon to cover expected losses. For a one-year bond bought at par:

For the $100,000 example, with the 1-year Treasury at 4.47% on October 28, 2026, the coupon must be 6.464%: a spread of 1.994 points. Check it: 97% of the time the holder receives $106,464, and 3% of the time $40,000; the average is $104,470, exactly what the Treasury pays. The spread depends heavily on the recovery assumption: 2.612 points at 20% recovery, 1.375 at 60%.

Using each rating's average annual default rate over ten years from S&P's table:

Spread that only pays for expected default losses, by rating AA 4 bp A 7 bp BBB 19 bp BB 71 bp B 162 bp CCC/C 469 bp Annual default rate from S&P 10-year averages; 40% recovery assumed; one-year par bond.
Annual default rate implied by S&P’s 10-year average cumulative rates (1981–2024); recovery assumed at 40%; one-year par bond against a 4.47% Treasury. Computed by Finabase.
Rating Average annual default rate Expected loss per year Break-even spread
AA 0.065% 3.9 bp 4.2 bp
A 0.103% 6.2 bp 6.7 bp
BBB 0.290% 17.4 bp 18.7 bp
BB 1.097% 65.8 bp 71.5 bp
B 2.457% 147.4 bp 162.4 bp
CCC/C 6.777% 406.6 bp 468.7 bp

Compare that with the market. The US Treasury's high-quality corporate (HQM) 10-year rate, built from bonds rated A or better, was 0.87 points above the 10-year Treasury yield in September 2026. Expected default losses at an A rating explain about 7 basis points of it. The rest pays for other things: uncertainty about when defaults will come (they cluster in bad times), the lower liquidity of corporate bonds, and taxes, since corporate interest is taxed by states and Treasury interest is not. For high-grade bonds, expected loss is a small part of the spread; for CCC bonds it is most of it. Yield spread measures: nominal, G-spread, I-spread, Z-spread and OAS covers how spreads are measured.

Pitfalls

  • Treating a rating's one-year default rate as the bond's risk. Over a 10-year bond, BBB's history implies about 2.86%, not 0.14%.
  • Treating default as total loss. Recovery is rarely zero; the loss is PD × LGD.
  • Treating recovery as a fixed number. It depends on seniority, collateral, the industry and the state of the economy at the time.
  • Reading a wide spread as a free return. A spread can be larger than expected losses and still be fair pay for losses that arrive all at once.
  • Ignoring distressed exchanges. A bond can default without missing a payment.

How Finabase does it

Finabase does not rate bonds or publish default probabilities. It shows how markets price default risk:

  • HYG holds high-yield bonds, the ratings where the default rates above are highest. Its price chart shows how the market reacts when investors expect more defaults.
  • LQD holds investment-grade bonds, where expected default losses are small and Treasury yields do more of the moving (Corporate and high-yield bonds: how company debt behaves compares the two).
  • The Treasury page gives the risk-free yields that the break-even formula starts from.

The default statistics are S&P Global Ratings' published averages. The cumulative-rate conversions and break-even spreads were computed by Finabase with the formulas shown; recovery is an assumption.

Frequently asked questions

What happens to my bond if the company goes bankrupt?

Interest payments usually stop and your bond becomes a claim in the bankruptcy case. In a Chapter 11 reorganization you may receive cash, new bonds or shares under a plan; in a Chapter 7 liquidation you are paid from the sale of assets in the order the law sets.

What is a typical recovery rate on a defaulted bond?

It varies with seniority, collateral and the economy. Secured bonds usually recover more than unsecured or subordinated ones. Use the rating agencies' published recovery studies for current figures; this lesson uses an assumed 40%.

How likely is a BBB-rated bond to default?

In S&P's data for 1981–2024, 0.14% of BBB-rated companies defaulted within a year, 1.36% within five years and 2.86% within ten.

What is the difference between default and bankruptcy?

Bankruptcy is one kind of default. A missed payment after the grace period, or a distressed exchange in which holders accept less than promised, is also a default, without any court case.

How do you calculate expected loss on a bond?

Multiply the probability of default by the loss given default (1 − recovery rate) and by the amount invested. A 3% chance of default with 40% recovery on $100,000 gives $1,800.

Summary

A bond defaults when a payment is missed beyond its grace period, when the issuer files for bankruptcy, or when holders are pushed into a distressed exchange, which made up most defaults in 2024. In S&P's data since 1981, about 3% of BBB issuers and 22% of B issuers defaulted within ten years, and a constant annual rate is a poor guide over such spans. What a holder loses is the probability of default times the loss given default, and the spread that only covers that loss is small for high grades, 7 basis points at A with 40% recovery, against market spreads many times larger.

Check your understanding

5 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. A company swaps its bonds for new ones with a lower coupon and a later maturity, without missing a payment. Does S&P count this as a default?

    1. No, because no payment was missed
    2. Yes, if the company is distressed and holders get materially less than promised: a distressed exchange
    3. Only if a court approves it
    Show the answer

    B. Yes, if the company is distressed and holders get materially less than promised: a distressed exchange S&P treats a distressed exchange as tantamount to default. In 2024 such exchanges were 59.3% of the defaults it counted.

  2. A bond has a 3% chance of default this year and an expected recovery of 40%. What is the expected loss on $100,000?

    1. $3,000
    2. $1,800
    3. $1,200
    Show the answer

    B. $1,800 Expected loss = PD × (1 − recovery) × exposure = 0.03 × 0.60 × $100,000 = $1,800.

  3. BBB-rated companies defaulted at 0.14% within one year in S&P’s 1981–2024 data. Repeating that rate for ten years gives 1.4%. The actual ten-year average was 2.86%. Why is the shortcut too low?

    1. Because many BBB companies are downgraded later and then default at a lower rating’s higher rate
    2. Because the data contain errors
    3. Because default rates are always constant
    Show the answer

    A. Because many BBB companies are downgraded later and then default at a lower rating’s higher rate A rating is a starting point. Over ten years some BBB issuers slide toward high yield, where annual default rates are many times larger.

  4. Which Chapter of the US Bankruptcy Code generally provides for reorganization of a business under a plan creditors vote on?

    1. Chapter 7
    2. Chapter 11
    3. Chapter 13
    Show the answer

    B. Chapter 11 Chapter 11 is reorganization; Chapter 7 is liquidation by a trustee who sells the assets and pays creditors.

  5. Expected default losses at an A rating come to about 7 basis points a year, yet high-grade spreads were about 87 basis points in September 2026. What does this suggest?

    1. High-grade bonds are mispriced and offer a free return
    2. Most of a high-grade spread pays for things other than average default losses, such as the timing of defaults in bad times, liquidity and taxes
    3. The default statistics must be wrong
    Show the answer

    B. Most of a high-grade spread pays for things other than average default losses, such as the timing of defaults in bad times, liquidity and taxes A spread is not only an expected loss. Investors also demand pay for losses that arrive together in recessions, for weaker liquidity and for state taxation of corporate interest.

Sources and further reading

  1. S&P Global Ratings, “Default, Transition, and Recovery: 2024 Annual Global Corporate Default And Rating Transition Study,” March 27, 2025 — Tables 1 and 24, definition of default (copy published by S&P Global Ratings Maalot)
  2. United States Courts — Chapter 11: Bankruptcy Basics
  3. United States Courts — Chapter 7: Bankruptcy Basics
  4. US Department of the Treasury — HQM 10-year corporate bond rate (FRED series HQMCB10YR) and Federal Reserve Board 10-year Treasury constant maturity (GS10)