Credit ratings: what AAA and junk actually mean
The answer
A credit rating is one agency's opinion of how likely a borrower is to pay you back on time. It measures the chance of not being repaid. It says nothing about whether the bond is a good buy, what return you will get, or what the price will do tomorrow.
Why this costs you money
An investor compares 2 debt mutual funds. Fund A returned 6.8% last year. Fund B returned 8.9%. Both sit in the same section of the app. The investor picks Fund B because the number is bigger.
What actually happened is that Fund B lent money to weaker borrowers. The extra 2.1% was not skill. It was payment for the risk that some of those borrowers would not pay. Most years nobody defaults and Fund B looks better. Then one year several borrowers default at once, and Fund B falls 12% in a week while Fund A does not move.
This is not rare. It is the defining event of Indian debt fund investing in the last decade. It happened after IL&FS defaulted in September 2018, again to funds holding DHFL paper in 2019, and it ended with 6 Franklin Templeton schemes being wound up in April 2020.
A higher yield in fixed income is almost never a better product. It is a different product, and the difference is written on the credit rating. Learning the scale is how you see the trade before you take it.
How it works
Every rated bond carries a grade from an agency. The grades run in a fixed order, and the most important line on the scale is not at the top.
| Grade band | S&P and Fitch | Moody's | What it means |
|---|---|---|---|
| Highest | AAA | Aaa | Repayment capacity is extremely strong |
| High | AA+, AA, AA− | Aa1, Aa2, Aa3 | Very strong |
| Upper medium | A+, A, A− | A1, A2, A3 | Strong, more sensitive to conditions |
| Lowest investment grade | BBB+, BBB, BBB− | Baa1, Baa2, Baa3 | Adequate. This is the line. |
| Speculative | BB+, BB, BB− | Ba1, Ba2, Ba3 | Faces major ongoing uncertainty |
| Highly speculative | B+, B, B− | B1, B2, B3 | Currently paying, little margin |
| Distressed | CCC, CC, C | Caa, Ca, C | Vulnerable, dependent on conditions |
| Failed | D | — | In default |
The line that matters is between BBB− and BB+. Above it is investment grade. Below it is speculative grade, high yield, or junk. All 3 words mean the same thing.
That line is not a matter of opinion for large parts of the market. Pension funds, insurance companies and many mutual fund mandates may hold investment grade only. When a bond crosses that line, those holders are forced to sell — not because they changed their view, but because their rules changed for them.
This is why a downgrade hurts the price even when nothing has defaulted. The company is still paying every coupon. The set of people allowed to own its bonds has shrunk, and the price falls to whoever is left.
Now the part that explains most rating failures. The borrower pays the agency for the rating. This is the issuer-pays model. A company that wants to sell bonds hires an agency, and the agency's revenue depends on being hired. Agencies argue that their reputation is the asset they protect. That is true over decades. It has repeatedly not been true in the years that mattered.
A rating is also slow by design. Agencies aim for ratings that hold through a cycle rather than ratings that move with the news, so the market often knows first. A bond trading 4 percentage points above government bonds of the same maturity is being priced as junk, whatever grade is printed on it.
What it tells you, and what it does not
A rating tells you about one risk only: default.
It says nothing about the risk that dominates bond returns, which is what happens to the price when interest rates move. A 30-year AAA government bond is the highest credit quality available and can still lose 30% of its value in a year. Article 4 is about that.
It says nothing about liquidity, which is whether you can sell at all. An AA-rated Indian corporate bond may not trade for weeks. A BB-rated US bond in a large index may trade every minute.
And a rating is an opinion at a point in time, with a direction called the outlook. "BBB, negative outlook" is a different statement from "BBB, stable outlook", and the outlook is free information most people never look at.
What a rating does give you is a ranking that has worked on average across thousands of bonds and many decades. Higher-rated bonds default less often. The error is treating that average as a promise about the 1 bond in front of you.
The decision rule
Read the rating as a question, not as an answer.
If a bond yields much more than a government bond of the same maturity, ask what the rating is. If the rating is high and the yield is high, one of the 2 is wrong, and it is usually the rating.
If you cannot say why the extra yield exists, do not accept it. If you cannot name the risk, you are being paid for something you have not identified.
Unless the money must be certain on a fixed date. Then take the government bond and stop looking at the yield column.
For debt funds the rule is simpler. Read the credit quality breakdown, not the past return. Two funds with the same past return and different credit breakdowns are not 2 versions of the same thing.
Try this now
Three minutes, on the fund you actually hold. Most people who own a debt fund have never seen this page.
- Find the monthly factsheet for your debt fund. Search the fund's exact name plus "factsheet". US bond funds publish the same thing as "credit quality".
- Find the section called Rating Profile, Credit Quality or Asset Allocation by Rating. It is usually a small pie chart or a short table.
- Write down 3 percentages: sovereign or government paper, AAA or A1+, and AA and below.
- Add the third number to anything marked "unrated" or "below investment grade". That is your risk bucket.
- Multiply your investment by that percentage. If you hold ₹4 lakh and 18% is AA and below, then ₹72,000 of your money is lent to borrowers the agencies do not consider very strong.
What you should see. A liquid fund or a gilt fund will show almost everything in sovereign and AAA. A corporate bond fund will show a large AAA block with some AA. A credit risk fund will show a large AA and below block, which is exactly what it is designed to do.
The number in step 5 is the one to keep. It is the part of your "safe" money that depends on somebody specific staying solvent.
Three real cases
1. Moody's and the 2006 mortgage securities, 2007 (United States) — the rating that was wrong at scale Mortgage bonds were assembled and rated during the housing boom. In 2007 Moody's downgraded 83% of the $869 billion of mortgage securities it had rated Aaa in 2006. Aaa is supposed to mean an extremely small chance of loss. The commercial context is documented: by 2006 Moody's earned more revenue from structured finance, $881 million, than from its entire business in 2001. On 3 February 2015, S&P agreed to pay $1.375 billion to settle US Department of Justice and state claims that it had misled investors about its mortgage ratings. It did not admit legal liability.
2. IL&FS, 17 September 2018 (India) — top grade to default in weeks IL&FS held high investment grade ratings from ICRA and CARE through August 2018, while carrying group borrowings of roughly ₹1 lakh crore and a large negative net worth. On 17 September 2018 both agencies cut it to D, the default grade. SEBI later penalised ICRA, CARE and India Ratings over their handling of IL&FS ratings, raising the penalty on ICRA and CARE to ₹1 crore each in September 2020. Investors in debt funds holding IL&FS paper had been paid a small extra yield to hold a default.
3. Greece, 27 April 2010 — the downgrade that moved the price by itself Standard & Poor's cut Greece's sovereign rating to BB+, below investment grade, on 27 April 2010. Greece had not defaulted and would not restructure until March 2012, nearly 2 years later. The downgrade alone forced institutional selling, because many European funds were not permitted to hold sub-investment-grade sovereign debt. The price fell on the letter, not on a missed payment.
The question that resolves it
Two bonds mature in 4 years. One yields 7.2%. One yields 11.5%.
A novice asks: which one pays more?
An expert asks: what does the market think will happen to the second one, and do I disagree?
The extra 4.3 percentage points is not free money somebody forgot to collect. It is a price, set by people who have read the accounts. You may think they are wrong. You may not think the difference is not there.
What would make this wrong
If ratings were meaningless, AAA bonds and B bonds would default at similar rates. They do not. Long-run default studies published by the agencies themselves show a clear gradient across the scale.
The honest limits are 3. Ratings have failed worst precisely where the exposure was largest, which is the opposite of what you want from a safety measure. A rating is about default probability, not about how much you lose when default happens; a secured and an unsecured bond from the same company can carry similar default risk and very different recovery. And the whole scale is calibrated to a market, which is the subject of the comparison below.
In India
Three domestic agencies dominate: CRISIL, ICRA and CARE. India Ratings, owned by Fitch, and a few smaller agencies also operate. All are registered with SEBI.
The scale runs AAA down to D, the same letters used globally. Short-term issues, under 1 year, carry A1+ down to A4 and D, where A1+ is the highest.
Two Indian features are worth knowing.
Side pocketing. Since 2018, SEBI has allowed a debt fund to move a defaulted or heavily downgraded holding into a separate segregated portfolio. Existing investors keep their claim on it, and new investors do not inherit it. This stops the fastest investors from redeeming at full value and leaving the loss behind.
The Potential Risk Class matrix, in force since 1 December 2021. Every open-ended debt scheme must display a 9-cell grid. One axis is interest rate risk, Class I to Class III. The other is credit risk, Class A to Class C. A fund labelled A-I takes low credit risk and low interest rate risk. A fund labelled C-III can take high amounts of both. It is the most honest disclosure in Indian mutual funds and almost nobody reads it.
In the United States
Three agencies dominate globally and all 3 are American: Moody's, S&P Global Ratings and Fitch. Together the Big Three rate the overwhelming majority of rated debt worldwide. They are registered with the Securities and Exchange Commission as Nationally Recognized Statistical Rating Organizations, or NRSROs. Many US regulations refer to NRSRO ratings directly, which is part of why they are so hard to displace.
The US high-yield market is the important structural difference. Bonds rated below BBB− trade in enormous volume and have their own indices, ETFs and analyst coverage. An American investor can buy a basket of several hundred junk bonds in one trade. The market prices those bonds continuously, so the market's opinion is visible alongside the agency's opinion, every day.
Fallen angels and rising stars are standard US terms. A fallen angel is a bond downgraded from investment grade to high yield. A rising star is the reverse. Both are tracked because the forced selling and buying around that boundary is predictable.
Where they differ, and what that tells you
An Indian AAA and an American AAA are not the same grade.
Indian domestic ratings are on a national scale: a ranking within India, with the Indian government as the reference point. Global ratings are on an international scale. India's own sovereign rating sits around the bottom of investment grade — S&P raised it to BBB on 14 August 2025, its first upgrade in 18 years.
A company almost never rates above the government of its own country on the international scale, because it operates under that government's currency, taxes and courts. So an Indian AAA company translates to something in the BBB region internationally. It is a genuine top domestic credit. It is not what an American investor means by AAA.
The second difference is depth. The US high-yield market is deep and continuously priced. India's is thin, and much lower-rated Indian paper barely trades at all.
What that tells you points in one direction. In the United States, buying lower-rated bonds is a considered decision with visible pricing and an exit. In India, it is a decision to hold something you may not be able to sell in a crisis. Credit risk and liquidity risk arrive together, on the same day, for the same reason. That is what happened in April 2020, and it is why the extra 2% of yield in an Indian credit risk fund is not comparable to the extra 2% in an American high-yield fund. You are being paid for 2 risks and quoted for 1.
Carry this
- A rating measures the chance of not being repaid. Nothing else.
- The line that moves prices is between BBB− and BB+, because rules force selling across it.
- Extra yield is a price for a named risk. If you cannot name the risk, do not take the yield.