A fallen angel's worst moment usually isn't the day the news breaks. It's the day an index rulebook, applied mechanically, pushes the bond out of the investment-grade universe and every manager benchmarked to that universe has to act at once. Ratings drift for months and then fall off a ledge. The ledge is regulatory, and the selling is scheduled.
The week Detroit fell out of investment grade
On May 5, 2005, Standard & Poor's cut both General Motors and Ford to junk. These were not obscure issuers. Counting their finance arms, the two companies carried roughly $450 billion of debt, and in a single afternoon a large slice of the US investment-grade corporate market became eligible for ejection.
The market had seen this coming. GM had issued a profit warning in March, its spreads had already blown out, and the rating agencies had kept both names on negative outlook for months. So why did the downgrade still matter? Because a credit opinion and a portfolio constraint are different things. Investors had already repriced the risk that GM might default. What they had not yet done was sell, since many of them were not allowed to hold junk and had not yet been forced to.
The aftershocks reached well beyond Detroit. Credit spreads widened across the high-yield market as dealers and hedge funds absorbed a supply wave the junk market was not sized for, and correlation trades built on the auto names unwound badly that spring. The lesson for anyone reading a rating action today is that the price damage from a fallen angel comes in two layers: the credit layer, which arrives gradually, and the mechanical layer, which arrives on a date.
How a rating is actually produced
A credit rating is a letter-grade opinion on the probability that an issuer pays its debt in full and on time. The three dominant agencies are S&P, Moody's and Fitch. S&P and Fitch use AAA down through BBB- for investment grade (IG), then BB+ and below for high yield (HY), also called junk. Moody's runs a parallel scale, with Baa3 as its lowest IG notch and Ba1 as its highest junk notch.
The rating does not come from a model output. It comes from a rating committee, a group of analysts who vote on a recommendation from the lead analyst after reviewing leverage, cash flow, business position and management guidance. Committees are deliberately conservative about movement. They signal first through an outlook (positive, stable or negative, typically a six-to-24 month view) and then through a watch or review placement, which flags that action is likely within about 90 days.
That process explains the shape of rating changes. Agencies smooth out noise, so ratings lag market prices, sometimes by quarters. Then, when the committee finally acts, the change is discrete: one notch, sometimes two, on a specific date. Credit spreads, the extra yield a bond pays over Treasuries, move continuously. Ratings move in steps.
The practical point is that the outlook and the watch placement are the forecast; the rating action itself is the confirmation, and the confirmation is what triggers the rules.
Where intuition gets it wrong: the second agency matters more than the first
Most readers assume the first downgrade to junk is the trigger. Usually it isn't, because the index providers that define "investment grade" for benchmark purposes don't rely on any one agency.
Bloomberg's widely used US corporate indices apply a middle-rating rule: if all three agencies rate a bond, the index uses the middle of the three; if only two, it uses the lower; if one, it uses that one. Work it through with three ratings on a hypothetical issuer.
Index rating = middle of (Moody's, S&P, Fitch)
- Starting position: Moody's Ba1, S&P BBB-, Fitch BBB. Converted to a common scale, that is junk, IG, IG. The middle rating is BBB-, so the bond stays in the IG index despite Moody's already calling it junk.
- S&P then cuts to BB+. Now the set is Ba1, BB+, BBB. The middle is BB+, and the bond exits the IG index at the next month-end rebalance.
The first junk rating (Moody's, in this example) produced headlines and a spread move. The second one produced forced selling. A split-rated BBB- issuer with one agency already at junk is therefore far more fragile than its index status suggests, because a single committee vote stands between it and ejection.
The timing detail compounds this. Index providers typically rebalance at month-end, so a downgrade on the 3rd of the month and one on the 27th both hit benchmark holders on the same day. Passive funds and benchmark-hugging active managers know that date in advance, and so does every dealer on the other side of the trade.
When you look at a BBB- name, count the agencies still holding it in investment grade. One is a cliff edge; two is a cushion.
Why the holders have to sell
The selling is forced by a stack of constraints, each with its own logic.
Index funds and ETFs tracking an IG benchmark must drop the bond when the index drops it, or accept tracking error they are paid to avoid.
Mandated active managers often have investment policy statements that cap or ban sub-IG holdings. Many mandates include a grace period, which softens the timing, but a grace period postpones the sale rather than cancelling it, and everyone knows the deadline.
Insurance companies face the sharpest incentive through capital charges. US insurers hold bonds under NAIC designations, a regulatory credit scale where NAIC 2 corresponds roughly to the BBB band and NAIC 3 to the BB band. The risk-based capital (RBC) charge, the regulatory capital an insurer must hold against an asset, rises several-fold across that boundary. As an illustrative example: if an insurer holds $100 million of a bond carrying a charge of around 1.3% as NAIC 2, it ties up about $1.3 million of capital. If the charge steps up toward roughly 4.5% at NAIC 3, the same position ties up about $4.5 million. Nothing about the bond's cash flows changed between those two numbers. Only its label did.
Bank capital rules, pension investment guidelines and central bank collateral frameworks all draw versions of the same line. That is why the IG/HY boundary behaves like a wall rather than a gradient. A move from A to A- changes almost nothing about who may own a bond. A move from BBB- to BB+ changes the entire buyer base.
Academic work on insurer behaviour supports the mechanism: research by Ellul, Jotikasthira and Lundblad found that capital-constrained insurers sold downgraded bonds more aggressively, and that prices of those bonds fell below fundamental value and then partially recovered. That reversal is the fingerprint of a fire sale: price pressure driven by who must sell, not by new information.
The takeaway: the size of the cascade depends less on how bad the credit news is and more on how much of the bond is held by people with rules.
The supply shock, and who absorbs it
The other side of the trade is the high-yield market, which is far smaller than the IG market. Roughly half of US investment-grade corporate debt sits in the BBB band, a share that grew substantially over the past decade, while the high-yield market is a fraction of the IG market's size. When a large BBB issuer falls, the junk market must absorb a supply of bonds that can rival a meaningful slice of its existing size, and it must do so on a known date.
Buyers exist: HY funds, distressed and crossover investors, dealers. They demand a concession for taking on the inventory, and they know the sellers are forced. That concession is the fallen-angel discount. It tends to open in the weeks around index exit and narrow afterward, once the bond has migrated from holders who must sell to holders who chose to buy.
This is why dedicated fallen-angel strategies exist. Funds such as the VanEck Fallen Angel High Yield Bond ETF (ticker ANGL) buy bonds after they have been downgraded out of IG, aiming to collect the discount that forced sellers leave behind. Comparing such a fund with a broad high-yield fund like iShares iBoxx High Yield Corporate Bond ETF (HYG) separates two things that are easy to conflate: the return from owning junk credit generally, and the return from being the buyer of last resort to rule-bound sellers.
For a reader, the downgrade discount is a liquidity premium for supplying capital on someone else's deadline; it is not a judgement that the issuer is about to default.
When the cascade doesn't happen as scheduled
The mechanism has failure modes, and they matter as much as the mechanism.
First, the market can front-run the cliff so completely that the rebalance itself is a non-event. When a name sits on negative watch for months with one agency already at junk, much of the selling migrates earlier, and the month-end can arrive with prices already through the floor.
Second, a policy backstop can break the link between the label and the buyer base. In March 2020, a wave of downgrades hit names including Ford, Occidental Petroleum and Kraft Heinz, producing one of the largest years of fallen-angel volume on record. The Federal Reserve's corporate credit facilities then explicitly made eligible issuers that had been rated investment grade as of March 22, 2020 and remained at least BB- afterward. The central bank declared itself a buyer for exactly the bonds the rules were forcing out, which compressed the fire-sale discount sharply.
Third, rating inertia can cut both ways. Agencies sometimes hold an issuer at BBB- longer than its leverage justifies, especially when management commits to deleveraging. Those rising stars in reverse, issuers that avoid the cut, deny the cascade entirely, and anyone positioned for forced selling gets nothing.
The practical rule is to treat the cascade as probable, not certain, and to check whether the selling has already happened or whether a large buyer stands ready to absorb it.
Reading a rating action the way the market does
The mechanism compresses into one sentence: a fallen-angel cascade occurs when a committee decision changes a bond's regulatory label across the IG/HY wall, forcing rule-bound holders to sell into a smaller market on a predictable date.
What changes once you see it that way is where you look. Don't focus on the downgrade headline. Watch the count of agencies still at investment grade, the middle-rating arithmetic, which holders own the bond (index funds and insurers versus unconstrained credit funds), and the next month-end rebalance date. A negative outlook on a split-rated BBB- issuer with heavy insurance ownership is a scheduled event in waiting. A downgrade that leaves two agencies at IG is mostly noise for forced flows, however loud the coverage.
