CLOs Explained: How Collateralized Loan Obligations Work

TL;DR: A collateralized loan obligation (CLO) is a securitization that buys a portfolio of leveraged corporate loans and funds itself by issuing rated debt tranches plus an unrated equity slice. Interest and principal cascade down a strict waterfall: senior tranches get paid first and absorb losses last, the equity tranche is paid last and absorbs losses first.

What is actually inside a CLO

Strip away the alphabet soup and a CLO is two things:

  1. A pool of leveraged corporate loans — typically 150 to 250 senior secured loans to speculative-grade (mostly single-B and double-B) borrowers, drawn from the broadly syndicated loan market that banks originate and then sell on.
  2. A capital stack issued against that pool — six or seven rated tranches from AAA down to single-B, plus an unrated equity tranche that absorbs first losses and collects the residual cash flows.

The New York Fed describes the structure plainly: CLOs "are securitization structures that allow syndicated bank lenders and bond underwriters to repackage business loans and sell them to investors as securities," and they "are actively overseen by a collateral manager that has the responsibility to trade loans in the portfolio to benefit from gains and mitigate losses from credit exposures." [NY Fed Liberty Street, 2016]

That active management is what separates CLOs from most other securitizations. A CLO manager — usually a credit-focused asset manager — selects which loans to buy at launch, then trades the portfolio during a reinvestment period of roughly five years. After that, the deal enters runoff and loan principal is used to pay down the rated tranches in order of seniority.

CLOs are structured as bankruptcy-remote special purpose vehicles, so the manager does not put its own balance sheet at risk. The investors in the tranches take the credit exposure to the loans.

The asset class is not small. The Federal Reserve's October 2023 Financial Stability Report put US leveraged loans outstanding at $1,358 billion as of the second quarter of that year, and the NY Fed reports that the volume of outstanding CLOs in the United States approached $647 billion at year-end 2019 — more than double the level a decade earlier. [Fed FSR Oct 2023] [NY Fed Liberty Street, 2021]

The tranche stack

Each tranche has three things that matter:

  • Its share of the capital stack (how much of the deal it represents)
  • Its rating (from AAA at the top to unrated equity at the bottom)
  • Its coupon, expressed as a floating spread over a reference rate (SOFR today, formerly LIBOR)

A representative US broadly syndicated loan (BSL) CLO looks roughly like this:

Tranche Rating Share of stack Spread over SOFR Position
Class A-1 AAA ~62–65% +130 to 160 bps First paid, last loss
Class A-2 / B AA ~10–12% +170 to 210 bps Senior mezz
Class C A ~5–7% +210 to 270 bps Mezz
Class D BBB ~5–6% +300 to 370 bps Mezz
Class E BB ~5–6% +550 to 700 bps Junior mezz
Class F (when issued) B ~1–2% +800 bps and up Deep subordinate
Equity Unrated ~8–10% Residual cash flows Last paid, first loss
Representative ranges for post-2017 US BSL CLOs. Actual structures vary by manager, vintage, and rate environment. Spreads are floating over SOFR. Source: NY Fed Liberty Street; Federal Reserve Financial Stability Reports.

How the cash flows actually move

Every payment period — typically quarterly — interest collected from the loan pool runs down a payment waterfall. The order matters more than any single line item, because money paid out at the top can never come back to make up shortfalls at the bottom.

A simplified interest waterfall in priority order:

  1. Senior trustee, administrative, and senior manager fees
  2. Class A (AAA) interest
  3. Class B (AA) interest
  4. Class C (A) interest
  5. Class D (BBB) interest
  6. Class E (BB) interest
  7. Class F (B) interest, if any
  8. Subordinated manager fee
  9. Equity — whatever is left

Critically, the waterfall has built-in stop signs called overcollateralization (OC) tests and interest coverage (IC) tests. If the loan pool deteriorates and the OC or IC ratios fall below specified thresholds, interest payments are diverted away from the equity and lower tranches and used to pay down senior debt until the tests cure. This is the structural mechanism that protects AAA holders.

CLO interest waterfallLoan pool interest cascades down through tranches in strict priority order, from senior fees and AAA at the top to equity at the bottom.Loan pool interest incomeSenior fees and expensesAAA (Class A) interestAA (Class B) interestA (Class C) interestBBB (Class D) interestBB (Class E) interestB (Class F) interest, if anySubordinated manager feeEquity (residual)Paid firstPaid lastLoses lastLoses firstOC/IC testsIf pool qualitydeteriorates,divert cash upto pay AAA early
Simplified interest waterfall in a typical CLO. The principal waterfall follows a similar order. Diagram: ECMSource. Mechanics described in NY Fed Liberty Street, "Are All CLOs Equal?" (2016).

A simplified worked example

Take a hypothetical $500 million US BSL CLO with the structure described above:

  • Loan pool: $500M of floating-rate leveraged loans, weighted-average coupon roughly SOFR + 350 bps
  • Capital stack: $310M AAA, $55M AA, $30M A, $30M BBB, $30M BB, $5M B, $40M equity

At a SOFR rate of 4.3%, the pool generates roughly $39M of gross interest in year one ($500M × 7.8%). Senior trustee and manager fees take the first few million. The rated tranches then collect their coupons in strict priority — AAA at SOFR plus around 145 bps (about $18M on the $310M AAA balance), with the AA, A, BBB, BB, and B tranches together consuming roughly $11M more at progressively higher spreads. The equity tranche sweeps the residual — a low-to-mid teens running yield on the $40M equity slice in a year without defaults.

Now assume year three brings a default cycle. Say $25M of loans (5% of the pool) default with a 60% recovery rate. The pool loses $25M × (1 − 0.60) = $10M of principal. Where does that loss go?

The equity holder, sitting at the bottom of the stack, absorbs the first dollar of loss. The $40M equity slice now has $30M of subordinated capital still protecting the rated tranches. The AAA holder remains nowhere near a loss — the entire equity, all six mezzanine tranches below AAA, and any OC cushion stand between the senior tranche and the next dollar of credit loss.

Loss absorption from the bottom up

The single most useful mental model for a CLO is the attachment point — the share of the pool that has to be wiped out before a given tranche starts taking losses. AAA is the last to take a loss because every cent of equity and every mezzanine tranche below it has to be destroyed first.

CLO tranche loss attachment pointsStacked diagram showing the share of the capital stack each tranche occupies and the order in which losses are absorbed, from equity at the bottom to AAA at the top.AAA — 62% of stackLoss only after 38% subordination wiped outAA — 11%A — 6%BBB — 5.5%BB — 5.5%B — 1.5%Equity — 8.5%FirstlossLastloss0%100%% of capital stackSizes are representative percentages, not specific deal terms.
AAA tranches sit on top of roughly 38% of subordination. For a AAA holder to lose principal, the equity, B, BB, BBB, A, and AA tranches must all be destroyed first. Diagram: ECMSource, based on tranche conventions described in NY Fed Liberty Street, 2016.

What CLOs actually survived in 2008 — and what they didn't

CLOs are often grouped together with the subprime CDOs that detonated the 2008 financial crisis. The structures look superficially similar, but they are not the same product. The senior tranches of cash-flow CLOs that were outstanding through the crisis performed dramatically better than CDOs of subprime mortgage-backed securities, and the difference came down to what was actually in the pool.

The New York Fed's research on CLOs originated between 2007 and 2011 found that manager skill and trading activity mattered — but the broad structural finding was that senior CLO tranches were largely insulated from credit losses, in contrast to the senior tranches of mortgage CDOs over the same period. [NY Fed Liberty Street, 2016]

The lesson is not that CLOs are safe by construction. It is that asset selection and correlation matter at least as much as the wrapper. CDOs of subprime mortgages failed because the underlying assets were correlated and mispriced, and the senior tranches were sized for losses that turned out to be far too optimistic.

Where the structure can still break down

CLO performance is loan-pool-dependent, not magic. Risks worth understanding:

  • Covenant-lite drift. The underlying leveraged loans have steadily lost the financial-maintenance covenants that historically forced borrowers to act early on distress. The Federal Reserve has repeatedly flagged this in its Financial Stability Reports. [Fed FSR]
  • Concentration in single loans or sectors. CLO documents cap exposure per borrower and per industry, but during a sector-specific recession the tail can still bite — energy in 2015 and 2020 was a textbook case.
  • Manager skill. Two CLOs with identical structures but different managers can produce very different equity returns. Trading and credit selection edge matter, and CLO managers are not interchangeable.
  • Equity loss in tail scenarios. Even when AAAs sit safely on top, the equity tranche can lose meaningful principal in a deep and prolonged default cycle.
  • Refinancing and reset cycles. CLO debt can be refinanced or reset, changing the economics for the equity holder. Late-cycle issuance with tight spreads can hand the equity a structurally worse deal than vintages issued in stress.

CLOs vs other structured credit

CLOs share the tranching DNA of the broader structured-credit family but differ on what matters most — the underlying asset.

  • CLO: Senior secured corporate loans, actively managed, senior tranches largely survived 2008.
  • CDO (of subprime ABS): Pools of mortgage-backed securities, mostly static, senior tranches suffered heavy losses in 2008.
  • CMBS: Commercial real-estate mortgages, pass-through structure, no active manager.
  • Consumer ABS (auto, credit card): Granular consumer receivables, very different default and prepayment dynamics from corporate loans.

If you remember one thing about structured credit, remember this: the wrapper does not determine the outcome. The asset does.

Related concepts to read next

Sources

Disclosure: This article was produced with AI assistance and reviewed before publication. It is for informational purposes only and is not investment advice.

Leave a Comment