CLO Trading (Collateralized Loan Obligations): How the Tranches Pay

A federal appeals court once had to settle, as part of an entirely different legal question, exactly how many CLOs failed in 2008. Its own finding: no more than six, against 435 defaulted CDOs. The two instruments get treated as cousins in casual conversation and as the same product in a lot of interview prep. They are not.

What follows is one CLO built in real numbers: a tranche stack sized off one asset manager's own 2025 research, the waterfall that pays it, and an overcollateralization test that fails and diverts cash, worked twice to check the arithmetic. Leveraged loans are the collateral here; MBS and ABS trade a separate desk, and how the underlying loans get originated in the first place is its own subject entirely.

What a CLO actually is

A collateralized loan obligation buys a pool of leveraged loans, usually more than 200 of them, and finances that pool by selling investors tranches with different claims on the cash it produces, rather than one uniform bond. Leveraged loans are senior secured debt to below investment grade companies, floating rate, and typically priced at SOFR plus 300 to 600 basis points.

Every tranche has a different claim on the cash flow the pool produces. The senior-most, AAA-rated tranche gets paid first and loses last. The equity tranche at the bottom is unrated, carries no fixed coupon, and simply collects whatever residual cash is left once every debt tranche has been paid.

One large asset manager's own 2025 research gives a typical stack, priced as of May 2025: AAA tranches usually run around 65 percent of the structure, each mezzanine tranche from AA down to BB usually runs 4 to 12 percent, and equity typically runs 8 to 10 percent. Real deals vary by vintage and manager; this is one dated source's typical range, not a rule.

Repaid from the top down. Each band is sized by its share of the structure.

AAA65% · SOFR+142bp, 6.9yr weighted average life

AA9% · SOFR+191bp, 8.9yr weighted average life

A4% · SOFR+228bp, 9.4yr weighted average life

BBB← you, as the mezzanine buyer6% · SOFR+353bp, 10.0yr weighted average life

BB7% · SOFR+650bp, 10.3yr weighted average life

Equity9% · unrated, residual cash flow only

Percentages and pricing follow one large asset manager's typical stack, priced in May 2025. They describe no single real deal, and the dollar amounts are built for this example.

More than half the entire loan pool would have to default before that AAA tranche takes a single dollar of loss, and that figure already assumes the loans that do default still recover a realistic share of par, the way senior secured loans typically do. Five layers of subordinate debt and equity sit beneath it and absorb losses first. That subordination, not the credit quality of any one loan, is what earns the top of the stack its rating.

Who trades with the desk, and what they want

Real money accounts, insurance companies, banks and money managers dominate the senior end of the stack, because a highly rated, floating rate bond is a natural match for a long dated liability. Specialist credit managers, hedge funds and other CLO managers take on more of the risk the further down the stack an investor sits.

AAA

8%

AA

36%

A

25%

BBB

29%

BB

68%

B

85%

Equity

75%

November 2024 notional data from one large asset manager, citing Bank of America. Real money holds most of the senior stack; specialists take on more risk further down.

Today's buyer base looks nothing like the one that existed before the crisis. Hedge funds, structured investment vehicles and bank trading desks running heavy leverage dominated CLO ownership before 2008. Post-crisis regulation has pushed most of that leverage out of the market, and today's buyer base is mostly large institutional asset managers, banks and insurers, far less prone to forced selling when markets move against them.

That shift matters for how the desk actually trades. A buy-and-hold senior investor base means investment grade CLO tranches turn over slowly with a stable bid-offer spread. The more opportunistic buyers further down the stack trade more often, and in smaller size.

A separate, older Federal Reserve breakdown of the domestically held CLO investor base found insurance companies holding 27.7 percent of tranches, mutual funds 15.5 percent, depository institutions 15.0 percent and pension funds 9.9 percent, with the rest split across other financial and non-financial holders. The tranche-level pattern above and this account-type pattern describe the same buyer base from two different angles.

How the desk makes money

Bid-offer spread on new-issue and secondary tranches is the base of the business, the same as any flow desk. A dealer that holds inventory funds it in repo, and the gap between what a tranche yields and what it costs to finance is carry, the same arithmetic that shows up on every other desk on this floor.

One dealer bank runs a public index business on top of that flow. J.P. Morgan built and still runs CLOIE, which it describes as the first rule-based total return benchmark for the broadly syndicated US CLO market, part of the same securitized index family as its ABS and MBS indices.

A J.P. Morgan pricing affiliate values the underlying tranches every day from secondary trading levels and new-issue pricing, and the index turns those prices into one number a portfolio manager can benchmark against.

The market by the numbers

$1.4tn

CLO market size

inside $13.3tn of structured credit

61%

of 2024 new loans CLOs bought

the dominant buyer, not a niche one

64%

of the loan market CLOs own

Bank of America data, 2025

CLOs are a $1.4 trillion asset class inside a $13.3 trillion US structured credit market that also includes mortgage and asset backed securities, as of the end of 2024. The market has grown from a post-crisis trough of $263 billion on the strength of investor demand for a floating-rate structure with a strong historical credit record.

That same investor demand is what makes CLOs the loan market's largest single buyer: they purchased 61 percent of all new-issue leveraged loans in 2024 and now own 64 percent of the loan market outstanding, more than any other buyer type.

Globally, one large asset manager separately puts the CLO market at $1.5 trillion by early 2026, roughly in line with the size of the US high yield bond market. Alongside the broadly syndicated market covered here, a smaller middle-market CLO segment adds roughly $131 billion outstanding, financing loans made directly to smaller companies rather than syndicated ones, a share that has roughly doubled over the past decade.

Middle-market managers are treated differently under the rules below: they remain subject to risk retention and typically do not issue below investment grade tranches, unlike the broadly syndicated managers this page mostly covers.

CLO vs CDO: the 2008 record, checked

Ask a candidate to separate a CLO from a CDO and a surprising number reach for "CLOs are just corporate-loan CDOs," which gets the history backwards. A CLO is backed specifically by senior secured corporate loans. A pre-crisis CDO was more often backed by subprime mortgage bonds, and frequently re-securitized into a "CDO squared" built from other CDOs.

A federal watchdog's own report, citing the Bank for International Settlements, puts almost 70 percent of newly issued CDO collateral in 2006 as subprime mortgage-backed securities, with another 15 percent as other CDOs. That structural gap showed up directly in performance, not just in theory.

A 2018 federal appeals court ruling, deciding an entirely unrelated risk-retention case, found in its own reasoning that no more than six CLOs defaulted during the 2008 crisis and all six carried features atypical for the product, against 435 defaulted asset-backed CDOs.

The same 2020 federal watchdog report puts a number on the CDO side of that gap, citing an academic study: AAA-rated CDO tranches lost an estimated $325 billion in the years after 2008. One major rating agency, by contrast, found zero losses on any AAA CLO tranche it rated, issued before or after the crisis, through September 2020.

BlackRock's own site states the same finding for a general reader: no AAA-rated CLO has ever defaulted, and CLOs took only minimal losses through the Global Financial Crisis. The structural reasons are specific, not lucky. CLOs are actively managed rather than static, are not commonly re-securitized, and are not widely linked to the kind of credit default swap exposure that multiplied losses inside crisis-era synthetic CDOs.

How risk retention changed, and the 2018 ruling that split the market

Dodd-Frank told regulators to make "any securitizer" of an asset-backed security retain a slice of its own credit risk. Six federal agencies wrote a rule in 2014 requiring five percent risk retention, phased in over the following two years, and applied it to CLO managers along with everyone else.

The Loan Syndications and Trading Association sued. Its argument: a CLO manager directs a special purpose vehicle to buy loans on the open market from independent third parties and never holds those loans itself, so it is not a "securitizer" transferring or retaining anything in the statute's own sense. In 2018, the D.C. Circuit agreed, and vacated the rule as it applies to open-market CLO managers.

The Volcker Rule shaped the desk in a different way. It bans short-term proprietary trading by a bank but explicitly carves out market making, underwriting and risk-mitigating hedging, and it separately excludes loan securitizations from its "covered fund" definition entirely. A bank can warehouse loans, structure a CLO and make markets in its tranches. What it cannot do is dress up a proprietary bet as any of those three things.

Warehousing: how a CLO gets built before it prices

Before a CLO exists, its manager needs a portfolio to put inside it, and that takes months to assemble one loan at a time. A warehouse period covers that gap: a warehouse lender finances the manager's purchase of leveraged loans, and the warehouse loan is expected to be repaid from proceeds once the CLO actually prices and closes.

• Warehouse period: three to nine months, financed by a warehouse lender, repaid from CLO proceeds at closing
• Ramp-up period: another three to six months, buying the rest of the portfolio with the new CLO's own proceeds
• Reinvestment period: up to five years, during which the manager can actively trade the pool and reinvest principal
• Amortization: cash pays down debt until the structure winds down, usually eight to ten years after it started

Non-call protection sits inside that reinvestment window too: equity investors cannot call or refinance the debt tranches for an initial period, typically six months to two years, before that option opens up.

A CLO that fails to price during its warehouse period leaves the warehouse lender and the equity investor holding the underlying loans directly, rather than a finished, tranched security. That downside is exactly why manager selection and warehouse terms get negotiated carefully well before a deal ever reaches the broader market.

The waterfall and the coverage tests

Cash from the loan pool pays the CLO's tranches in strict order: senior-most first, junior tranches next, and equity last. The industry calls that distribution method the waterfall. Two coverage tests exist specifically to protect that order from breaking down when the underlying loans stop performing.

An overcollateralization test, or OC test, checks that the loan pool's par value still exceeds the outstanding balance of a given tranche and everything senior to it. An interest coverage test, or IC test, checks the same relationship for interest income rather than principal.

If either test fails on a given tranche, cash that would otherwise flow to equity and any junior tranche is redirected instead, applied to pay down the senior-most debt until the ratio clears its trigger again.

An overcollateralization test failing, worked in numbers

Take the $400 million structure above. At closing, the loan pool's par value matches the $400 million of issued notes, so the AAA-only OC ratio is $400 million divided by the $260 million AAA balance, or about 154 percent, comfortably above an illustrative 130 percent trigger. Covering every rated tranche through BB, the ratio is $400 million divided by $364 million, or about 110 percent, above an illustrative 105 percent junior trigger.

Now assume loan defaults erode the collateral pool by $50 million, down to $350 million. The AAA-only test still passes, at roughly 135 percent. The junior test does not: $350 million divided by $364 million comes to about 96 percent, below the 105 percent trigger, and the test fails.

Once that happens, cash that would have gone to equity gets redirected to pay down the AAA tranche first, since AAA sits inside every OC test's denominator. Curing the junior test back to exactly 105 percent takes a specific amount: divide $350 million by 1.05 to get a target debt balance of about $333.3 million, then subtract that from the current $364 million balance. The required paydown comes to about $30.7 million.

Check the paydown two ways. Solve forward: $364 million minus $30.7 million leaves a $333.3 million balance, and $350 million divided by $333.3 million comes back to almost exactly 1.05. Solve backward: $350 million divided by 1.05 is $333.3 million, and $364 million minus $333.3 million is the same $30.7 million. Both routes land on the identical number.

Notice what the stress did and did not touch. The same $50 million loss cut the senior test from about 154 percent to about 135 percent, still well clear of its trigger, while the junior test fell straight through its floor. A loss hits the thinnest cushion first, which is why junior tests trip long before a senior one does.

CLO managers, and why tiering moves tranche prices

Large multi-strategy managers run CLO platforms alongside credit specialists most readers have never heard of. BlackRock, PIMCO, Ares, Blackstone and Carlyle all manage collateralized loan obligations, on top of everything else they do.

Manager selection matters more the further down the stack an investor sits. Trade press covering both the European and US markets describes a widening "tiering" effect: tranches from a manager with a strong record of avoiding problem credits price tighter than tranches from a manager whose older deals carry more defaults, sometimes by a meaningful margin on the same rating in the same week.

In the middle-market segment, manager selection matters more still, because deal flow runs on relationships and the underlying loans are far less transparent than in the broadly syndicated market.

A manager's fee is earned on the portfolio it runs, which is exactly why a deal-by-deal track record, not any single loan pick, is what a buyer is actually pricing when tiering shows up in a new tranche's spread.

Who the dealers are

The same ten dealer banks named throughout this site structure new deals, finance warehouses and make secondary markets in already-issued tranches. No reliable, current ranking of dealer market share in CLOs exists to repeat here; the one league table that still circulates is years out of date and names Credit Suisse, which UBS absorbed in 2023.

A junior's day on the desk

The day looks different depending on which side of the market a junior sits on. On the bank structuring desk, most of the work is building and marketing a new deal: running the debt stack through pricing models, producing the equity and debt materials investors see before a deal prices, and coordinating with the warehouse lender on timing.

On the manager side, a junior credit analyst spends more time on ongoing portfolio work. That means covering a specific set of loan issuers to support buy and sell decisions, and monitoring the OC and IC tests, concentration limits and eligibility criteria that keep an existing CLO in compliance day to day.

Backgrounds in bank-side CLO issuance work are described as good preparation for moving to the manager side of the same market later on, since building the debt stack for a new deal and monitoring an existing one are two views of the same underlying mechanics.

What the interview tests

Four things come up on almost every CLO or structured credit interview: the tranche and subordination structure worked through above, the OC and IC test mechanics, the CLO-versus-CDO distinction and its 2008 record, and open-market risk retention and why the 2018 ruling only reaches part of the market.

Being able to rebuild the OC test example above with different numbers, not just recite that a test failed, is what separates a strong technical answer from a memorized one. Bond math shows up here too, since a CLO tranche still prices off a yield and a spread even though its cash flows are more complicated than a plain bond's.

• Explain why a CLO's AAA tranche is more loss-remote than a similarly rated corporate bond
• Walk through what happens, step by step, when an overcollateralization test fails
• Separate a CLO from a CDO without being prompted, including what happened to each in 2008
• Know which CLO managers are subject to risk retention today, and which are not

A strong answer to the overcollateralization question names the direction as well as the mechanism: cash moves up the stack, not down, and equity feels it first, because its distributions are the ones redirected.

The broader fixed income interview preparation, and a bond calculator that works price, yield and duration together, round out the desk-agnostic half of the prep.

Where CLO desk experience leads

Career guides describe the most common lateral move from a CLO analyst seat as another CLO manager, since the skill set, reading a coverage-test calendar and pricing tranche risk, transfers directly. Credit-focused hedge funds are a real but more selective option: they tend to want analysts drawn to deep, single-name research rather than the broader, shallower coverage a CLO portfolio demands.

Distressed desks and credit trading draw on adjacent but distinct skills, walking a capital structure rather than a tranche stack. Bank-side structuring experience is itself described as good preparation for a later move to the manager side of the same market, since the two jobs are really two views of the same mechanics.

How to prepare

• Rebuild the OC test example above with a different collateral loss and a different trigger level
• Practice explaining CLO versus CDO out loud in under a minute, including the 2008 record
• Read a real CLO's coverage-test language in an offering memorandum if one is available
• Know which dealer banks name a specific CLO or structured credit business on their own site, and which do not

A structured run through the interview readiness tool is a reasonable last step before a first round, once the mechanics above can be rebuilt from memory rather than read off the page.

The bottom line

CLOs are not a niche corner of fixed income. They are the leveraged loan market's largest single buyer, and the desk that trades them sits on a genuinely different instrument from the CDOs that defined 2008. The tranche stack, the waterfall and the coverage tests worked through above are the mechanics. Getting the CLO-versus-CDO history right, unprompted, is what actually separates a strong answer from a rehearsed one.