A tranche in finance is a slice of a pooled debt security, carved out with its own payment priority, risk level, and yield. The word comes from the French for “slice.” When a bank or asset manager bundles thousands of mortgages, auto loans, or corporate loans into a single pool, the resulting cash flows get divided into layers so that a pension fund needing predictable income and a hedge fund chasing high yields can invest in the same underlying pool, each buying the layer that matches their appetite for risk.
Where Tranches Come From
Tranches are the product of securitization. Individual loans get bundled into a single pool, and claims on that pool’s cash flows are sold to investors. The assets might be residential mortgages, auto loans, credit card balances, or corporate debt. Once bundled, the loans generate a stream of monthly payments from borrowers, and that stream becomes the raw material for the tranches.
The pooled assets are transferred to a Special Purpose Vehicle, usually a limited liability company or trust set up for the sole purpose of holding those assets and issuing securities. The SPV has no employees and makes no business decisions. It exists to keep the asset pool legally separate from the company that originated the loans, so that if the originating bank goes bankrupt, its creditors can’t reach the assets inside the SPV. That insulation is what makes the securities viable for outside investors. The SPV then issues multiple classes of securities against the same pool. Those classes are the tranches.
The Payment Waterfall
The defining feature of any tranche is where it sits in the payment order, commonly called the waterfall. Cash collected from borrowers flows into the SPV and gets distributed by strict hierarchy. The waterfall determines who gets paid first, who absorbs losses first, and how much yield each investor earns for their level of risk.
Senior, Mezzanine, and Equity
Tranches are organized into three broad tiers.
Senior tranches sit at the top. They receive interest and principal before anyone else, which gives them the strongest protection against losses in the underlying loan pool. Credit rating agencies frequently assign senior tranches AAA or AA ratings even when the individual loans in the pool carry lower credit quality.
Mezzanine tranches occupy the middle. They get paid only after senior obligations are met, typically carry ratings in the A to BBB range, and offer higher yields to compensate for the added risk.
Equity tranches, sometimes called junior or first-loss tranches, come last. They collect whatever is left after everyone above them is paid. When borrowers default and the pool takes losses, the equity tranche absorbs them first. Its size effectively sets the loss cushion protecting the tranches above it. Equity tranches are usually rated below investment grade or not rated at all.
How Losses Flow: A Simple Example
Imagine a $100 million loan pool divided into $80 million of senior debt, $15 million of mezzanine debt, and $5 million of equity. Each month, borrower payments flow into the SPV. Administrative fees come off the top. Then interest goes to the senior tranche holders, followed by the mezzanine holders, with whatever remains flowing to the equity holders.
Principal follows the same seniority. In many structures, all principal repayments retire the senior tranche first. Only after the senior tranche is fully paid off does the mezzanine start receiving principal. The equity collects last.
Now suppose defaults cause $4 million in losses. The equity tranche absorbs the entire hit and is left with $1 million. Mezzanine and senior are untouched. If losses reach $8 million, the equity is wiped out and the mezzanine takes a $3 million loss. The senior tranche still comes through whole. Losses would need to exceed $20 million before senior holders lose a dollar.
This is why the same pool of below-average loans can produce a senior tranche rated AAA. The structural protection, not the underlying loan quality alone, drives the rating. That distinction matters because many institutional investors, including banks, insurance companies, and pension funds, face regulatory restrictions on buying anything below investment grade.
Protection Beyond the Waterfall
Subordination is the most visible form of credit enhancement, but structures typically stack several others on top of it.
- Overcollateralization. The face value of the loan pool is set larger than the total value of securities issued against it. If $100 million in loans backs only $90 million in securities, that extra $10 million creates a buffer that absorbs defaults before bondholders feel them.
- Excess spread. Borrowers in the pool pay a higher interest rate than the weighted-average coupon owed to tranche holders. That gap generates additional cash each month that can cover losses or build up the overcollateralization cushion. If borrowers pay 7% and the securities carry a blended coupon of 4%, the 3% difference is available to absorb shortfalls.
- Coverage tests. Many structures include tests that redirect cash flows when the pool’s health deteriorates. If the principal value of the loan pool drops below a threshold relative to the outstanding debt, payments that would have gone to the equity tranche get diverted upward to pay down senior debt. These triggers act as automatic circuit breakers.
In a well-structured deal, subordination, overcollateralization, and excess spread all need to be exhausted before senior tranche holders see a loss.
Where You Encounter Tranches
Mortgage-Backed Securities and CMOs
Mortgage-backed securities were among the earliest and most widespread applications of tranching. Residential mortgages are pooled, and homeowner payments form the cash flow that gets sliced. Government-sponsored enterprises like Fannie Mae acquire mortgages from lenders and securitize them, and Fannie Mae guarantees timely payment of principal and interest on its MBS, which shifts credit risk away from investors.
Collateralized Mortgage Obligations, or CMOs, extend tranching to address prepayment risk. When interest rates drop, homeowners refinance and investors get principal back earlier than expected. When rates rise, prepayments slow and investors sit on below-market yields longer than planned. CMOs carve up that prepayment uncertainty. Sequential-pay CMOs direct all principal to the first tranche until it retires, then to the second, and so on. Investors wanting shorter maturities buy early tranches; those willing to wait buy later ones.
Planned Amortization Class tranches, or PACs, push the idea further. A PAC is designed to receive principal at a predetermined pace within a specified prepayment band. Companion or “support” tranches absorb the variability. If prepayments run faster than expected, the support tranches soak up the extra principal, keeping the PAC’s cash flows stable. If prepayments stay high long enough to pay off the support tranches entirely, the PAC “busts” and starts behaving like a regular sequential CMO.
Collateralized Loan Obligations
CLOs apply the tranche structure to corporate loans, typically leveraged loans issued to companies with below-investment-grade credit. The U.S. CLO market has grown to roughly $1 trillion in outstanding securities. Unlike a static mortgage pool, a CLO is actively managed: a fund manager buys and sells loans within the portfolio throughout the deal’s life, subject to constraints. Interest payments flow through the waterfall to tranche holders in order of seniority.
CLOs include structural tests that monitor the health of the loan pool. An overcollateralization test checks whether the aggregate principal of the loans still exceeds the outstanding CLO debt. An interest coverage test checks whether interest collected from borrowers is sufficient to cover interest owed to tranche holders. If either test is breached, cash that would have gone to the equity tranche gets redirected upward to pay down senior debt.
Collateralized Debt Obligations
CDOs are a broader category. Where CLOs are backed by bank loans, CDOs can pool corporate bonds, other asset-backed securities, or even other CDO tranches. In the years before the 2008 crisis, CDOs frequently repackaged the lower-rated mezzanine tranches of mortgage-backed securities into new structures, with rating agencies assigning fresh AAA ratings to roughly 80% of the resulting CDO tranches.1Financial Crisis Inquiry Commission. The CDO Machine – FCIC Final Report Chapter 8
A synthetic CDO doesn’t hold any actual loans or bonds. Instead, it uses credit default swaps to create exposure to a reference portfolio of debt. Investors are essentially selling insurance against defaults in that portfolio. This unfunded structure allowed Wall Street to build virtually unlimited exposure to the mortgage market without needing to find new loans to securitize.
How Tranches Are Priced
Investors demand compensation for the credit risk, liquidity risk, and prepayment risk embedded in a tranche. That compensation shows up as a spread over a benchmark interest rate, currently the Secured Overnight Financing Rate, or SOFR, in the U.S. dollar market.
A senior AAA-rated CLO tranche might price at SOFR plus around 100 to 150 basis points. A mezzanine tranche rated BBB could demand SOFR plus 300 to 500 basis points. The equity tranche, which has no fixed coupon and receives only residual cash flows, might target total returns of 12% to 18%. Higher ratings compress spreads because a wider pool of institutional buyers can participate, while lower-rated or unrated tranches draw a narrower set of specialized investors willing to accept illiquidity in exchange for yield.
Financial modeling drives the pricing. Analysts project cash flows under various economic scenarios, estimating default rates, recovery rates, and prepayment speeds. For senior tranches with relatively predictable cash flows, the modeling is straightforward. For equity tranches, where the outcome depends on what’s left after everyone else is paid, small changes in assumptions produce wildly different return projections. That difficulty is one reason equity tranches are illiquid and tend to be held to maturity by specialized funds.
What Can Go Wrong: The Lesson of 2008
The financial crisis of 2007–2008 was in many ways a story about tranches gone wrong. Between 2003 and 2007, Wall Street issued nearly $700 billion in CDOs backed by mortgage-backed securities. Rating agencies rated CDO tranches based on their own earlier ratings of the mortgage-backed securities inside the CDOs, without examining the actual underlying mortgages. Moody’s acknowledged it lacked meaningful historical default data for these structures and, as one internal assessment put it, essentially made up the correlation assumptions that drove its models.1Financial Crisis Inquiry Commission. The CDO Machine – FCIC Final Report Chapter 8
Roughly 80% of CDO tranches received AAA ratings despite being built primarily from the lower-rated mezzanine tranches of mortgage-backed securities. Major financial institutions retained exposure to what they believed were the safest portions of these CDOs. When housing prices fell nationally and mortgage defaults spiked, the underlying securities turned out to be far more correlated than anyone had modeled. They stopped performing at roughly the same time. In 2007, 20% of U.S. CDO securities were downgraded. In 2008, that figure was 91%.1Financial Crisis Inquiry Commission. The CDO Machine – FCIC Final Report Chapter 8
Synthetic CDOs amplified the damage. Because they used credit default swaps rather than actual loan pools, there was no natural limit on how much exposure could be created. Goldman Sachs alone packaged 47 synthetic CDOs with a combined face value of $66 billion between 2004 and 2007. When losses hit, the sellers of credit default swap protection, most notably AIG, faced obligations they couldn’t meet. The FCIC concluded that declining demand for the riskier tranches of mortgage-backed securities drove the creation of CDOs, which in turn fueled demand for more subprime lending in a self-reinforcing cycle.1Financial Crisis Inquiry Commission. The CDO Machine – FCIC Final Report Chapter 8
The lesson applies to any tranche, in any structure. A credit rating is not a guarantee. It reflects the rating agency’s model-driven estimate of loss probability under certain assumptions. When those assumptions are wrong, every tranche in the structure reprices, sometimes catastrophically.
The Risks Worth Understanding Before You Buy
Knowing where a tranche sits in the waterfall is necessary but not sufficient. Several risks cut across the capital structure.
Credit risk. The risk that borrowers in the underlying pool default at rates exceeding the loss cushion for your tranche. Senior tranches have thick buffers; equity tranches have none. Collateral quality and underwriting standards are the primary drivers.
Prepayment risk. When borrowers pay off loans early, investors get their principal back sooner than expected and must reinvest at potentially lower rates. This hits mortgage-backed tranches hardest, since homeowners routinely refinance when rates drop. PAC tranches insulate against some of this; support tranches absorb the volatility.
Extension risk. The flip side. When rates rise, borrowers hold onto their loans longer, and investors sit on below-market yields. This is particularly painful for longer-dated tranches.
Liquidity risk. Senior tranches with high ratings trade frequently among institutions. Equity and mezzanine tranches trade infrequently and at wider bid-ask spreads. Selling a junior tranche quickly almost always means accepting a discount.
Model risk. Pricing and rating tranches depends on assumptions about default correlations, prepayment speeds, and recovery rates. Correlation assumptions in particular are difficult to estimate from limited historical data, and small errors can cause large mispricings, especially in mezzanine and equity tranches. That is exactly what surfaced in 2008.
The appeal of tranching is real. It lets a single pool of loans serve investors with genuinely different needs, and it channels credit into the economy that a single-class security might not attract. But the same slicing that creates useful products also concentrates risk into specific layers, and the layers deepest in the waterfall behave very differently from the AAA slice on top of the same collateral. Reading a tranche means reading the waterfall, the enhancement, the collateral, and the model assumptions behind all three.