What it means
The starting point is a tranche nobody especially wants. Mezzanine slices of ordinary securitisations are awkward to sell, so arrangers gathered large numbers of them into a new pool and issued new tranches against it.
On paper this looked like diversification. The new pool contained pieces of dozens of separate deals, each of which already held hundreds of loans, which suggested that no single borrower could matter very much.
The arithmetic worked very differently. Because each mezzanine slice sits above an equity layer, it only begins to lose money once losses in its own pool pass a threshold, and once that threshold is crossed the slice loses value extremely quickly.
That makes a CDO2 highly sensitive to small changes in assumptions. A modest rise in losses across the underlying pools, not enough to trouble the senior tranche of any individual deal, can destroy most of the value of every mezzanine slice the second structure owns at the same time.
There is also the overlap problem. The same mortgage or loan frequently appeared, indirectly, in many of the underlying deals, so the apparent spread across dozens of structures was far narrower than it looked, and the ratings assigned to the top slices proved far too generous.
In practice
Real-world examples.
Example
An arranger struggling to place mezzanine tranches from six separate mortgage deals pools them into a new structure. The top slice of the new deal receives a high rating and sells easily, even though its collateral is the paper nobody wanted individually.
Example
An insurance company buys the senior slice of a second-layer structure believing it owns a diversified claim on thousands of loans. A later review finds that four of the eight underlying deals hold loans from the same two originators in the same two regions.
Example
A risk team is asked to stress-test a portfolio containing one of these structures. They find that moving the assumed loss rate on the underlying pools from 6% to 8% takes the position from a full recovery to losing more than half its value.
Formula
Calculation
Loss on the CDO squared collateral = Total of the losses suffered by each underlying tranche it holds, which then flows through its own tranche ranking
Suppose a CDO squared of $200,000,000 holds eight mezzanine tranches of $25,000,000 each, and each of those tranches sits in a $500,000,000 pool behind an equity layer of $25,000,000. Each mezzanine tranche therefore absorbs losses between 5% and 10% of its own pool. If every pool loses 8% of its value, that is $500,000,000 x 0.08 = $40,000,000 of losses, of which $25,000,000 falls on equity and $15,000,000 on the mezzanine tranche, a 60% loss on its $25,000,000. The CDO squared has therefore lost 8 x $15,000,000 = $120,000,000 of its $200,000,000 of collateral. If its own structure is a senior tranche of $140,000,000 and a junior tranche of $60,000,000, the junior tranche is wiped out and the remaining $60,000,000 of losses hits the senior tranche, cutting it to $80,000,000, a loss of $60,000,000 / $140,000,000 = 43% on a slice that was marketed as low risk.Case study
Seen in the real world.
Vellacourt Capital is an illustrative, fictional investor used to show how the layering behaves. Its mandate allowed only highly rated holdings, so it bought the senior slice of a second-layer structure on the strength of its rating and a two-page summary describing a pool of eight separate deals.
When a risk analyst finally traced the collateral through to the underlying loans, the picture changed. The eight deals shared originators, geography and loan vintage, and the whole position behaved as though it were a single concentrated bet with two extra layers of paperwork on top of it.
In the illustrative outcome the fund sold at a steep discount before losses arrived and rewrote its investment rules: no instrument whose collateral is itself a tranche of another structure, unless the underlying loan pool can be examined directly. The lesson in the story is not that layering is fraud, but that each layer multiplies the effect of being slightly wrong about the layer beneath it.
Watch out
Common mistakes.
- Reading a long list of underlying deals as proof of diversification, without checking whether those deals share originators, regions or vintages.
- Assuming the second structure behaves like the first, when the thin mezzanine slices it owns turn a small pool loss into a near-total loss.
- Relying on a credit rating as a substitute for tracing the collateral through to the loans that actually generate the cash.
Questions
People also ask.
Why were these structures created at all?
Mainly to find buyers for mezzanine tranches that were hard to place individually, by repackaging them into a pool and issuing a new highly rated slice against it.
What makes a CDO squared more fragile than a CDO?
Its collateral is already subordinated, so it starts taking losses only after a threshold is crossed and then loses value very quickly, which compresses the range between full payment and near-total loss.
Are they still sold?
They are largely gone from mainstream markets, partly through regulation on re-securitisation and partly because investors have little appetite for collateral they cannot look through.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related