How a Pile of Bad Mortgages Became a Triple A Bond
The 2008 crisis is usually explained as reckless lending. The more useful explanation is the machinery that turned risky loans into securities the whole world was willing to buy.
Start With the Loan
A mortgage is a promise to make monthly payments for thirty years. On its own it is illiquid, sitting on a bank's balance sheet consuming capital until it is repaid.
Securitization changed that. A bank sells thousands of mortgages to a separate legal entity, which pools them and issues securities backed by the combined payments. The bank recovers its capital immediately and lends again. Investors get exposure to mortgage payments without originating loans.
None of this is inherently dangerous. It had worked for decades.
The Tranche
The critical step is that the pool's cash flows are divided into layers called tranches, arranged in priority. The senior tranche is paid first. Below it sit mezzanine tranches, and at the bottom the equity tranche.
When homeowners default, losses hit the bottom first. The equity tranche is wiped out before the mezzanine takes a cent, and the mezzanine is exhausted before the senior tranche is touched.
That structure is what allowed a pool of risky loans to produce a highly rated bond. If losses would have to exceed some large threshold before the senior tranche lost anything, the senior tranche could plausibly be rated as safe as government debt, even though every underlying loan was risky.
The rating was never a statement about the mortgages. It was a statement about how much of the pool had to fail before the top slice took a loss.
The Assumption That Failed
The models supporting those ratings rested on correlation. If mortgage defaults across different regions were largely independent, then a pool spread across many states was genuinely diversified, and simultaneous nationwide default was extremely unlikely.
Historical data supported that. House prices had not fallen nationwide on a sustained basis in modern American history, and regional downturns had been just that, regional.
The assumption failed because the loans were no longer independent. They shared a common driver: national house prices, loose underwriting standards, and the availability of refinancing. When prices stopped rising, borrowers everywhere lost the ability to refinance at once. Correlation went toward one exactly when the structure depended on it staying low.
The Second Layer
The mechanism compounded. Mezzanine tranches, which were harder to sell, were themselves pooled into new securities called collateralized debt obligations, and those pools were tranched again.
So a senior tranche of a CDO was built from mezzanine slices of mortgage pools. The math produced a highly rated security from ingredients that were already the risky middle of something else. Each layer of repackaging made the final instrument more sensitive to the correlation assumption, not less.
Why Nobody Stopped
The incentives all pointed one direction. Originators earned fees and sold the loans onward, so they bore no default risk. Banks earned fees structuring deals. Rating agencies were paid by issuers and competed for that business. Investors reached for yield in a low rate environment and relied on the ratings rather than analyzing the pools.
Every participant behaved rationally within their own incentives, and the aggregate outcome was a system where nobody held the risk they were creating.
The Bottom Line
The crisis was not simply bad lending. It was a structure that converted correlation assumptions into credit ratings, distributed the result globally, and left no one in the chain holding the consequence of being wrong.