A capital stack is a queue for losses, and the order is the whole product
The Big Short · 2015
The idea
Put a pool of loans into a structure and the structure does one thing: it decides who loses money first. The bottom slice absorbs the first defaults, the next slice the next, and the top slice is touched only once everyone below it is wiped out. That ordering is what lets the same pool of risky loans produce both a high-yield slice and a slice rated as safe. Nothing about the loans has changed; only the queue has. Understanding this makes most structured credit questions mechanical: ask how thick the slices below yours are, and what rate of loss would eat through them.
Seniority is a queue for losses; the thickness of the slices beneath you is your protection.
When it works
Any structured credit, securitisation or capital structure question, in an interview or on the desk.
When it may fail
Subordination protects against expected losses in a known distribution. It does nothing when the distribution itself was wrong, which is what the correlation assumption insight is about.
How to use it
Run the Capital Stack lab: set a loss rate on the pool and watch which tranche is hit. Then state, in a sentence, the loss rate at which your tranche starts to lose, which is the number a buyer of that tranche actually cares about.
Where it came from
Film The Big Short · 2015
What the source says. The film's explanations of the CDO show a pool of mortgages sliced by seniority, with the top slices rated highly because the slices below would absorb the early losses.
Connected ideas
Worth knowing first
A rating is a model output, and the model rests on a correlation assumptionAsk what a rating assumes about defaults happening together before trusting what it says about the pool.
Worth knowing first
A capital structure is a set of rights people will litigate overIn distress, a claim is worth what its documents let it argue for, not where it sits on the slide.
Concepts
Put it to work on L3VLUP
More like this
- LessonRiskLetter located
Beware of geeks bearing formulas: a model built on calm history cannot see the storm
Ask what period a model was fitted to, and whether that period contained the event you are worried about.
From Berkshire Hathaway's 2008 letter to shareholders · Warren Buffett
- FrameworkCreditLetter
The duration and covenants of a loan matter as much as the amount of leverage
Ask when the debt comes due, what it can reach, and what trips it, before asking how much there is.
From Brookfield Asset Management's 2008 letter to shareholders · Bruce Flatt
- ArgumentArtificial intelligenceBook
A model trained on a biased outcome will reproduce it at scale
A model predicts the past decisions in its training data; if those were biased, the model is a bias machine with a straight face.
From Weapons of Math Destruction · Cathy O’Neil
- FrameworkRiskFilm
A risk limit is a decision about who has to be woken up
A limit is an escalation rule: it decides when a problem must move up the firm, not just how big it may be.
From Margin Call · 2011