Financial Theory: Lecture 25 Transcript
December 8, 2009
Professor John Geanakoplos:
So I'm going to talk about what led up to the crisis and how the crisis makes
us realize that the things about the theory that we've been learning all semester that ought to be slightly
different, and I've been working on a new version of the theory for the last ten years, what I call the Leverage
Cycle, which I have to say didn't get very much attention ten years ago, but which now seems to be, I don't
know, it seems to be getting more attention.
It's almost like life imitated art, really. It couldn't have turned out more like the theory than I would have
expected. So I'm going to tell you the facts and then try to fit a theory to it, and probably will spill over until
Thursday, and I think Thursday I won't spend that much time finishing. I'll have a review class if anyone's
interested for the whole material.
And I think when you see the critique of the theory it might help put the theory in a little bit more
perspective. But anyway, nothing of what I say these last two days you should feel is going to be on the
exam. It's just more information that you might find interesting.
So we talked last time, remember, at one of those special classes about the sub-prime mortgage market, and
the idea was that you would pool together a bunch of loans. These are the--I can do this, I guess. You'd pool
together a bunch of individual mortgages like this, and then you'd put them in a giant pool, and then cut the
pool up into a triple-A bond, a double-A bond, a single-A bond, a triple-B bond, and this is the over
collateralization and the residual.
And I'm not going to go through the whole structure, but suffice it to say that the idea of this is you take
individual loans which are very risky, after all these are sub-prime borrowers so everyone knew that they
were risky, and you create what's supposed to be an incredibly safe bond. The definition of triple-A bond is
that it has a 1 in 100 chance of going bankrupt in 10 years, of losing any principal in 10 years.
That was the definition they claimed. I heard a talk about power failures last night, by the way, the power grid
which went out all across from Cleveland to Manhattan in 2003, and they said that was an event that should
happen once every 10 years. And then when you hear what happened in the event, someone didn't trim a tree,
and the line sagged into the tree, and so then it shut itself off, and then the city next door they had the same
problem with someone not trimming a tree, and that line shut off, and once you have two or three lines shut
off the system automatically goes down. It seems like it could happen much more often than once every 10
But anyway, it's the same thing here. This is supposed to be once every 10 years, 1 in 100 chance, and it
happened in a giant way. So the idea of the triple-A, so they're supposed to very safe because if the first house
defaults, the first homeowner defaults, instead of paying back the 100 he owes, let's say he just stops paying,
and then you throw the guy out of the house, but then you get to sell the house.