11. Institutions and Incentives in Mortgages and Mortgage-Backed Securities

Prof: Okay let’s begin.
We’ve got a terrific guest
today and I’m just going to briefly introduce him and then
leave forty-eight minutes for you to enjoy what he has to say.
Will Goetzmann,
a graduate of this college in what year?
Will Goetzmann: 1978.
Prof: The class of 1978,
with a professional and grown up career in art history and in
finance, and that’s not an easy combination.
He does it with grace and with
ease. He, like Sharon Oster,
is a standout teacher in the School of Management MBA
program. He directs The International
Center for Finance and is– the catholicity of his
interests, which we’ll see in his slides and in his analysis
is truly impressive. Please help me welcome Will
Goetzmann. Will Goetzmann:
Well, thanks a lot for the introduction,
and I’m–I was thrilled with the idea of this course so was
really looking forward, as Doug was describing,
the idea of the course and his plans for it and what you all
have been doing. It seems like a fantastic thing.
So the question in my mind was,
what could I add today that would be of interest to you,
and I thought what I would do is focus a little bit on the
historical background to the current crisis.
I know you’ve been doing some
reading about the current crisis and its relationship to ideas of
capitalism and so forth. I thought a little bit of
historical foundation would be a good thing to lead off with and
give you an idea of how old these financial crises are.
This is a cuneiform tablet from
the old Babylonian period, roughly 1600 B.C.
It’s a loan,
so that it was created with the terms of the loan on one side,
and typically, although you can’t see it from
this side, you’ll have people that are
witnesses to the loan on the other side.
So debt has been around for a
long time. One interesting feature about
debt is that usually when you–if you’re a lender,
you like to have some collateral.
The collateral for debts in
this time period often were human capital.
In other words,
if you defaulted on your loan, the borrower could seize you
for three years and your family and you’d have to work the loan
off. That sort of debt slavery was
something that began quite early in human culture.
You had–when you had–with
that debt slavery there would be periods when people would get
themselves into a terrible mess, and let’s say crops failed and
lots of farmers wouldn’t be able to pay off their loans,
you would have lots of the society in deep debt.
And the king,
periodically, would decree all debts null and
void, so wiping the slate clean. So here is a fragment by edict
Samsuiluna, who was the son of Hammurabi,
who, by the way, lived in the city that Yale has
had a long time expedition to excavate called Tal Alon.
Anyway this is releasing
everybody from bankruptcy in the form of debt slavery.
But the issue there,
of course, is where you draw the line?
Which debts do you release,
and if you just want to take people out of slavery,
but you want to maintain a financial system,
how do you differentiate? The solution to that problem,
at least from the view of writers in antiquity,
particularly Aristotle, who wrote about Solon,
comes in the form of another major debt crisis.
When Solon of Athens became the
autarch, or whatever you want to call it, the leader of Athens,
it was about 600 B.C. or so, somewhere in that
timeframe, and he was brought in
because there was a terrible crisis,
much like the crises in the Middle East–
it was a period when Athenian citizens had been seized for
debts, and these were debts incurred
typically by farmers, and then they’d be sold off
outside of Athens, they’d just sell them offshore
and export these debt slaves, and it created terrible crisis.
So, Solon was from a wealthy
family, he was elected by the–sort of
the rich people in Athens, they thought that he would
stand up for property rights and he wouldn’t let the financial
system go to hell just because of the pain inflicted on the
poor portion of society. What he did made nobody happy.
He said, “We’re going to
preserve property rights, we’re going to preserve the
right to contract, but the only thing we’re going
to forbid is the right to sell yourself into slavery,”
so he drew the line at that debt slavery point.
Ever afterwards,
people–Athenians would look at him,
he sort of was like the George Washington of Athens,
and 200 years later, people would write about this
moment in time when he made the differentiation between the
financial system and the system of human slavery.
It was not to say that slaves
didn’t exist, but the ability to contract on
your freedom was not allowed after Solon.
I’m going to skip forward
through a lot of exciting financial history.
Those of you that like that
kind of stuff might consider taking–
I teach a course seminar, senior seminar on financial
history in the spring, so if you are interested in
those topics, I would love to have you
participate. I’m going to skip forward to
our debt crisis because I worry–
I stay up late nights worrying that we’re getting ourselves
back into a debt slavery situation,
and the recent code that we–the bankruptcy code has made
it hard for people to shift off their debts,
or shake off their debts, which is the term that Solon
used. It was called
seisachtheia or something like this, a shaking off,
like an earthquake. But I’m going to skip forward
to 1892 and have you read this quote because it lets you have
some sense of where the policy for real estate,
mortgage lending in the United States might have come from.
The occasion for this was a big
festive dinner. It might have been at
Delmonico’s Restaurant, I’m not sure exactly where,
I don’t think it says here, but in New York and a bunch of
guys in the real estate business in New York were getting
together and they were really kicking off an attempt to create
a real estate exchange in New York City.
And the exchange–there had
been something like a real estate exchange where properties
were listed and brokers were admitted and so forth–
the idea that they were coming to was that they needed an
exchange for real estate securities.
The real estate securities that
we have now are– a lot of them are things
like–you’ve heard lots about Fannie Mae and Freddie Mac,
and subprime loans and commercial mortgage backed
securities and things like that, well these guys were thinking
about setting up exchange to trade debt,
maybe some equity too actually, in New York.
And New York would be the
center of real estate trading. What were they going to trade?
Well here’s a picture,
just a piece of a bond of the kind they were going to trade.
These things are incredibly
beautiful. The New York–the U.S. Bank
Note Company made bonds for–made stocks and bonds and
money for all the countries in the world,
and they had incredibly detailed filigrees and these
beautiful motifs and so forth. It wasn’t to make these things
more attractive to customers. All of this detail was really
to prevent counterfeiting, because if you could
counterfeit a bond like this you could take it in and get
coupons. So what’s this bond?
It’s a $500 bond,
it’s issued to finance something called 44 Wall Street
Corporation, and it’s The Manhattan Company building.
Well, 44 Wall Street is still a
building in Manhattan. I don’t know if anybody here
might have worked there during the summer,
but there was a corporation that was set up to build and
operate one single building, and there was some equity
capital in the corporation and there was some debt in the
corporation. The debt was issued to the
public in bonds. You and I could go out and we
could buy these bonds and then the coupons,
the money that would come in from the bonds is money that
would come in based upon the rents that were being paid by
people that leased space in this 44 Wall Street building.
This is a fairly extraordinary
kind of security, particularly because we don’t
have securities like this anymore.
I mean there are some bonds
that have been issued against buildings,
and I’m thinking now of things like–
in London you had this vast project in the Canary Wharf,
and that was financed with some really interesting debt.
But one to one,
building to bond, really we don’t have in the
modern day. This one, by the way,
was issued one month or so, right after–not long after the
great crash. Here’s a little bit about this
market. The market was nothing
much–this is the total cumulated amount of outstanding
debt based on the face values. The face value of that one was
$500, you’d multiply that times the
total number of bonds, you add all of those bonds that
existed up in this– these are the numbers you’re
getting: $100 million, $200 million and so on.
You see that even in 1913 there
were some of these bonds being issued and then there are two
companies here that we’re tracking,
New York Title & Mortgage Company and Lawyer’s
Mortgage Company. These two companies actually
served as guarantors for debt. In particular,
actually, what they would do, about eight companies got
together–or didn’t get together, but eight companies
offered a service. They would take a residential
mortgage, put a guarantee on it, and then sell that mortgage to
somebody else. Just like you have bond–one
commercial building offered through bonds,
you also had one mortgage for one house then guaranteed and
offered for sale. These firms also began to pool
these mortgages, and you’d pool together 100 or
so of these mortgages, or a thousand of these
mortgages, put a guarantee on them and then sell those pooled
claims out as bonds. So this picture represents not
just those commercial bonds that I showed you,
but also residential securities.
And you can see what happened
is that they were coming along doing fine,
small part of the business, and then in the ’20s both of
them picked up dramatically. As a matter of fact the First
World War was an important hinge point in the history of this
particular kind of financial innovation.
The First World War was the
first U.S. war that was financed by
massive issues of war bonds, and it became your patriotic
duty to buy war bonds to finance the effort.
When the war was over and those
bonds were reclaimed by the government,
they were paid off by the government,
people had been used to owning securities.
A whole operation to sell these
securities had existed, a sales operation,
so then the brokers said, “What else are we going to
sell?” Real estate bonds became a
really important new product. So you can see this might have
been driven by demand, actually, for investment
product in a marketing system that had grown for savings bonds
earlier during the war. Nevertheless,
you could see this stuff really take off and peak around
19–late ’20s. That was cumulative,
this is just new issues. There you see the drop around
the crash time even more dramatically.
This is a picture of
those–plotting those number of new issues for commercial
bond–for commercial properties, against new buildings.
Here we only took a look at new
buildings that were seventy meters high or so,
so we’re looking at really tall buildings.
This is a time period when–the
skyscraper was really born in the 1890s, and really got going
in the early part of the twentieth century.
So a question that we’ve begun
to ask ourselves is, maybe the skyscraper was a
response to the emergence of a new capital market for fixed
income securities. Maybe the financing drove the
desire for big buildings rather than the other way around.
I mean New York and other
cities had existed for a long time without skyscrapers and
there’s a lot of interesting theory about how skyscrapers
were kind of a result of changes in zoning laws in some places
and so forth. But if you think about it,
London did fine without skyscrapers for many years.
Paris, London,
other great capitals of the world did fine without
skyscrapers. Why would you suddenly have
these sort of immediate blossoming of skyscrapers
largely coincident with an emergence of a capital market
and ability to sell these bonds? By the way how would–is there
any way you can think of testing that theory about which–
if my theory was finance made skyscrapers or finance led to
big buildings, how would you test that theory
if you had some data? Prof: Come on guys,
let’s take a shot. Will Goetzmann: Anybody?
I’m going to cold call on
somebody, but your chance of being cold called–yes,
back there. Student: Maybe you
could do a comparison, look at other countries where
for some reason– perhaps there is a case where
there’s another country where the capital markets didn’t
mature in the same way. I mean there’s going to be some
endogeneity issues anyway with this study,
but that’s a start, and in comparing if–
how their timing was with their–the start of skyscrapers.
Will Goetzmann:
Yeah, so look across country,
and cross-sectionally, you’d like to find conditions
that maybe didn’t have this financial explosion.
Student: Perhaps you
might find some exogenous reason why they didn’t have the capital
markets that are completely independent of the other
factors, but that’s hard to do.
Will Goetzmann:
Okay, does everybody know what endogeneity is?
Everybody who is an Econ.
major and you’re a senior or a
junior you probably know. Other than that you probably
think it’s some kind of horrible skin disease.
Endogeneity means that the
factors that you’re studying could actually be–
it could be some reverse causality going on,
or there could be some–a common unidentified factor that
could be driving both of them, and you have to find some way
of sorting this out. The question is–so the one way
to do this is to find some independent phenomena to
observe. That’s what’s being proposed.
Any other suggestions?
Here’s some college–
Student: I was just going to say that you could
always not only look at the rest of the world but look at,
in the United States, the development of these types
of bonds and the number of skyscrapers like with this,
and I would wager that probably the first skyscrapers were not
financed with these types of bonds,
which would seem to indicate that there was some other factor
at least initially driving skyscraper construction.
Will Goetzmann:
Okay, so if you can find a skyscraper that was–
the first skyscraper was–if you can show that it wasn’t
financed by these bonds than you could prove that it’s possible
to have skyscrapers without the financing,
so that’s a way to test the basic proposition.
It may be–it may not–you can
prove that it’s not a necessary condition if you find one
example. One more idea–who’s an SOM
student? Got one here–no.
You here in the–yes,
gray–charcoal t-shirt. Student: Other than
taking a survey looking at multiple cities and how–
when skyscrapers started to be built,
and when this capital market developed,
I don’t think I can really add anything else other–
Will Goetzmann: That’s good.
Take a look cross-sectionally
at the different cities, see if they have,
maybe, access to capital markets differently across the
cities, and see what the timing is.
That’s a good idea.
I’ll tell you what we have been
doing. We figured that one interesting
issue here is–has to do with the size of the bond issue.
If you have a really–if you
have a tiny building and you want to go to the public capital
markets, do you think an investment bank
is going to give you the time of day?
Absolutely not;
there are basic issues of economies of scale that have to
do with financing. So what we did is we said,
“Let’s take a look at the size of the bond issue,”
we also looked at the height of the building,
but the size of the bond issue is the legitimate thing to look
at, and presuming that the bigger
the bond issue, the bigger the building.
Then we asked,
“Is the interest rate on the bond,
that is the rate at which the bond was issued,
was that lower or higher for bigger buildings?”
What we found pretty strikingly
is that the bigger the building the lower the yield,
the lower the cost of capital. So if you’re thinking about–if
you’re a developer and you’ve got a plot,
and you could build two small buildings or one big building,
or you could finance it all at once or one at a time and you
know that you’re going to be able to have a lower cost of
capital, what are you going to do?
You’re going to build one
really big building. You can see finance has this
potential, maybe to possibly even distort
the way that the– the way that cities develop,
because of simple issues like the cost of capital,
and those guys sitting around the–toasting themselves that we
saw in 1892, the creation of that market for
those securities, may have had an influence on
the New York skyline during that time period.
Here’s just a picture,
by the way, of the tall buildings in New York and when
they were all constructed. You see huge boom during this
period, and than a long malaise during the middle of the
century, during wartime, and then the market picked back
up. I’ll tell you one sort of
shocking thing to me is that the residential mortgage backed
security market completely disappeared in the ’30s and that
commercial one building/one bond market completely disappeared at
least by 1940 or so. I mean there’s still bonds that
may exist, but when I say disappeared,
they stopped new issues and you couldn’t easily trade these
things, they didn’t change hands.
So there are bonds that are
still paying out. I have a friend whose
mother–whose family has some securities that were used to
build the Empire State Building, they’re still paying them money
and they’re proud to own these things,
but these are the fossils of the financial world.
What you had is this
extraordinary period of innovation in financing of real
estate. The shift in the way that
people were able to access mortgage money,
and a change in the way the developers thought about how
they would build buildings, and all of it came to a
grinding horrible end as a result of The Great Depression.
Did that market cause the crash
or was it–did the fallout from the crash destroy that market?
That’s the question we’re
asking ourselves today, right?
We’re saying,
“Well, subprime obviously caused the big problem that
we’re in and so the foolish bankers,
idiots that wanted to own their own home and were willing to
borrow to the hilt, this horrible evil brew of
people that couldn’t plan ahead and bankers that were willing to
give them enough rope that they could hang themselves,”
that’s sort of the theme, the drumbeat we’re hearing.
You know sort of wonder,
maybe the mortgage market is a casualty rather than a cause of
this crash. Doug has given you some
readings that– kind of a hard reading to slog
through, but because–although,
again, if you’re an Econ major you should be able to snap your
way right through this thing because it’s got a bunch of
regressions, and tables, and plots.
Let me just tell you a little
bit about the idea of this paper.
I’m doing it with Liang Peng,
who is a colleague of mine, he’s now at Colorado and Jackie
Yen, who is a doctoral student at Yale.
Jackie is studying to be a
finance professor but she also worked in Wall Street for quite
some time and knows a lot about these capital markets,
particularly about mortgage data.
The question we’re asking is a
simple one, like many scholars right now,
we want to try and help the world understand what the causes
of this crisis are. We’re in a position that we
don’t–I’m not quite sure I know what the causes are,
and so the way that we’re dealing with that is we’re
gathering some data and we’re looking at it and putting
questions to this data. The data that we got are
housing indexes. Those come from Bob Shiller and
Chip Case who developed these when I was a graduate student
here, Bob and Chip Case had just
started building these indexes of housing.
We really have data going back,
I think they stretch back–their data start from
about, well, late ’80s or so. What they do is they take sales
of housing; they take houses that have been
sold twice in a given city, and they get thousands of
those, and you can sort of figure out what’s the best
measure of return to explain all of those repeated sales,
so it’s actually transactions based measures of housing.
We’ve got those for all the
city–the big cities in the country, and metropolitan
statistical areas, about 320 or 330 of these.
Then we have some additional
city level information we can compare that to,
then we’ve got mortgage issuance by city,
and finally, and this is pretty amazing,
for one year for 2006 we took all the mortgages that were
issued in the entire country and we have detailed data about
those mortgages. We figured somewhere buried in
this data we’re going to be able to get some sense of what the
problem was. Here’s a chart.
On this axis is past price
growth from 1999 to 2005–end of 2005.
What did we do?
We took each one of those
cities, we look at that index that Bob Shiller had created,
and we said–we just plotted the total growth that that index
had undergone over the period– over the 2000s, up through 2005.
We wanted to look at the hot
markets. Which were the really hot
markets? Which ones had really grown a
lot? You could see some of them had
grown by almost a factor of three;
Las Vegas, there are parts of California and Nevada,
and Florida that were just off the charts, having grown an
amazing amount. By the way there are a lot of
cities where there wasn’t much growth at all,
the whole cluster here of cities that had grown by a
factor of– well, they had grown 30% but
that was over a long length of time.
They’re probably growing at
less than 10% per year. Well 10% per year over this
time period was actually a pretty good investment when you
think about it. The stock market was doing
terribly. We had just gotten off this
horrible bender from the tech bubble.
People said,
“Stocks are terrible, I don’t want to invest in them
anymore,” bonds were a decent investment,
but they were–people were really wondering what are we
going to put my money in. A lot of people thought
housing, although a modest growth in many of these cities,
housing might not be such a bad thing.
When you think about it,
if you’re sitting in the year 2000 having just been burned by
the tech bubble, and you’re saying,
“How am I going to save for myself and my family?”
You know, why not put your
money in your house? At least you can watch the
asset, you can take care of it, you can improve it,
granted you have to pay takes on it and you have to fix it up,
and it’s not very liquid. But you balance these things
off, it might be–it was the new asset class of the era.
It was an idea–it was not a
speculative asset, it was a savings related asset.
Okay, how about this other axis?
Subprime approvals in 2006;
the log of the dollar amount in thousands.
We looked at all of these–we
looked at the rates city by city of subprime approvals.
That is, people that had
applied for a subprime mortgage and then been approved.
A subprime mortgage means you
don’t qualify to get a prime mortgage.
To qualify for a prime mortgage
you have to have steady source of income,
you have to have a good credit rating,
there has to be a good loan to value ratio,
that means you aren’t trying to borrow too much,
you have to have a house that’s appraised in such a way that the
loan to value ratio is a legitimate measure that a bank
can trust. If you’re not in that
circumstance then you’re in sub–the world of subprime.
What we see here is subprime
approvals were more frequent in markets where the prices had
gone up. Why would that be?
Okay I’ll ask this as a general
question. Why would you expect that?
Student: When you
have–when the prices are higher more people are going to have–
when the prices are higher, assuming that incomes in the
cities didn’t go up with that, you’re going to have people
that have lower incomes relative to the value of the house that
they’re buying because the average price is higher and the
amount they’re going to borrow is also going to need to be
higher relative to the price of the house,
so it’s going to be a higher ratio.
Will Goetzmann: Okay.
This sounds like endogeneity;
good. People–the houses would go up
too fast, they–a normal loan–they can’t
get a normal loan to buy a house they might not have been able to
buy ten years ago with a prime loan,
so therefore you’ve got necessity of more demand for
subprime loans. These are approvals though.
That means the bankers had to
say yes, so bankers had to go along with this to get this
graph. Yeah.
Student: When prices
are rising there’s this almost– there’s a market contagion that
prices can’t fall down, they can just go upwards,
so people in general just start taking more risks and financial
institutions which grant these subprime loans also become
overly optimistic. They all feel that prices have
to go up, and they feel that the real value of houses have
increased, and that is– Will Goetzmann:
That sounds like irrational exuberance.
It’s easy to say that they’re
irrational. It’s harder to say that they’re
rational. It would be nice to know
if–first of all Bob Shiller is a good friend of mine,
and I love his book, and I love his wonderful
intuition about when the cart is going off the track.
He tends to be right so it’s
hard to deny that irrational exuberance is behind a lot of
the big messes we’re in. However, as an economist,
you sort of want to see how much you can get with the
rational stories first, and then have the remainder be
the exuberance. We’re going to push on that a
little bit–yeah. Student: Yeah,
I interpret this chart as, if prices of asset values are
going up and subprime approvals are going up at the same time
that means the structure has to be increasingly leveraged,
that’s the only way to fill the gap.
You’re not saying that subprime
incomes are going up, but you’re saying the approvals
are going up, so something has to fill that
gap in the purchase price and that means they’re being more
heavily leveraged. Will Goetzmann:
Yeah, I think that there’s some evidence consistent with
your observation. It’s amazing how much you can
get off staring at a graph like that because buried behind it
are a whole bunch of decisions by individuals applying and by
bankers deciding. I’m going to show you a picture
just from one city, this is from Seattle,
and what we’ve done here is we’ve gone from 2006–
what we’ve done is we’ve said, “Let’s extrapolate the
housing price in 2006, let’s extrapolate it and then
use some econometrics to figure out how far down we think it
could go and how far up could go.”
Those are those two lines
there, and then there’s wiggly line which shows you exactly
what housing prices did. Actually I’m going to show you
more in just a second. I want you to understand the
geography of this first. These are confidence bands and
if you’re thinking– if you’re a banker and you’re
going to make a loan the first thing you want to know is,
is this loan to value ratio going to change a lot over the
time period of the loan? A lot of subprime loans sort of
reset after three years and so forth.
You’d like to have some
confidence, or at least know how far down
the value could go, so that you–because you’re
worried about the loan value being higher–
the loan being higher than the house value.
That’s called underwater.
So you’re worried about the
value of the asset, so this picture is really
putting confidence bounds on the value of the asset,
and what this is saying is that, in 2006,
if you just extrapolated the prices from that time you’d get
kind of a flat line, and then you didn’t–you
wouldn’t find any reason to expect that over the next three
years prices could drop more than 20% or so.
Only one time out of twenty
would you get a shock of more than 20%.
How would you get that?
Here we are;
now here’s a whole–here are a whole bunch of these–the same
exercise. I’ll tell you a little bit
about the exercise. Here’s where the econometrics
comes in. We take those indexes that
Shiller gave us for all the different cities,
these major cities, and than we did something that
any red-blooded econometrician would do.
We did an auto-regression which
estimates the relationship between past growth and future
growth. What you see from all these
housing indexes is that past trends seem to be followed–
there seems to be a huge amount of momentum and inertia in these
movements of these markets, even when they’re going down.
Once they start–here’s
Minneapolis, it was going up like this,
incredible auto correlation in past trends,
and then of course once it starts going down it just keeps
going down. This is not a random walk.
Housing doesn’t follow a random
walk by any stretch of the imagination.
It’s very, very predictable.
You could predict,
the models tell you that you could predict the housing
prices, with a huge degree of
confidence for three years out just given about twenty–
fifteen or twenty years worth of past data.
All of these pictures show you
that for each city there is a line which is the past price
increase. That’s the–not Las Vegas–went
up like crazy at one point. Then there are the two
confidence bands, then there’s the orange bar
saying what do we expect the price trend to be?
But it’s not the expectation
that’s so important, it’s this lower bound;
how bad can it get? If you’re a banker that’s the
thing that tells you, “Is this loan going to be
a disaster?” Every single one of these
cities, virtually, the actual price just plunged
right through the banker’s confidence bands.
If you’ve got a model that’s an
econometric model that’s based on past price trends,
and you ran that model in 2006 you wouldn’t have been able to
predict any of this crash. You would have felt pretty
confident that when you were writing those mortgages,
even the subprime mortgage, that those were good bonds,
that was a good loan. A subprime loan doesn’t mean
that the house is bad; it just means a borrower is bad.
It’s a subprime borrower,
not a subprime house. So you say, “Look,
what’s the worst that could happen?
My borrower can’t–we get into
some trouble, my borrower can’t make the
payments on the house, then the bank has to repossess
the house, but if the house price isn’t
going to move much, then what the heck,
we’ll just turn around and sell the house so we’ve got good
collateral.” We all know what happened.
That model turned out–that
econometric model turned out to be a complete disaster.
Actually one more point on this.
What do you need for an
econometric model aside from an econometrics textbook?
SAS, I use R,
I did this from R, Stata–some people use that.
What else do you need to run an
econometric model? This isn’t a trick question.
You have a blue scarf on,
what do you need to estimate a model like this?
Student: Data.
Will Goetzmann: Data.
You need data,
and so the data come from Robert Shiller.
The fact is,
unless you have those housing indexes you can’t run this
model. The irony of these indexes is
that although they opened a world to us of what housing
prices do, they also made econometricians
believe that they could estimate the risk associated with the
housing. Before we had those indexes we
couldn’t do any prediction, we couldn’t do any of this
sophisticated mathematical calculations.
We couldn’t walk into the
president of the bank and say, “Look, we’ve done a value
at risk calculation using these fantastic indexes that tell us
that the probability of us losing more than 20% of our
capital is .005.” You couldn’t do it.
Suddenly, with this data,
a little bit of information is a really dangerous thing when
you put it in front of somebody that knows how to crank through
a regression. It’s a terrible thing to blame
Bob Shiller, the person that forecast this
crash, to blame him on the other hand,
for providing the matches that allowed people to burn down the
house. It was a failure of models but
you can’t run the models without the data.
I’ll just give you a little bit
more flavor for what we do. We run a whole bunch of
regressions. We’re looking for
relationships, but mainly what we do is we
divide things up into the world of the demand side for loans and
the supply side for loans. What we want to ask is,
was it the stupid or avaricious banker that drove the crash?
Or was it the stupid and
avaricious homeowner, borrower that drove the crash,
neither, or both. Here we have subprime mortgages
and prime mortgages. Roughly when we look at
relationships, here’s what we find.
We find for both subprime and
prime, the numbers of applications and
measured by– dollars and numbers of
applications were increasing in those home prices.
So when the home price goes up
you make your forecast of what the worst scenario might be,
and if you had a positive trend, you would think that
lenders would be more comfortable with the bottom
line, that they’d be able to get
their money out. And borrowers,
if they think, “Hey look,
I’m buying a house, I’m putting every dime I have
into it, but the trends look pretty
good, I should be able to get my money out at the other
end,” then you ought to see a positive relationship between
past price increases and applications,
however you want to slice it. Somebody mentioned–you
mentioned leverage. That’s exactly what happened.
There was a greater loan to
income requests for–higher leverage requests for both of
those two. People actually also tended
to–this is loan to income, measure of leverage,
not loan to value. The loan to value actually
also–the value to loan went up. Why is that?
People felt comfortable putting
more of their money into houses and than people bought more
expensive houses, but the endogeneity issue–who
mentioned that? That was a really important
point. There were more expensive
houses so that was inevitable. So were people being foolish?
Well that’s a hard thing to say.
They looked at it as a savings
opportunity. You could say this was a crisis
driven by over too much demand for savings.
It’s just that they didn’t save
in–just that the savings vehicle was the house.
Were they irrationally
exuberant? Well if you look at those
trends maybe I would have been fooled by the same trend as
well. How about on the banker’s side?
Now the banker’s job is to say
no, to reject loans. What do we find here?
Now the weirdest thing that we
completely– that completely shocked us is
that approvals didn’t really– the rate of approvals for
subprime, once we controlled for a lot of
things, didn’t change as much as we had
expected. In fact the rate for approvals
for prime actually went down, which is strange.
That means bankers were getting
tougher. Well, they were only getting
tougher because the demands were increasing.
You have to look at this as,
the environment was changing because the demand for loans was
increasing a lot, but so that there were–their
approval rates were actually going down.
It really doesn’t look like
that–it doesn’t look like irrational exuberance on the
bankers’ part for prime loans. Mostly on the prime side,
bankers appeared to be behaving really pretty rationally,
less so on the subprime side. The loan-to-income ratio,
that was going up; the value-to-loan ratio was
going down; in other words,
people were having to put up less of their own cash to own
the home, and people were getting higher loans with lower
incomes. If anything I think what we’re
beginning to convince ourselves, using this loan data,
that the subprime market and the prime market were two kind
of slightly disjointed markets with different decision
processes in operation. Both sides had securitization
going on and we haven’t attacked the question of whether
securitizable loans were driving all of this,
but other people have been asking that question.
We certainly found evidence of
high–of increasing demand in prices.
It could be that people are
chasing after–chasing the trends.
But it could be also that the
expectation of the liquidation value of the house was going up
and made them feel safer about the borrowing.
The riskiness of mortgage
applications is increasing in past price.
Everybody wants to buy more
expensive homes, but–that we documented,
and as I mentioned, there’s this disjunction
between the two. That’s sort of a–kind of a
flavor for how a professor and some graduate students look at
the crisis. You’ve been reading Posner and
you see how somebody tries to put the whole thing together;
well, our goal was to kind of deconstruct it.
Let me see if there’s–that’s
just a–kind of more documentary evidence of what I was talking
about. I don’t know,
I’ve looked at history, been interested in that,
and looked at current data, so with that–I don’t know if
we have time for questions or not but–
Prof: Yeah I think there’s time for one or two
questions. Will Goetzmann:
Back there–Notre Dame. Student: On one of your
previous slides, you mentioned that the supply
side seemed to be almost constrained,
and you seem to attribute that to, I guess,
potentially greater responsibility on the part of
the bankers in not approving the increased number of
applications, and I was wondering if you
think that that is the case, or if there’s evidence that it
could just be that there was a constraint on the supply where
banks just were not able to issue more loans due to capital
requirements or other financial regulation.
Will Goetzmann:
The beautiful part about this work is that we’re taking
as given that people– everybody, the banks and the
borrowers are operating in their own best interests.
So that we’re looking at this
as sort of a equilibrium outcome where they’ve taken into account
the potential for the capital constraints and what have you,
so that negative coefficient actually–
you see it here, this negative coefficient on
approvals for prime versus subprime.
My tone of voice may have led
you to believe that I thought that was a foolish thing.
It certainly no evidence that
it’s foolish at all, but it’s an interesting puzzle.
I mean, we expected to see
everybody lifting the floodgates and make–writing all these
loans. The notion that the prime
lenders were actually clamping down suggests that they were
exercising some judgment about quality of loan that resulted in
this negative rate. Prof: One last question.
Will Goetzmann:
Question here? Student: On the same
slide you also implied that the supply side,
the bankers, were not accepting these
overleveraged mortgage applications as much as you
would expect, but does your research indicate
anything about how overleveraged the banks were themselves in how
much capital they were taking out of the market?
Will Goetzmann:
The question you’re asking really implies a point of view
about bank capital requirements. Certainly, I’ll just use the
data that I showed you, so we didn’t use that–any
information. As you know,
this is a big spider web of different relationships that
scales all the way up to the world,
the government, and the role of capital
requirements and regulators. We just didn’t–we didn’t pay
attention to all of that. Our job is not to put the thing
together but to see if we can find certain pieces that help us
understand whether decisions were good or bad.
Prof: Thanks,
Will, it was a terrific lecture.

Leave a Reply

Your email address will not be published. Required fields are marked *