2006 Economic Outlook Roundtable:
Scenarios for the Next U.S. Recession
Whitney, Executive Director, CIBC World Markets, a subsidiary of
Canadian Imperial Bank of Commerce.
MR. BROWN: The last two presentations have really highlighted the importance
of the housing sector to the economic outlook and the banking outlook,
and that's something that Meredith Whitney is going to comment
on in her presentation. I would like to mention that her recent
paper, "The State of the U.S. Consumer: 10 Percent at Risk," really
made an impression upon the staff here at the FDIC. A copy of the
paper is in your folder. We are glad that she accepted our invitation
to come and tell you about it today.
With that I'll turn it over to Meredith.
MS. WHITNEY: Thank you. That's music to my ears, because I analyze banks
and financial institutions for a living, and I place particular focus
in my career on the state of the U.S. consumer. It's of key interest
to me, and, particularly because of what I do for a living, it's an enormous
honor to present to all of you.
So with that said, the paper that we put out in October was really inspired
by an analysis of Hurricane Katrina. So many Americans saw a side of
the U.S. economy that I don't think many of us had seen before or had
really digested as far as how it fit into the total U.S. economy. It
challenged me to drill down deeper into the analysis of who was at risk
in terms of any type of consumer softening, or a potential recession.
My conclusion was that I believe 10 percent of the U.S. economy will
actually experience a segmented consumer recession over the next 24 months
(see Chart 36).
Looking at [Chart 36], which we'll get back to in a second, my analysis
divides the U.S. consumer market into three main buckets: those who have
had newfound access to credit, those who have had a long-time access
to credit, and those who, over the past decade, have had access to unprecedented
amounts of credit.
The underlying thesis
to our recession scenario is that an interruption in liquidity always
a recession (see Chart 37). You can look
at the S&L crisis, especially, or, as Kathy did, to the Long-Term
Capital Management crisis or to the Enron/WorldCom bankruptcies in 2002.
We believe that the next interruption of liquidity will affect the consumer
Factors that will precipitate this event could include a continued rise
in the fed funds rate, a capital strain caused by the ratings agencies,
or the change in minimum payment guidelines for the credit card market.
Let's start with higher rates. We believe that the Fed is surgically
targeting the housing market in its efforts to raise rates (see Chart
38). Let me just give you an example. Thirty-five percent of the housing
market's new purchase originations in 2004 were adjustable-rate
mortgages. Now, 30 percent of new home originations are adjustable-rate
Just from an anecdotal perspective, if you know short-term rates are
rising, and long-term rates are declining or at least staying static,
wouldn't it be prudent to lock in a long-term mortgage? Why is 30 percent
of the mortgage market still booking into adjustable-rate mortgages?
Because we believe they couldn't otherwise afford to squeeze themselves
into a home or they're squeezing themselves into a more expensive home
for affordability factors.
I want to take out one major portion of the economy first, which is
the corporate market. And this is not incongruous to what my esteemed
colleagues have said already in the presentation. Today, 17 percent of
corporate funding comes from banks. The vast majority of corporate funding
comes from the public market. It's more efficient to go directly to the
public market, and it's more cost effective. So this is a 38 percent
decline since the last cycle. So we think that higher short-term rates
have much less of an impact on the U.S. corporate market than they ever
have before, and that Greenspan and the Fed governors are well aware
of this (see Chart 39).
In addition, as you
look at cash levels on U.S. corporate balance sheets, they have never
better (see Chart 40). Cash as a percentage of long-term
debt is at 42 percent. The last time they were at this level was when
they were about 40, 41 percent in 1987. And as you remember, those were
the days of Michael Milken, leveraged buyout fever, and tremendous consolidation.
Because what are you going to do with piles of cash? You're going to
invest, and you're going to consolidate. You're going to invest, and
you're going to buy. Extra cash—go shopping.
Okay. So the idea here is that U.S. corporations will probably extend
a 14-month cycle of capital investment, and you will probably see near-record
merger and acquisition levels this year. So in our analysis, we're not
worried about the corporate market at all. It was exclusive of exogenous
events that will actually put risk-based pricing back into the spread
market. We don't see that, believe it or not.
At any rate, let's focus on the consumer, because this is really what
our analysis is about. And this will be probably the most important thing
I say today. You have to look at three separate buckets of the U.S. consumer
market: those with traditional, good quality credit; those folks who
have not had any access to credit; and then this segment of the market
that has had new access, new and unprecedented access to credit (see
I want to focus on this statistic, which is pretty alarming. We estimate
that 26 percent of the U.S. population lives at, near, or below the poverty
level (see Chart 42). The number for the poverty level is living below
$23,000 a year, and that covers 13 percent of the population. We go on
to estimate, actually, that 26 percent of the population lives at, near,
or below the poverty level. Now, these guys have no access to credit.
Now, my estimate seemed alarming when I came up with it. I've been told
by leading consumer lenders that actually my estimate is low. It could
be as high as 30 percent or higher. So what you need to do when you look
at the U.S. consumer is extract this portion from the total pie at risk,
because these guys don't have access to credit. These are the folks that
shop at dollar stores primarily, where they use cash and food stamps
for purchases. Dollar stores are nearby. They're not driving a car, so
they're less impacted by commodity prices, in terms of prices at the
pump, and, as a result, have not gotten wealthier or poorer over the
last decade; they've stayed steady-state. And if you look at a
lot of the dollar store results, they have essentially flatlined, which
supports this point.
The next segment of the market, which we also extract from this pie,
is the 64 percent of the market made up by what I will call "fat
cats" (see Chart 43). Sixty-four percent was the traditional homeownership
rate as far back as the 1970s to about 1994. This is as far back as the
data show, so 64 percent is what we really call the "natural rate
Of course, you know that 69 percent of the U.S. population today owns
homes. Greenspan, in his paper on the mortgage industry in September,
targeted 5 percent of the U.S. population at risk in terms of very low
equity levels in homes. Sixty-nine minus 64 is 5 percent. So these are
the new, first-time homebuyers, which we see at risk.
Kathleen also talked about how amazing it has been if you have owned
a home over the last decade. If you own a home, you've been sitting on
the single-greatest wealth creator in the market, as opposed to equities
or bonds. Homeownership has had a 10-year rate of return of around 8
or 9 percent. And last year, as you know, home prices grew by 10 percent.
In terms of equity in homes, it has declined by 9 percent. This is average
equity in homes. So, in 1990, the average equity in homes was 60 percent;
today it's 56 percent. So we think, actually, that has created
a barbell between those folks who own their homes outright versus the
new buyers, who have put very little down on their homes.
Now, [Chart 44] shows the 10 percent of the population that we're talking
about, that we believe is at risk. They have had access to unparalleled
levels of credit.
In 1994, the standards through which an individual could qualify for
a home really relaxed considerably. And from that time, 15 million new
homeowners were created in the United States. So it's not just a matter
of money going from the equity market into the housing market. It's also
the fact that 15 million new households became homebuyers, and an increased
demand versus supply clearly sent home prices higher. So an expansion
of 8 percent in the size of the market cannot be overestimated in terms
of its impact on the market (see Chart 45).
Since 1996, subprime lending has grown five-fold (see Chart 46). And
these numbers actually may be understated, as a lot of large banks are
reticent to admit that they actually have any type of subprime loans.
Instead, it has been called non-prime, or non-traditional. What this
shows is a five-fold expansion in the subprime lending volume.
[Chart 47] is an example of how lenders have loosened their underwriting
standards. One obvious element that you are well aware of is lower documentation,
in terms of credit scoring and also in terms of proof of income, proof
of employment, et cetera.
It's interesting that the National Association of Realtors recently
came out and said that 43 percent of all new homes are purchased with
no money down. You wouldn't have heard about that prior to 1994. So the
ability to access credit has expanded dramatically.
Something I referenced a little bit earlier relates to the hazards of
adopting an adjustable-rate mortgage and the negative amortization that
is now associated with a lot of the adjustable-rate mortgage products
(see Chart 48).
Again, I want to talk about why someone would do this to themselves,
why someone would take on an adjustable-rate product when you know short-term
rates are rising. That doesn't make any sense. The only conclusion that
you can draw from this is that, otherwise, these homes would not be affordable
to these consumers.
Home equity, which has really been the consumer's ATM over the last
certainly five to ten years, has enabled a lot of this affordability.
A lot of the new home mortgages that are originated come with piggyback
home equity loan features.
Now, I want to point out that [Chart 49] is only talking about revolving
home equity lines. The total home equity number is closer to about $800
billion. So a lot of this home equity is acting like sort of a new version
of a credit card for consumers.
This is something that Art discussed previously. Bank exposure to mortgages
and home equity is now at peak levels, having risen dramatically (see
Chart 50). If you look at 1998, the total exposure to mortgages and
home equity loans was about 25 percent. In the last quarter, the third
quarter, it had risen to 37 percent.
Now, there are a couple of reasons why this is the case. Many U.S. companies
have managed to carry on without needing banks. We showed this in the
slide that shows that only 17 percent of corporate funding is delivered
by the banks.
The banks need to
originate some product. One of the biggest problems for the bank industry
is tremendous deposit growth and very little
asset growth. So you have institutions that are liability-rich and asset-poor,
and they're going to put anything on the balance sheet they can. For
the past five years or so, mortgages have provided the best vehicle for
asset growth and any type of margin gain whatsoever—although we
would also argue that that margin has been increasingly challenged over
the last year.
There are a couple of things that we see happening to banks over the
next 24 months, really triggered by liquidity strains. And I want to
talk about the liquidity strains for the mortgage sector in two parts.
First, you have a liquidity strain because of what I will call the denominator
effect, which is caused by higher short-term interest rates (see Chart
Over the past two months, home equity lines have declined for the first
time in almost a decade because the yield spread between the 3-month
LIBOR and prime has inverted (see Chart 52). This is undesirable for
consumers, and it's undesirable for banks making these loans.
So there is a liquidity squeeze for consumers as rates rise, which is
related to the affordability factor. In 2004, as I said, adjustable-rate
mortgages were 35 percent of total purchase originations. Today, they're
30 percent. And I can't emphasize enough that the vast majority of subprime
purchases are coming through adjustable-rate mortgage products.
The second factor which is significant is that the ratings agencies
have required higher subordination levels associated with subprime mortgage
origination (see Chart 53). So it's less profitable for banks and independent
institutions to originate these mortgages. In the second quarter, capital
required against subprime mortgage originations increased by 10 percent.
That's a lot of dough, and that's a big impact to profitability.
So what all of this creates, then, is smaller loan balances, and a smaller
denominator makes your loss ratio become all of a sudden a lot higher,
and your profitability become much more challenged.
We believe a shrinking adjustable-rate mortgage market is a prelude
to decelerating subprime mortgage originations. Yesterday we saw that
a subsidiary of First Franklin, which is one of the largest subprime
mortgage originators, reported that it still has over 70 percent of its
mortgage originations coming through adjustable-rate mortgages.
So as you see adjustable-rate mortgages decline in terms of total purchases,
you will see subprime mortgages decline in terms of total originations.
You already see declining home equity balances. You will begin to see
declining mortgage balances.
The factor through which liquidity is challenged in the credit card
market is the change in minimum payment guidelines, which prior to 2003
took the independent credit card lenders out of negative amortization.
In 2005, the OCC required all of the big banks to become compliant with
negative amortization extraction in that, if you were a customer, you
had to start paying down your principal. You could no longer continue
to contribute to your debt burden each month. Well, that's a liquidity
squeeze for the consumer, and we'll get into that in a second. It also
shrinks balances at the credit card lenders.
I'm going to give
a little background here. We believe credit cards have become the loan
of adverse selection, based on the expansion
in homeownership (see Chart 54). The age-old rule by banks is that 80
percent of your losses come from 20 percent of your borrowers—the
so-called "80/20 rule."
Well, if you've owned a home over the past decade, particularly over
the past five years, and you all of a sudden refinanced your mortgage,
you've either paid down your credit card debt or you've refinanced your
mortgage and taken a home equity line out, which has tax advantages just
like a mortgage does. It also carries a lower interest rate, because
it's collateralized. So we believe that the home equity market has cannibalized
the best credit card customers, leaving those folks who have revolving
balances, which is about 40 percent of all U.S. credit card customers,
as the folks to be worried about.
Now the growth in the credit card industry is in single digits. We're
in an environment that's flush with cash. People have either not wanted
to maintain revolving debt or they've transferred it, if they could,
into home equity loans, or they're simply not revolving debt at all.
So what is left in the credit card industry is those folks who have
nowhere else to go, those folks who are playing credit card roulette,
because it is certainly a source of borrowed funds. And I want to point
out that those folks have been plied with seven times the rate of capital
than the standard credit card industry. In other words, the growth in
the subprime market has exceeded the growth in the traditional credit
card market by a factor of seven times (see Chart 55).
Why is minimum payment compliance such a significant factor? Just as
you've heard about mortgage payment shock that occurs when your mortgage
reprices, this is effectively a repricing of a consumer debt burden,
a monthly consumer debt burden. And this change in minimum payment, at
least in the example that we provide with a $10,000 beginning balance,
increases a consumer's minimum payment by almost a factor of two times.
By the end of 2005, everyone except for J.P. Morgan and Citi (who are
two of the top three largest credit card issuers in the United States)
was compliant with the minimum payment guidelines (see Chart 56).
So let's say, for example, you're a Capital One credit card customer.
Your minimum payment has already been revised to meet the new guidelines,
as of 2003. Let's say you have a Capital One card, on which you're paying
a higher minimum payment as required by Capital One.
Suppose you've also got an American Express card. Similarly, those
issues apply. Then you also have a J.P Morgan card and a Citigroup card.
And by 2006, when these payment standards are adopted across the board,
all of a sudden your credit card payment doubles on two of your cards.
And most likely, in the subprime sector, you're not just carrying two
cards. You're carrying several cards.
This is a massive liquidity squeeze. At a very minimum, it will shrink
the growth in balances at the credit card issuer level. So, as I said
with the denominator effect, these loan balances are going to shrink
for Citi and J.P. Morgan, because customers are paying more of their
debt off. At another level, your liquidity as a consumer is also strained,
and that is what we believe will set up higher defaults at the bank issuers.
Now, Art had talked about how overall credit losses for banks are at
a 15-year low. And from a corporate perspective, we think credit spreads
are going to remain very narrow for an extended period of time because
of the savings environment that we're in (see Chart 57).
And credit losses would have to deteriorate significantly from today's
levels to have a major impact on banks. We estimate a 40 percent increase
in credit losses would still not have a meaningful impact to earnings
for a lot of these banks. But for banks like Citi and J.P. Morgan, where
they have 20 to 25 percent of their earnings coming from credit cards,
higher losses are going to create a very challenging environment for
them to show earnings growth.
So I want to summarize
all of this, and I hope I didn't go too quickly. I'm a New Yorker,
been told I speak quickly. If you segment the
consumer market into three very important buckets, then take out the
folks that have no access to credit and take out the folks that are what
I would call the fat cats—who already had access to credit, have
gotten a lot richer because they've owned homes, and may have taken on
more debt but remain well positioned from a financial balance sheet—then
we can really focus on the 10 percent of consumers that are at risk (see
Five percent of the U.S. consumer market is made up of folks that are
new homebuyers. This is the 5 percent of the U.S. consumer market that
Greenspan is worried about, that have such little equity in their homes
that if home prices decline in any way, they owe more than they own.
That's a bad situation, particularly as most of these guys are in adjustable-rate
mortgage products, and they are going to be squeezed as rates continue
Then we add the other 5 percent of households, including the folks that
are playing credit card roulette, who are going to be squeezed by this
minimum payment guideline, who have never had access to credit before,
who have gotten a lot richer over the past decade, and are about to get
a lot poorer because of liquidity strain.
So when you look at this 10 percent of the market which we qualify as
subprime, what does it mean to the total economy at large? Well, 40 percent
of discretionary spending is done by the top 20 percent of the consumers.
We don't think the subprime market correlates to the top 20 percent of
The way I refer to this market is that these are probably the Wal-Mart
customers. It's one level above the dollar stores, and probably one level
below a Target store. And you've already seen spending at Wal-Mart contract
somewhat. Many of these numbers can be confirmed by just looking at general
trends in retail sales. You see no blip whatsoever at Nordstrom's and
Neiman Marcus, all of the higher-end retailers, that have really been
carrying a lot of the consumer spending. So we think consumer spending
in a finite portion of the market will slow, and credit deterioration
will increase dramatically in a finite segment of bank portfolios.
Again, we think 10 percent of the U.S. consumers will definitely go
into a segmented recession, but the rest of corporate and consumer America
looks pretty good.