2008 Financial Crisis: A Ten-Year Review Conference – Short Term Debt Liquidity

2008 Financial Crisis: A Ten-Year Review Conference – Short Term Debt Liquidity


Thank you very much
for those kind words. It’s a great pleasure to
be part of this conference. I learned a lot yesterday. I enjoyed Tim Geithner’s
discussion a lot. So I actually– unlike the
people two sessions from now– didn’t get any
firsthand knowledge of the financial
crisis, but managed to talk to a lot of people. And I’m going to have
basically one point that’s going to be about liquidity,
whether there’s such a thing as too much liquidity. But I’m going to start
with a very standard– at least in this room, based
on what I heard yesterday– a very standard
narrative on the crisis, and a somewhat standard one on
the responses to the crisis. So what cau– there
was a housing shock. Housing prices went down more
than some people expected. And presumably, you can
compare that in the value to the stock market drop in
2000, or something like that. So it was a similar-size
shock, but we got a crisis. So obviously, it
had something to do with the financial system,
and the architecture of the financial system. And a lot of regulation
and law across the world came in to try to say,
let’s try to reduce some of the vulnerabilities
so we don’t get a crisis just like the last one. And there were many narratives–
we heard several of them yesterday– there were
many narratives on what it was about the architecture
of the system that caused the vulnerability
to the crisis. So one big one– the one that
I’m going to focus on today– is runs, and runs due to things
that look like short-term debt. Collateral calls at AIG. That looks a lot
like short-term debt, because you have to come
up with money in a hurry. There’s also this
interconnectedness thing. Too interconnected, or too
partially interconnected, if you think about
Franklin stuff. And it turns out that actually
ends up being about liquidity, in many cases. The other extreme is
anticipated bailouts. We have crises because people
anticipate if there’s a crisis, they’ll get bailed out. And if we could just commit
never to have a bailout, we’d never have any crises. That one, I have a hard
time taking seriously, but one of my former colleagues
takes that one seriously. So there’s also this
thing about regulation, that there was the
unregulated shadow banking sector, that a lot of the
maturity and liquidity mismatch was there. That was clearly important. There was also,
clearly, very low levels of capital in the
broker dealers, and in a number of
the commercial banks. OK, so the regulations
didn’t really– if you look at them,
you’re going to say, well, which idea did they have in
mind when they wrote Dodd-Frank? And it was like,
well, they imagined it could be any of these. Let’s try to fix
them all, or let’s do some things that deal with all
of these in and of themselves. So every talk I’ve given
in the last 10 or 15 years that isn’t a paper
and was about crises has this slide in it. The word “private”
got introduced after the euro crisis. Private financial crises
are everywhere and always due to the problems
of short-term debt. So that says, it’s about
short-term debt, the need to raise funding quite
rapidly, and the notion that there can be panic
self-fulfilling prophecy problems if people think,
for whatever reason– but particularly potential
insolvency– if they think, for whatever reason,
a firm that holds, or a bank that holds a bunch
of assets that aren’t perfectly liquid is going to have a hard
time refinancing its debt. So if you have short-term
debt, most of your debt has to be refinanced. So this is sort of
what a crisis is. So going back to
what I was about before, a lot of these
things are either ways of implementing
this fragility through short-term debt
backing illiquid stuff, or some of these
other things are about potential vulnerabilities,
potential losses that might get you into the
range where the bank is so– or the firm is so poorly
capitalized that people are on a hair trigger
to have a fear of a run. So I’m going to say that
a lot of the things that were done were about
addressing vulnerabilities, but if we’re actually–
you could imagine you had vulnerabilities
with lots and lots of long-term
finance, and you could then have drops in
consumption and things like that similar
to what you can get in the financial crisis. But if you’re actually
thinking about the crisis, if you read essentially
all the books I read by people who were
dealing with this day-to-day, when the crisis actually
became a crisis, it was the runs, the
panics, the liquidity that people started to focus on. And then they needed
to get that in place so they could think about some
longer-term things, including recapitalization, which had to
be dealt with in the very, very short term as well,
because solvency, even without the illiquidity
problems, was there. So that’s pretty standard. So then I could ask– and I’m going to give out very
different-than-normal answer to this one. How do we think about
evaluating the regulations? That’d be nice. I can say, based on
10-years’ experience, how well did these
regulations do? So one thing we can say
is we didn’t have a crisis after the last one. We had continuing things
going on in Europe, but I’m going to say
sovereign crises that– governments can do things that
the private sector can’t, and they can have
crises of their own. So we could say, well, we
didn’t have a crisis, therefore one of two things. Either the regulations worked,
or we didn’t need any stinking regulations because crises– that was so 10 years ago. The point I’m going to make
is a bit different from that. I’m going to say
that, in addition to the long-term things
that are supposed to address
vulnerabilities to crisis, the thing you do
in the short-term, and the thing that
persists today, is there’s lots and
lots of liquidity in the corporate system and
in the financial system. So liquidity doesn’t
just mean cash. It doesn’t just mean lines
of credit, borrowing ability. It doesn’t just
mean Fed programs. It means, if we think
of Nobu Kiyotaki stuff, we think there’s high net
worth, high financial capacity in the corporate sector today. And some of that liquidity
has been due to policy, things like QE. But the point I’m
going to make– and I’m going to make it
in two different ways– is that this liquidity,
the lots of liquidity essentially provides a tailwind
for the financial sector, sort of like a shock absorber
for the financial sector. So we put in these long-term
things that potentially could be good,
potentially could be bad, that address vulnerability. But at the same
time, we injected, and the corporate
sector chose to keep lots and lots of liquidity. And lots and lots
of liquidity leads to the opposite of
fire-sale prices for real and financial
assets, but I’m going to focus on real assets. So real assets,
when they’re selling and they’re anticipated
to keep selling at their long-term
fundamental value, there’s no fire-sale pricing
today, or none’s anticipated tomorrow, that provides lots
of potential extra borrowing capacity to firms and to
financial intermediaries. It does more than
that, it turns out. So that one reason that
looks– that even if we put the absolute wrong– unless we went too far– in the absolute wrong set of
vulnerability regulations, the fact that we have lots
and lots of liquidity floating around in the system today
could be, and probably is the one main reason that
things look stable today and that we haven’t
had another crisis. So liquidity is a
important short-term way of fighting crises. There’s a problem of
too little liquidity. There’s a problem of
too much liquidity, I’ll talk about in a minute. And if we’re in the
period where there’s lots and lots of liquidity,
that short-term thing will prevent, or at least
reduce the vulnerability in the short run to a
crisis, so you can’t measure the long-term stuff very well. So let me just talk about– so the liquidity, this
thing about liquidity providing a tailwind, that’s
going to be the theme. So let me just say two things. So there are some
good regulations. And what we can say
is, essentially, there are regulations
that did something. They actually changed behavior. That’s pretty much
what we can conclude. And then we’ve pretty
much had some ideas from looking what
happened in the crisis, or just thinking
about how crises work, on what’s a good regulation,
what’s a bad regulation. Crises regulation
did some things– they were not just irrelevant. People complain, so that also
means they did something, because if the
regulation prevents you from doing something you want
to do, you will complain. So here’s one thing. Here’s a good regulation. This is purely about runs. And it was an
interesting thing that– I think the way it worked
was sort of an unintended– unless it was very
subtle, because I was talking to people when
these things were coming in. There was an
unintended consequence of the way the
regulation worked. So there was one
run-prone thing, which was a great shock to people,
was runs on institutional money market funds. So in fact, when
Phil Dybvig and I were giving the early versions
of the Diamond-Dybvig paper, we said, yeah, the
run-free thing, it’s sort of like
a money market fund because they have liquid
assets and liquid liabilities. Well, the point is, it’s like,
one-day safe liquid assets, that would be,
essentially, no mismatch. But there was enough
mismatch here. And as we know– as I didn’t know at the
time, but as we learned, there hardly is a very
active secondary market for commercial paper. So even money
market funds, if you have people who can think
about fear of fear itself, which is what the
institutions are good at– or maybe one thing
they’re good at– we can see that there
was, essentially, a run right around Lehman. You see this big drop in
outstanding institutional prime. People hold commercial
paper, money market funds. And there was another one
that was almost as fast during the eurozone crisis. And if you look at the
retail stuff– which is the bottom line in the red–
that looked relatively smooth. Because this idea of thinking
strategically about getting out before the other
guys get out, that wasn’t something that was on the
radar screen of the household. And then empirical
work probably suggests this is probably
about the difference between institutional
holders and retail holders, because you could see
this on money market funds from the same originator
that had the same portfolio. So it wasn’t differences
in the portfolio. It was differences
in the holdings. So what do we do? There were lots of ideas that–
so think of money market funds as repos with no haircut, but
on a diversified portfolio. So some people said, well,
let’s put a haircut in there. It’s like a 2% capital
requirement or something. That would have been very hard
to implement, it turned out. So instead, what
they did was say, OK. If you’re an institutional
money market fund, you have to have a
floating net asset value, unless you’re a government one. And you have to put suspension
of convertibility in. Gates. That if people are
pulling money out so fast that you’re
running out of cash, then you have to force people
to stay in for a while. So what that did
is essentially kill the institutional prime
money market fund. So you see there, sitting
around a trillion, and they go at around
$200 billion in assets. That’s the red. And then you can see
this offsetting thing. The blue, the increase in the
government institutional asset. So the corporate
treasurers basically said, gee, if this thing can
have a floating net asset value, we’re only allowed
to hold safe stuff, so now we can’t hold it anymore. So it’s the same assets. So if you just change the way
you do the accounting for it, it’s [INAUDIBLE]. So that basically
says, these were what– like Andrei Shleifer and Rob
Vishny call fake safe assets. And this was– or maybe to get
around regulation, or something like that. So they made these guys stable
by killing them, essentially. And that will usually
stop fluctuations. And it works– well, never mind. You can think of other
things where that’s true. And then you can see– the stuff
on the bottom, the green line and the purple line,
something similar went on for the retail
money market funds. But nothing as extreme–
and you can see, [INAUDIBLE] looking at the red line. The red line was the
institutional prime funds. And that falls in total
value below the retail ones. And so it started way
above, went way below. You can see it was a
bigger thing going on. So that’s a regulation. It changed behavior. So is that good? Is that bad? We don’t know. So now I’m going to
switch back to liquidity. And we do have a lot of
evidence that liquidity was a problem during the crisis. And if you look at what actually
happened fast, in the Federal Reserve very quickly,
and then later, the ECB, introduced lots and
lots of very, very creative liquidity interventions. So what this chart shows– that
you can see on the screen– is the secondary market
price of leveraged loans. So these are syndicated
high-yield syndicated loans. And the default rate on these
didn’t change very much, it turned out, in the crisis. So let’s imagine everybody
believed that ex-ante, which is obviously too
far, so that the bid price of these things goes
from somewhere in the 90 to 95 range, down to 65 by the
month and a half after Lehman. So think about that
as fire-sale pricing. So that says there’s limited
liquidity in this market. People could work–
well, unless you think people thought the
world was coming to the end. There’s some other
evidence that suggests that that was actually not
the world coming to the end. It was probably something about
liquidity, at least in part. So the illiquidity is the
thing that makes runs occur. So if you think about
for a single bank, it’s the reason you care
about other people getting out in a hurry. If you think about
a network of banks, it’s the fire sale that other
people are losing funding, so they’re selling
assets, pushing them down. One’s within an institution. One’s across institutions. It’s the same thing. It’s the same thing. So clearly, there was way too
little liquidity in the crisis. It’s not like that we didn’t
inject enough liquidity ex-ante. The fact that there was
a shortage of liquidity was what made the
crisis a crisis. Why things happen so fast. That’s what runs are about. And the point I’m
going to make today is one, the one I already made. That there’s a lot
of liquidity today. That works like
a shock absorber, and makes regulations look
either unneeded or successful. But Raghu Rajan, [INAUDIBLE]
and I have a theory– we actually have two
of them right now, we have one
intermediary leverage that we just finished– that basically says,
too much liquidity can be something that removes
the incentives for firms to maintain their
future debt capacity. Call it financial being careful. They’re going to be
financially careless if there’s lots and lots of liquidity. And basically, the idea is,
if asset prices are selling at their full value, there’s
no underpricing expected, or it’s unlikely that
tomorrow there’ll be underpricing, actions
you take voluntarily today to improve your
and others’ ability to borrow against an
asset can’t increase the value of the asset above
its full fundamental value. But they can increase
the value of the asset above its fire-sale price,
because it allows people buying it during the fire
sale to finance a bigger fraction of their
bid, and bid a higher amount. So essentially, too much
liquidity in a boom. This means net worth
within the firms. It may have something to do
with monetary policy and things like that, but given that
Ben was talking to Raghu about that last night, I’m not
going to get into that part, not having ever been
a central banker. So too much net worth, too
much liquidity in a boom, high asset prices– I’ll show you a
picture of this later– removes this incentive
to take these things. Put covenants into loans,
improve voluntary accounting standards, to choose
monitored or screened lending versus market lending. And we also have a new
paper that basically says, this leads
to an equilibrium, less needed skin in the
game for securitizations. So securitizations
can end up being almost like they sell
off the equity piece as well, essentially
because there’s not screening or monitoring
needed during these times. So it’s not the way
it works, but given that Tim Geithner
was here yesterday, I thought of an
interesting analogy. So he has this foam
on the runway analogy, a metaphor of thinking
about what you do to fight a financial crisis. So it’s not the way
it works, but the idea is if there’s a lot
of foam in the runway, no profit-maximizing airline
would put any maintenance into their landing
gear, because you can land just fine without them. And then during
this period where there’s foam on the runway,
and lots and lots of it, you will see plenty
of safe landing for firms who didn’t put their– banks– banks, firms– planes who didn’t put
down their landing gear. So that’s essentially our– so
if you want to give incentives for people to be careful,
you’ve got to make them– they have to be– and careful means
putting covenants and increasing your future
borrowing capacity– it has to be the
case that if you don’t do that, your assets
don’t sell at a very high price, you can’t borrow as
much in the future. And if you can’t refinance,
inefficient things will happen. So we just take this as
an outcome of our model. As it turns out
that Janet Yellen and other central
bankers around the world have been a little
bit on the same page. We came from this from
theory, rather than really understanding what’s going on. So basically, this model says
that in these high liquidity periods where assets are not
selling at anything close to fire-sale values,
there’s going to be a boom in
covenant-lite lending. And we’ll see that
there has been. I’ll show you a picture
of it in a minute. And if you think about this
as corporate governance does, is sort of a waste in the sa–
improved corporate governance, and improved
accounting, auditing is a waste in the same way
that landing gear maintenance would be if there was lots
of foam on the runway. We see there was a recent
uptick in US audits from the– US regulations were
regulated– where auditors report major weaknesses in
internal control in earnings restatements. So here’s the covenant light. Raghu showed a similar
picture last night. But essentially, if
you look at the– the blue is the billion-dollar
values of covenant-lite loans coming out in the US. So leverage loans. Syndicated loans are high yield. Covenant-lite means
that there’s not a control covenant, like
a net worth covenant, that trips something and
turns the thing, if it’s uncollateralized,
then collateralized, or gives the lender the right
to accelerate payment to today. So there was essentially
none of this before 2004. There was a little blip
of it up in 2006 to 2007. Sort of went away for a while,
and it came back in spades. And it’s very, very large today. So that’s the blue. And then the gold
line is the fraction, the percentage of
covenant-lite leverage loans. It’s around 50% at the
end of this period. It’s actually gone
up a bit since. This is through the
third quarter of 2017, when we put this data together. So I also mentioned
this thing on weakness of internal control. This is the percentage
of firms that reported with weak internal
control and earnings restatements the year or two
subsequent years after that. So we saw a big increase
of this right in the run-up before the crisis. It went down to almost nothing
right after the crisis, and it’s sort of going back up. That’s, again, a
little more suggestive. The covenant-lite thing
is a direct measure of what we’re talking about. The weakness of
internal control is a consequence of
previous periods of financial carelessness. So then Raghu talked
about this one last night, that basically,
the point of this is that, if you correlate these
things like covenant-lite, things that look like not lots
and lots of financial care being taken, not lots of
monitoring and screening going on, it does
correlate quite highly with these measures of financial
looseness within countries. OK. Great. So again, with moderate amounts
of liquidity in the system, the market encourages covenants. Market forces naturally
limit leverage, encourage covenants, et cetera. We’re doing some empirical work
on the pricing of covenants and things like that. There’s some evidence that
suggests this is true. With huge amounts of
liquidity in the system, market forces don’t
naturally encourage covenants and low leverage. The cross-section
pricing of more covenants is fairly flat so that
if you don’t like them, you don’t get them. Only if we get a negative
shock after the period of high liquidity, we’ll find
that the firms didn’t maintain their financial capacity. They didn’t put
covenants in their loans. They didn’t improve their
corporate governance, and bad things will happen. And this explains Warren
Buffett’s quote in our mind. Only when the tide goes
out do you discover who’s been swimming naked. And so our idea is that
swimsuits are endogenous, and people would really like it
if other people wore swimsuits. But in periods where the tide,
the liquidity is very high, they’re basically a waste of
resources, because there’s no– well, there’s too
much transparency after the tide goes out. OK. So that’s our basic idea. And I had a little
picture of how this works, but I don’t have time
to show you the picture. Basically, liquidity goes up. Fire-sale pricing goes away. If it’s going to stay away
with a high-enough probability, leverage goes up in a way
that removes your incentive to improve pledgeability. Look at the paper. The original paper’s
on my website, and the second paper
will be on my website as soon as I get
back to Chicago. Thank you very much. [APPLAUSE] Doug is happy to
take some questions. Yeah. Yes. Isn’t another way to– [INAUDIBLE] There’s a microphone. Hi, I’m Bengal. Isn’t another way to think about
it that, simply during a time when there’s a lot of
liquidity, there are basically more cash chasing fewer assets,
and people just cut corners because they got to
put money to work? That’s sort of an alternative– so essentially,
that’s going to say, the quality of firms
that get financed is going to go way down. So if you imagine there was– lots of liquidity
going around basically means that the discount rate
that people use to discount future cash flows changes. So if there’s too
much liquidity, you start going more and more
down the margin of bad firms. So that could be true. In fact, there’s also
this Shleifer-Vishny thing about excessive optimism
during these times. That could be true as well. We have a point which is
separate from those, and works in addition to those,
that in some sense, the actual contracting,
governance, covenant-monitoring and things like that, market
forces don’t naturally encourage those things
during such periods. So we actually– in
our models, the amount of liquidity that the buyers
have is never a constraint. This is purely liquidity
in the hands of firm– like Rob Vishny and
Andrei Shleifer’s stuff– in the hands of firms
who could acquire the specific real
assets of firms. When those guys have a lot of
net worth, we got our effect. Lots of money floating around
among financial intermediaries, or private equity firms,
and things like that. We don’t have that effect. I think, in practice,
they’re both important. But our thing through governance
is not an implication, I think, of what you just said. Wilson? [INAUDIBLE] Oh. Question about– there’s a
lot of enforced liquidity now in the financial system. So LCR, NSFR put trillions
of enforced liquidity in banks’ balance sheets. Does that mean banks
have more liquidity, or does that mean their actual
buffers of real liquidity– the distance to regulatory
constraint– is now lower, and there’s less liquidity? How do you view that conundrum? OK. So first thing, the
liquidity in our model is sitting on corporate
balance sheets. And then we have
this separate thing that determines how much
leverage and liquidity the financial
intermediaries have. So that’s just in terms
of relating it directly. So let me put on
a different hat. Anil Kashyap and I have
thought about the point of requiring some liquidity
that can’t be used. So the point that we thought– the only way you
can understand that is that it’s not like a
fixed amount of liquidity, but it’s a fixed
proportion of liquidity as a function of
your liabilities. So the idea is that since
there’s some you can’t use, liquidity requirements like the
LCR or the net stable funding ratio encourage banks
to hold liquidity in excess of that’s
required, so that they hold the buffer over
the minimum, which increases the
amount of liquidity held in the entire system. And then we had the idea
that if the system as a whole holds this extra liquidity,
then the lender of last resort, at the first little
bit, is lending against required
liquidity in these firms, so it actually could
improve the stability of the whole financial system. If you think aggregate
liquidity matters as much as individual bank
liquidity, if you think about systemic liquidity. Liquidity requirements, as
long as there’s not a number, as long as it’s a proportion,
actually makes perfect sense. Doug. Bob Pozen from MIT. So is there any
evidence that having a fluctuating in a V on
institutional money funds actually reduces the
chance of a liquidity problem or financial crisis? As I’m sure you know, in Europe,
they had fluctuating in a V before 2008, and
people still fled– So I completely,
completely agree. I was part of this thing
called the Squam Lake Group, and we had a recommendation that
essentially required capital require– if you wanted to
have a non-floating NAV, you could only do that if you
had at least 2% capital left, and it had to float after
that, and then you essentially couldn’t take any new money in. So basically, if you think
about commercial paper, there’s not an active secondary
market for commercial paper. So if there’s no market to
mark to, marking to market is a fairly stupid concept. So that’s why I
said, the thing that worked on the money market
funds was the fact that they had to mark to market meant
that the corporate treasurers couldn’t treat them as safe. And marking to market wouldn’t
work, but killing them did work. So that’s my take on this. Thank you. OK, well, let me thank
Doug for his keynote. Thank you very much. [APPLAUSE]

Leave a Reply