Collective Moral Hazard, Maturity Mismatch, and

Systemic Bailouts
Emmanuel Farhi and Jean Tirole

February 8, 2011
Abstract
The paper shows that time-consistent, imperfectly targeted support to distressed
institutions makes private leverage choices strategic complements. When everyone
engages in maturity mismatch, authorities have little choice but intervening, creating
both current and deferred (sowing the seeds of the next crisis) social costs. In turn, it is
protable to adopt a risky balance sheet. These insights have important consequences,
from banks choosing to correlate their risk exposures to the need for macro-prudential
supervision.
Keywords: monetary policy, funding liquidity risk, strategic complementarities,
macro-prudential supervision
JEL numbers: E44, E52, G28.

Farhi: Harvard University, department of Economics, Littauer 318, 1875 Cambridge Street, Cambridge
MA 02138, Tel: 617-496-1835, Fax: 617-495-8570, [email protected], and Toulouse School of Economics.
Tirole: Toulouse School of Economics, Manufacture des Tabacs, 21 allees de Brienne, F-31000 Toulouse,
Phone: (33 5) 61 12 86 42, Fax: (33 5) 61 12 86 37, e-mail: [email protected]. We thank Fernando Alvarez, Markus
Brunnermeier, Doug Diamond, Luigi Guiso, John Geanakoplos, Bob Hall, Pricila Maziero, Martin Schneider,
Rob Shimer, Jeremy Stein, Nancy Stokey, and Mike Woodford. We also thank seminar participants at the
Bank of France, Bank of Spain, Bocconi, Bonn, Chicago Booth School of Business, IMF, Kellog, LSE,
Maryland, MIT, MIT Sloan School of Business, New York Fed, UCLA Anderson School Management,
University of Houston, Yale and at conferences (2010 AEA Meetings in Atlanta, Banque de France January
2009 conference on liquidity, Banque de France - Bundesbank June 2009 conference, Columbia conference on
Financial Frictions and Macroeconomic modelling, European Symposium on Financial Markets in Gersenzee,
SED in Montreal, XIIth Workshop in International Economics and Finance in Rio, XIIth ECB-CFS Network
conference in Rome). Jean Tirole is grateful to Banque de France and Chaire SCOR for nancial support
to research at IDEI and TSE.
1
One of the many striking features of the recent nancial crisis is the extreme exposure of
economically and politically sensitive actors to liquidity needs and market conditions:
Subprime borrowers were heavily exposed to interest rate conditions, which aected
their monthly repayment (for those with adjustable rate mortgages) and conditioned
their ability to renance (through their impact on housing prices).
Commercial banks, which traditionally engage in transformation, had increased their
sensitivity to market conditions. First, and arbitraging loopholes in capital adequacy
regulation, they pledged substantial amounts of o-balance-sheet liquidity support to
the conduits they designed.
1
These conduits had almost no equity on their own and
rolled over commercial paper with an average maturity under one month. For many
large banks, the ratio of asset backed commercial paper to the banks equity was
substantial (for example, in January 2007, 77.4% for Citibank and 201.1% for ABN
Amro, the two largest conduit administrators).
2
Second, on the balance sheet, the
share of retail deposits fell from 58% of bank liabilities in 2002 to 52% in 2007.
3
Third,
going forward commercial banks counted on further securitization to provide new cash.
They lost an important source of liquidity when the market dried up.
Broker-dealers (investment banks) gained market share and became major players in
the nancing of the economy. Investment banks rely on repo and commercial paper
funding much more than commercial banks do. An increase in investment banks
market share mechanically resulted in increased recourse to market nancing.
4
The overall picture is one of a wide-scale maturity mismatch. It is also one of substan-
tial systematic-risk exposure, as senior CDO tranches, a good share of which was held by
commercial banks, amounted to economic catastrophe bonds.
5
This paper argues that this wide-scale transformation is closely related to the unprece-
dented intervention by central banks and treasuries.
6
As described more in detail when we
map out the interpretation of our model in terms of actual policies, roughly two categories of
1
See e.g. Figure 2.3 (page 95) in Acharya-Richardson (2009), documenting the widening gap between
total assets and risk-weighted assets.
2
See Table 2.1 (page 93) in Acharya-Richardson for the numbers for the 10 largest administrators.
3
Source: Federal Reserve Boards Flow of Funds Accounts.
4
Source: Federal Reserve Boards Flow of Funds accounts.
5
To use Coval et al (2007)s expression.
6
Strikingly, by March 2009, the Fed alone had seen its balance sheet triple in size (to $ 2.7 trillion) relative
to 2007.
2
interventions were pursued in order to facilitate nancial institutions access to renancing.
For lack of better words, we term them respectively interest rate and transfer policies. By
interest rate policies, we have in mind various forms of government intervention which eec-
tively lower borrowing costs for banks: lowering the Fed Funds rate to zero, extending debt
guarantees to a wide range of nancial institutions, accepting low-quality assets as collateral
with low haircuts in loans or repurchasing agreements, purchasing commercial paper in the
primary market etc. By transfer policies, we refer to interventions that primarily boost the
net worth of nancial institutions without lowering their borrowing cost: recapitalizations,
the purchase of legacy assets at inated prices etc. The distinction between these two cate-
gories is sometimes blurred in practice. From our perspective, the key distinguishing feature
is whether the intervention under consideration reduces banks borrowing costs or simply
acts to boost their net worth.
In a nutshell, the central argument of the paper is that private leverage choices depend
on the anticipated policy reaction to the overall maturity mismatch. Dicult economic
conditions call for public policy to help nancial institutions weather the shock. Policy
instruments however are only imperfectly targeted to the institutions they try to rescue. For
example, the archetypal non-targeted policy, lowering the Fed Funds rate, benets nancial
institutions engaging in maturity mismatch, but its eects apply to the entire economy. An
accommodating interest rate policy involves (a) an invisible subsidy from consumers to banks
(the lower yield on savings transfers resources from consumers to borrowing institutions),
(b) current costs, such as the (subsidized) nancing of unworthy projects by unconstrained
entities, and (c) diered costs (the sowing of seeds for the next crisis, both through incentives
for maturity mismatch, going forward, and the authorities loss of credibility).
While the rst cost is proportional to the volume of renancing, the other two are not and
are instead akin to a xed cost. This generates strategic complementarities in balance-sheet
riskiness choices. It is ill-advised to be in a minority of institutions exposed to the shock, as
policymakers are then reluctant to incur the xed cost associated with active interest rate
policy. By contrast, when everybody engages in maturity transformation, the central bank
has little choice but intervening. Refusing to adopt a risky balance sheet then lowers banks
rate of return. It is unwise to play safely while everyone else gambles.
7
The same insight applies when some players expose themselves to liquidity risk either
7
As Charles Prince, then CEO of Citigroup, famously stated in the summer of 2007: as long as the
music is playing, you have to get up and dance.
3
because they are unsophisticated
8
or because they engage in regulatory arbitrage.
9
Strategic
complementarities then manifest themselves in the increased willingness of other actors to
take on more liquidity risk due to the presence of unsophisticated players or regulatory arbi-
trageurs. A reinterpretation of our analysis is thus in terms of an amplication mechanism.
The papers rst objective is to develop a simple framework that is able to capture and
build on these insights. Corporate entities (called banks) choose their level of short-term
debt (or, equivalently, whether to hoard liquid instruments in order to meet potential liquid-
ity needs). In the basic model liquidity shocks are correlated, and so there is macroeconomic
uncertainty. Maturity transformation is intense in the economy when numerous institutions
take on substantial short-term debt. The issuance of short-term debt enables banks to in-
crease their leverage and investment, but exposes them to a potential renancing problem in
case of a shock. When privileging leverage and scale, bankers thereby put at risk banking
stakeholders, a designation regrouping those agents who would be hurt in case banks have
to delever: bankers themselves, industrial companies that depend on bank loans for their
nancing, and employees of those banks and industrial companies.
10
Authorities maximize a weighted average of consumer surplus and banking stakeholders
welfare. Focusing in a rst step on interest rate policy, they can, in case of an aggregate shock,
facilitate troubled institutions renancing by lowering the eective interest rate at which
banking entrepreneurs borrowing. However, loose interest rate policy, besides transferring
resources from consumers to banks with renancing needs, might for example facilitate the
nancing of unworthy projects (in the basic version) or entails future costs (future illiquidity
of institutions or loss of credibility). This distortion is akin to a xed cost, which is worth
incurring only if the size of the troubled sector is large enough. We obtain the following
insights:
Excessive maturity transformation. The central bank supplies too much liquidity in
the time-consistent outcome. Our theory therefore brings support to the view that
authorities in the recent crisis had few options when confronted with the fait accompli,
and that the crisis should have been contained ex ante through more careful prudential
policies. While prudential supervision is traditionally concerned with the solvency of
8
Such players may for instance miscalibrate the risk involved in relying on funding liquidity or on secu-
ritization to cover their future needs, and thereby mistakenly engage in maturity mismatch.
9
As was the case with largely underpriced liquidity support to conduits.
10
Note that consumers may have multiple incarnations: As taxpayers/savers, they should oppose an
intervention, while as employees of these corporate entities, they might welcome it. All these eects are
taken into account in our welfare analysis.
4
individual institutions, our framework suggests the potential value of a new, macro-
prudential approach, in which prudential regulators consider not only the individual
institutions transformation activities, but also the overall transformation of the strate-
gic institutions.
11,12
Optimal regulation. In our model, optimal regulation takes the form of a liquidity
requirement or equivalently of a cap on short-term debt. Importantly, breaking down
banks into smaller banks would achieve no benet in our framework. The basic problem
here is not too big too fail, but rather that the banks as a whole are doing too much
maturity mismatch, and are taking on too much correlated risk.
Regulatory pecking order. If regulation is costly, our model suggests that regulation
should be conned to a subset of key institutions, the ones that authorities are the
most tempted to bail-out ex post.
13
Endogenous macroeconomic uncertainty. We relax the correlated-shock assumption
and let banks choose the correlation of their shock with that of other banks. We nd
that they actually choose to maximize the correlation of their shocks due to the nature
of the policy response. This result runs counter to conventional wisdom. Financial
theory (CAPM) predicts that, when faced with a choice among activities, a rm will
want to take as much risk as possible in those states of nature in which the economy
is doing well. That is, it will strive to be as negatively correlated as possible with the
market portfolio.
Sowing the seeds of the next crisis. Loose interest rate policy today increases the likeli-
hood of future crises. First, they signal the central banks willingness to accommodate
maturity mismatches, and deprive it of future credibility. Second, they stimulate new
11
Although extremely imperfect, liquidity regulation does exist at the micro level (both through stress
tests under Basel II, and through the denition of country-specic liquidity ratios).
12
These questions are at the forefront of the regulatory reform agenda. The Financial Stability Forum
(2009) calls for a joint research program to measure funding and liquidity risk attached to maturity transfor-
mation, enabling the pricing of liquidity risk in the nancial system (Recommendation 3.2) and recommends
that the BIS and IMF could make available to authorities information on leverage and maturity mismatches
on a system-wide basis (Recommendation 3.3).
13
These strategic institutions correspond to large retail banks (where size matters indirectly because of the
disruption in the payment and credit systems, or because of the greater coverage in the media), or to other
large nancial institutions that are deeply interconnected with them through opaque transactions (as was
the case recently with AIG or the large investment banks). They also include those with close connections
with the central bank; in the latter respect, while starting with Barro-Gordon (1983) the literature on central
bank independence as a response to time-inconsistency has emphasized political independence, our analysis
stresses the need for independence with respect to the nancial industry.
5
maturity mismatches through a price eect: They make short-term debt cheaper, en-
couraging maturity mismatches; and they provide a subsidy to capital, encouraging
overall leverage.
Interest rate policies are rough instruments because they entail distortions. By contrast
transfer policies do not entail similar distortions and instead involve only a subsidy. One
might therefore conjecture that interest rate policy is a dominated instrument when such
transfer policies are available. Relatedly, we need to check the robustness of the insights
stated above to optimal policy interventions.
Accordingly, the second objective of the paper is to analyze the optimal bailout mix using
a mechanism design approach. We allow authorities to operate direct transfers to institutions.
However when implementing transfer policies, they face an asymmetry of information (they
are unsure which banks are distressed or intact); consequently, direct transfers entail a
dierent set of distortions, associated with wasted-support costs. We characterize the optimal
policy intervention (interest rate and transfers) given informational constraints. We show
that:
Interest rate policy is actually always used in equilibrium; indeed transfers are not even
used unless the crisis aects a large fraction of the banks, in which case interest rate
policy and transfers are used in conjunction. The key insight is that interest rate policy
is a market-driven solution, in that it benets primarily those institutions with actual
borrowing needs; put more technically, it helps screen out opportunistic institutions
with limited renancing needs. While transfers better focus on strategic actors, they
entail a greater waste of resources by supporting entities that have no need for, or
should not engage in renancing.
The insights gleaned for pure interest rate bailouts carry over to optimal bailouts:
strategic complementarities in the size and quality of liquidity positions, excessive
maturity transformation, pecking order of regulation and endogenous macroeconomic
uncertainty.
The paper is organized as follows. Section I sets up the model. Section II analyses the
commitment benchmark, where the central bank can announce and stick to an interest-rate
policy. Section III performs the same exercise for the time-consistent outcome. Section IV
draws the implications for regulation. Section V provides the two foundations for the hazards
6
of low interest rate policies as sowing the seeds for the next crisis. Section VI allows for the
full range of policy instruments and derives the optimal bailout policy. Finally, Section VII
concludes.
Relationship to the literature. Our paper is related to several disjoint bodies of literature.
The importance of keeping interest low in recessions is classic in macroeconomic theory. Our
paper contributes to this literature rst by pointing out a new channel through which interest
rate policy suers from time inconsistency (of the kind emphasized by Finn E. Kydland and
Edward Prescott 1977), and second by viewing interest rate policy in a broader bailout
context in which support to institutions and asset prices are alternative instruments.
Potential macroeconomic shortages of liquidity exist if corporations are net lenders (Michael
Woodford 1990) or if corporations are net borrowers and face macroeconomic shocks (Bengt
Holmstr om and Jean Tirole 1998). The literature on aggregate liquidity has emphasized
the role of governments in providing (possibly contingent) stores of value that cannot be
created by the private sector. Like in Holmstr om and Tirole, liquidity support is viewed
here as redistribution from consumers to rms in bad states of nature; it however is an ex
post redistribution rather than a planned one, and it emphasizes the role of interest rates
and more generally borrowing costs in enabling renancing.
Time-inconsistency from rescuing banks and the resulting moral hazard problems in a
single-bank context have been emphasized by numerous works, starting with Walter Bagehot
(1873).
Through its emphasis on strategic complementarities, our paper is reminiscent of the
wide body of literature on multiple equilibria in macroeconomics, starting with Peter A.
Diamond (1982) and Russell W. Cooper and Andrew John (1988) (see, e.g., Cooper 1999
for a review). Our paper emphasizes the idea that strategic complementarities stem from
the governments policy response. In that, it is particularly related to Stephen Morris and
Hyun Song Shin (1998), Martin Schneider and Aaron Tornell (2004) and Romain Ranci`ere,
Aaron Tornell and Frank Westermann (2008). Morris and Shin, and Schneider and Tornell
are concerned with exchange rates while Ranci`ere, Tornell and Westermann focus on a
risky technological choice. These papers posit that the government accommodates private
agents once the latter have reached some exogenous threshold of private involvement (in
speculation, currency mismatch or realized returns). This threshold gives rise to strategic
complementarities. An important dierence in our model is that the incentives to bail out
and hence the policy reaction function are endogenized. This puts the time-inconsistency of
7
policy at the center stage and has important positive implications for comparative statics,
as well as normative consequences by allowing us to study the optimal design of regulation.
Viral Acharya and Tanju Yorulmazer (2007, 2008) study the incentives for banks to
correlate the risks inherent in their investment choices. In Acharya-Yorulmazer (2007), the
possibility for one bank to acquire the other pushes banks to minimize their correlation.
However, they assume that when both banks fail, both banks are bailed out. If the bailout
guarantee when both banks fail is worth more than the rent obtained by the surviving
bank when only one bank fails and is sold to the other, then banks seek to maximize their
correlation. Acharya-Yorulmazer (2008) introduces a richer model with re-sales and makes
the point that from an ex-post perspective, bailing out failed banks and subsidizing intact
banks to take over failed banks have similar eects, but that the latter is preferable ex ante
because it induces banks to dierentiate their risks. There are important dierences with
our paper: rst, in our setup, bank managers are indispensable to the project, so that intact
banks are at no comparative advantage over outside investors when renancing failed banks;
second, these papers do not emphasize the role of untargeted policy instruments; third, they
do not allow banks to vary the amount of risk that they take.
14
Humberto N. Ennis and Todd Keister (2009, 2010) study a modication of the model of
Douglas W. Diamond and Phillip H. Dybvig (1983) where a policy maker can only choose
policies (such as deposit freezes) that are contingent on the fraction of agents who have
already withdrawn their deposits, and that are ecient ex post. They point out that bank
run equilibria can exist together with the ecient equilibrium. Our paper shares the idea
that policy responses (under no commitment) can generate multiplicity. However, while they
focus on the incentives of depositors to run, we analyze instead the ex ante choices of banks.
Diamond and Raghuram Rajan (2009) emphasize, as we do, that interest rate policy
is time inconsistent and that low-interest rate policies may encourage excessive leverage.
Interestingly, in their framework, because of an assumed form of market incompleteness (non-
contingent deposits) absent in our model, optimal interest rate policy under commitment
involves both low interest rates in bad times and high interest rates in good times.
Varadarajan V. Chari and Patrick J. Kehoe (2010) study a model in which inecient
ex-post liquidations are required for ex-ante eciency, but the possibility of ex-post bailouts
introduces a time-inconsistency problem. They show, as we do, that regulation in the form
of specic ex-ante restrictions on private contracts can increase welfare. Interestingly, in
14
In our model, this occurs through the choice of short term debt, leverage, and maturity mismatch.
8
their setup, the ex-post cost of bailouts is that they trigger a bad continuation equilibrium
of the policy game. In the best equilibrium, any deviation triggers the worst equilibrium,
introducing a xed cost from bailouts. This reputational mechanism is therefore similar to
ours. We elaborate on this analogy in Section V.
Finally, the optimal regulation in our model bears some resemblance to Anil K. Kahsyap,
Rajan and Jeremy Stein (2008). They propose replacing capital requirements by mandatory
capital insurance policy, whereby banks are forced to hoard liquidity, in the form of T-bills.
I. The model
A. Banks
There are three periods, t = 0. 1. 2. Banking entrepreneurs have utility function l =
c
0
+ c
1
+ c
2
, where c
t
is their date-t consumption. They are protected by limited liability
and their only endowment is their wealth at date 0. Their technology set exhibits constant
returns to scale. At date 0 they choose their investment scale i and a level of short-term
debt (see below). At date 1, a safe cash ow :i accrues, that can be used to pay back the
short-term debt. Uncertainty bears on the investment project: It is intact with probability
c. and distressed with probability 1c. Whether the project is intact or distressed depends
on the realization of an aggregate shocka crisis. In other words, the shocks impacting
the dierent banking entrepreneurs are perfectly correlated.
15
If the project is intact, the investment delivers at date 1; it then yields, besides the safe
cash ow :i, a payo of j
1
i, of which j
0
i is pledgeable to investors.
16
If the project is
distressed, the project yields no payo at date 1, except for the safe cash ow :i. It yields a
payo at date 2 if fresh resources , are reinvested. The project can be downsized to any level
, i. It then delivers at date 2 a payo of j
1
,, of which j
0
, is pledgeable to investors.
17
The
15
Later we will allow entrepreneurs to choose the correlation of their shock with those faced by other
entrepreneurs.
16
As usual, the agency wedge j
1
j
0
can be motivated in multiple ways, including limited commitment,
private benets or incentives to counter moral hazard (see Section I.B; see also Holmstr om and Tirole 2010).
17
Note that we are assuming that the manager is indispensable to the project. As a result, intact banks
are at no advantage over consumers in buying or operating distressed banks. This assumption turns o a
channel that could generate strategic substituabilities, whereby some institutions (banks, or other specialist
buyers) overhoard liquidity to secure available resources when a lot of banks are distressed and attractive
opportunities arise. On the other hand if banks are expected to be rescued (as is the case in the paper),
specialist buyers have no incentive to hoard liquidity. See Acharya-Yorulmazer (2007, 2008), and chapter 7
9
following assumption will guarantee that the projects are attractive enough that banking
entrepreneurs will always invest all their net worth.
18
Assumption 1 (high return) j
1
1 : + 1 c.
The interest rate is a key determinant of the collateral value of a project. It plays an
important role in determining the initial investment scale i as well as the reinvestment scale
,. We explain how interest rates are determined in Section I.B. In sum, the gross rate of
interest is equal to 1 between dates 0 and 1. Between dates 1 and 2, the interest rate is equal
to 1 in the absence of a crisis, and to 1 1 otherwise.
19
For the rest of the paper, we adopt
the convention that 1 refers to the interest rate between dates 1 and 2 if there is a crisis.
At the core of the model is a maturity mismatch issue, where a long-term project requires
occasional reinvestments. The bank has to compromise between initial investment scale i
and reinvestment scale , in the event of a crisis. Maximizing initial scale i requires loading
up on short-term debt and exhausting reserves of pledgeable income. This in turn forces the
bank to downsize and delever in the event of a crisis. Conversely, limiting the amount of
short-term debt to mitigate maturity mismatch requires sacricing initial scale i.
The bank issues state-contingent short-term debt. It is always optimal to set short-
term debt in event of no crisis equal to :i. We denote di (where d :) the amount of
short-term debt in the event of a crisis; we refer to it simply as short-term debt through-
out the paper. The excess ri (: d)i of the safe cash ow :i over debt payments di
represents cash available at date 1 in the event of a crisis (r is the analog of a liquidity
ratio). We assume that any potential surplus of cash over liquidity needs for reinvestment
max {(: d) i , (1 j
0
,1) . 0} is consumed by banking entrepreneurs. The policy of
pledging all cash that is unneeded for reinvestment is always weakly optimal. Pledging less
is also optimal (and leads to the same allocation) if the entrepreneur has no alternative use
of the unneeded cash to distributing to investors. However, if the entrepreneur can divert
(even an arbitrarily small) fraction of the extra cash for her own benet, then pledging the
entire unneeded cash is strictly optimal.
At date 1, in the adverse state, the bank can issue new securities against the date-2
in Holmstrom-Tirole (2010).
18
This condition is intuitive: investing 1 at t = 0 and 1 at date 1 if a crisis occurs yields a return j
1
+
and costs 1 + (1 c) .
19
In all the cases that we consider, it is always optimal for the central bank to set the interest rate to 1 at
date 0 (see Section V), and also at date 1 if there is no crisis, but to some 1 1 at date 1 if there is a crisis.
10
pledgeable income j
0
,, and so its continuation , [0. i] must satisfy:
, (: d)i +
j
0
,
1
yielding continuation scale:
, = min

r
1
j
0
1
, 1

i.
This formula captures the fact that lower interest rates facilitate renancing. A banking
entrepreneur would never choose to have excess liquidity and so we restrict our attention to
d [: (1 j
0
,1) . :] or equivalently r [0. 1 j
0
,1].
The banks needs to raise i from outside investors at date 0. Because the bank
returns di + (: d) i + j
0
i to these investors in the good state and only di in the bad one,
its borrowing capacity at date 0 is given by:
i = c(:i +j
0
i) + (1 c)di
i.e.
i =

1 : cj
0
+ (1 c)r
.
The banking entrepreneur therefore maximizes over d [:(1 j
0
,1) . :] or equivalently
r [0. 1 j
0
,1]:
(j
1
j
0
) [ci + (1 c),] = (j
1
j
0
)
"
c + (1 c)
a
1j
0
1
1 : cj
0
+ (1 c)r
#

The banking entrepreneur loads up on short-term debt (r = 0, i.e., d = :) if and only if

c +: 1 +cj
0

1
1
1

.
and takes on just enough short-term debt to be able to continue full scale (r = 1 j
0
,1,
i.e., d = : 1 +j
0
,1) otherwise.
We assume that the banking entrepreneurs prefer to limit the amount of short-term debt
to have enough liquidity to continue at full scale in the adverse state of nature in the relevant
range of interest rates (which will be [j
0
. 1]):
20
20
Note that Assumption 2 implies that 1 cj
0
0 which guarantees that investment is nite.
11
Assumption 2 (demand for liquidity): c +: < 1.
B. Rest of the economy
Consumers born at date t {0. 1} consume at date t +1; so their utility is n
t
= c
t+1
. They
are endowed with a large amount or resources (savings) : when born.
A short-term storage technology yields 1 in the next period for 1 invested today. In
particular the natural rate of interest (the marginal rate of transformation) between dates
1 and 2 is 1 = 1. For the date-1 interest rate to be 1 6= 1, the storage technology must be
taxed at rate 1 1 (see later for interpretations). The proceeds are rebated lump sum to
consumers at date 2. Throughout the paper, we assume that : is large enough to nance all
the necessary investments in the projects of banking entrepreneurs at each date t. As a result,
consumers always invest a fraction of their savings in the short-term storage technology.
This modelling device is a way to capture a range of policy interventions that reduce bor-
rowing costs for banks. For instance, taxing the short-term storage technology and rebating
the proceeds lump-sum to consumers is essentially equivalent to subsidizing investment in
the banks and nancing this subsidy by a lump-sum tax on consumers. We elaborate on this
analogy below and propose several interpretations for which this way of modelling interest
rate policy could be a convenient reduced form. For now, we do not introduce any other
instrument. In Section VI, we allow for the full range of policy instruments and derive the
optimal bailout policy.
Assumption 3 (interest rate distortion): The set of feasible interest rates is [j
0
. 1]. Fur-
thermore, there exists a xed distortion or deadweight loss 1(1) 0 when the interest rate
1 diverges from its natural rate: 1(1) = 1
0
(1) = 0. and 1 is decreasing on [j
0
. 1].
The upper bound at 1 for the interest rate 1 is not crucial but simplies the analysis. As
we shall see below, it will be used to normalize the optimal interest rate under commitment to
1 = 1. One can justify this assumption by positing arbitrage (foreigners or some long-lived
consumers would take advantage of 1 1) or by assuming that marginal distortions 1
0
(1)
are very high beyond 1. But again, we want to emphasize that this particular assumption
only simplies the exposition and plays no economically substantive role in the analysis.
The lower bound at j
0
for the interest rate 1 is also without loss of generality. Indeed, as
will become apparent, the central bank will never nd it optimal to lower the interest rate in
12
times of crisis below j
0
. At that interest rate, projects can continue at full scale even when
no liquidity has been hoarded. Lowering the interest rate below j
0
would only increase the
distortion associated with interest policy at no gain.
Assumption 4 (consumers): Suppose that date-0 investment is equal to i, that banks hoard
liquidity r and so can salvage , = ri, (1 j
0
,1) in case of crisis. Then
(i) if there is a crisis at date 1, date-1 consumer welfare is \ = 1(1) (1 1)j
0
,,1;
(ii) if there is no crisis at date 1, date-1 consumer welfare is \ = 1(1) = 0.
In (i), the second term in \ stands for the implicit subsidy from savers to borrowing
banks. Indeed date-1 consumers return on their savings e : is 1e : + (1 1) (e : j
0
,,1)
(the last term representing the lump-sum rebate on the e : (j
0
,,1) invested in the storage
technology), or e : (1 1)j
0
,,1.
21
Note that welfare \ does not include the additional
indirect benets that rms managers and workers might derive from banks functioning at
high scale. More on this below. Finally, we ignore the welfare of date-0 consumers as they
have constant utility n
0
= :.
Our modelling of interest rate policy deserves some comments. It is a stylized represen-
tation of some actual interest rate policies. Their common feature is to reduce borrowing
costs for banks. We list a few of those below.
Interpretation 1. One case in point is unconventional monetary policy.
22
Extended debt
guarantees by the government reduce the rate (1) paid both by constrained institutions that
the government wants to help (the banks) and by other borrowers. The subsidy is paid by
taxpayers who end up bearing the risk of debt.
23
Similarly, accepting assets as collateral at
low haircuts in loans or repurchasing agreements and directly purchasing commercial paper
at favorable terms lower the eective interest rate faced by borrowers. Such interventions
21
Note that we use the notation : instead of : for the savings of date-1 consumers. This is because under
our Interpretation 1 below, some of the savings : of date-1 consumers are invested in alternative wasteful
investment projects. As a result, only a part : of their savings are split between reinvestment in banks and
the short-term storage technology.
22
See, e.g., Gertler-Karadi (2009) and Gertler-Kiyotaki (2009) for models with both conventional and
unconventional monetary policies.
23
One way to formalize this is as follows. Imagine that in case of a crisis, the pledgeable part of the
return j
0
from reinvesment is the expectation of a random variable realized at t = 2 that takes the value
j
+
0
j
0
with probability ` and 0 with probability 1 `. Banking entrepreneurs can issue (defaultable)
debt with nominal value j
+
0
. A guarantee from the government to deliver a fraction c of the value of every
debt contract in case of default then reduces the (gross) interest rate demanded by creditors by a proportion
factor `, [` + (1 `) c] < 1.
13
involve (in expectation) a subsidy from savers to borrowers, reduce the marginal borrowing
cost of banks, and are therefore captured by our model.
With this in mind, the deadweight loss may for instance result from the date-1 nancing
of projects that have negative net present value at the natural rate. Suppose that there is a
distribution of nancially unconstrained rms with projects that have unit cost and return
i with cumulative distribution function H(i). Then the deadweight loss if consumers and
project owners are weighted equally is 1(1) =
R
1
1
(1 i)dH(i). If the projects owners
receive welfare weight ,
&
1 relative to consumers instead, then 1(1) = (1 1)[H(1)
H(1)] ,
&
R
1
1
(i 1)dH(i) still satises our assumptions.
24
Moreover, with our assumed preferences for consumers, a tax on the storage technology
between dates 1 and 2 combined with a lump sum rebate is exactly equivalent to a subsidy
on reinvestment in banks between dates 1 and 2. The distortion behind 1(1) arises because
these projects are subsidized at the same rate as reinvestment in banks (this implicitly
assumes that the government cannot screen out these projects from genuine positive net
present value bank projects).
Interpretation 2. Another interest rate policy captured in a stylized way by the model
is conventional monetary policy. An interpretation closely related to Interpretation 1 relies on
a view of the monetary transmission mechanism whereby higher reserves allow banks to lever
more through access to cheap retail deposits, as deposit insurance tends to be underpriced
(at least during hard times since it is not indexed on the banks riskiness).
25
This involves
an implicit subsidy to banks since this deposit insurance is backed by taxes on consumers.
Increasing reserves (or reducing reserve requirements) therefore both reduces the borrowing
cost of banks (i.e. lowers the eective interest rate face by banks) and involves a subsidy
from taxpayers to borrowing banks.
Interpretation 3. The deadweight loss function 1 can also be interpreted as a re-
duced form of a more standard distortion associated with conventional monetary policy, as
emphasized in the New-Keynesian literature. Here we have in mind not a short-term inter-
vention, but a prolonged reduction of interest rates (a year to several years, think of Japan).
Even though our model is entirely without money balances, sticky prices or imperfect com-
petition, it captures a key feature of monetary policy in New-Keynesian models routinely
used to discuss and model monetary policy. In New-Keynesian models, the nominal interest
24
Note that in this case, project owners are lumped with consumers, and their welfare is included in
consumer welfare.
25
See Stein (2010) for a detailed exposition of this view.
14
rate is controlled by the central bank. Prices adjust only gradually according to the New-
Keynesian Phillips Curve, and the central bank can therefore control the real interest rate.
The real interest rate regulates aggregate demand through a version of the consumer Euler
equationthe dynamic IS curve. Without additional frictions, the central bank can achieve
the allocation of the exible price economy by setting nominal interest rates so that the
real interest rate equals to the natural interest rate. Deviating from this rule introduces
variations in the output gap together with distortions by generating dispersion in relative
prices. To the extent that these eects enter welfare separately and additively from the
eects of interest rates on banks balance sheetsarguably a strong assumptionour loss
function 1(1) can be interpreted as a reduced form for the loss function associated with
a real interest rate below the natural interest rate in the New-Keynesian model.
2627
Under
this interpretation, monetary policy works both through the usual New-Keynesian channel
and through its eects on banks via a version of the credit channel.
28
C. Welfare and policy-making
The authorities (the central bank) control the date-1 real rate of interest.
Assumption 5 (welfare function): At date 1, the central banks objective function is a
weighted average \ of consumer welfare \ and continuation scale , (, = i if there is no
crisis): \ = \ +,,. At date 0, the central banks objective function is the expectation of its
date-1 objective function.
The second term ,, in the social welfare function deserves some comments. One possible
interpretation is as follows. Imagine that, say, three categories of banking stakeholders ben-
et from the banks ability to continue. First, and most obviously the banking entrepreneurs
themselves: They receive rent :
b
,, where :
b
is the banks stake in continuation. Second, the
26
Yet another cost, absent in cashless New Keynesian models, is the so called ination tax and arises when
money demand is elastic.
27
Because they are not our focus, we imagine here that the traditional time-inconsistency problems asso-
ciated with monetary policy in the New-Keynesian model have been resolved. As is well known, this is the
case if a sales subsidy is available to eliminate the monopoly price distortion.
28
There are two versions of the credit channel (see Bernanke-Gertler 1995 for a review): the balance
sheet channel and the bank lending channel. Our model is consistent with the former in its emphasis on
the eect of interest rates on collateral value. It is consistent with the latter in that low interest rates boost
the real economy by facilitating bank renancing and thereby increasing the volume of loanable funds to the
economy.
15
higher ,, the better o their borrowers. Third, the workers working in banks and industrial
companies; to the extent that they are better o employed (e.g., they receive an eciency
wage) and that preserved employment is related to ,, then workers welfare grows with ,.
Thus if :
)
and :
&
denote the stakes of the industrial rms and the workers, and if ,
b
, ,
)
and
,
&
denote the three categories of stakeholders welfare weights or political inuence, then
,, =

,
b
:
b
+,
)
:
)
+,
&
:
&

,. Assumptions 1 and 6 (see below) imply that ,

b
1: this guar-
antees that the central bank never seeks to lower interest rates more than what is necessary
to guarantee reinvestment at full scale since the freed-up cash is appropriated by banking
entrepreneurs and hence represents an unattractive transfer to them from consumers.
This can be formalized further along the lines of Holmstr om-Tirole (1997). We only
sketch it here: At date 0, the bank makes an investment in knowledge/sta so as to be
able to monitor a mass i of rms at date 1. These rms enter in a relationship with the
bank at date 0; from then on, they share available resources in coalition with the banks.
Independently of whether or not rms succeed, they each produce short-term prots : at
date 1 (hence the mass i of rms produces short-term prots :i). Firms succeed (return : per
rm) or fail (return 0). Success is guaranteed if the bank monitors, and the rm managers
as well as the workers in the rm do not shirk. Otherwise success accrues with probability
0. Shirking on monitoring yields benet /. shirking for a rm manager brings benet ,, and
shirking for a rm worker brings benet n. Therefore incentive payments /, , and n per
rm are required to discipline the bank, the rm manager and the workers. For simplicity,
we assume that workers are cashless. Firms are cashless at date 0; at date 1, each rm has
resources :. If there is no crisis, then the return of rms (: or 0 per rm) occurs in period
1. A crisis means that rms to be monitored must invest 1 each at date 1 and the return on
reinvested funds occurs in period 2. This model is summarized by the equations in Sections
I.A and I.B with j
0
= : (/ +, +n) . j
1
= : (, +n), :
b
= /, :
)
= , and :
&
= n. The
only dierence is that the total return on a unit of successful investment is : j
1
. This
dierence, however, makes no dierence to our analysis, since only a fraction j
0
of the return
can be pledged to outside investors, and banking entrepreneurs get to keep the dierence
between j
1
and the fraction of the return pledged to outside investors.
A roadmap. We will analyze two situations: one where the central bank can commit at
date 0 to a specic contingent policy at date 1. and the (probably more likely) alternative
where the central bank lacks commitment and instead determines its policy at t = 1 with
no regard for previous commitments. In both cases, banking entrepreneurs and consumers
form expectations regarding the interest rate 1 [j
0
. 1] that will be set if a crisis occurs.
16
Note that we have not included the interest rate 1
0
between dates 0 and 1 in the set
of policy instruments. In Section V.A we relax this assumption and allow for the storage
technology between dates 0 and 1 to be taxed, with the proceeds rebated lump sum to
consumers. We show that both under commitment and under no commitment, 1
0
= 1 is
always chosen. This justies proceeding under the assumption that 1
0
= 1.
II. Commitment solution
This Section analyzes the equilibrium when the date-1 interest rate is chosen at date 0.
Ex-ante welfare is
\
ex ante
(1) c[\ (1) +,i(1)] + (1 c)[\ (1) +,,(1)]
where
\ (1) [1(1) + (1 1)
j
0
,(1)
1
]
and
,(1) = i(1) =

1 : cj
0
+ (1 c)(1
j
0
1
)
.
Using \ (1) = 0, we can write
\
ex ante
(1) =

, (1 c)
1 1
1
j
0

i(1) (1 c)1(1).
An increase in the interest rate 1 [j
0
,1] in case of crisis reduces the distortion 1(1).
It also reduces the banks leverage and therefore investment i(1), which involves a redis-
tribution from banking stakeholders to the rest of the population. We assume that ,
1 c + 1 : j
0
, which implies that [, (1 c) (1 1) j
0
,1] i(1) is non-decreasing in
1 and hence that the optimum under commitment is 1 = 1.
Note that this condition is equivalent to that making socially undesirable a date-0 unit
lump-sum transfer from consumers to banks in the absence of interest rate policy (1 = 1):
such a transfer would have social welfare cost 1 , [i(1),] = 1 ,,(1 c + 1 : j
0
),
where the term in brackets is the leverage multiplier.
Assumption 6 (No ex-ante wealth transfer): , 1 c + 1 : j
0
.
17
Proposition 1 The optimal interest rate policy under commitment features 1
c
= 1.
III. No-commitment solution
Let us now assume that the interest rate is set at date 1. without commitment. At date 0.
investors and banking entrepreneurs form an expectation for the interest rate 1

[j
0
. 1]
that the central bank will set if a crisis occurs. Based on this expectation, the representative
bank invests at scale i (1

) and hoards just enough liquidity r

i (1

) to be able to reinvest
at full scale in the event of a crisis, where r

= 1 (j
0
,1

).
The central banks decision. At date 1, the central bank is not bound by any previous
commitment and is free to set the interest rate to maximize welfare from date 1 on. We have
to distinguish two cases depending on whether or not the economy is in a crisis.
If there is no crisis, it is optimal to set 1 = 1. There is no point in lowering the interest
rate since all banks are intact. If there is a crisis, the central bank is confronted with the
following trade-o. By setting a low interest rate, it can limit the amount of downsizing
that banks have to undergo. But this comes at the cost of a large interest rate distortion.
An additional cost comes together with the implicit subsidy to banks, in the form of a
redistribution of resources from consumers to banking stakeholders.
Because ,
b
1, the central bank would never set 1 < 1

. Indeed, lowering 1 below

1

does not increase continuation scale but merely redistrbutes resources from consumers to
banking entrepreneurs, and comes at the cost of a greater interest rate distortion. However,
the central bank might be tempted to set 1 1

. In this case, banks are forced to downsize.

The reinvestment scale is determined according to:
, =
r

1
j
0
1
i(1

) , =
1
j
0
1

1
j
0
1
i (1

) .
Proceeding as in Section II, we can compute ex-post (date-1) welfare \
ex post
(1; 1

) in
case of crisis when the central bank sets the interest rate to 1 1

and agents contracted

at date 0 anticipating an interest rate of 1

:
(1) \
ex post
(1; 1

) = 1(1) +
h
, (1 1)
j
0
1
i
1
j
0
1

1
j
0
1
i (1

) .
18
At date 1. in the event of a crisis, the central bank sets 1 [1

. 1] so as to maximize
\
ex post
(1; 1

) . Denote by R(1

) the set correspondence dened by

(2) R(1

) arg max
1
\
ex post
(1; 1

)
and let
n , (1 j
0
) .
The rst term on the right-hand side of (1) is increasing in 1 with 1
0
(1) = 0. The behavior
of the second term on the right-hand side of (1) depends crucially on the sign of n. If n 0.
then it is increasing in 1 with a positive derivative at 1 = 1. In this case, R(1

) = {1}:
There is no commitment problem. If n 0 on the other hand, then this term is strictly
decreasing in 1 with a negative derivative at 1 = 1 so that R(1

) [1

. 1]. We will focus

on this latter case in the rest of the paper.
Assumption 7 (Ex-post bailout temptation): n = , (1 j
0
) 0.
Note that this assumption is consistent our previous assumptions, and in particular with
Assumption 6: it is more tempting to transfer wealth to banks ex post than ex ante.
Equilibria. We are now in position to describe the set of equilibria of the no-commitment
economy, parametrized by the interest rate 1
ac
set by the central bank in the event of a
crisis. The equilibrium set {1
ac
} corresponds to the set of xed points of:
(3) 1
ac
R(1
ac
) .
Proposition 2 To every solution 1
ac
of equation (3) corresponds an equilibrium where in-
vestors and banking entrepreneurs correctly anticipate that the central bank will set 1 = 1
ac
if a crisis occurs, invest at scale i (1
ac
), and issue short-term debt (: 1 +j
0
,1
ac
) i (1
ac
).
Moreover, there exists 0 such that [1 . 1] {1
ac
}.
The equilibrium of the commitment economy 1
ac
= 1 is always an equilibrium of the
no-commitment economy. However, there are always other equilibria with 1 1
ac
j
0
.
The condition for 1
ac
to be an equilibrium, namely 1
ac
R(1
ac
) . is equivalent to the
following condition
(4)
nj
0
1
j
0
1

1
1
ac

1
1

i (1
ac
) 1(1
ac
) 1(1) for all 1 [1
ac
. 1] .
19
The left-hand side of equation (4) represents the cost in terms of a lower reinvestment
scale of setting a higher interest rate 1 1
ac
. The right-hand side of equation (4) represents
the gain in terms of a lower interest rate distortion of setting such a higher interest rate.
The interest rate 1
ac
is an equilibrium if and only if the cost exceeds the gain for all interest
rates 1 1
ac
. The fact that a neighborhood [1 . 1] of 1 is always part of the equilibrium
set {1
ac
} follows direction from the fact that 1
0
(1) = 0. Intuitively, the right-hand side of
equation (4) is small compared to the left-hand side for 1
ac
close enough to 1.
It is illuminating to examine the necessary and sucient condition for 1
ac
= j
0
to be an
equilibrium. In this case, the banks hoard no liquidity and the optimal policy is either to let
the banks fail (and set 1 = 1) or to make continuation self-nancing (1 = j
0
). A bailout is
chosen if 1(j
0
) (1 j
0
)i(j
0
) +,i(j
0
) 0, or:
(5)
n
1 : cj
0
1(j
0
) .
Corollary 1 Suppose that condition (5) holds. Then 1
ac
= 1 and 1
ac
= j
0
are equilibria
of the no-commitment economy.
In words, if agents expect the central bank to adopt a tough stance by setting 1 = 1 in
case of crisis, then banks choose a small scale i(1) and hoard enough liquidity (1 j
0
) i (1)
to withstand the shock even if the central bank sets 1 = 1. In turn, the central bank has no
incentive to lower the interest rate below 1. Conversely, if agents expect the central bank to
adopt a soft stance by setting 1 = j
0
in case of a crisis, then banks choose a large scale i (j
0
)
and hoard no liquidity. Then if a crisis occurs, banks can continue at a positive scale only
if the central bank sets the interest rate at its lowest possible level 1 = j
0
and engineers an
extreme bailout. In turn, this extreme bailout is the optimal course of action for the central
bank.
Strategic complementarities. The possibility of multiple equilibria illustrates that
banks leverage decisions are strategic complements. These strategic complementarities re-
sult from the interaction of three ingredients: imperfect pledgeability on the banks side,
untargeted instruments and time inconsistency on the policy side. Each banks leverage
decision has an eect on the other banks through the policy reaction function in case of a
crisis.
To see this more formally, let r [0. 1 j
0
] be the liquidity choice of a particular bank
and r [0. 1 j
0
] the choice of other banks. The central bank would never choose an
20
interest rate 1 lower than j
0
, (1 r). It sets the interest rate 1 [j
0
, (1 r) . 1] in order
to maximize
(6) 1(1) +
, +j
0

j
0
1
1
j
0
1
r
1 : cj
0
+ (1 c) r
.
Without further hypotheses, the objective function in this equation is not necessarily
concave in 1. For the sake of this discussion, assume that there is enough convexity in the
loss function 1(1) so that the objective function (6) is indeed strictly concave in 1 for all
values of r. This guarantees the existence of a unique maximizer 1

( r) [j
0
, (1 r) . 1].
Because the objective function (6) has a negative cross partial derivative between 1 and r,
and that j
0
, (1 r) is increasing in r, there exists r [0. 1 j
0
] such that: for r < r we have
1

( r) j
0
, (1 r) and 1

( r) is decreasing in r; for r r, we have 1

( r) = j
0
, (1 r)
and 1

( r) is increasing in r.
The particular bank under consideration chooses r [0. 1 j
0
,1

( r)], where its objec-

tive function is
l (r. r) = (j
1
j
0
)

c + (1 c)
a
1

0

( )

1 : cj
0
+ (1 c)r
.
The best response of this particular bank is therefore given by r( r) 1 j
0
,1

( r). It
has the following properties: for r < r, we have r( r) r and so r( r) is decreasing in r;
for r r, we have r( r) = r and so r( r) is increasing in r. The best response r( r) is
not increasing over the whole range of liquidity choices [0. 1 j
0
]. However, because it is
increasing over [ r. 1j
0
], there are strategic complementarities in liquidity choices over that
range. Note also that [ r. 1 j
0
] is the relevant range, since all equilibria are in that range.
29
Comparative statics. There are two ways to perform comparative statics when there
are multiple equilibria. One possibility is to use a selection criterion. For example, one could
select the banks preferred equilibrium, i.e., the one associated with the lowest interest rate
min{1
ac
}. Another, more ambitious approach, which we pursue here, is to establish the set
monotonicity (with respect to the inclusion order) of {1
ac
} with respect to parameters.
Below, we will establish a number of results of the following form: the set of equilibrium
interest rates {1
ac
} is expanding with respect to some parameter j. By this we mean
29
In fact, with the extra assumptions that we have made for this discussionthat the the objective function
in equation (6) is strictly concavethe set of possible equilibrium liquidity choices is simply [ r, 1 j
0
], and
the set of equilibrium interest rates is [j
0
, (1 r) , 1]. The set of equilibria need not be an interval when
this extra assumption is not veried.
21
that for j < j
0
. the set of equilibrium interest rates associated with j is included in the
one associated with j
0
. The minimum of the set of equilibrium interest rates min{1
ac
} is
weakly decreasing in j. By contrast, the maximum of the set of equilibrium interest rates
1 = max {1
ac
} is invariant to j.
Corollary 2 The set {1
ac
} of equilibrium interest rates is expanding in the relative weight
, of banking stakeholders in the central banks objective function as well as with the size (as
measured by ) of banks.
Proof. Suppose that 1
ac
R(1
ac
) - i.e. that (4) holds- when , and are set to some
initial value. Then if , and are increased to ,
0
and
0
. (4) still holds. As a result it is still
the case that 1
ac
R(1
ac
) .
Strategic complementarities associated with bigger, more powerful and more strategic
banks are stronger. The xed cost of a bailout is independent of the characteristics or choices
of banks. By contrast, for any given interest rate anticipated by the banks, the benets of
a bailout increase with the size (as measured by ), the inuence and the importance (as
measured by ,) of banks.
It is also interesting to perform comparative statics on the set of equilibria with respect
to the severity of the crisis. To this end, we now consider an extension of the basic model
where only a fraction of banks are distressed in the event of a crisis. The parameter
indexes the severity of the crisis. We keep the probability 1 c of a crisis constant and
let the probability of being intact c c + (1 c) (1 ) adjust. The logic of the model
is essentially unchanged. The only dierence is that date-1 welfare \
ex post
(1; 1

) is now
given by
(7) \
ex post
(1; 1

) = 1(1) +
h
, (1 1)
j
0
1
i
1
j
0
1

1
j
0
1

1 : cj
0
+ (1 c)(1
j
0
1

)
.
Corollary 3 The set {1
ac
} of equilibrium interest rates is expanding in the severity of the
crisis .
One interpretation of Corollary 3 (whose proof is similar to that of Corollary 2) is that one
could observe banks increasing their leverage i, and decreasing their liquidity hoarding r
as the severity of a crisis increases. This is particularly interesting since exactly the opposite
would happen in a model with xed or pre-committed interest rate. Indeed, the opposite
22
conclusion is usually obtained in corporate nance. Strategic complementarities can therefore
result in perverse comparative statics.
Endogenous correlation. We have so far assumed that the correlation of distress
shocks across banks was exogenous. We now relax this assumption and allow banks to choose
the correlation of their distress risk with other banks distress risk. There is a continuum
of states of the world indexed by o [0. 1] . The distribution of o is uniform on [0. 1] . Each
bank faces a given probability of distress 1c but can decide how to spread this distress risk
across the states of the world. That is, each bank can choose a probability of being intact
c
0
in state o subject to the constraint that c
R
c
0
do. Let 1
0
[j
0
. 1] be the interest rates
that banking entrepreneurs expect the central bank to set in state o.
The banking entrepreneur commits to repay d
0
i in state o if c
0
< 1 and :i if c
0
= 1.
In this case, banks will always choose d
0
[: (1 j
0
,1
0
) . :]. The continuation scale in
state o when pledging debt d
0
i is given by ,
0
= i (: d
0
) , (1 j
0
,1
0
) and the investment
scale by
30
(8) i =

1
R
1
0
[(1 c
0
) d
0
+c
0
(: +j
0
)] do
.
The banking entrepreneurs expected payo is (j
1
j
0
)
R
1
0
[c
0
i + (1 c
0
) ,
0
+ (1 c
0
)i] do.
Proposition 3 (i) (general structure of strict equilibria) All the strict equilibria have the
following properties: (a) there exists a set of crisis states
crisis
[0. 1] of measure 1 c
such that 1
0
< 1 if o
crisis
and 1
0
= 1 otherwise, (b) c
0
= 0 if o
crisis
[0. 1] and
c
0
= 1 otherwise, and (c) d
0
= : (1 j
0
,1
0
) and i is given by equation (8).
(ii) (particular class of strict equilibria) To every set of crisis states
crisis
[0. 1] of measure
1 c and solution 1
ac
of equation (3) corresponds a strict equilibrium where: (a) for o

crisis
, we have c
0
= 0, 1
0
= 1
ac
, (b) for o ,
crisis
, we have c
0
= 1. 1
0
= 1, and (c)
d = : (1 j
0
,1
ac
) and i = i (1
ac
).
Proof. Assumption 2 implies that d
0
= : (1 j
0
,1
0
) when c
0
< 1. The results then
follow easily from the fact that the derivative with respect to c
0
of the objective function
obtained by replacing these values of d
0
in the objective function of banking entrepreneurs
is higher, the higher is 1
0
.
30
The investors date-0 breakeven condition is i =
R
1
0
[(1 c
0
)d
0
i + c
0
i + j
0
i] d0.
23
Banks want to fail when the largest possible number of other banks are failing and
correlate their risks with those of other banks. Because interest rate policy is non-targeted,
bailouts take place in states of the world where a large number of banks are in distress,
making it cheaper to renance in these states. Proposition 3 illustrates the presence of
strategic complementarities in correlation choices. In equilibrium, banks coordinate on a
given set of crisis states
crisis
which is completely indeterminate up to the constraint that
it be of measure 1 c. This proposition also validates our choice of focusing on aggregate
shocks as opposed to idiosyncratic shocks: this is the stochastic structure that prevails when
correlation choices are endogenized.
It is important to contrast Proposition 3 with the standard prescriptions of the CAPM.
In a CAPM world, the cost of capital associated with an investment project is negatively
related to the correlation of its cash ows with the market return. As a result, a bank
would always choose a minimal correlation with the aggregate risk in the economy. In our
economy, just the opposite occurs. Banks maximize their correlations in order to fail when
all the other banks are failing and the central bank lowers interest rates.
IV. Welfare and regulation
The time inconsistency of policy introduces a soft budget constraint problem and creates
moral hazard on the banks side. In this context, banks leverage and liquidity hoarding
choices at time 0 can be inecient.
Welfare. The equilibria in {1
ac
} can be ranked in terms of welfare. Indeed, under our
assumptions, ex ante welfare \
ex ante
(1) is increasing in 1. As a result, equilibria with a
higher interest rate 1
ac
feature higher welfare. The equilibrium with the highest welfare is
the equilibrium that prevails under commitment with no bailout and the interest rate equal
to 1. Moreover, the banking entrepreneurs perspective is exactly the opposite: the lower the
interest rate 1
ac
, the better the equilibrium for banks.
Role for regulation. In this context, regulation of banks leverage and liquidity hoard-
ing choices at time 0 can be welfare improving. Indeed, consider putting a cap on short-term
debt: d :(1j
0
), or equivalently regulating liquidity hoarding by imposing r 1j
0
.
31
At t = 1. there is then no incentive for the central bank to proceed to a bailout: there would
31
In our simple model, no matter what interest rate is expected at the contracting stage, this is equivalent
to regulating leverage by imposing that i, ,[1 c + 1 j
0
].
24
be a distortionary cost and no benet to lower the interest rate below 1 since all banks are
able to continue at full scale when 1 = 1. Therefore, this regulation reduces the set of
equilibrium interest rates {1
ac
} down to a singleton {1}, i.e. the no bailout, commitment
solution.
Proposition 4 With limited commitment, the optimal regulation of banks choices at t = 0
takes the form of a liquidity requirement r 1j
0
or equivalently of a maximum short-term
debt d : (1 j
0
). With this regulation, there is only one equilibrium, which coincides
with the commitment solution 1
ac
= 1. By contrast, there is no role for such regulation
under commitment.
Remark 1 Subsidizing liquidity, a form of intervention that is sometimes put forward, would
be counterproductive in our model. It would only allow banks to increase their scale and
aggravate the time-inconsistency problem of policy, rendering bailouts more likely.
32
Remark 2 Regulations are hard to enforce and banks try to circumvent them. In Farhi-
Tirole (2009), we used : = 0 so that banks liquidity came exclusively in the form of hoarded
assets. We introduced the possibility for banks to engage in regulatory arbitrage to fool the
regulator by purchasing toxic assets instead of safe assets to fulll their liquidity requirement.
These toxic assets are cheaper but run the risk of a bad performance. We showed that in
this context, there are strategic complementarities in regulatory arbitrage. The more banks
engage into regulatory arbitrage, the lower the interest rate set by the central bank in the event
where toxic assets are worth 0. and the more each bank can aord to hoard toxic assets. The
very insights gleaned with respect to the quantity of liquidity hoarded hold just the same with
respect to their quality.
Macroprudential versus microprudential regulation. This soft budget constraint
rationale for regulation is also present in microeconomic principal-agent models when the
principal lacks commitment. The dierence in our setting is that the actions of the central
bank (the interest rate) aect all banks at the same time. If one bank were to take idiosyn-
cratic risks that would materialize only when none of the other banks are in distress, there
32
To formalize this insight, assume that the government subsidizes the short-term hoarding of liquidity so
that it returns 1 per unit in the bad state. So ( d) = 1

0
1
. The analysis of Section III carries over.
In particular the characterization of equilibria is literally identical. The only dierence is that i (1) is now
given by i (1) = , [1 cj
0
+ (1, c,) , (1 j
0
,1)]. It is then easy to verify that the set {1
nc
} of
equilibrium interest rates is expanding in the return of liquidity . Note that by having a corner solution we
shut down a possible channel through which subsidizing liquidity hoarding may help (a substitution eect).
25
would be no soft budget constraint problem. Because the only policy instrument, interest
rate policy, is not targeted, the central bank would not be tempted to lower interest rates to
bail out this individual bank when its individual risk is realized. As a result, our framework
suggests that the focus of regulation should be on aggregate leverage and liquidity hoarding
and not only on individual risk-taking. In other words, in our model, the optimal regulation
is macroprudential and not only microprudential.
It is important to stress that in our model, breaking down banks into smaller banks
would be ineective. The set of equilibria would be unaected. The problem here is not so
much that banks are too big to fail, but that the nancial sector as a whole might take on
too much correlated risk and too much short-term debt. This irrelevance result would break
down if big banks (with a high ) carried a higher welfare weight (,) than small banks per
unit of investment, say because big banks failures have bigger systemic consequences, or
because the bankruptcy of a large bank is disproportionately reported in the media, creating
pressure for a bailout.
The pecking order of regulation. So far, we have assumed that regulation is costless
and that banks are homogenous. We now relax those assumptions. Banks are allowed to
dier on size and weight , in the central banks objective function. These characteristics
are assumed to be distributed according to an arbitrary distribution 1 (,. ). We assume
that the costs c (i) of regulating a bank increase with the scale i of the bank, where c is
homogenous of degree ` : c (i) = ci
A
with c 0 and ` 0. In this context, regulation
involves a trade-o and the regulator might nd it optimal to regulate certain banks but not
others. In order to analyze this trade-o formally, we characterize the minimal amount of
aggregate resources devoted to regulation
1 =
Z
:(,. ) c

1 : cj
0
+ (1 c) (1 j
0
)

A
d1 (,. )
required to ensure that {1
ac
} = {1}. Here the authorities regulate a fraction :(,. ) [0. 1]
of banks of size and weight ,.
Proposition 5 Suppose that Assumptions 2, 6 and 7 hold for every , in the support of 1.
Then the minimal cost of insuring that {1
ac
} = {1} is achieved by regulating banks for which
[, (1 j
0
)]
1A
is greater than a certain threshold .
Proof. See the appendix.
26
When regulation is costly, optimal regulation is characterized by a pecking order de-
termined by a summary statistic [, (1 j
0
)]
1A
which combines their size and their
weight , in the central banks objective function. This summary statistic can be thought of
as a cost-of-regulation-adjusted systemic importance. For a given size , the higher ,. the
higher the bank in the regulatory pecking order (recall that , is the sum of the weights placed
on the dierent categories of bank stakeholders, i.e. banking entrepreneurs, rm managers,
and workers). Whether the rank of the bank in the regulatory pecking order increases or
decreases with its size for a given , depends on the returns to scale in regulation `. With
increasing returns to scale in regulation (` < 1), the rank of the bank in the regulatory
pecking order increases with , and the opposite holds true with decreasing returns to scale
in regulation (` 1). With constant returns to scale in regulation (` = 1), size per se is
irrelevant: the costs and benets of regulation scale up exactly at the same rate with the
size of bank.
V. Sowing the seeds of the next crisis
In Section I, we gave some examples of immediate deadweight losses 1 associated with low
interest rates. This section focuses on deferred costs, associated with the incentive for new
borrowers to lever up and increase maturity mismatch, or with the central banks loss of
reputation. We derive alternative microfoundations for the distortions associated with an
interest rate bailout. We extend the model to an overlapping generations structure with two
successive generations G
1
and G
0
of banking entrepreneurs. We derive a loss function 1
from an interest rate bailout of generation G
1
at date 0. This loss function originates in the
perverse consequences on the subsequent generation G
0
of banking entrepreneurs, who end
up levering up more and hoarding less liquidity, resulting in an interest rate bailout at date
1 (if a crisis occurs). In order to present these extensions, it is useful to rst rst extend our
basic setup by allowing policy to also determine the date-0 interest rate 1
0
. Throughout
Section V, we assume that there are no immediate date-0 distortions from lowering 1
0
, so
that any xed cost from an interest rate bailout at date 0 will be a deferred one. In Section
V.A, we allow for immediate distortions from lowering the interest at date 1, but in Sections
V.B and V.C, we assume them away.
27
A. Date-0 Interest Rate
We assume that the government can tax the rate of return on the storage technology be-
tween dates 0 and 1. The proceeds are rebated lump sum to consumers. This is rigorously
equivalent to assuming the government can subsidize investment in the banks at date 0, a
subsidy nanced by lump sum taxes on consumers. We show that it is optimal, both under
commitment and under no commitment, to set the date-0 interest rate 1
0
to 1, even without
assuming any distortion from lowering 1
0
.
33
The borrowing capacity of banking entrepreneurs at date 0 is now given by
(9) i =

1

1
0

cj
0
1
0
+
(1c)
1
0
r
where r, as earlier, denotes the liquidity ratio. Banking entrepreneurs hoard liquidity r =
1 j
0
,1 at date 0 if and only if it expects a date-1 interest rate 1 satisfying
(10) : +c 1
0
cj
0

1
1
1

.
and hoards no liquidity (r = 0) otherwise. This generalizes our analysis of Section II to
arbitrary 1
0
1. Denote by i (1. 1
0
) the corresponding investment scale chosen by banking
entrepreneursobtained by replacing the optimal liquidity choice r of banking entrepreneurs
in equation (9), a choice governed by equation (10). Similarly, let , (1. 1
0
) denote the
corresponding reinvestment scale in case of a crisis.
Ex-ante welfare under commitment is given by
\
ex ante
(1. 1
0
) ,i (1. 1
0
) (1 c)
1 1
1
j
0
, (1. 1
0
) (1 c) 1(1) (11)
[i (1. 1
0
) ] (1 1
0
) .
The last term on the right-hand side is new and reects the implicit subsidy to bank invest-
ment at date-0 (consumers invest a total amount i (1. 1
0
) ). The following proposition
shows that under the assumptions that we have maintained throughout the paper, the opti-
mal interest rate policy under commitment is passive both at date 1 and at date 0.
33
These results would only be reinforced if we were to introduce a loss function 1
0
(1
0
) as we did for the
date-1 interest rate.
28
Proposition 6 The optimal interest rate policy under commitment features 1
c
0
= 1
c
= 1.
The intuition is straightforward. Assumption 6 guarantees that redistributing resources
from consumers to banking entrepreneurs is welfare reducing. Lowering the interest rate 1
0
or 1 below 1 induces such a redistribution of resources (with additional distortions in the
case of 1). Setting these interest rates equal to 1 is therefore optimal.
Analyzing the no-commitment solution requires solving a dynamic game. We focus on
subgame-perfect equilibria. We will need to impose some renement. We know from our
previous analysis that the set of equilibrium interest rates of the continuation game is ex-
panding in the date-0 investment scale. We therefore nd it natural to focus on subgame
perfect equilibria that satisfy the following monotonicity requirement:
Assumption 8 The continuation equilibrium of the no-commitment game is such that the
date-1 interest rate is non-decreasing in the date-0 interest rate.
This assumption would be automatically satised if for example, we always chose the
worst possible continuation equilibrium (the one with the lowest 1) for any value of 1
0
.
Corollary 4 In every equilibrium of the no-commitment game, 1
ac
0
= 1.
This corollary follows directly from Proposition 6. Indeed, lowering the interest rate 1
0
below 1 only entails larger costs compared with the commitment solution, because it leads
banking entrepreneurs to hoard less liquidity anticipating a lower interest rate 1 at date 1.
B. Leverage decisions, going forward
Consider a longer-horizon model, say the overlapping-generations version of this model, in
which banking entrepreneurs live, like in this model, for three periods. A bailout at date
t + 1 of those banks that borrowed at date t then also aects the nancing decisions of the
next generation of banks, which borrow at date t + 1. Interestingly, this operates through
two channels: increased leverage and increased maturity mismatch.
34
To preview the results, bailing out generation G
t
induces leverage and maturity mismatch
for generation G
t+1
, sowing the seeds for a date-t +2 crisis and bailout. Indeed, an interest
34
Arguably, both channels seem to have operated during the long episode of very low interest rates in the
wake of the 2000 Internet bubble crash.
29
rate bailout of generation G
t
makes generation G
t+1
(a) more willing to take an illiquid posi-
tion by loading up on short-term debt and (b) increase the size of its investment. These two
eectsthe increased maturity mismatch and the increased leverage channels respectively
distort generation G
t+1
s incentives and generate a social cost that is a xed cost (in the
sense that it does not depend on generation G
t
s size or illiquidity) when contemplating a
rescue of generation G
t
.
The logic of the argument can be grasped by appending to the model of Section V.A a
prior generation of entrepreneurs, generation G
1
, living at dates 1. 0 and 1.
35
Generation
G
1
is in all respects similar to generation G
0
born at date 0 and studied in this paper, except
that its short-term (date-0) prot :
0
is suciently large that not hoarding any liquidity is
a dominant strategy (for any 1
0
[j
0
. 1]), which boils down to the condition that :
0
+c
1 + c(1 j
0
). This assumption merely shortens the analysis by ensuring that generation
G
1
in case of a date-0 crisis is unable to withstand the shock unless the date-0 interest rate
is brought down to 1
0
= j
0
.
The analysis of the generation G
0
born at date 0 is exactly as in Section V.A. We make
the following assumption, which is consistent with our previous assumptions.
36
Assumption 9 (no liquidity hoarded when 1
0
= j
0
): : (1 c)j
0
.
We rule out any deadweight loss from inecient investments, so as to starkly illustrate
that 1 can come from impaired incentives for the subsequent generation of entrepreneurs.
We maintain this assumptions throughout Sections V.B and V.C.
From our previous analysis and Assumption 8, we know that the optimal date-0 policy
if one ignores the welfare of generation G
1
consists in setting 1
0
= 1. Suppose that 1
0
= 1
is actually chosen. There are a number of possible continuation equilibria, corresponding
to dierent expectations regarding the interest rate 1. Because we have assumed away any
current distortion, all interest rates 1 [j
0
. 1] correspond to a possible continuation equi-
librium, where banking entrepreneurs hoard liquidity r = 1 j
0
,1. Using the denition in
35
The model can be extended to innite-horizon overlapping generations; we oer here the simplest illus-
tration.
36
The analysis remains valid when this assumption is violated, but the cost of bailing out generation G
1
in terms of generation G
0
s incentives then comes solely from an increased investment scale, not from an
increased maturity mismatch.
30
equation (11), this yields an ex-ante welfare for generation G
0
\
ex ante
(1. 1) =

, (1 c) (1 1)
j
0
1

1 : cj
0
+ (1 c)

1
j
0
1

where \
ex ante
(1. 1) increases with 1. In our analysis, we do not need to take a stand on
which equilibrium is actually selected.
Any 1
0
(j
0
,1) is dominated: by Assumption 8, such interest rates reduce welfare from
generation G
0
and do nothing to help generation G
1
. Therefore, we only have to analyze
whether 1
0
= j
0
is preferred to 1
0
= 1. When 1
0
= j
0
, no liquidity is hoarded by
generation G
0
, and because there are no immediate distortions from interest rate policy, a
full interest rate bailout with 1 = j
0
follows. Using the denition in equation (11), welfare
from the date-0 generation becomes
\
ex ante
(j
0
. j
0
) =
[, (1 c) (1 j
0
)]
1 : cj
0
(1 j
0
)
"
1
1

j
0
c
1
#
.
Turning now to the decision of whether to rescue generation G
1
at date 0, and remem-
bering that choices outside {j
0
, 1} are dominated, we conclude that there exists a xed cost
of bailing out generation G
1
by setting 1
0
= j
0
. Depending on which continuation equilib-
rium is selected when 1
0
= 1, with corresponding welfare \
ex ante
(1. 1), this xed cost is
given by 1(j
0
) \
ex ante
(1. 1) \
ex ante
(j
0
. j
0
). It is maximized when following 1
0
= 1,
the continuation equilibrium with 1 = 1 is selected, and minimized when the continuation
equilibrium with 1 = j
0
is selected. It can be veried that even in the latter case, we have
1(j
0
) 0 as long as Assumption 6 holds. The selection of the continuation equilibrium with
1 = j
0
isolates the increased leverage channel mentioned above. By contrast, the selection
of the continuation equilibrium with 1 = 1 combines both the increased leverage and the
increased maturity-mismatch channels.
Proposition 7 When contemplating whether to rescue the generation G
1
banks, the gov-
ernment faces a xed cost equal to 1(j
0
) = \
ex ante
(1. 1) \
ex ante
(j
0
. j
0
) 0, where
1 [j
0
. 1] corresponds to the selected continuation equilibrium when 1
0
= 1. This cost is
xed in that it does not depend on generation G
1
s investment to be rescued.
31
C. Central banks reputation
Yet another deferred cost of bailouts is the loss of reputation by the central bank. This can
be modelled by introducing a tough type and a soft type. A bailout then reveals the type of
the central bank to be soft, raising the likelihood of future bailouts and pushing banks to take
on more risk, hoard less liquidity and lever up, resulting in increased economywide maturity
mismatch and in turn larger bailouts. Even a central bank of the soft type internalizes this
reputation cost and is reluctant to engage in a bailout in the rst place.
To show how reputation concerns generate yet another xed cost, we follow Section
V.B but introduce uncertainty about the central bankers preferences. With prior (date-0)
probability 1 ., the central banker is bailout-prone as he puts weight , 1 j
0
(so
n 0) on investment as earlier. With probability ., the central bank is tough as he puts
no or little weight on investment and therefore always chooses interest rates equal to 1.
The situation is otherwise the same as in Section V.B: Generation G
1
optimally hoards
no liquidity; when it faces a crisis, the rational choice for a bailout-prone central banker is
again between 1
0
= 1 and 1
0
= j
0
. A choice of 1
0
= j
0
reveals that he is bailout-prone.
Suppose rst that the central banker sets 1
0
= j
0
to bail out generation G
1
. The
equilibrium for generation G
0
is then as in Section V.B. Welfare is \
ex ante
(j
0
. j
0
) for that
generation. By choosing 1
0
= 1 by contrast, the central banker creates posterior beliefs
b . ..
37
If . is large enough ( . .), then hoarding liquidity is a dominant strategy for a
generation-G
0
bank and welfare for generation G
0
is \
ex ante
(1. 1).
38
Proposition 8 There exists . < 1 such that for . ., the bailout-prone central banker faces
a xed reputation-loss cost equal to 1(j
0
) \
ex ante
(1. 1) \(j
0
. j
0
) 0 when rescuing
the generation G
1
banks. This cost is xed in that it does not depend on generation G
1
s
investment to be rescued.
37
b . = . in a pooling equilibrium, b . . in a partially revealing one, including a separating equilibrium for
which b . = 1.
38
The threshold . is the solution of
:o+[1:]
1to
0
=
1
1to
0
+(1o)(1
0
):
. From Assumption 2, .< 1. To
understand this condition, note that the tough central banker (who has posterior probability b .) will not bail
out banks in case of a crisis, and so the probability of continuation is at most b .c + [1 b .] in the absence
of liquidity hoarding. Conversely, hoarding liquidity (1 j
0
)i is needed in order to be able to continue in a
crisis when the central banker is tough; but if the central banker turns out to be bailout-prone and lowers
the interest rate at date 1 in case of crisis to 1 = j
0
(this is the most optimistic hypothesis, which arises
if other bankers hoard no liquidity), then this unneeded liquidity can be returned to investors. The cost of
liquidity is therefore only (1 c)(1 j
0
)b . on average, which explains the term on the right-hand side.
32
The derivation of equilibrium behavior is then straightforward. For . ., the bailout-
prone central banker pools with the tough one if and only if n
1
i
1
1(j
0
) is veried,
where n
1
= ,
1
(1 j
0
) and ,
1
and i
1
are the weight on, and the investment of
generation G
1
. If n
1
i
1
1(j
0
), the equilibrium is separating. Finally, for . <., the
details of equilibrium behavior depend on equilibrium selection, but the overall pattern (a
trade-o between the benet of bailout and a cost of reputation loss) remains the same.
VI. Optimal ex-post bailouts
So far, we have restricted the set of policy instruments to the interest rate. We emphasized
that this instrument was not targeted. Together with time inconsistency, this feature was a
key ingredient in generating strategic complementarities in leverage choices. In this section,
we relax the assumption of an exogenously specied policy instrument set. Instead, we follow
a mechanism design approach and characterize the optimal ex-post bailout where policy tools
are endogenous to the constraints of the economic environment.
As we noted in the introduction, one may wonder whether interest rate bailouts, which
involve both an implicit subsidy to banks and various distortions, are still desirable when
other interventions are possible. For example, purchases of legacy assets, liquidity support
and recapitalizations also involve direct transfers from consumers to banksboosting their
net worth and allowing them to renance at a larger scale. However, these transfers do not
reduce borrowing costs at the margin, and do not generate similar distortions.
Interestingly, we show that interest rate policy still plays an important role within the
optimal bailout scheme: It is always part of the optimal ex-post package, and over a range
of parameters, bailouts boil down to pure interest rate policy. Furthermore, we recover a
key insight from our previous analysis, the existence of strategic complementarities: to some
extent, optimal bailouts are themselves untargeted.
Setup. We allow the fraction of banks that are distressed in a crisis to be less than
1. Denoting the probability of a crisis by 1 c. we maintain the convention that c
c+(1 c) (1 ) represents the probability of being intact. In what follows the probability
c of a crisis is kept constant and the dependence of c with respect to is left implicit.
Moreover, we introduce an informational friction: the central bank can observe which
banks are distressed, but the underlying auditing technology is imperfect. More precisely,
we assume that the probability of generating a false positive when assessing if a bank is
33
distressed is equal to i.
39
As a result, in a crisis, a fraction i (1 ) of banks are mistaken
by the authorities as banks that need liquidity. These banks are aware that they belong to
the false positive group.
We assume that banking entrepreneurs and their investors form perfect coalitions, and
that banks have full bargaining power in these coalitions.
Instruments. We assume that the government cannot directly hold bank securities.
The only available instruments are the interest rate (the borrowing cost of banks) and direct
transfers to banks perceived as being distressed. We have already commented on the possible
interpretations of interest rate policies as policies that lower the borrowing cost of banks.
Direct transfers capture policies used in practice to boost the net worth of banks, such as
the purchase of legacy assets at inated prices.
The assumption that the government cannot directly hold bank securities deserves some
comments, as it rules out some forms of government intervention that are used in practice.
For example, a recapitalization involves a transfer from the government in exchange of shares
or warrants. However in most practical cases, the government usually sells its stake relatively
quickly. What remains is a transfer from the government to the bank. At a theoretical level,
this assumption limits the ability of the government to screen between intact and distressed
banks. In Farhi-Tirole (2009) we analyzed the case where the government could hold stakes
in banks and use this as a screening device. We showed that our insights were robust in this
environment. However the analysis was quite involved. The assumption that the government
cannot directly hold a stake in banks allows us to considerably streamline the analysis.
If direct transfers could be perfectly targeted to distressed banks, they would dominate
interest rates as a policy instrument. However, the government cannot perfectly recognize
if a given bank is distressed and some banks might engage in rent seeking by successfully
portraying themselves as distressed. Moreover, low interest rates benet distressed banks
comparatively more than intact ones. As a result, there is a non-trivial policy tradeo be-
tween interest rate policy and direct transfers to institutions perceived as being distressed.
40
Although we will consider only symmetric equilibria, we analyze the general case where
the authorities face an arbitrary distribution 1 (i. r) of banks with scale and liquidity (i. r).
39
Note that for simplicity, we assume that there are no false negatives.
40
The possibility of false positives is crucial for the following reason. If i = 0 or = 1, then the authorities
do not face any information extraction problem. As long as Assumption 7 holds, then banks in distress are
always rescued through a direct transfer and are allowed to continue at full scale even if banks are completely
illiquid. In equilibrium, banks would then choose to be completely illiquid, interest policy would not be used.
34
A bailout species an interest rate 1. and for every scale and liquidity (i. r) a transfer
t (i. r) i 0 for banks perceived as being distressed. This implies the following reinvestment
scale for distressed banks
, (i. r) = min

r +t (i. r)
1
j
0
1
. 1

i.
Timing. The timing within period 1 is as follows: (1) the government announces a
rescue scheme {1. t (i. r)}; (2) each banking entrepreneur accepts the plan if and only it
makes him better o given that investors must be as well o as they would in the absence
of participation in the scheme.
To simplify the expressions, we assume that the government places no Pareto weight on
banking entrepreneurs ,
b
= 0. Transfers to banks that are perceived to be distressed but
are in fact intact therefore represent a pure welfare loss.
Planning problem. In order to solve for the optimal ex-post bailout, we use the
variables {1. , (i. r)} rather than {1. t (i. r)} . The transfers t (i. r) can be inferred as follows:
t (i. r) = (1 j
0
,1) , (i. r) ,i r 0 since without loss of generality r 1 j
0
,1.
Up to a constant, the optimal ex-post bailout maximizes
41
1(1) +
Z

, (1 1)
j
0
1

, (i. r) d1 (i. r) (12)

[ + (1 ) i]
Z
h
1
j
0
1

, (i. r) ri
i
d1 (i. r)
s.t.
i , (i. r)
r
1
j
0
1
i.
To obtain equation (12), note that for distressed banks, the total subsidy is given by
t (i. r) + (1 1) j
0
, (i. r) ,1. For intact banks that are perceived as distressed, the subsidy
is t (i. r) as those banks do not need to borrow more funds. The rst two terms in equation
(12) are exactly as in Section III. The third term corresponds to the additional implicit
subsidy associated with the direct transfer t (i. r). This is a transfer to distressed rms and
intact rms that are perceived as distressed. Consumers get a rate of return equal to 0 (and
not 1 as when they reinvest in the banks) on this transfer.
41
Because of the linearity of the objective and the constraints in t (i, r) , we need to take a stand on
government actions when the government is indierent. We assume that it sets t (i, r) equal to its minimal
possible value in case of indierence.
35
The government could achieve a given pattern of continuation scales , (i. r) entirely
through direct transfers by setting t (i. r) = , (i. r) r. This would economize on interest
rate distortions. However, it increases undesirable transfers to intact banks perceived to
be distressed. Lowering the interest rate 1 increases the collateral value j
0
, (i. r) ,1 of
distressed rms, allowing them to continue at a larger scale, while at the same time reducing
the funds transferred to intact rms perceived to be distressed. By aecting the terms at
which banks can borrow on the market, the government is able to better target rms that
are actually distressed. However, it pays a cost in terms of distortions 1(1). We need to
ensure that the latter is convex enough.
Assumption 10 (enough convexity). The function 1
2
1
0
(1) is decreasing in 1 [j
0
. 1].
Optimal ex post bailout. The optimal bailout policy is summarized in Figure 1. Due
to the linearity of the program, there can be two equilibrium congurations. We discuss
them informally here and refer the reader to the appendix for the full details.
(i) No liquidity hoarding (r = 0 for all banks). In this equilibrium conguration, banks
load up on short-term debt as they expect to be rescued in case of a crisis. Given the absence
of liquidity in the banking sector, the least-cost rescue policy for the government is a mix
of transfer and interest rate policies. To see this, note that low interest rates reduce the
wasteful transfer t to intact banks given that distressed ones require a unit renancing from
both sources: 1 = t + j
0
,1. Basically, distressed banks can lever up the direct transfer to
renance at a greater scale. However low interest rates also increase the deadweight loss
1(1); the marginal distortion is 0 at 1 = 1, and increases as the interest rate is reduced.
When the severity () of the crisis increases, undue transfers to intact banks become less of
a concern, and the optimal interest rate

1() increases to reach 1 when there are no false
positives ( = 1).
42
It must also be the case that the government wants to bail out the banks, that is that
there is not too much adverse selection. Thus if the crisis is not very severe ( < in the
gure), the no-liquidity-hoarding equilibrium disappears.
(ii) Liquidity hoarding. In this equilibrium conguration, banks hoard just enough liq-
uidity (issue just as little short-term debt) as to be able to continue at full scale in case of a
crisis. This requires two conditions.
42
The function

1() is dened by the equation

1()
2
1
0

1()

=
.(1~)i
0
1to
0
.
36
First, given the expected date-1 interest rate, an individual bank must not expect that the
government will make up for insucient liquidity through a targeted transfer. Otherwise it
would hoard no liquidity at all. Put dierently, liquidity hoarding relies on the expectation of
a pure interest rate policy bailout. Whether the government is willing to bail out a bank with
insucient liquidity depends on the trade-o, discussed in (i), between wasteful transfers to
intact banks and rescues of distressed ones but in reverse: when 1 increases, the needed
transfer t = 1 j
0
,1 increases and so bailouts become less attractive. This implies that 1
has to be greater than or equal to 1(), where 1() is an increasing function of as the
government is more tempted to rescue banks as the crisis becomes more severe. It reaches 1
when = .
43

Figure 1: pure and mixed bailouts
Second, because the bailout is a pure interest rate intervention, the equilibrium must be
an equilibrium when no transfers are feasible; that is, it must satisfy equation (3). So the
equilibrium set (the shaded area in Figure 1) is the set of interest rates in Proposition 2 that
satisfy 1 1().
43
A simple inspection of equation (12) shows that this threshold 1() is dened as the solution of the
following equation in 1:
1

u+1

=
~
~+(1~)i
. The threshold 1() is increasing in with 1(0) = j
0
and
1( ) = 1 where [0, 1] is the solution of the following equation in :
~
~+(1~)i
=
1
0
u+1
0
.
37
Once again banks leverage and liquidity choices are strategic complements, and there
can be multiple equilibria. More generally, in this region, the insights gleaned from our
analysis of pure interest rate bailouts in Sections III and IV carry over to optimal bailouts.
There is one dierence in the form of an additional lower bound 1() j
0
on the interest
rate. For these equilibria disappear.
The function

1() is increasing in . For close to ,

1() is below 1() and the
opposite is true for close to 0 provided that

1(0) j
0
. The following assumption ensures
that these two functions cross only once. We denote by the crossing point.
Assumption 11 The function

1() dened by the solution of the equation 1
2
1
0
(1) =
(1 ) ij
0
, [1 : cj
0
] crosses the function 1() once from above on the interval [0. ].
Proposition 9 The symmetric equilibria of the no-commitment economy are as follows:
(i) if then there is an equilibrium with 1 =

1(); it features i, = 1,(1 : cj
0
).
r = 0. and d = :;
(ii) if then there is an equilibrium associated with each xed point of the equation
1 R(1) [1(). 1]; it features i, = 1,[1 : cj
0
+ (1 c)(1 j
0
,1)]. r = 1 j
0
,1,
and d = : (1 j
0
,1).
To sum up, interest rate policy is always used in equilibrium; indeed direct transfers are
not even used in some regions of the parameter space; in the remaining regions, interest rate
policy and direct transfers are used in conjunction. Interest rate policy is a market-driven
solution, in that it benets primarily those institutions with actual borrowing needs. Direct
transfers better focus on strategic actors, but they entail a greater waste of resources by
supporting entities that have no need for, or should not engage in renancing. When direct
transfers are not used (for equilibria of type (ii)), the insights from Sections III and IV carry
over to optimal bailouts: multiple equilibria, macroprudential regulation, and endogenous
macroeconomic uncertainty.
44
44
Marvin Goodfriend and Robert G. King (1988) argue that the central bank should just provide liquidity
anonymoulsy through open market operations and leave it to the market to allocate this liquidity eciently.
They argue against providing liquidity to specic institutions. Our interpretation of interest rate policy is
broader than open market operations, complicating the comparison with their view. Nevertheless, in our
framework, optimal policy under commitment is entirely passive and can therefore be seen as vindicating
their view. Even under limited commitment, the central bank nds it optimal ex post to use only interest
rate policy for equilibria of type (ii). However for equilibria of type (i), a combination of interest rate policy
and direct transfers to (perceived to be) distressed institutions is used. Direct transfers are used only when
38
VII. Conclusion
We have built a simple framework to jointly analyze leverage and maturity mismatch in the
banking sector, bailouts, and optimal regulation. In order to derive a simple yet general
model, we have made a number of simplifying assumptions. Rening our analysis would
require a richer model.
For example, we have argued that one interpretation of interest rate policy in our frame-
work was as a reduced form for conventional monetary policy. We could introduce an explicit
nominal structure with sticky prices and imperfect competition to model conventional mon-
etary policy as in New-Keynesian models. The framework could be further enriched to study
the consequences of bailouts in a setup where interactions between the central banks bal-
ance sheet, ination, and the government budget are non trivial. We could also introduce
the possibility of sovereign default. We have also maintained the assumption of risk neutral-
ity. Introducing risk aversion would allow studying how dierent policy interventions can
impact banks net worth and borrowing costs by aecting risk premia. Finally, in our model,
liquidity is costless in the sense that there are no liquidity premia. If liquidity were costly
instead, interest rate policy under commitment would not necessarily be passive: it could be
optimal for the government to provide liquidity in bad times. However, our insights would
carry over: authorities would provide too much liquidity in the time-consistent outcome.
45
We leave these questions for future research.
the losses from mistaken transfers to intact institutions are smaller than the distortions arising from interest
rate policy. Note that these policies are not Pareto improving ex post since they transfer resources away from
consumers, but they do lead to an increase overall ex post welfare. These policies reduce ex ante welfare.
45
This could be formalized along the lines of Holmstr om-Tirole (1998). Imagine that consumers cannot
commit at t = 0 to provide funds to banks at t = 1. Banks then have to hoard liquidity by purchasing
assets at t = 0 and selling them at t = 1 in the even of a crisis. Banking entrepreneurs are willing to pay a
liquidity premium (in the form of a lower return) on these assets, which can happen if there is a scarcity of
stores of value. Then, the lack of commitment of consumers is a form of market incompleteness which can be
alleviated through government intervention (even under the commitment solution). For example, through a
commitment to a loose interest rate policy in crisis. See Farhi-Tirole (2009) for a discussion.
39
References
[1] Acharya, Viral, and Tanju Yorulmazer. 2007. Too Many to Fail An Analysis of Time-
Inconsistency in Bank Closure Policies. Journal of Financial Intermediation, 16(1):
131.
[2] Acharya, Viral, and Tanju Yorulmazer. 2008. Cash in the Market Pricing and Optimal
Resolution of Bank Failures. Review of Financial Studies, 21(6): 27052742.
[3] Bagehot, Walter. 1873. Lombard Street: a Description of the Money Market. London:
H.S. King.
[4] Bernanke, Ben S., and Mark Gertler. 1995. Inside the Black Box: The Credit Channel
of Monetary Policy Transmission. Journal of Economic Perspectives, 9(4): 2748.
[5] Chari, Varadarajan V., and Patrick J. Kehoe (2010). Bailouts, Time-Inconsistency,
and Optimal Regulation. Mimeo Minneapolis Fed and Princeton.
[6] Cooper, Russell W. 1999. Coordination Games: Complementarities and Macroeco-
nomics. Cambridge: Cambridge University Press.
[7] Cooper, Russell W., and Andrew John. 1988. Coordinating Coordination Failures in
Keynesian Models. Quarterly Journal of Economics, 103(3): 441463.
[8] Diamond Peter A. 1982. Aggregate Demand Management in Search Equilibrium.
Journal of Political Economy, 90(5): 881894.
[9] Diamond, Douglas W., and Philip H. Dybvig. 1983.Bank Runs, Deposit Insurance,
and Liquidity. Journal of Political Economy, 91(3): 40119.
[10] Diamond, Douglas W., and Raghuram Rajan. 2009. Illiquidity and Interest Rate Pol-
icy. Mimeo University of Chicago.
[11] Ennis, Huberto M., and Todd Keister, 2009. Bank Runs and Institutions: The Perils
of Intervention. American Economic Review, 99(4): 15881607.
[12] Ennis, Huberto M., and Todd Keister, 2010. Banking Panics and Policy Responses.
Journal of Monetary Economics, 57(4): 404419.
[13] Farhi, Emmanuel, and Jean Tirole 2009. Collective Moral Hazard, Maturity Mismatch,
and Systemic Bailouts. http://ideas.repec.org/f/pfa207.html#works.
40
[14] Gertler, Mark, and Peter Karadi. 2009. A Model of Unconventional Monetary Policy.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.163.6175&rep=rep1&type=pdf
[15] Gertler, Mark, and Nobuhiro Kiyotaki. 2009. Financial Intermediation and Credit
Policy in Business Cycle Analysis, in preparation for the Handbook of Monetary Eco-
nomics.
[16] Goodfriend, Marvin, and Robert G. King. 1988. Financial deregulation, monetary
policy, and central banking. Federal Reserve Bank of Richmond Economic Review,
74(3): 3-22.
[17] Holmstr om, Bengt and Jean Tirole. 1997. Financial Intermediation, Loanable Funds,
and the Real Sector. Quarterly Journal of Economics, 112(3): 66391.
[18] Holmstr om, Bengt, and Jean Tirole. 1998. Private and Public Supply of Liquidity.
Journal of Political Economy, 106(1): 140.
[19] Holmstr om, Bengt, and Jean Tirole. Forthcoming. Inside and Outside Liquidity. MIT
Press.
[20] Kashyap, Anil K., Raghuram Rajan, and Jeremy Stein. 2008. Rethinking Capital Reg-
ulation. In Maintaining Stability in a Changing Financial System, Federal Reserve
Bank of Kansas City, pp. 431471.
[21] Morris, Stephen, and Hyun Song Shin. 1998. Unique Equilibrium in a Model of Self-
fullling Currency Attacks. American Economic Review, 88(3): 587597.
[22] Ranci`ere, Romain, Aaron Tornell, and Frank Westermann. 2008. Systemic Crises and
Growth. Quarterly Journal of Economics, 123(3): 359406.
[23] Schneider, Martin, and Aaron Tornell. 2004. Balance Sheet Eects, Bailout Guarantees
and Financial Crises. Review of Economic Studies, 71(7): 883913.
41