Unlock document.
This document is partially blurred.
Unlock all pages and 1 million more documents.
Get Access
1.1.1.1.1Chapter Twenty-One
Managing Liquidity Risk on the Balance Sheet
1.1.1.2 I. Chapter Outline
1. Liquidity Risk Management: Chapter Overview
2. Causes of Liquidity Risk
3. Liquidity Risk and Depository Institutions
a. Liability-Side Liquidity Risk
b. Asset-Side Liquidity Risk
c. Measuring a Depository Institution’s Liquidity Exposure
d. Liquidity Risk, Unexpected Deposit Drains, and Bank Runs
e. Bank Runs, the Discount Window, and Deposit Insurance
4. Liquidity Risk and Insurance Companies
a. Life Insurance Companies
b. Property-Casualty Insurance Companies
c. Guarantee Programs for Life and Property-Casualty Insurance Companies
5. Liquidity Risk and Investment Funds
Appendix 21A: Sources and Uses of Funds Statement: Bank of America, June 2013 (available on
Connect or from your McGraw-Hill representative)
Appendix 21B: New Liquidity Risk Measures Implemented by the Bank for International
Settlements (available on Connect or from your McGraw-Hill representative)
II. Learning Goals
1. Identify the causes of liquidity risk.
2. Define the two methods financial institutions use to manage liquidity risk.
3. Describe how depository institutions measure liquidity risk.
4. Examine the components of a liquidity plan.
5. Explain why abnormal deposit drains occur.
6. Consider the extent to which insurance companies are exposed to liquidity risk.
7. Clarify the extent to which investment funds are exposed to liquidity risk.
1.1.1.3 III. Chapter in Perspective
This chapter discusses sources of liquidity risk and how these risks can be managed with both
assets and liabilities. Liquidity risk arises from the need to obtain cash before funds from
maturing assets are available. Sources of funds are decreases in an asset or increases in a
liability or equity account. Liquidity can thus be ‘stored’ by holding cash and near cash assets
(sometimes called primary and secondary reserves) or liquidity can be obtained by borrowing
additional funds as needed. Measuring prior period expected and unexpected liquidity needs can
help FI managers plan for future expected and unexpected liquidity requirements. All DIs
operate on a fractional reserve system where they retain only a small portion of deposits and
other borrowings in the form of liquid assets. Each institution is dependent upon the public’s
belief in the soundness and safety of the individual institution and the financial system. A
perceived erosion of the safety of deposits can quickly generate bank runs and liquidity crises
although federal deposit insurance and the Fed’s role as lender of last resort limit the likelihood
of banks runs in the U.S. Deposit insurance in particular has largely eliminated bank runs by the
general public, but DI liquidity crises still occur and are a normal part of market discipline.
Insurance companies and mutual funds normally face lower amounts of liquidity risk than DIs,
but liquidity problems can still occur at these institutions.
1.1.1.4
1.1.1.5 IV. Key Concepts and Definitions to Communicate to Students
Fire sale prices Peer group ratios
Core deposits Liquidity index
Net deposit drains Financing Gap
Financing Requirement Full pay vs prorated claims
Purchased liquidity Liquidity Plan
Stored liquidity Bank runs and bank panics
Excess reserves Surrenders and surrender value
Net liquidity statement Insurance guaranty funds
Net stable funding ratio Contagion
Credit Crunch Subprime Crisis
Primary Dealer Credit Facility Liquidity Coverage Ratio
1.1.1.6 V. Teaching Notes
1. Liquidity Risk Management: Chapter Overview
All FI managers must deal with liquidity planning and liquidity risk on a daily basis, although
DIs have substantially more liquidity risk than other types of FIs. The main goal of liquidity
management is to maintain ‘just enough’ liquid assets in combination with liability funding
sources to be able to meet expected and unexpected liquidity needs. FIs do not wish to hold
excessive amounts of liquid assets because they earn low rates of return. Banks and DIs
generally have more liquidity risk than insurers, mutual funds and hedge funds. Nevertheless
several hedge funds have gone bankrupt recently. Hedge funds and securities brokers pledge
their security holdings for collateral on short term loans used to provide liquidity. When the
subprime problems reduced the value of mortgage backed securities lenders to these funds and
dealers refused to renew loans without better collateral. Two Bear Stearns hedge funds collapsed
as a result, eventually bringing Bear down with them. As the credit problems spread throughout
the economy liquidity problems emerged as well. Interbank offering rates such as LIBOR soared
from 2.57% in September 2008 to 6.88% on September 30, 2009. Without reasonable cost
funding banks curtailed lending to non-bank customers further exacerbating liquidity problems
in other markets. Central banks around the world pumped liquidity into markets to limit the
crisis.
Teaching Tip:
There is an old saying that I heard from a fund manager several years ago, “Liquidity doesn’t
matter until it matters, and then it is the only thing that matters.” Liquidity is a necessary
condition for well functioning markets and is a necessary component for successful hedging of
risk. Since virtually all hedging models assume adequate liquidity, when liquidity dries up all
models of risk fail and outcomes can be much more extreme than anticipated. This is a lesson
that investors who rely on math based modeling to assess risk must learn.
2. Causes of Liquidity Risk
Unexpected withdrawals of liabilities
Unexpected withdrawals of deposits or unanticipated policy claims can force FIs to sell
assets or borrow more funds. If the FI does not have enough liquid assets to sell, or cannot
borrow enough additional funds at short notice they may have to liquidate longer term
investments, perhaps at prices below market value (at so called ‘fire-sale’ prices). If the
liquidated assets must be marked down to market, balance sheet losses occur and equity write
downs would result.
For example, a bank faces net deposit withdrawals of $30 million of uninsured deposits
as word hits the street that the bank faces large loan losses from a regional collapse in real
estate values. 1 The bank liquidates $15 million in liquid assets at fair market value,
borrows an additional $10 million in short term debt markets, and liquidates longer term
investments at below book value, and even below fair market value because it needs the
money now. The book value of the long term investments is $7 million but the bank
obtains only $5 million net of transaction costs. The bank must bear a $2 million loss due
to its illiquidity.
Unexpected increases in assets
Unexpected drawdowns on credit lines and unanticipated loan demand are two sources of
asset side liquidity risk. Unanticipated defaults on loans can also generate additional cash
needs, as can unexpected payments on contingent items such as bankers’ acceptances and
financial standby letters of credit.
3. Liquidity Risk and Depository Institutions (DIs)
a. Liability-Side Liquidity Risk
DIs have large amounts of transaction and savings deposits that customers can make due
immediately if they choose. These accounts give depositors a put option with the exercise price
equal to the amount of their deposit. Banks estimate the amount of core deposits that are usually
relatively stable on a day to day basis and estimate expected growth in deposits. Core deposits
are low turnover accounts that are at the bank for reasons other than the interest rate earned.2
They may be placed at the bank for convenience needs, or because the customer has some other
1Amounts over the $250,000 insurance limit are uninsured.
2Core deposits typically include all consumer accounts, some business accounts and retail CDs.
relationships with the institution. Net deposit withdrawals are called net deposit drains.
Although net deposit drains usually have a seasonal component, increasing at Christmas and
vacation time for example, they are usually quite predictable on a daily basis, particularly if a FI
has a substantial core deposit component.
1.1.1.7 Purchased liquidity
Banks can obtain funds by borrowing additional cash in the money markets. This practice is
termed ‘purchasing liquidity’ or sometimes ‘liability management.’ Purchased liquidity
sources were harder to obtain during the financial crisis. It is riskier for banks to overly depend
on purchased or wholesales funds sources to provide liquidity.
Teaching Tip: The practice of purchasing liquidity is fairly recent. It began in the 1960s with the
advent of a secondary market for negotiable CDs and it has been spurred on by the growth in the
fed funds market. Purchasing liquidity can be expensive and can increase the interest rate
sensitivity of a bank’s liabilities because the bank adds interest rate sensitive funds to meet
liquidity needs, thus reducing the proportion of funding from core deposits. The tradeoff is that
if a bank is willing to rely on purchased liquidity sources, it need not hold as many low earning
liquid assets. More funds can then be placed in riskier investments and loans that promise higher
rates of return. Purchased liquidity allows a bank to maintain a given size and distribution asset
portfolio while still allowing the institution to obtain the cash needed to fund withdrawals or
additional loan demand.
1.1.1.8 Stored Liquidity
Liquidity can be stored by investing in cash and/or liquid securities that earn a rate of return.
Primary reserves are vault cash, CIPC, correspondent balances and deposits at the Federal
Reserve. Recall that the Fed imposes minimum liquidity requirements on DIs (basically 10% on
transaction deposits), but banks generally hold substantial excess reserves (reserves beyond the
Fed requirements) that can be used for liquidity purposes.
Banks normally utilize both purchased and stored liquidity. The costs of each can be easily
illustrated via an example:
NorthView Bank (NVB)
Assets Liabilities and Equity
Amount
(mill$)
Rate of
Return
Amount
(mill$)
Cost
Rate
Cash $ 20 0% Deposits $ 560 4%
Securities 230 7% Borrowings 160 6%
Loans 550 10% Equity 80
Total $800 Total $800
NVB is expecting a $35 million deposit drain and only $5 million in cash is available for
liquidation since required reserves are $15 million. NVB faces the choice of purchasing liquidity
by borrowing or by liquidating cash and securities. Let’s examine the costs of each alternative:3
1. Borrow $35 million to replace lost deposits: Deposit cost is 4% and borrowing cost is assumed
3 All examples ignore changes in required reserves resulting from the change in deposits.
to remain at 6% so the pre–tax change in net income from the deposit drain is 2% *
$35,000,000 = –$700,000. The advantage of borrowing is that no part of the asset portfolio
has to be liquidated.
2. a) Pay off depositors with $5 million in cash excess reserves and liquidate $30 million in
securities on which the bank is earning 7%. The change in pre–tax net income in this case is
($35,000,000 * 0.04) – ($30,000,000 * 0.07) = –$700,000. In this case the costs of
alternatives 1 and 2 are identical, but alternative 2 decreases the bank size by $35 million and
decreases the amount of leverage. The drop in size may be a concern if the bank loses
economies of scale.
2. b) Pay off depositors with $5 million in cash excess reserves and liquidate $30 million in
securities on which the bank is earning 7%. This alternative is the same as 2. a), but in this
case suppose the securities liquidity index is 97% (the liquidity index is described below).
This implies that the bank can only receive 97 cents per dollar of fair market value on the
securities liquidated because they must be liquidated rapidly. The bank has to liquidate $30
million / 0.97 = $30,927,835 in securities to raise $30 million. This results in an additional
loss of $927,835. The change in pre–tax net income in this case is ($35,000,000 * 0.04) –
($30,927,835 * 0.07) – $927,835 =
–$1,692,783. The loss represented by the sale below fair market value reduces equity as well.
b. Asset-Side Liquidity Risk
Exercise of loan commitments by borrowers can also generate liquidity needs. Loan
commitments at banks grew tremendously in the 2000s. An unused loan commitment provides
fee income to the bank. The ratio of unused loan commitments to cash was about 529% in 1994,
and rose to 1014.6% in October 2008. The crisis led to a decline to about 609%. This can be
dangerous if the bank has not planned properly because net unexpected asset increases lead to
immediate funding requirements. As before the FI can choose to meet the need by purchasing
liquidity (and allowing the bank’s assets to grow) or by using stored liquidity (maintaining the
same amount of assets). Text Tables 21-6 and 21-7 illustrate two possible adjustments to a $5
million exercise of a loan commitment. Table 21-6 illustrates the immediate effect of the loan
exercise and Table 21-7 illustrates two possible adjustments, first, the bank could borrow an
additional $5 million (Purchased Liquidity Management) or, second, the bank could instead
reduce cash assets by $5 million. The instructor may wish to encourage a student discussion of
the pros and cons of each alternative. The second require holding low earning cash assets but is
safer, while the first may increase interest expense, expense volatility and may be riskier in stress
scenarios where purchased funds may be more expensive or not available.
c. Measuring a Depository Institution’s Liquidity Exposure
Tools to measure liquidity exposure include the following (two more methods are presented in
Appendix 21B):
Net liquidity statement
A net liquidity statement is a report of net available liquid sources of funds. For example:
Net Liquidity Position (millions $)
Sources
1. Total near cash assets $ 5,000
2. Excess cash reserves $ 2,000
3. Maximum new borrowings $ 9,000
Total $16,000
Uses
1. Funds already borrowed $ 8,000
2. Discount Window loans that
must be repaid quickly $ 1,000
Total $ 9,000
1.1.1.8.1 Total Net Liquidity
$ 7,000
The FI can handle unanticipated liquidity needs of $7,000 millions.
Teaching Tip: The FI management must decide if the amount of liquidity coverage ($7,000.) is
reasonable in light of the likely amount of net deposit drains. Examining the historical
distribution of drains adjusted for any seasonality can help the FI ascertain the likely amount of
drains. The FI does not want to hold excessive amounts of liquid assets because their low return
is a drag on profitability and competitiveness.
Peer group ratios
Banks will often monitor key liquidity ratios such as
March 20084March 2011 Dec 2013
Loans to deposits 81.33% 71.66% 71.77%
Loans to core deposits 102.84% 78.64% 77.55%
Short Term Non-Core Funding to
Assets
17.08% 5.76% 4.94%
Core deposits to total liabilities &
equity
65.24% 77.75% 78.81%
Commitments to lend to assets 8.57%
Liquidity Index
The liquidity index is the ratio of the fire sale price required to liquidate assets in an
emergency situation divided by the fair market value of the assets liquidated. The lower
the index the greater the liquidity risk. For instance, suppose a securities portfolio
contains two securities with the following data:
Securities
Value if liquidated
immediately
Fair market value
if liquidated in 1
month
% invested in
each (at FMV)
Treasury Bills $ 9,700,000 $ 9,850,000 38.58%
Bonds $15,000,000 $15,675,000 61.42%
The liquidity index is calculated as:
4Source: FFIEC; all banks in the nation, Peer Group Average Distribution report, report date s
3/31/11, 3/31/08 and 12/31/13.
[38.58% * ($9.7 mill / $9.85 mill)] + [61.42%*($15 mill / $15.675 mill)] = 96.76%
Teaching Tip: Discount instruments increase in price as they approach maturity but non-discount
instruments receive interest income. The liquidity index should measure not only any loss in fair
market value, but also any loss in income due to a required change in FI behavior. For example,
a T-bill may be priced at fair market value at 99% of par prior to maturity. Nevertheless, if the FI
planned to hold the bill until maturity but had to sell it to meet liquidity needs, the required sale
at the fair market value of $99 per $100 of par still represents a loss due to liquidity risk. Thus,
the index should account for lost interest as well as losses in current fair market value.
Teaching Tip: The index is a better measure of the cost of liquidity risk than the likelihood of
occurrence of liquidity problems.
Financing Gap and the Financing Requirement
(Uses) (Sources)
Financing Gap = Average loans – Average (core) deposits
If the financing gap is positive, (as it is for the typical bank) the DI must obtain additional
financing either by borrowing or liquidating assets.
The Financing Requirement is the amount of funds that must be borrowed and it is found as:
Financing Requirement = Financing Gap + Required liquid asset holdings5
An increasing financing requirement may indicate future liquidity problems for a bank since
this indicates greater borrowing requirements for the DI.
New Liquidity Risk Measures Implemented by the Bank for International
Settlements (BIS)
The BIS has created two new liquidity requirements, the liquidity coverage ratio (LCR)
and the Net Stable Funding Ratio (NSFR). The LCR is the ratio of the stock of high
quality assets that can be liquidated at short notice to the total net cash outflow over the
next 30 days. This ratio must be ≥ 100%, but the requirement is being phased in from
2015 to 2019. The total net cash outflow in the denominator is estimated under an acute
stress scenario that includes institutional and systemic shocks as developed by the
regulators. The NSFR must be reported quarterly beginning in 2018. This ratio is
amount of available stable funding over 1 year divided by the required amount of stable
funding over the year. The NSFR ratio must be > 100% and it is meant to limit the
reliance on short term funding for longer term assets. In addition as of 2013 the BIS is
requiring internationally active banks to more robustly measure and understand their
intraday liquidity requirements.
5The textbook does not indicate that the assets must be required although this is implied in a
footnote. Logically the bank could liquidate its liquid assets and reduce its financing
requirement. Note that in this formulation these numbers are not flows, they are balance sheet
levels. Because these are levels, there is also an implicit assumption that the level of non-earning
assets equals the amount of equity.
Liquidity Plan
A liquidity plan should include the following key components:
oManagerial guidelines and assignment of responsibilities
oList of fund providers ranked by likelihood of withdrawal (Institutional and corporate
investors are more likely to withdraw funds quickly.)
oEstimation of seasonal components of liquidity (Christmas, planting time, harvest
time, vacation season, etc.)
oEstimation of amounts of withdrawals over specified time intervals.
oInternal limits on subsidiary and branch borrowings from parents and maximum
borrowing rates.
oPlanned order of disposition of assets in the event liquidations become necessary.
An example liquidity plan may look like the following:
Potential Deposit Withdrawals and Associated Required Asset Liquidations (Mill $)
Potential Deposit
Withdrawals
From most likely to withdraw to least
likely
Mutual Funds $ 70
Pension Funds $ 40
Correspondent banks $ 50
Large corporations $ 45
Small businesses $ 25
Consumers $ 75
Total $305
1.1.1.9 Expected total withdrawals per
period
Average Maximum Likely
1.1.1.10 One week $ 60 $100
One month $ 70 $150
Three months $130 $220
Total $260 $470
Sequence of funding
options as needed One Week One month Three month
New deposits $ 15 $ 35 $ 75
Sale liquid assets $ 15 $ 25 $ 55
Sale investment portfolio $ 30 $ 40 $ 50
Borrowings from other FIs $ 30 $ 40 $ 35
Borrowings from Fed $ 10 $ 10 $ 5
Total $100 $150 $220
In the event the maximum likely withdrawals actually occur, the bank has already determined
how the withdrawals will be funded in the bottom panel. The numbers in the bottom panel are
developed in conjunction with the necessary strategies that can be used if needed to bring about
the increases shown. For instance, in the one week period, deposit rates may have to be
increased 15 basis points to attract $15 million in new deposits.
d. Liquidity Risk, Unexpected Deposit Drains and Bank Runs
Abnormal deposit drains can threaten a FI’s solvency. These usually arise due to problems in the
management of some other area of risk such as credit or interest rate risk.
Demand and other deposits are first-come, first-served contracts that are full pay or no pay
contracts. They are not pro-rata claims that are apportioned based on a fair distribution of the
liquidation value of the DI’s assets. Hence, there is always a possibility of a bank run when
banks maintain only partial reserves to back deposits because only the first depositors to demand
their money receive anything. A bank run occurs when the fundamental assumption underlying
fractional reserve banking is violated; namely, that all depositors do not wish to obtain their
money at the same time. Since all deposits in all institutions are this way, failure, or fear of
failure, at one or more institutions can quickly spread (the dreaded contagion effect) potentially
causing widespread bank panics or system wide runs on banks. Contagion effects are
particularly serious in countries or situations where there is no credible deposit insurance.6
In 2008 IndyMac faced a bank run after Senator Schumer’s letters warning of problems at the
bank became public. Over the 11 days following the public release of his letter depositors
withdrew over $1.3 billion from IndyMac. Schumer was right; the bank was in trouble due to its
mortgage holdings. This is another case where problems in credit spilled over into liquidity
problems when investors lost confidence in the bank’s ability to meet its obligations.
e. Bank Runs, the Discount Window and Deposit Insurance
The two major stabilizing factors that limit bank runs are the discount window and deposit
insurance.
Deposit Insurance
In the U.S. deposits are currently insured up to $250,000 per account. The amount was
increased from $100,000 during the financial crisis. Actually unlimited insurance was
temporarily provided during the crisis. When an institution is deemed too big to fail and a
bailout or buyout is arranged then all depositors receive defacto 100% insurance,
regardless of the size of their deposits. This removes a market discipline requiring large
depositors to evaluate the riskiness of large institutions.
When deposit insurance was established in 1933, bank runs were virtually eliminated at
federally insured institutions. State insurance is not sufficient to prevent widespread bank
runs because the insurance funds do not have enough reserves to maintain public
confidence in a crisis. The FDIC now assesses risk based deposit insurance premiums.
Capital adequacy and supervisory judgment are used to assign DIs to risk categories. DIs
have to pay more to maintain deposit liquidity when they take on more risk.
6A bank run can still occur even if there is credible deposit insurance if inflation is high enough
or if depositors fear upcoming restrictions on repatriation. Panics may occur because the value
of money is its purchasing power. Any threat to the purchasing power of the money could
conceivably cause a run. Moreover there can be payment delays in the event of bank failure and
concerned depositors (insured or not) may withdraw their funds as a result.
The Discount Window
The Fed provides short term emergency lending to qualifying DIs. The Fed has shown a
willingness to open the discount window during risky events such as the 2001 terrorist
attacks, during several stock market crashes and most recently the subprime crisis. In the
2001 attacks on the World Trade Center phone and computer outages, grounding of plans
that carried checks and building evacuations led to many disruptions of payment systems.
These problems led to unexpected shortages at other institutions expecting to be paid by
New York banks. On September 11th, the Fed announced the window was open and
encouraged all FIs to borrow as needed to cover unexpected shortfalls. It was
particularly important that the Fed offered discount window services to banks and
securities dealers who finance their substantial securities inventory with short term call
loans. If banks had called in large numbers of these loans some of the major investment
banks could have been in danger of severe liquidity crises forcing them to liquidate their
securities inventories and causing sharp declines in asset prices.
Typically DIs must pledge short term, high quality assets as collateral that are
‘discounted,’ hence the term discount window loans. The discount rate used to be kept
below open market rates and at that time the Fed actively discouraged use of the discount
window except as an emergency source of short term borrowing. The Fed has now
changed the discount window policy. See Chapter 4 for details but basically the Fed
operates three types of loan programs. The first is termed primary credit. Primary
credit is available to sound institutions on a short term basis at a rate 100 basis points
above the FOMC target fed funds rate. Primary credit loans may be used for any purpose
and loan terms can be as long as several weeks. Secondary credit is available for
overnight loans to sound institutions that are having temporary funding problems at a rate
150 basis points above the FOMC target fed funds rate. Secondary credit may not be
used to finance institutional growth. Finally, seasonal credit is available on a longer
term basis at a rate below the target FOMC fed funds rate. The borrower must
demonstrate seasonality.
In response to the liquidity problems caused by the credit crunch in 2007 and 2008 the
Fed announced in March 2008 that it would lend up to $200 billion to both commercial
and investment banks through its new Primary Dealer Credit Facility (PDCF). Under the
PDCF, firms borrowed an average of $31.3 billion per day from the Fed in the first three
operating days of the facility. The borrowers could swap mortgage backed securities for
Treasuries. The borrower could swap some securities that the Fed would not ordinarily
have accepted. The Fed took this extraordinary step because many institutions were
unable to borrow against mortgage securities, creating a liquidity crunch. Not all agree
that this was a sound move by the Fed. Some analysts believe the bailout of Bear Stearns
and the intervention into the markets will create or exacerbate the moral hazard problem
over the long run and encourage other institutions to take excessive risks believing that
the Fed will come to their rescue if needed. New borrowing programs emerged over the
succeeding months providing funding to money market mutual funds, commercial paper,
insurance companies and others. The Fed also lowered interest rates to near zero and
reduced the spread between the discount rate and the Fed funds rate.
Teaching Tip: One of the original functions of the Fed was to serve as a lender of last
resort to DIs. If the Fed was willing to supply unlimited amounts of loans to a DI facing
insolvency, there would theoretically be no need for deposit insurance to prevent bank
runs. The Fed could create whatever money was needed to prevent a DI from becoming
insolvent and the public would have no reason to withdraw their deposits. The conditions
the Fed imposes on discount window loans limit its effectiveness as a deterrent to bank
runs. Indeed the Fed was around during the Crash of 1929 and it was either unable or
unwilling to prevent the widespread bank runs that led to the failure of thousands of
banks at that time. Several aspects of normal Fed policy limited the usefulness of the
‘lender of last resort’ in preventing bank runs. These include:
a) The requirement to pledge high quality assets to back the loan eliminates the ability of
most failing institutions to obtain a sufficient amount of discount window loans. The Fed
has weakened this requirement due to the crisis however.
b) The Fed does not automatically grant discount window loans for extended periods, so
depositors cannot count on this method as a sufficient means of financing to ensure that
the value of all deposits will be preserved even with the Fed’s new policies.
c) The purpose of the discount window is to provide short term financing to solvent
institutions not to keep afloat failing institutions. Indeed, loans to troubled,
undercapitalized institutions are specifically limited to no more than 60 days in any 120
day period unless the FDIC and any other primary regulator certify that the bank is viable.
The discount window is designed to limit bank’s need to liquidate assets at fire sale prices
in order to fund required liquidity needs, not to protect depositors.
The Fed evidenced a willingness to go beyond the normal functions of the Discount Window
during and after the financial crisis.
