Debt Limit
Delays Create Debt Management Challenges and Increase Uncertainty in the Treasury Market
Gao ID: GAO-11-203 February 22, 2011
GAO has prepared this report to assist Congress in identifying and addressing debt management challenges. Since 1995, the statutory debt limit has been increased 12 times to its current level of $14.294 trillion. The Department of the Treasury (Treasury) recently notified Congress that the current debt limit could be reached as early as April 5, 2011, and the Congressional Budget Office (CBO) projects that under current law debt subject to the limit will exceed $25 trillion in 2021. This report (1) describes the actions that Treasury traditionally takes to manage debt near the limit, (2) analyzes the effects that approaching the debt limit has had on the market for Treasury securities, and (3) describes alternative mechanisms that would permit consideration of the link between policy decisions and the effect on debt when or before decisions are made. GAO analyzed Treasury and market data; interviewed Treasury officials, budget and legislative experts, and market participants; and reviewed practices in selected countries.
The debt limit does not control or limit the ability of the federal government to run deficits or incur obligations. Rather, it is a limit on the ability to pay obligations already incurred. While debates surrounding the debt limit may raise awareness about the federal government's current debt trajectory and may also provide Congress with an opportunity to debate the fiscal policy decisions driving that trajectory, the ability to have an immediate effect on debt levels is limited. This is because the debt reflects previously enacted tax and spending policies. Delays in raising the debt limit create debt and cash management challenges for the Treasury, and these challenges have been exacerbated in recent years by a large growth in debt. In the past, Treasury has often used extraordinary actions, such as suspending investments or temporarily disinvesting securities held in federal employee retirement funds, to remain under the statutory limit. However, the extraordinary actions available to the Treasury have not kept pace with the growth in borrowing needs. For example, unlike the past, the amount potentially provided by the extraordinary actions for 1 month in fiscal year 2010 was less than the monthly increase in debt subject to the limit for most months of the year. As a result, once debt reaches the limit, Congress will likely have less time than in prior years to debate raising the debt limit before there are disruptions to government programs and services. This trend is likely to continue given the long-term fiscal outlook. Failure to raise the debt limit in a timely manner could have serious negative consequences for the Treasury market and increase borrowing costs. Also, some of the actions that Treasury has taken to manage the amount of debt near the limit add uncertainty to the Treasury market. In the past, Treasury has postponed auctions and dramatically reduced the amount of bills outstanding, which compromised the regularity of auctions and the certainty of supply on which Treasury relies to achieve the lowest borrowing cost over time. GAO's analysis suggests that borrowing costs modestly increased during debt limit debates in 2002, 2003, and most recently in 2010. In addition, managing debt near the debt limit diverts Treasury's limited resources away from other cash and debt management issues at a time when Treasury already faces challenges in lengthening the average maturity of its debt portfolio. Observers and participants suggested improving the link between the spending and revenue decisions that drive debt and changes in the debt limit. Better alignment could be possible if decisions about the debt level occur in conjunction with spending and revenue decisions as opposed to the after-the-fact approach now used. This practice, which is similar to practices used in some other countries, might facilitate efforts to change the fiscal path by highlighting the implications of tax and spending decisions on changes in debt. To avoid potential disruptions to Treasury markets and help inform fiscal policy decisions in a timely way, Congress should consider ways to better link decisions about the debt limit with decisions about spending and revenue. Treasury provided technical comments on a draft of this report, which GAO incorporated as appropriate.
Recommendations
Our recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.
Director:
Susan J. Irving
Team:
Government Accountability Office: Strategic Issues
Phone:
(202) 512-9142
GAO-11-203, Debt Limit: Delays Create Debt Management Challenges and Increase Uncertainty in the Treasury Market
This is the accessible text file for GAO report number GAO-11-203
entitled 'Debt Limit: Delays Create Debt Management Challenges and
Increase Uncertainty in the Treasury Market' which was released on
February 22, 2011.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as
part of a longer term project to improve GAO products' accessibility.
Every attempt has been made to maintain the structural and data
integrity of the original printed product. Accessibility features,
such as text descriptions of tables, consecutively numbered footnotes
placed at the end of the file, and the text of agency comment letters,
are provided but may not exactly duplicate the presentation or format
of the printed version. The portable document format (PDF) file is an
exact electronic replica of the printed version. We welcome your
feedback. Please E-mail your comments regarding the contents or
accessibility features of this document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
United States Government Accountability Office:
GAO:
Report to the Congress:
February 2011:
Debt Limit:
Delays Create Debt Management Challenges and Increase Uncertainty in
the Treasury Market:
GAO-11-203:
GAO Highlights:
Highlights of GAO-11-203, a report to the Congress.
Why GAO Did This Study:
GAO has prepared this report to assist Congress in identifying and
addressing debt management challenges. Since 1995, the statutory debt
limit has been increased 12 times to its current level of $14.294
trillion. The Department of the Treasury (Treasury) recently notified
Congress that the current debt limit could be reached as early as
April 5, 2011, and the Congressional Budget Office (CBO) projects that
under current law debt subject to the limit will exceed $25 trillion
in 2021.
This report (1) describes the actions that Treasury traditionally
takes to manage debt near the limit, (2) analyzes the effects that
approaching the debt limit has had on the market for Treasury
securities, and (3) describes alternative mechanisms that would permit
consideration of the link between policy decisions and the effect on
debt when or before decisions are made. GAO analyzed Treasury and
market data; interviewed Treasury officials, budget and legislative
experts, and market participants; and reviewed practices in selected
countries.
What GAO Found:
The debt limit does not control or limit the ability of the federal
government to run deficits or incur obligations. Rather, it is a limit
on the ability to pay obligations already incurred.
While debates surrounding the debt limit may raise awareness about the
federal government‘s current debt trajectory and may also provide
Congress with an opportunity to debate the fiscal policy decisions
driving that trajectory, the ability to have an immediate effect on
debt levels is limited. This is because the debt reflects previously
enacted tax and spending policies.
Delays in raising the debt limit create debt and cash management
challenges for the Treasury, and these challenges have been
exacerbated in recent years by a large growth in debt. In the past,
Treasury has often used extraordinary actions, such as suspending
investments or temporarily disinvesting securities held in federal
employee retirement funds, to remain under the statutory limit.
However, the extraordinary actions available to the Treasury have not
kept pace with the growth in borrowing needs. For example, unlike the
past, the amount potentially provided by the extraordinary actions for
1 month in fiscal year 2010 was less than the monthly increase in debt
subject to the limit for most months of the year. As a result, once
debt reaches the limit, Congress will likely have less time than in
prior years to debate raising the debt limit before there are
disruptions to government programs and services. This trend is likely
to continue given the long-term fiscal outlook.
Failure to raise the debt limit in a timely manner could have serious
negative consequences for the Treasury market and increase borrowing
costs. Also, some of the actions that Treasury has taken to manage the
amount of debt near the limit add uncertainty to the Treasury market.
In the past, Treasury has postponed auctions and dramatically reduced
the amount of bills outstanding, which compromised the regularity of
auctions and the certainty of supply on which Treasury relies to
achieve the lowest borrowing cost over time. GAO‘s analysis suggests
that borrowing costs modestly increased during debt limit debates in
2002, 2003, and most recently in 2010.
In addition, managing debt near the debt limit diverts Treasury‘s
limited resources away from other cash and debt management issues at a
time when Treasury already faces challenges in lengthening the average
maturity of its debt portfolio.
Observers and participants suggested improving the link between the
spending and revenue decisions that drive debt and changes in the debt
limit. Better alignment could be possible if decisions about the debt
level occur in conjunction with spending and revenue decisions as
opposed to the after-the-fact approach now used. This practice, which
is similar to practices used in some other countries, might facilitate
efforts to change the fiscal path by highlighting the implications of
tax and spending decisions on changes in debt.
What GAO Recommends:
To avoid potential disruptions to Treasury markets and help inform
fiscal policy decisions in a timely way, Congress should consider ways
to better link decisions about the debt limit with decisions about
spending and revenue.
Treasury provided technical comments on a draft of this report, which
GAO incorporated as appropriate.
View [hyperlink, http://www.gao.gov/products/GAO-11-203] or key
components. For more information, contact Susan J. Irving at (202) 512-
6806 or irvings@gao.gov or Gary T. Engel at (202) 512-3406 or
engelg@gao.gov.
[End of section]
Contents:
Letter:
Background:
Increased Borrowing and Limited Borrowing Capacity Provided by
Extraordinary Actions Create Debt Management Challenges:
Approaching the Debt Limit Can Add Uncertainty in the Treasury Market:
Experts and Practices of Other Countries Offer Insights for Better
Linking Policy Decisions with Their Effect on Debt:
Conclusions:
Matter for Congressional Consideration:
Agency Comments:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Detailed Methodology and Findings of Statistical Analysis
of Treasury Borrowing Costs near the Debt Limit:
Appendix III: GAO Contacts and Staff Acknowledgments:
Tables:
Table 1: Extraordinary Actions Taken by Treasury to Manage Debt near
the Debt Limit:
Table 2: Estimated Borrowing Capacity Provided by Extraordinary
Actions:
Table 3: Average Operating Cash Balance Less Supplementary Financing
Program Account Balance, Fiscal Years 2003-2010:
Table 4: Postponed Auctions and Delayed Auction Announcements, 1995-
2010:
Table 5: Variables Used in Multivariate Regression:
Table 6: Regression Results for Debt Limit Event Periods from Fiscal
Year 2002 to 2010:
Figures:
Figure 1: Extraordinary Actions Taken by Treasury Prior to Debt Limit
Increases, 1995-2010:
Figure 2: Estimated Borrowing Capacity Provided by Extraordinary
Actions Based on a 1-Month DISP Relative to Actual Monthly Changes in
Debt Subject to the Limit for Select Years:
Figure 3: Daily Change in Net Cash Flows during Recent Debt Limit
Event Period, August 7, 2009, to February 12, 2010:
Figure 4: Supply of Treasury Bills Decreased Sharply during Debt Limit
Event:
Figure 5: Spreads between Yields of 3-Month Commercial Paper and 3-
Month Treasury Bills Were Lower during the 2009-2010 Debt Limit Event
Period:
Figure 6: Trends in Treasury's CDS Spreads Were Consistent with Trends
in the Yield Spread for the 2009-2010 Debt Limit Event:
Abbreviations:
BPD: Bureau of the Public Debt:
CBO: Congressional Budget Office:
CDS: credit default swap:
CPFF: Commercial Paper Funding Facility:
CM bill: cash management bill:
CSRDF: Civil Service Retirement and Disability Fund:
DISP: debt issuance suspension period:
ESF: Exchange Stabilization Fund:
FFB: Federal Financing Bank:
G-Fund: Government Securities Investment Fund of the Federal
Employees' Retirement System:
GDP: gross domestic product:
IMF: International Monetary Fund:
LIBOR: London interbank offer rate:
Moody's: Moody's Investors Service:
ODM: Office of Debt Management:
OECD: Organisation for Economic Co-operation and Development:
OFP: Office of Fiscal Projections:
Recovery Act: American Recovery and Reinvestment Act of 2009:
Secretary: Secretary of the Treasury:
SFP: Supplementary Financing Program:
SLGS: State and Local Government Series:
TARP: Troubled Asset Relief Program:
TIPS: Treasury Inflation-Protected Securities:
Treasury: Department of the Treasury:
VIX: Chicago Board Options Exchange's Volatility Index:
[End of section]
United States Government Accountability Office:
Washington, DC 20548:
February 22, 2011:
Report to the Congress:
Congress and the President have enacted laws to establish a limit on
the amount of federal debt that can be outstanding at one time. The
debt limit does not restrict Congress' ability to enact spending and
revenue legislation that affect the level of debt or otherwise
constrain fiscal policy; it restricts the Department of the Treasury's
(Treasury) authority to borrow to finance the decisions enacted by the
Congress and the President. As a result, as the government nears the
debt limit, Treasury often must deviate from its normal cash and debt
management operations and take a number of extraordinary actions such
as temporarily disinvesting securities held as part of federal
employees' retirement plans to meet the government's obligations as
they come due without exceeding the debt limit. Once the extraordinary
actions are exhausted, Treasury is not authorized to issue new debt
and could be forced to delay payments for government services or
operations until funding is available and could eventually be forced
to default on legal debt obligations.
Since 1995, the statutory debt limit has been increased 12 times to
its current level of $14.294 trillion. Treasury recently notified
Congress that the current debt limit could be reached as early as
April 5, 2011, and the Congressional Budget Office (CBO) projects
that, if current laws remain in place, debt subject to the limit will
exceed $25 trillion in 2021. Meanwhile, GAO's long-term simulations
show that absent policy changes, federal debt will increase
continually over the next several decades. The medium-and long-term
outlook for the federal budget makes an understanding of the
operations and implications of the debt limit important.
GAO has prepared this report under the Comptroller General's authority
to conduct evaluations on his own initiative as part of a continuing
effort to assist Congress in identifying and addressing debt
management challenges. This report examines the challenges associated
with managing cash and debt near the limit. Specifically, the
objectives of this report are to (1) describe the actions that
Treasury has taken to manage debt near the limit and challenges that
arise, (2) analyze the effects that approaching the debt limit has on
the market for Treasury securities, including Treasury's borrowing
costs, and (3) in light of the disconnect between the debt limit and
the policy decisions that have an effect on the size of federal debt,
describe alternative triggers or mechanisms that would permit
consideration of the link between policy decisions and their effect on
debt when or before decisions are made.
To answer our first objective, we analyzed publicly available data on
Treasury cash and debt transactions from the last 16 years (1995-2010)
to identify factors that could create challenges for Treasury in
managing debt near or at the limit. We reviewed documents provided by
Treasury, interviewed Treasury officials, and obtained estimates of
the time and staff involved in planning for when the debt limit will
be reached and related operations. We conducted a check for
reasonableness of these estimates.
To determine what effect approaching the debt limit had on the
Treasury market, we analyzed changes in the size and timing of
Treasury auctions when debt was near or at the limit. Our review
generally covered the last 16 years--or as many of those years for
which data were readily available--in order to include both a
particularly disruptive debt limit debate in 1995-1996 that required
Treasury to take a number of extraordinary actions and the most recent
debt limit increase. We interviewed several market participants and
observers, including primary dealers, money market fund managers, and
credit rating agencies, to obtain their views on how the debt limit
and the actions Treasury takes to manage the amount of debt when it is
near the debt limit affect the Treasury market. We analyzed changes in
the yields for Treasury securities before, during, and after five of
the debt limit debates in the past 10 years, including the most recent
in 2009-2010, to determine how proximity to the debt limit affected
Treasury's borrowing costs. See appendix II for more details on how we
estimated increased borrowing costs including limitations to our
analysis.
To identify alternative triggers or other mechanisms, we interviewed
budget and legislative experts including former congressional staff;
former CBO, Office of Management and Budget, and Treasury officials;
and other congressional observers from a range of policy research
organizations. We also reviewed information from select member
countries of the Organisation for Economic Co-operation and
Development (OECD) and received input from budget or debt office
representatives from Canada, Denmark, New Zealand, Sweden,
Switzerland, and the United Kingdom about mechanisms used in their
countries to manage aggregate levels of debt. We selected these
countries based on a review of relevant reports and other information
published by the OECD and International Monetary Fund (IMF). While
selected countries offer illustrative examples, their experiences are
not always applicable to the United States given differences in
political systems and economies.
To assess the reliability of the data used in this report, including
financial markets data downloaded from Thomson Reuters' proprietary
statistical database, Datastream, and IHS Global Insight and publicly
available data from Treasury and the Federal Reserve, we examined the
data to look for outliers and anomalies and, when possible, compared
data from multiple sources for consistency. In general, we chose
databases that were commonly used by Treasury and researchers to
monitor changes in federal debt and related transactions. On the basis
of our assessment we believe the data are sufficiently reliable for
the purpose of this review.
We conducted our work from December 2009 to January 2011 in accordance
with generally accepted government auditing standards. Those standards
require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings
and conclusions based on our audit objectives. We believe that the
evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
Background:
Congress and the President first enacted a statutory limit on federal
debt during World War I to eliminate the need for Congress to approve
each new debt issuance and provide Treasury with greater discretion
over how it finances the government's day-to-day borrowing needs.
Federal debt subject to the limit includes both debt held by the
public and debt held by government accounts (intragovernmental debt
holdings). The majority of debt held by the public consists of
marketable Treasury securities, such as bills, notes, bonds, and
Treasury Inflation-Protected Securities (TIPS), that are sold through
auctions and can be resold by whoever owns them. Treasury also issues
a smaller amount of nonmarketable securities, such as savings
securities and special securities for state and local governments.
Debt held by the public primarily represents the amount the federal
government has borrowed to finance cumulative cash deficits.
Intragovernmental debt holdings represent balances of Treasury
securities held in government accounts such as the Social Security
trust funds. It increases when these accounts run a surplus or accrue
interest on existing securities.[Footnote 1] Debt subject to the limit
increased from roughly $43 billion in 1940 to more than $13,000
billion (or $13 trillion) in 2010.
In the past, Congress has sought to link decisions about the level of
debt to those about the level of federal spending and revenue by
integrating changes to the debt limit into the larger budget process.
For example, the Congressional Budget Act of 1974 requires that
Congress include the levels of debt implied by the spending and
revenue levels in the budget resolution for the next 5 years and
allows for specific estimates of the increase in debt subject to the
limit.[Footnote 2] Until recently, the House of Representatives had a
rule that automatically generated a joint resolution considered to
have been passed in the House changing the debt limit by the amount
recommended in the budget resolution for the next fiscal year.
[Footnote 3] The Congressional Budget Act as amended also established
an alternative procedure for changing the debt limit through
reconciliation legislation that is subject to expedited procedures.
Despite these rules and procedures, Congress usually votes on the debt
limit after fiscal policy decisions affecting federal borrowing have
begun to take effect.
Debt limit increases frequently involve protracted debate, regardless
of prior votes on the budget resolution or other legislation that
increases the need to borrow. This debate often occurs when federal
debt is near or at the debt limit. Three pieces of legislation enacted
to respond to the financial market crisis and economic downturn are
recent exceptions--in each of these the debt limit was increased by
roughly the amount the legislation was expected to increase debt. For
example, in addition to spending and revenue provisions, the American
Recovery and Reinvestment Act of 2009 (Recovery Act) increased the
debt limit by $789 billion from $11,315 billion to $12,104 billion
[Footnote 4]. Federal debt at the time of enactment was more than $600
billion below the limit.
Treasury's normal cash management operations involve ensuring that
there is enough cash on hand to pay government obligations as they
come due. Treasury has two primary sources of funds to finance these
obligations: (1) revenue collections, such as federal tax revenues and
other fees the federal government imposes and (2) cash borrowed from
the public through auctions of marketable securities. One of
Treasury's goals is to finance the government's borrowing needs at the
lowest cost over time by, among other things, issuing a wide range of
securities in a regular and predictable pattern.[Footnote 5] Treasury
currently issues bills that mature in a year or less, notes with
maturities of 2 to 10 years, and bonds with maturities of greater than
10 years. Treasury also issues 5-year, 10-year, and 30-year TIPS,
which offer inflation protection to investors who are willing to pay a
premium for this protection in the form of lower interest rates.
[Footnote 6] Treasury does not "time the market"--or seek to take
advantage of low interest rates--when it issues securities. Instead,
Treasury strives to lower its borrowing costs over time by relying on
a regular preannounced schedule of auctions.
Treasury holds cash in its operating cash balance in an account at the
Federal Reserve and in accounts at depository institutions across the
country. Treasury can draw down its operating cash balance as debt
approaches the limit, which allows Treasury to temporarily make
payments without increasing the amount of debt subject to the limit.
However, cash balances in its account at the Federal Reserve must be
kept at a sufficient level to avoid overdrawing this account since the
Federal Reserve System cannot legally lend directly to the Treasury.
The issuance of cash management bills (CM bills) provides another way
for Treasury to manage more closely the amount of additional debt
subject to the limit. CM bills are flexible securities that Treasury
issues outside of its regular preannounced auction schedule. Treasury
sets the amount and time to maturity to meet its immediate borrowing
needs at the time.[Footnote 7] Issuing CM bills allows Treasury to
borrow cash for shorter periods than regular bills to help manage the
uncertainty around the timing of increases to the debt limit. However,
our past work showed that Treasury paid a premium, in the form of
higher yields, for CM bills relative to regular bills.[Footnote 8]
There are also a number of extraordinary actions currently available
to Treasury to avoid exceeding the debt limit. These actions reduce
uncertainty over future increases in debt subject to the limit or
reduce the amount of debt subject to the limit. Table 1 provides an
overview of each one. Two of these actions relate to the Civil Service
Retirement and Disability Fund (CSRDF), which is the trust fund for
two federal retirement plans that hold nonmarketable securities. To
take these actions, Treasury must declare in advance a debt issuance
suspension period (DISP)--a period in which Treasury determines that
it cannot issue debt without exceeding the debt limit. Another four
actions can be taken without first declaring a DISP.
Table 1: Extraordinary Actions Taken by Treasury to Manage Debt near
the Debt Limit:
Extraordinary actions that do not require the declaration of a DISP:
Suspension of new issuances of State and Local Government Series
(SLGS) Securities; Extraordinary actions that do not require the
declaration of a DISP: SLGS are special securities offered to state
and local governments and other issuers of tax-exempt bonds.
Suspending new SLGS issuances reduces uncertainty over future
increases in debt subject to the limit. Suspending SLGS issuances
eliminates a flexible, low-cost option that state and local government
issuers have frequently used when refinancing their existing debt
before maturity. Suspending new SLGS issuances is generally the first
extraordinary action Treasury takes to manage debt near the debt limit.
Exchanging Federal Financing Bank (FFB) debt for debt subject to the
limit; Extraordinary actions that do not require the declaration of a
DISP: FFB is a government corporation under the general supervision
and direction of the Secretary of the Treasury, which borrows from the
Treasury to finance purchases of agency debt and agency guaranteed
debt. It can also issue up to $15 billion of its own debt--FFB 9(a)
obligations--that is not subject to the debt limit. This debt can be
exchanged with other federal debt (e.g., securities held by the CSRDF)
to reduce the amount of debt subject to the limit.
Suspension of investments to the Government Securities Investment Fund
of the Federal Employees' Retirement System (G-Fund)[A]; Extraordinary
actions that do not require the declaration of a DISP: The G-Fund
contains contributions made by federal employees toward their
retirement as part of the Thrift Savings Plan program, which are
invested in special one-day nonmarketable Treasury securities that are
subject to the limit. As debt nears the limit and the Secretary
determines that the G-Fund may not be fully invested without exceeding
the debt limit, Treasury can suspend investment for the entire amount
or a portion of the G-Fund on a daily basis to reduce debt subject to
the limit. Treasury is required to restore lost interest on the G-
Fund's uninvested funds after the debt limit has been increased.[B]
Suspension of Exchange Stabilization Fund (ESF) Investments;
Extraordinary actions that do not require the declaration of a DISP:
The ESF is used to help provide a stable system of monetary exchange
rates. Dollar-denominated assets of the ESF not used for program
purposes are generally invested in one-day nonmarketable Treasury
securities that are subject to the debt limit. When debt approaches
the limit, Treasury can suspend investment for the entire amount or a
portion of the ESF's maturing nonmarketable Treasury securities.
Treasury is not authorized to restore lost interest to the ESF when
the debt limit is increased.
Extraordinary actions that require the declaration of a DISP:
Suspension of new CSRDF investments[C]; Extraordinary actions that do
not require the declaration of a DISP: Once debt reaches the debt
limit, Treasury is able to suspend investment of new receipts to the
CSRDF. To do so, Treasury must send a letter notifying Congress that
CSRDF receipts cannot be invested without exceeding the debt limit
(i.e., declaring a DISP). Treasury is required to make the CSRDF whole
after the DISP has ended.
Disinvestment of securities held by CSRDF[C]; Extraordinary actions
that do not require the declaration of a DISP: Once debt reaches the
debt limit, Treasury is able to disinvest Treasury securities held by
the CSRDF. To do so, Treasury must send a letter notifying Congress
that it will not be able to issue debt securities without exceeding
the debt limit and provide the expected length of the DISP, which
Treasury uses to determine the amount of CSRDF investments that can be
disinvested. Treasury is required to restore lost interest after the
DISP has ended.
Source: GAO.
[A] Under 5 U.S.C. § 8438(g), to repay lost interest on suspended G-
Fund investments, Treasury must declare a separate DISP unrelated to
the actions taken involving the CSRDF. The Secretary is required to
notify Congress when a G-Fund DISP begins but is not required to
determine the length of a G-Fund DISP in advance. For the purposes of
this report, we use the term DISP to refer to the authorities that
Treasury has related to the CSRDF under 5 U.S.C. § 8348(j) unless
otherwise specified.
[B] 5 U.S.C. § 8438(g).
[C] 5 U.S.C. § 8348(j).
[End of table]
Since 1995, the debt limit has been increased 12 times. Prior to 6 of
these, Treasury had to take one or more extraordinary actions to avoid
exceeding the debt limit. Figure 1 shows when the debt limit was
increased and the extraordinary actions that were used.
* Prior to five of the six debt limit increases between 1996 and 2006,
Treasury took extraordinary actions, including declaring a DISP.
* During the period immediately preceding the debt limit increase in
August 1997, Treasury did not take any extraordinary actions.
* The federal government ran budget surpluses in fiscal years 1998
through 2001. Debt subject to the limit increased by $293 billion
during this period, but no increases to the debt limit were required.
* During the period immediately preceding the debt limit increase in
September 2007, Treasury suspended the issuance of SLGS but did not
take any other extraordinary actions or declare a DISP.
* In 2008 and 2009, three laws that were expected to increase the
amount of debt held by the public included corresponding increases in
the debt limit at the time of enactment.
* In December 2009 and February 2010, Treasury avoided taking
extraordinary actions as debt approached the limit in part by allowing
the Treasury securities it issued for the Supplementary Financing
Program (SFP)--a temporary program begun in 2008 at the request of the
Federal Reserve to drain reserves from the banking system and assist
with its emergency liquidity and lending initiatives--to mature
without rolling them over.
Figure 1: Extraordinary Actions Taken by Treasury Prior to Debt Limit
Increases, 1995-2010:
[Refer to PDF for image: illustrated table]
Extraordinary actions that do not require the declaration of a DISP:
Suspension of new issuances of SLGS:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Check];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Check];
September 29, 2007: [Check];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 6.
Exchanging FFB debt for debt subject to the limit[A]:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Empty];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Empty];
September 29, 2007: [Empty];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 3.
Suspension of G-Fund investments:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Check];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Check];
September 29, 2007: [Empty];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 5.
Suspension of ESF investments:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Empty];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Check];
September 29, 2007: [Empty];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 4.
Extraordinary actions that require the declaration of a DISP:
Suspension of new CSRDF investments:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Check];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Check];
September 29, 2007: [Empty];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 5.
Disinvestment of securities held by CSRDF:
March 29, 1996: [Check];
August 5, 1997: [Empty];
June 28, 2002: [Check];
May 27, 2003: [Check];
November 19, 2004: [Check];
March 20, 2006: [Check];
September 29, 2007: [Empty];
July 30, 2008: [Empty];
October 3, 2008: [Empty];
February 17, 2009: [Empty];
December 28, 2009: [Empty];
February 12, 2010: [Empty];
Total number of times used since 1996: 5.
Source: GAO.
[A] The 1996 transaction did not involve FFB issuing FFB 9(a)
obligations. Instead, FFB exchanged other financial assets--namely
loans to federally chartered entities--that were also exempt from the
debt limit for securities held by the CSRDF that were subject to the
debt limit.
[End of figure]
If debt is at the limit and the extraordinary actions are exhausted,
Treasury may not issue debt without further action from Congress and
could be forced to delay payments until sufficient funds become
available. In the past, Congress and the Secretary of the Treasury
have taken additional actions beyond those described above when
necessary to ensure that the government paid its obligations as they
came due without breaching the debt limit. For example, in 1996,
Congress passed and the President signed legislation allowing Treasury
to issue securities temporarily excluded from the debt limit in an
amount equal to the March 1996 Social Security payments to ensure that
benefit payments were made on time.[Footnote 9]
Treasury has never been unable to pay interest or principal on debt
held by the public because of the debt limit. Treasury, credit rating
agencies, and others agree that failure to pay principal or interest
on Treasury securities because of the debt limit could have costly
consequences for the U.S. government and financial markets including
higher future borrowing costs for Treasury and the public; stress on
the value of the dollar in currency markets; and major disruptions in
capital markets due to the repricing of products, services, and
transactions dependent on an efficiently functioning Treasury market.
Increased Borrowing and Limited Borrowing Capacity Provided by
Extraordinary Actions Create Debt Management Challenges:
Extraordinary Actions Provide Less Borrowing Capacity Relative to
Borrowing Needs Than They Did in the Past:
The borrowing capacity provided by the extraordinary actions has grown
in size but has not kept pace with the growth in Treasury's borrowing
needs. The amount potentially provided by the extraordinary actions
for a 1-month DISP in fiscal year 2010 was less than the monthly
increase in debt subject to the limit for most months of the year. As
of August 31, 2010, the extraordinary actions available to Treasury
could provide about $147.5 billion in additional borrowing capacity
without a DISP and an additional $7.7 billion per month based on the
length of the DISP declared. As table 2 shows, the amounts available
from suspending G-Fund investments, suspending ESF investments and the
disinvestment of CSRDF funds have all grown--with the bulk of the
growth in the G-Fund. G-Fund growth results from an increase in
federal employee retirement funds being invested in Treasury
securities. However, the estimated total borrowing capacity provided
by extraordinary actions available without a DISP is still $15 billion
below Treasury's average monthly borrowing needs in fiscal year 2010,
which was over $162 billion, and only 44 percent of the largest single
monthly increase in debt subject to the limit, which was over $330
billion. Treasury officials stated that there are no additional
extraordinary actions within their legal authorities that could be
prudently used in the future to create additional borrowing capacity.
Table 2: Estimated Borrowing Capacity Provided by Extraordinary
Actions:
Extraordinary actions that do not require the declaration of a DISP:
Exchanging FFB debt for debt that is subject to the limit[A];
Fiscal year: 2002: $0.0 billion;
Fiscal year: 2005: $1.0 billion;
Fiscal year: 2006: $1.0 billion;
Fiscal year: 2010: $4.8 billion.
Suspension of G-Fund investments;
Fiscal year: 2002: $44.0 billion;
Fiscal year: 2005: $62.6 billion;
Fiscal year: 2006: $72.2 billion;
Fiscal year: 2010: $122.3 billion.
Suspension of ESF investments;
Fiscal year: 2002: $9.8 billion;
Fiscal year: 2005: $15.2 billion;
Fiscal year: 2006: $15.6 billion;
Fiscal year: 2010: $20.4 billion.
Subtotal--extraordinary actions available without declaring a DISP;
Fiscal year: 2002: $53.8 billion;
Fiscal year: 2005: $78.8 billion;
Fiscal year: 2006: $88.8 billion;
Fiscal year: 2010: $147.5 billion.
Extraordinary actions that require the declaration of a DISP (amount
per month based on the length of the DISP declared):
Suspension of new CSRDF investments[B];
Fiscal year: 2002: $1.7 billion;
Fiscal year: 2005: $2.0 billion;
Fiscal year: 2006: $2.1 billion;
Fiscal year: 2010: $2.0 billion.
Disinvestment of securities held by the CSRDF;
Fiscal year: 2002: $4.0 billion;
Fiscal year: 2005: $4.6 billion;
Fiscal year: 2006: $4.8 billion;
Fiscal year: 2010: $5.7 billion.
Extraordinary actions: Total;
Fiscal year: 2002: $59.6 billion;
Fiscal year: 2005: $85.3 billion;
Fiscal year: 2006: $95.8 billion;
Fiscal year: 2010: $155.2 billion.
Source: GAO and Department of the Treasury.
Notes: These estimates represent an approximation of the additional
borrowing capacity provided by the extraordinary actions as of August
31st of each year--the last month in the fiscal year for which data
are typical of most months of the year. They do not reflect the actual
amount of borrowing capacity Treasury obtained by taking extraordinary
actions in any given year.
[A] Some or all of the $15 billion in FFB 9(a) securities that FFB can
issue were already exchanged for debt subject to the limit.
[B] Treasury can also suspend large investments to the CSRDF that are
made three times a year. In June and December, Treasury makes
semiannual interest payments to the CSRDF and in September, Treasury
makes a onetime investment in the CSRDF for financing the unfunded
liability of new and increased annuity benefits. These amounts would
be added to the monthly averages calculated above.
[End of table]
Some of the options used in the past are either more limited or no
longer available. FFB has the authority to issue up to $15 billion in
securities that are not subject to the debt limit that it can exchange
for other Treasury securities to reduce the amount of debt subject to
the limit. However, some or all of these FFB securities may be
outstanding from previous transactions, including those made to manage
the amount of debt subject to the limit in the past, and therefore
unavailable. For example, as of August 31, 2010, the exchange of FFB
securities for other Treasury securities could provide less than $5
billion in additional borrowing capacity under the debt limit. In the
past, FFB reversed these transactions by redeeming FFB 9(a)
obligations prior to maturity once the debt limit was raised. However,
Treasury officials said they no longer reverse these transactions
because of the potential costs FFB and its counterparties could incur
as a result.[Footnote 10] Also, until March 2004, Treasury kept
"compensating balances" in non-interest-bearing accounts at banks to
compensate them for collecting federal receipts for the Treasury. This
allowed Treasury to call back tens of billions of dollars when needed
to pay obligations and avoid breaching the debt limit.[Footnote 11]
Since these compensating balances were replaced in March 2004 by
direct payment to banks for services, this option is no longer
available.
Assuming current borrowing trends, our estimates show that the
borrowing capacity provided by the extraordinary actions would be
sufficient to meet the government's borrowing needs for as little as a
few days to a few weeks during certain times of the year. This means
that once debt approaches the debt limit, Treasury may not be able to
manage the amount of debt subject to the limit for as long a period of
time as it had in the past before the debt limit must be increased or
payments must be delayed. Figure 2 below shows the estimated borrowing
capacity provided by these actions for a 1-month DISP relative to the
monthly change in debt subject to the limit for fiscal year 2010 and 3
previous fiscal years in which Treasury took extraordinary actions.
The amount potentially provided by the extraordinary actions for a 1-
month DISP in fiscal year 2010 was less than the monthly increase in
debt subject to the limit in 8 of the 12 months. In contrast, in
earlier years, the potential borrowing capacity provided by the
extraordinary actions was greater than the monthly increase in debt
subject to the limit in almost all months.
Figure 2: Estimated Borrowing Capacity Provided by Extraordinary
Actions Based on a 1-Month DISP Relative to Actual Monthly Changes in
Debt Subject to the Limit for Select Years:
[Refer to PDF for image: illustrated table]
Change in total federal debt subject to the limit:
Fiscal year 2003:
October: $11.7 billion;
November: $72.3 billion;
December: $54.6 billion;
January: ($5.5 billion);
February: $67.3 billion;
March: $2.0 billion;
April: ($20.3 billion); Estimated borrowing capacity provided by
extraordinary actions as of August 31: $59.6 billion;
May: $35.2 billion;
June: $108.3 billion;
July: $33.7 billion;
August: $50.8 billion;
September: $18.6 billion.
Fiscal year 2005:
October: $50.6 billion;
November: $80.8 billion;
December: $70.9 billion;
January: $32.1 billion;
February: $85.0 billion;
March: $62.8 Estimated borrowing capacity provided by extraordinary
actions as of August 31: $85.3 billion;
April: ($11.5 billion);
May: $13.5 billion;
June: $60.6 billion;
July: $50.9 billion;
August: $39.4 billion;
September: $2.6 billion.
Fiscal year 2006:
October: $93.7 billion;
November: $64.1 billion;
December: $78.1 billion;
January: $25.3 billion;
February: $51.7 billion; Estimated borrowing capacity provided by
extraordinary actions as of August 31: $95.8 billion;
March: $97.5 billion;
April: ($18.7 billion);
May: $1.1 billion;
June: $66.8 billion;
July: $22.0 billion;
August: $70.7 billion;
September: ($3.0 billion).
Fiscal year 2010:
October: ($16.8 billion);
November: $220.7 billion;
December: $197.2 billion;
January: ($32.0 billion); Estimated borrowing capacity provided by
extraordinary actions as of August 31: $155.2 billion;
February: $161.2 billion;
March: $332.8 billion;
April: $220.7 billion;
May: $256.8 billion;
June: $248.0 billion;
July: $35.6 billion;
August: $231.6 billion;
September: $112.0 billion.
Source: GAO analysis of Treasury data.
[End of figure]
The actions available without declaration of a DISP could potentially
provide $147.5 billion (as of Aug. 31, 2010), but the amount of time
that these actions provide before debt reaches the limit depends on a
number of factors. For instance, debt subject to the limit increases
sharply certain days of the year. Treasury makes semiannual interest
payments on a large amount of debt held in government accounts on the
last day of June and December.[Footnote 12] During the recent debt
limit debate in early fiscal year 2010, debt increased by more than
$165 billion in a single day--December 31--because of $81 billion in
net nonmarketable securities issuances, including interest payments to
government accounts,[Footnote 13] as well as $84 billion in net
marketable securities issuances.
Another factor that determines the amount of time that Treasury is
able to manage debt near the debt limit is the size of Treasury's
operating cash balance. Treasury can draw down its operating cash
balance to pay obligations rather than increase borrowing. While the
size of Treasury's operating cash balance routinely fluctuates
throughout the year depending on the timing of withdrawals and
deposits, table 3 shows that Treasury's average operating cash balance
was roughly twice as high in fiscal year 2009 and fiscal year 2010 as
it was in the previous 6 fiscal years. From December 15, 2009, to
February 11, 2010, when debt was approaching the debt limit,
Treasury's operating cash balance (excluding the SFP account balance)
rarely fell below $90 billion.[Footnote 14] Higher cash balances
helped ensure that Treasury had enough cash available to make large
disbursements on short notice. Treasury officials explained that
higher cash balances were not related to the debt limit but rather to
regular and predictable financing patterns coupled with large receipts
and expenditures related to the Troubled Asset Relief Program (TARP),
Recovery Act, and other legislation to address the financial crisis
and the economic downturn.
Table 3: Average Operating Cash Balance Less Supplementary Financing
Program Account Balance, Fiscal Years 2003-2010:
Average operating cash balance:
Fiscal year: 2003: $17.9 billion;
Fiscal year: 2004: $20.5 billion;
Fiscal year: 2005: $25.9 billion;
Fiscal year: 2006: $26.4 billion;
Fiscal year: 2007: $30.6 billion;
Fiscal year: 2008: $24.9 billion;
Fiscal year: 2009: $58.9 billion;
Fiscal year: 2010: $57.7 billion.
Source: GAO analysis of Treasury data.
[End of table]
The amount of additional borrowing capacity provided by disinvesting
CSRDF securities depends on the length of the DISP declared by the
Secretary.[Footnote 15] For past DISPs, the Secretary determined the
amount of disinvestments based on the length of the DISP and the
estimated monthly CSRDF benefit payments that would occur during this
time. For example, Treasury declared a DISP from May 16 to June 28,
2002, and disinvested about $4 billion in Treasury securities held by
the CSRDF. This amount was roughly equal to the amount that would have
been needed to make 1 month's worth of Civil Service benefit payments.
Similarly, Treasury declared a 12-month DISP in November 1995 and
disinvested $39.8 billion in Treasury securities held by the CSRDF,
roughly the equivalent of 12 months' worth of benefit payments. The
statute does not require that disinvestments be made only for the
purpose of making CSRDF benefit payments. However, Treasury cannot
disinvest additional securities later to make those benefit payments.
As a result, the amount provided by the CSRDF declines over the period
of the DISP. In the past 16 years, the Secretary of the Treasury
declared DISPs ranging from 14 days to 14 months. The period of time
between the declaration of a DISP and the debt limit increase ranged
from 1 day to 4-½ months.
Treasury Diverts Resources from Other Priorities to Manage Debt near
the Limit:
Debt and cash management require more time and Treasury resources as
debt nears the debt limit. The size and timing of auctions must be
adjusted when nearing the debt limit; cash and borrowing needs must be
forecasted and monitored with increasing frequency and in increasing
detail; and contingency plans and alternative scenarios for the
possible implementation of extraordinary actions must be developed,
reviewed, and tested. These activities divert time and Treasury
resources from other cash and debt management issues. We reviewed
estimates provided by the Office of Debt Management (ODM), the Office
of Fiscal Projections (OFP), and the Bureau of the Public Debt (BPD)
that overall indicated they devoted as much as several hundred hours
per week to managing debt near the debt limit.
Treasury's operational focus on the debt limit begins as early as 6 to
9 months before the debt limit is expected to be reached and increases
as debt nears the limit. Since this work involves contingency
planning, it is undertaken whether or not the debt limit is raised
prior to the use of extraordinary actions or the declaration of a
DISP. For example, Treasury staff develop projections under multiple
scenarios of when debt might reach the debt limit. As debt nears the
debt limit, these projections and scenarios are developed weekly, then
daily, and finally as often as multiple times a day. According to
Treasury, these projections and scenarios may take 3 of OFP's 11 staff
members between 2 to 4 hours per day to produce.
While Treasury needs accurate cash-flow forecasts to project changes
in the amount of debt subject to the limit, the precision and
frequency increases when debt is near or at the limit. While large
regular and predictable payments and receipts--such as Medicare and
Social Security payments and receipts from corporate taxes--cause
predictable swings in daily deposits and withdrawals, an official from
OFP said that it was uncertainty about other revenue and irregular
payments that made planning and forecasting more difficult as debt
approached the debt limit in fiscal year 2010. For example, as figure
3 shows, Treasury received an influx of repayments of more than $90
billion from financial institutions under TARP in December 2009.
However, since Treasury did not know for certain when these payments
would be received, Treasury officials ran multiple projections of when
the debt limit would be reached.
Figure 3: Daily Change in Net Cash Flows during Recent Debt Limit
Event Period, August 7, 2009, to February 12, 2010:
[Refer to PDF for image: vertical bar graph]
Date: 8/7/09: -$4 billion;
-$1 billion;
-$8 billion;
-$7 billion;
-$12 billion;
-$2 billion;
-$18 billion;
-$7 billion;
-$7 billion;
-$6 billion;
-$6 billion;
-$2 billion;
-$6 billion;
-$8 billion;
-$9 billion;
-$2 billion;
-$5 billion;
Date: 9/1/09: -$30 billion; ($15 billion Medicare payment);
-$1 billion;
-$29 billion;
-$6 billion;
-$3 billion;
-$12 billion;
-$6 billion;
-$4 billion;
$3 billion;
$27 billion; ($29 billion corporate tax deposit);
$2 billion;
-$4 billion;
-$1 billion;
$5 billion;
$3 billion;
-$9 billion;
-$10 billion;
-$6 billion;
0;
-$3 billion;
-$6 billion;
Date: 10/1/09: -$31 billion; ($15 billion Medicare payment);
-$24 billion;
-$2 billion;
-$9 billion;
0;
-$8 billion;
-$3 billion;
-$3 billion;
-$14 billion;
-$3 billion;
-$3 billion;
-$3 billion;
-$6 billion;
-$4 billion;
-$7 billion;
-$1 billion;
$1 billion;
-$7 billion;
-$10 billion;
-$7 billion;
-$31 billion; ($15 billion Medicare payment);
Date: 11/2/09: -$4 billion;
-$27 billion;
$2 billion;
-$7 billion;
-$4 billion;
0;
-$16 billion;
-$4 billion;
-$12 billion;
-$11 billion;
-$8 billion;
-$9 billion;
-$8 billion;
-$4 billion;
$1 billion;
-$8 billion;
-$8 billion;
-$6 billion;
-$8 billion;
Date: 12/1/09: -$32 billion;
-$2 billion;
-$30 billion;
-$3 billion;
-$5 billion;
-$7 billion;
$37 billion; ($46 billion TARP deposit);
-$7 billion;
-$4 billion;
-$1 billion;
$31 billion; ($37 billion corporate tax deposit);
-$1 billion;
-$6 billion;
$1 billion;
$4 billion;
-$6 billion;
$29 billion; ($46 billion TARP deposit);
-$5 billion;
$3 billion;
$1 billion;
$31 billion; ($44 billion FDIC insurance deposit);
-$46 billion; ($22 billion Social Security payment and $15 billion GSE
investment payment);
Date: 1/4/10: -$7 billion;
-$9 billion;
$3 billion;
-$5 billion;
$2 billion;
0;
-$6 billion;
-$9 billion;
-$6 billion;
-$3 billion;
$13 billion;
$11 billion;
$1 billion;
-$5 billion;
$3 billion;
-$5 billion;
-$7 billion;
-$8 billion;
-$10 billion;
Date: 2/1/10: -$24 billion;
-$4 billion;
-v23 billion;
-$8 billion;
-$26 billion;
$1 billion;
-$8 billion;
-$3 billion;
-$7 billion;
Date: 2/12/10: -$32 billion ($29 billion IRS tax refund payment).
Source: GAO analysis of Treasury data.
Note: Net cash flow is equal to deposits less withdrawals. This
excludes federal debt transactions including noncash transactions
involving debt held in government accounts such as large regularly
occurring interest payments in late December.
[End of figure]
Treasury uses the projections of debt subject to the limit not only
for operational scenarios but also in meetings to inform senior
Treasury officials--including the Secretary. These meetings also
increase in frequency from monthly to as often as daily as debt
approaches the limit. The meetings, which have included 10 or more
executives and senior career staff, are used to discuss strategies for
managing debt near the debt limit including the potential use of
extraordinary actions. According to Treasury, these meetings can
require several hours of preparation. While Treasury officials and
staff can draw on previous experiences managing debt near the debt
limit, they told us that each debt limit event presents new and
different issues to be considered and addressed; in addition, there
are often senior officials who have not been through the experience
and must be fully briefed and prepared.
BPD--the bureau within Treasury that is responsible for implementing
the extraordinary actions and the accounting associated with those
transactions--also dedicates extensive resources on operations related
to the debt limit. BPD estimates that a 2-month DISP results in
roughly 1,900 hours of work including the time spent before, during,
and after the debt limit increase. This includes more than 400 hours
in the 6 weeks prior to the implementation of any extraordinary
actions spent on meetings to prepare for when the debt limit is
reached, preparation of parallel accounts and spreadsheets in the
event that extraordinary actions involving the G-Fund and CSRDF are
used, tests of the accounting system, and a mock auction to practice
and verify procedures for potential auction postponements. BPD also
estimates that it spends in excess of 140 hours on debt limit-related
activities each week once the first extraordinary action is taken, and
over 270 hours on activities such as unwinding past transactions and
preparing reports after the debt limit has been increased. The
increased workload could result in overtime hours for BPD employees.
Treasury officials said that the increased focus on debt limit-related
operations in the months and weeks approaching the debt limit can
divert time and attention from other tasks that could improve Treasury
operations. For example, according to Treasury, OFP is able to spend
less time working to update or improve the models it uses in routine
forecasting of tax receipts, expenditures, and borrowing needs.
Similarly, Treasury officials said that ODM is able to spend less time
analyzing short-term financing needs that could help inform auction
amounts. Both of these activities help Treasury more accurately
project future borrowing needs to avoid the following:
(1) Borrowing more than is needed to fund the government's immediate
needs, which results in increased interest costs.
(2) Borrowing less than is sufficient to maintain Treasury's operating
cash balance at a minimum level through regularly scheduled issuances
of marketable Treasury securities. This may require Treasury to issue
CM bills with little advance notice to the market, resulting in
potentially higher interest costs for the federal government.
Approaching the Debt Limit Can Add Uncertainty in the Treasury Market:
Postponed Auctions and Other Disruptions May Lead to Increased
Borrowing Costs:
Some of the actions Treasury takes to manage the amount of debt as it
approaches the debt limit disrupt the regular and predictable auction
schedule that Treasury relies on to promote liquid markets and finance
the government's borrowing needs at the lowest cost over time. Regular
and predictable auctions provide investors greater certainty and
better information with which to plan their investments. Meanwhile, a
liquid market allows investors to more easily buy and sell Treasury
securities in the secondary market. Market participants that we spoke
with said that any actions that Treasury takes to manage debt as it
approaches the limit that cause Treasury to deviate from its otherwise
regular and predictable schedule or reduce liquidity introduce
uncertainty into the Treasury market and have the potential to
increase Treasury's borrowing costs.
Since 1995, Treasury delayed the announcement of 17 regularly
scheduled auctions by 1-½ hours to 7 business days and postponed the
auction date for 11 auctions by as many as 8 business days (see table
4). Treasury also reduced the offering size of a 13-week bill by $7
billion after the initial auction announcement in October 1995 in
order to stay under the debt limit. These actions introduced
uncertainty into the market for Treasury securities, and in some
circumstances may have increased borrowing costs.
Table 4: Postponed Auctions and Delayed Auction Announcements, 1995-
2010:
Type of Security: 3-year note;
Originally scheduled announcement date: 11/1/1995;
Date of actual auction announcement: 11/13/1995;
Original auction date: 11/7/1995;
Actual auction date: 11/20/1995;
Length of delay in announcement (business days): 7;
Length of auction postponement (business days): 8.
Type of Security: 10-year note;
Originally scheduled announcement date: 11/1/1995;
Date of actual auction announcement: 11/13/1995;
Original auction date: 11/8/1995;
Actual auction date: 11/21/1995;
Length of delay in announcement (business days): 7;
Length of auction postponement (business days): 8.
Type of Security: 52-week bill;
Originally scheduled announcement date: 11/2/1995;
Date of actual auction announcement: 11/13/1995;
Original auction date: 11/9/1995;
Actual auction date: 11/15/1995;
Length of delay in announcement (business days): 6;
Length of auction postponement (business days): 3.
Type of Security: 2-year note;
Originally scheduled announcement date: 6/19/2002;
Date of actual auction announcement: 6/28/2002;
Original auction date: 6/26/2002;
Actual auction date: 6/28/2002;
Length of delay in announcement (business days): 7;
Length of auction postponement (business days): 2.
Type of Security: 13-week bill;
Originally scheduled announcement date: 6/27/2002;
Date of actual auction announcement: 6/28/2002;
Original auction date: 7/1/2002;
Actual auction date: 7/1/2002;
Length of delay in announcement (business days): 1;
Length of auction postponement (business days): 0.
Type of Security: 26-week bill;
Originally scheduled announcement date: 6/27/2002;
Date of actual auction announcement: 6/28/2002;
Original auction date: 7/1/2002;
Actual auction date: 7/1/2002;
Length of delay in announcement (business days): 1;
Length of auction postponement (business days): 0.
Type of Security: 13-week bill;
Originally scheduled announcement date: 5/15/2003;
Date of actual auction announcement: 5/19/2003;
Original auction date: 5/19/2003;
Actual auction date: 5/20/2003;
Length of delay in announcement (business days): 2;
Length of auction postponement (business days): 1.
Type of Security: 26-week bill;
Originally scheduled announcement date: 5/15/2003;
Date of actual auction announcement: 5/19/2003;
Original auction date: 5/19/2003;
Actual auction date: 5/20/2003;
Length of delay in announcement (business days): 2;
Length of auction postponement (business days): 1.
Type of Security: 4-week bill;
Originally scheduled announcement date: 5/19/2003;
Date of actual auction announcement: 5/19/2003;
Original auction date: 5/20/2003;
Actual auction date: 5/21/2003;
Length of delay in announcement (business days): [Empty];
Length of auction postponement (business days): 1.
Type of Security: 13-week bill;
Originally scheduled announcement date: 5/22/2003;
Date of actual auction announcement: 5/27/2003;
Original auction date: 5/27/2003;
Actual auction date: 5/28/2003;
Length of delay in announcement (business days): 2;
Length of auction postponement (business days): 1.
Type of Security: 26-week bill;
Originally scheduled announcement date: 5/22/2003;
Date of actual auction announcement: 5/27/2003;
Original auction date: 5/27/2003;
Actual auction date: 5/28/2003;
Length of delay in announcement (business days): 2;
Length of auction postponement (business days): 1.
Type of Security: 2-year note;
Originally scheduled announcement date: 5/22/2003;
Date of actual auction announcement: 5/27/2003;
Original auction date: 5/28/2003;
Actual auction date: 5/29/2003;
Length of delay in announcement (business days): 2;
Length of auction postponement (business days): 1.
Type of Security: 4-week bill;
Originally scheduled announcement date: 11/15/2004;
Date of actual auction announcement: 11/19/2004;
Original auction date: 11/16/2004;
Actual auction date: 11/19/2004;
Length of delay in announcement (business days): 4;
Length of auction postponement (business days): 3.
Type of Security: 13-week bill;
Originally scheduled announcement date: 11/18/2004;
Date of actual auction announcement: 11/19/2004;
Original auction date: 11/22/2004;
Actual auction date: 11/22/2004;
Length of delay in announcement (business days): 1;
Length of auction postponement (business days): 0.
Type of Security: 26-week bill;
Originally scheduled announcement date: 11/18/2004;
Date of actual auction announcement: 11/19/2004;
Original auction date: 11/22/2004;
Actual auction date: 11/22/2004;
Length of delay in announcement (business days): 1;
Length of auction postponement (business days): 0.
Type of Security: 2-year note;
Originally scheduled announcement date: 11/18/2004;
Date of actual auction announcement: 11/19/2004;
Original auction date: 11/23/2004;
Actual auction date: 11/23/2004;
Length of delay in announcement (business days): 1;
Length of auction postponement (business days): 0.
Type of Security: 13-week bill;
Originally scheduled announcement date: 3/16/2006;
Date of actual auction announcement: 3/16/2006;
Original auction date: 3/20/2006;
Actual auction date: 3/20/2006;
Length of delay in announcement (business days): 0[A];
Length of auction postponement (business days): 0.
Type of Security: 26-week bill;
Originally scheduled announcement date: 3/16/2006;
Date of actual auction announcement: 3/16/2006;
Original auction date: 3/20/2006;
Actual auction date: 3/20/2006;
Length of delay in announcement (business days): 0[A];
Length of auction postponement (business days): 0.
Source: GAO analysis of Treasury data:
[A] Treasury's auction announcement release time was delayed from
11:00 a.m. to 12:30 p.m.
[End of table]
Treasury officials and market participants both stated that postponing
the announcement or auction of longer-dated securities such as notes
and bonds is more disruptive to the Treasury market than postponing
shorter-dated securities such as bills. This is due in part to the
fact that dealers generally require additional time to work with
customers and secure financing for note and bond auctions. There is
generally about a week between the announcement of a note auction and
the actual auction. The announcement of a 2-year note auction was
delayed 7 business days in 2002, reducing the time that dealers had to
prepare bids from 5 business days to less than 1. This auction had the
lowest bid-to-cover ratio--the ratio of dollar value of bids at
auction to the amount accepted--for any 2-year note auction since
these data began to be recorded in 1998. It is difficult to say with
certainty how much in additional interest Treasury paid because of the
postponement of this auction. Based on one estimate that assumes that
the rate on the note would have been roughly equal to the constant
maturity rate, or the closing rate on actively traded Treasury
securities maturing at the same time on the day that the auction
actually took place, Treasury paid an additional 7 basis points--or
0.07 percentage points--on the $27 billion in 2-year notes issued that
day. This equals almost $19 million in additional interest costs each
year. Based on an alternative estimate that assumes that Treasury
would have received the prevailing interest rate on 2-year notes in
the secondary market on the day that the auction was originally
scheduled to take place, the increased interest costs would be even
greater.
The level of disruption resulting from a postponed auction depends in
part on how early Treasury forewarns the market. Unlike the 2-year
note auction in 2002, Treasury discussed the prospect of postponing a
3-year note and a 10-year note auction with members of the Treasury
Borrowing Committee at its regular committee meeting a week before the
original auction date and released the auction announcement a week
before the postponed auction took place. The yields at the subsequent
auctions on November 20 and 21, 1995 were roughly the same as the
constant maturity rate for each maturity on the day of the auction.
Auctions that are postponed beyond the maturity date of a previously
issued security can cause significant disruptions. Treasury generally
makes the actual exchange of Treasury securities for cash--the
settlement--a few days or more after the auction. Settlements
frequently occur around the same time that previously issued
securities are maturing. This allows Treasury to refund maturing
securities--to use some or all of the cash that it raised in the
auction to redeem maturing securities. It also allows investors to
easily roll over, or reinvest, cash received from maturing securities
in newly issued securities. Treasury stated that when the settlement
date of a new security is moved past the maturity date of a previously
issued security, investors are unable to rollover their investments in
a timely way. As a result, they may choose to invest in a financial
instrument other than Treasury securities. This could affect auction
demand. Most of the postponed auctions in the past 16 years were
delayed by only 1 to 3 days and therefore did not affect the refunding
of maturing securities. However, in November 1995, the 3-and 10-year
note auctions intended to refund notes maturing on November 15, 1995,
were postponed past the maturity date. In this instance, Treasury
bridged the gap between the maturity and settlement date by auctioning
a short-term CM bill.
Overall, Treasury issued 20 CM bills totaling more than $300 billion
during DISPs in the past 16 years to manage the amount of debt near
the limit. In some cases, these were used to augment Treasury's
regular schedule of bill auctions. For example, because of debt limit
constraints, Treasury delayed the 4-week bill auction scheduled for
November 16, 2004. Treasury then auctioned a 5-day CM bill for $7
billion on November 17, 2004. As mentioned previously, our prior work
found that Treasury may have paid a premium, in the form of higher
interest, on these CM bills compared to bills of a similar maturity.
[Footnote 16]
Treasury Sharply Reduced the Supply of Bills during Recent Debt Limit
Debate:
While Treasury did not postpone any of its regularly scheduled
auctions during the most recent debt limit debate, it sharply reduced
the total dollar amount of bills outstanding primarily to manage the
amount of debt as it approached the limit. The total amount of
Treasury bills outstanding dropped by $350 billion (or about 17
percent) from September 23, 2009, to February 12, 2010, (see figure
4). Roughly $200 billion of this was related to reductions to the SFP.
The decline in Treasury bills outstanding was accompanied by a decline
in short-term rates paid by Treasury. However, Treasury officials and
market participants stated that a sharp and irregular bill reduction
in such a short period of time could affect liquidity in the near term
and add uncertainty in the market over the longer term. While several
different types of investors use bills to invest their funds in the
short term in a safe and highly liquid asset, market participants we
spoke with said that money market funds were likely most affected by
the reduction in bill supply. Market participants also noted that if
Treasury had to make similar reductions to its issuance of notes or
bonds because of proximity to the debt limit, the effect on the
Treasury market would be greater.
Figure 4: Supply of Treasury Bills Decreased Sharply during Debt Limit
Event:
[Refer to PDF for image: line graph]
Dollars in billions:
Date: May 2009:
$2,023;
$2,023;
$2,023;
$2,023;
$2,023;
$2,037;
$2,032;
$2,032;
$2,032;
$2,032;
$2,048;
$2,048;
$2,048;
$2,048;
$2,065;
$2,065.
Date: June 2009:
$2,065;
$2,065;
$2,065;
$2,076;
$2,076;
$2,076;
$2,076;
$2,076;
$2,076;
$2,076;
$2,050;
$2,050;
$2,050;
$2,044;
$2,044;
$2,044;
$2,044;
$2,009;
$2,006;
$2,006;
$2,006;
$2,006.
Date: July 2009:
$2,006;
$1,980;
$1,980;
$2,010;
$2,010;
$2,010;
$2,020;
$2,020;
$2,020;
$2,020;
$2,020;
$2,036;
$2,036;
$2,036;
$2,036;
$2,036;
$2,046;
$2,046;
$2,046;
$2,046;
$2,046;
$2,020;
$2,020.
Date: August 2009:
$2,020;
$2,020;
$2,020;
$2,020;
$2,020;
$2,020 (8/7/2009: Treasury notifies Congress that the debt limit needs
to be raised);
$2,020 ;
$2,020;
$2,024;
$2,024;
$2,024;
$2,060;
$2,060;
$2,061;
$2,061;
$2,061;
$2,061;
$2,061;
$2,068;
$2,068;
$2,068.
Date: September 2009:
$2,068;
$2,068;
$2,063;
$2,063;
$2,063;
$2,063;
$2,075;
$2,075;
$2,075;
$2,055;
$2,055;
$2,036;
$2,036;
$2,036;
$2,036;
$2,036;
$1,993;
$1,993;
$1,993;
$1,993;
$1,993.
Date: October 2009:
$1,952;
$1,952;
$1,952;
$1,952;
$1,952;
$1,924;
$1,924;
$1,934;
$1,934;
$1,909;
$1,909;
$1,909;
$1,909;
$1,909;
$1,853 (9/24/09: SFP reduction begins);
$1,853;
$1,853;
$1,853;
$1,853;
$1,834;
$1,859.
Date: November 2009:
$1,859;
$1,859;
$1,859;
$1,862;
$1,862;
$1,862;
$1,862;
$1,867;
$1,867;
$1,867;
$1,867;
$1,867;
$1,847;
$1,847;
$1,846;
$1,846;
$1,846;
$1,850;
$1,850.
Date: December 2009:
$1,850;
$1,850;
$1,853;
$1,853;
$1,853;
$1,853;
$1,853;
$1,865;
$1,865;
$1,853;
$1,853;
$1,853;
$1,816;
$1,816;
$1,816;
$1,816;
$1,816;
$1,814;
$1,814;
$1,799;
$1,804 (12/30/09: SFP is reduced to $5 billion);
$1,793.
Date: January 2010:
$1,793;
$1,793;
$1,793;
$1,770;
$1,770;
$1,770;
$1,770;
$1,770;
$1,745;
$1,745;
$1,745;
$1,745;
$1,714;
$1,714;
$1,714;
$1,714;
$1,714;
$1,689;
$1,689.
Date: February 2010:
$1,689;
$1,689;
$1,689;
$1,679;
$1,679;
$1,669;
$1,679;
$1,679 (2/12/2010: Debt limit increase is signed into law);
$1,686;
$1,686;
$1,686;
$1,686;
$1,701;
$1,701;
$1,701;
$1,701;
$1,701;
$1,736;
$1,736.
Date: March 2010:
$1,736;
$1,736;
$1,736;
$1,710;
$1,770;
$1,770;
$1,770;
$1,770;
$1,774;
$1,799;
$1,799;
$1,799;
$1,799;
$1,820;
$1,820;
$1,820;
$1,820;
$1,820;
$1,843;
$1,843;
$1,843;
$1,843;
$1,843.
Date: April 2010:
$1,850;
$1,850;
$1,871;
$1,871;
$1,871;
$1,902;
$1,902;
$1,902;
$1,902;
$1,902;
$1,900;
$1,900;
$1,883;
$1,883;
$1,883;
$1,868;
$1,868;
$1,868;
$1,868;
$1,868;
$1,848;
$1,848.
Date: May 2010:
$1,848;
$1,848;
$1,848;
$1,833;
$1,833;
$1,833;
$1,833;
$1,833.
Source: GAO analysis of Treasury data.
[End of figure]
This sharp reduction in Treasury bills was unique to this most-recent
debt limit debate and largely related to the SFP, though large TARP
repayments in December, a reduction in Treasury's operating cash
balance, and the transition to longer-dated securities were also
factors. When Treasury approached the debt limit earlier in the
decade, the amount of Treasury bills outstanding tended to fluctuate
within a narrower range. To finance the SFP program, Treasury auctions
a series of CM bills, and places the proceeds in a special account at
the Federal Reserve Bank of New York. These outstanding CM bills count
against the debt limit. On September 16, 2009, Treasury announced it
intended to reduce the SFP account from $200 billion to $15 billion to
preserve flexibility in debt management. This amount dropped to as low
as zero in December 2009. As late as February 2, 2010, Treasury
announced that both the outlook for Treasury bills issuance and the
future of the SFP were uncertain. Shortly after the debt limit was
raised on February 12, 2010, Treasury returned the SFP account to
approximately $200 billion.
Evidence Suggests Borrowing Costs Increased during Some Debt Limit
Event Periods:
Our analysis suggests that the general uncertainty surrounding some
debt limit events including the most recent in 2009-2010 increased
Treasury's borrowing costs in the months immediately prior to a debt
limit increase.[Footnote 17] To measure changes in Treasury's
borrowing costs, we examined the spread between 13-week (i.e., 3-
month) Treasury bill yields and 3-month commercial paper yields around
debt limit events since 2001.[Footnote 18] Rates for Treasury bills,
commercial paper, and other financial assets can fluctuate from day to
day in response to changes in the broader economy. By focusing on a
yield spread rather than changes in individual interest rates, we are
able to better measure changes in the relative risk of 3-month
Treasury bills and identify potential risk premiums. A narrowing of
the spread indicates that the market perceives the risk of Treasury
bills to be closer to that of commercial paper, while a widening of
this spread means that Treasury bills are perceived to be less risky
relative to commercial paper. We found that Treasury paid a premium on
3-month Treasury bills issued during debt limit events in 2001-2002
and 2002-2003, but not in 2004-2005 and 2005-2006.[Footnote 19] For
the most-recent debt limit event--which lasted from August 7, 2009, to
February 12, 2010--we found that Treasury again paid a premium on 3-
month Treasury bills. Figure 5 shows that the average spread narrowed
during the debt limit event in 2009-2010 compared to the average for
the proceeding 3 months, implying that Treasury paid a premium on 3-
month Treasury bills issued during this time period.
Figure 5: Spreads between Yields of 3-Month Commercial Paper and 3-
Month Treasury Bills Were Lower during the 2009-2010 Debt Limit Event
Period:
[Refer to PDF for image: line graph]
Percentage points:
Preevent:
Date: 5/7/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.24;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.24;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.23;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17.
Date: 5/19/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.22;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.07;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.05;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.26.
Date: 6/1/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.22;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.19;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14.
Date: 6/11/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.33;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.22;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.43.
Date: 6/23/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.25;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13.
Date: 7/6/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.26;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14.
Date: 7/16/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12.
Date: 7/28/09;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.16;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13.
End Preevent, begin Event:
8/7/09: Treasury notifies Congress that the debt limit need to be
raised;
Date: 8/7/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08.
Date: 8/19/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14.
Date: 8/31/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11.
Date: 9/11/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1.
Date: 9/23/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.03;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12.
Date: 10/5/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13.
Date: 10/16/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17.
Date: 10/28/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16.
Date: 11/9/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2.
Date: 11/20/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15.
Date: 12/3/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17.
Date: 12/15/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14.
Date: 12/28/09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.07;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15.
Date: 1/11/10;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1.
Date: 1/22/10;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09.
Date: 2/3/10;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
2/12/10: Debt limit increase is signed into law:
End Event; Begin Postevent:
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06;
Average: 0.125;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1.
Date: 2/16/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06.
Date: 2/26/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.07;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.05;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.07;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.04;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.07.
Date: 3/10/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.05;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.04;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.04;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.03;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.05;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.06.
Date: 3/22/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.1;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.16;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17.
Date: 4/5/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.09;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.08;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.13.
Date: 4/15/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.11;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.12;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.14;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.24.
Date: 4/27/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.21;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.17;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.2;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.15;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.19;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.24.
Date: 5/7/10;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.18;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.33;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.21;
Average: 0.126;
Yield spread on 3-month commercial paper and 3-month Treasury bills:
0.33.
Source: GAO analysis of Board of Governors of the Federal Reserve
System data.
[End of figure]
After controlling for other factors that could have affected the yield
spread, such as economic uncertainty and liquidity in the bill market,
we estimated that the debt limit added a premium of about 4 basis
points during the debt limit event period in 2009-2010.[Footnote 20]
Applying this premium to all 3-month Treasury bills issued during this
period, we estimate that Treasury paid $78 million in additional
borrowing costs as a result of the debt limit. We did not estimate the
effects of nearing the debt limit on yields of other Treasury
securities, and therefore, do not know whether a premium of the same
size would apply to Treasury securities with longer or shorter terms
to maturity. However, even a smaller premium on the large amount of
Treasury securities issued during the debt limit event period would
result in a notable increase in borrowing costs. For instance, for
each additional basis point paid on bills issued during the 2009-2010
debt limit event period, Treasury's borrowing cost would increase by
roughly $92 million. For more information on our statistical analysis,
including a discussion of the limitations, see appendix II.
Trends in U.S. credit default swap (CDS) premia, or "spreads" also
suggest that Treasury securities were perceived to be relatively more
risky as debt approached the limit in 2009-2010. While CDS are used in
a variety of ways other than to insure against a default, such as
hedging against counterparty risk, or the risk that another party will
not fulfill its contractual obligation, CDS spreads serve as an
indicator of changes in the market's perception of risk in the
Treasury market. U.S. CDS spreads increase when risk in the Treasury
market increases and therefore have an inverse relationship to the
yield spread. Not long after Treasury notified Congress that the debt
limit would need to be raised and Treasury began reducing issuance of
CM bills for the SFP, rates for 1-year U.S. CDS spreads increased
while the yield spread trend decreased (see figure 6). Similarly, in
the weeks after the debt limit was raised, U.S. CDS spreads decreased
sharply. Some of the increase in U.S. CDS spreads likely reflects
general uncertainty about the global economy rather than the debt
limit; CDS spreads increased from late 2009 to early 2010 for other
countries, not just for the United States. However, 1-year U.S. CDS
spreads tended to increase more than comparable spreads for Germany,
Japan, and the United Kingdom during the event period.
Figure 6: Trends in Treasury's CDS Spreads Were Consistent with Trends
in the Yield Spread for the 2009-2010 Debt Limit Event:
[Refer to PDF for image: multiple line graph]
Graph depicts the following:
For the time period 5/7/2009 through 5/12/2010:
1-year Treasury CDS spread (in basis points);
Yield spread on 3-month commercial paper and 3-month Treasury bills
(in percentage points).
CDS basis points range: from 0 to 50;
Yield spread on 3-month commercial paper and 3-month Treasury bills
percentage points range: from 0 to 0.5.
The following kep dates are depicted on the graph:
8/7/09: Treasury notifies Congress that the debt limit needs to be
raised.
9/24/09: SFP reduction begins.
2/4/10: Legislation increasing debt limit is passed by Congress.
2/12/10: Debt limit increase is signed into law.
[Specific data points are available upon request]
Source: GAO analysis of Board of Governors of the Federal Reserve
System data and Datastream data.
[End of figure]
Overall, the debt limit requires Treasury to deviate from its regular
debt management practices in ways that add uncertainty to the Treasury
market. While there are limitations to our analysis of changes to the
auction schedule and to Treasury yields around debt limit events,
collectively the analyses provide strong evidence that this
uncertainty does not come without a cost.
Beyond the immediate costs associated with the debt limit, some market
participants and others that we spoke with said that failing to raise
the debt limit in a timely manner added an additional level of risk to
the Treasury market. As noted above, the extraordinary actions provide
less borrowing capacity and therefore less time for debate about
raising the debt limit once debt reaches the limit than they have in
the past. Analysts and observers--including former congressional
staff--expressed concern that a miscalculation in when the debt limit
needs to be increased by could trigger a financial crisis in the
Treasury market. Despite these concerns, most of those whom we spoke
with believe that Congress will raise the debt limit before there is a
significant market disruption and that Treasury will continue to
successfully manage debt near the debt limit as it has done in the
past. Treasury continues to consistently borrow at relatively low
interest rates, and demand for Treasury securities both in the United
States and abroad remains strong during periods of economic
instability because of their liquidity and low risk.
Experts and Practices of Other Countries Offer Insights for Better
Linking Policy Decisions with Their Effect on Debt:
Tying Debt Limit Increases to Annual Budget Decisions Is Similar to a
Practice Used in Some Other Countries:
We spoke with participants, observers, analysts, and other experts
representing a range of backgrounds and political perspectives--
including former congressional staff, former CBO Directors, former
Treasury officials, and representatives from credit rating agencies--
about the role of the debt limit in the U.S. budget process. Many
cited the failure to link fiscal policy decisions to changes in the
debt limit as a weakness in the process. Credit rating agencies, for
example, consider a number of different factors when assigning a
credit rating to sovereign nations' debt, including the strength of
the national economy, overall levels of government debt, and monetary
policy, as well as the budgetary framework. While the debt limit has
not compromised the United States' AAA credit rating, credit rating
agencies expressed concern about separating the vote for spending
increases and revenue decreases that increase debt from the vote for
additional borrowing authority. One credit rating agency described
this delinking as a weakness in the U.S. budgetary framework.[Footnote
21] Another credit rating agency said that anything that has the
potential to delay the timely redemption of federal debt is viewed as
a negative.[Footnote 22] In 1996, Moody's Investors Service (Moody's)
indicated a possible downgrade for some Treasury securities with
interest payments coming due in part because Treasury had nearly
exhausted its options for managing debt near the debt limit. According
to Moody's this was the only time that Moody's has officially taken a
negative rating action related to U.S. Treasury securities. None of
the experts that we spoke with said that the existence of the debt
limit served to restrain spending and tax decisions prior to the debt
limit debate, but some believe the debt limit has served a useful
purpose in highlighting the growth of federal debt.
Many of the experts that we spoke with identified possible changes to
the legislative process that would better link decisions about fiscal
policy and debt. Some suggested that since the budget resolution
reflects aggregate fiscal policy decisions, it should be used to
consider the level of debt implied by those decisions. The budget
resolution generally sets out the level for spending, revenues, and
debt for the next fiscal year and the following 4 fiscal years. Some
favored a mechanism similar to what was House Rule XXVIII in the 111th
Congress. This rule provided for the automatic engrossment and
transmittal to the Senate of a joint resolution changing the debt
limit by the amount specified in the budget resolution. This joint
resolution was considered to have passed the House and was then sent
to the Senate. The Senate did not have a similar rule; it sometimes,
though not always, passed the joint resolution from the House, albeit
with a lag. In the last 16 years, this lag has ranged from 1 month to
more than 10 months. Others believed that this process still gave too
little visibility to the implications of spending and tax decisions on
federal debt and preferred separate votes on stand-alone legislation.
They too believed that the vote should be timed to go with the budget
resolution--in part to keep the link and in part to avoid reaching the
debt limit later in the year after the spending and tax decisions had
been made.
Opinions varied on how to address increases in borrowing needs not
contemplated in budget resolutions. The actual amount of debt can
differ from the amount anticipated in the budget resolution because of
newly enacted legislation or because of the automatic stabilizers
through which changes in the economy affect government spending and
revenue.[Footnote 23] Some supported a formal or informal process
whereby any legislation that would increase debt beyond that
envisioned in the resolution would contain a separate title raising
the debt limit by the appropriate amount. Congress took this approach
with three pieces of legislation enacted in 2008 and 2009 in response
to the financial market crisis and economic downturn. The Housing and
Economic Recovery Act of 2008, the Emergency Economic Stabilization
Act of 2008, and the Recovery Act each included a separate provision
increasing the debt limit. Some congressional observers pointed out
that while this would tie spending and revenue decisions to the debt-
level effect of those decisions, it would not address increases in
debt arising from an economic downturn. Some of the ideas for dealing
with both the policy linkage and any increases in debt driven by
economic conditions were (1) considering further changes to the debt
limit at the time that the annual mid-session review is released;
[Footnote 24] (2) setting aside a reserve fund in the budget
resolution for unanticipated borrowing needs; and (3) delegating
additional authority to Treasury to borrow for intrayear financing
needs that resulted from changes in the economy rather than direct
policy decisions.
The practice of approving borrowing authority in connection with
approval of the annual budget is used in other countries we examined.
For example, in Canada, the Ministry of Finance is provided with a
fixed amount that it is authorized to borrow for the fiscal year. When
necessary, the Ministry of Finance can request increased borrowing
authority from the executive branch of government to fund unforeseen
borrowing needs. In Sweden, the legislature approves borrowing
authority annually; however, it is limited to purpose--to finance
current deficits, provide loans, and redeem national debt, for
example--rather than by amount.
Some Suggested Delegating Broader Authority to Treasury--a Practice
Used in Some Other Countries:
Some budget experts and a former Treasury official said that Congress
could delegate authority to Treasury to borrow as needed to fund
congressionally approved expenditures subject to a periodic review.
They suggested that Congress could vote to renew this authority at the
start of a new Congress or a new legislative session. This would
preserve Congress's ability to have periodic debates over the current
path of federal debt, they argue, but change the trigger for debate
from proximity to the debt limit to another point in the legislative
process to minimize disruptions to debt management and the Treasury
market.
Providing finance departments with broad authority to borrow is
consistent with practices in four of the countries we examined. In the
United Kingdom, for example, the Treasury is given broad authority to
raise money in a manner it "considers expedient for the purpose of
promoting sound monetary conditions." In New Zealand, the Minister of
Finance is given similarly broad borrowing authority to borrow in the
public interest. However, when comparing borrowing authority across
countries, it is important to recognize that the division of power
between the legislative and executive branches varies among different
political systems. In parliamentary systems, the government is
generally formed by the political party that has the support of the
majority of the parliament; therefore, the interests of the
legislative and executive branches are likely more aligned, making the
delegation of borrowing authority more of a formality and not a
subject of extensive deliberation. Of the countries that we reviewed,
only Denmark has a fixed nominal debt limit that is raised through
legislation outside the annual budget process comparable to the U.S.
debt limit. According to a Danish official, the limit is set high
enough that it does not impede debt managers' ability to issue debt.
Some of those with whom we spoke said that tying delegation of
borrowing authority to a fixed nominal debt limit creates an action-
forcing event that draws attention to the growth in federal debt. They
noted that previous debt limit debates provided opportunities for the
Congress and the President to consider the implications of past fiscal
policy decisions on federal borrowing and sometimes played a role in
the enactment of budget process agreements intended to slow the growth
of future federal borrowing. For example, debt limit increases were
passed jointly with budget controls legislation five times between
1985 and 1997[Footnote 25]--and again in February 2010 with the
reenactment of a statutory pay-as-you-go, or PAYGO, rule.[Footnote 26]
Meanwhile, others said that the debate over debt limit increases
played a smaller role in fiscal policy discussions in recent years
than it had in the past and expected that the debt limit would not be
needed to trigger debate over federal borrowing in the future given
the already increasing attention to debt and deficits. In addition,
some thought that risks associated with the debates--such as the
potential for Congress to delay or to miscalculate the timing of a
debt limit increase given the small amount of borrowing capacity
provided by the extraordinary actions--outweighed the benefits.
Some Countries Have Mechanisms to Increase Attention to or Control
over Fiscal Policy Decisions That Lead to an Increase in Debt:
The United States is unusual among the countries we reviewed in using
the authorization of additional borrowing authority as an occasion to
draw attention to past fiscal policy decisions. Other countries that
we reviewed generally use fiscal rules or targets to increase
attention to or control over fiscal policy decisions that lead to an
increase in debt. Fiscal rules generally refer to permanent or
multiyear constraints on fiscal policy through simple numerical limits
on budgetary aggregates. For example, Switzerland has adopted a
constitutional "debt brake" that places a limit on expenditures that
is equal to the expected revenue for the year adjusted to reflect the
country's place in the current business cycle. Differences between
estimated and actual numbers are recorded in a separate account that
must by law be reduced if it reaches a certain level. Germany has
adopted a "golden rule" limiting net borrowing to the amount of
investment spending. Germany also approved a constitutional amendment
in 2009 requiring that structural deficits not exceed 0.35 percent of
gross domestic product (GDP)--or very close to balance.
Instead of budget constraints, some countries use debt targets to
establish an acceptable outcome for policymakers to work toward when
making fiscal policy decisions. These can be either statutory
requirements or political commitments by the current government. For
example, in part to keep debt at a sustainable level, Sweden targets a
net surplus of 1 percent of GDP over the course of the business cycle.
In New Zealand, the government is required to maintain debt at a
"prudent level" and set specific short-term and long-term targets for
meeting this goal. If the government deviates from these targets, the
Minister of Finance must explain the approach the government intends
to take to return to prudent levels. Members of the European Union
agreed to target a ratio of gross general government debt to GDP of 60
percent and budget deficits of not more than 3 percent of GDP. Several
nations have struggled to meet these targets in recent years, raising
concerns about the effectiveness of the current enforcement
mechanisms. In general, budget experts and other observers have noted
that the success of fiscal rules depends on effective enforcement
along with a sustained commitment by policymakers and the public.
Specific fiscal rules and targets used in other countries may not be
appropriate for the United States given differences in the national
economies and political institutions. Nevertheless, some of the fiscal
rules and targets that we reviewed shared some common features that
distinguish them from the U.S. debt limit and offer insights for
increasing attention to or control over fiscal policy decisions that
lead to an increase in debt. These rules and targets:
(1) measure debt in relation to the overall size of the economy (e.g.,
debt-to-GDP ratio),
(2) take into consideration whether the economy is in a period of
expansion or contraction,
(3) provide a near-term or medium-term debt target, as opposed to a
ceiling, for policymakers to work toward.
In recent years, several bipartisan and nonpartisan groups have
suggested the establishment of fiscal rules with one or more of these
features in the United States such as a debt-to-GDP target.[Footnote
27]
Conclusions:
The debt limit does not control or limit the ability of the federal
government to run deficits or incur obligations. Rather, it is a limit
on the ability to pay bills incurred. The decisions that create the
need to borrow are made separately from--and generally earlier than--
the decision about the debt limit. Debates surrounding the debt limit
may raise awareness about the federal government's current debt
trajectory and also provide Congress with an opportunity to debate the
fiscal policy decisions driving that trajectory. However, since this
debate generally occurs after tax and spending decisions have been
enacted into law, Congress has a narrower range of options to effect
an immediate change to fiscal policy decisions and hence to federal
debt.
Failure to raise the debt limit could lead to serious negative
consequences in the Treasury market and for the ability of the United
States to finance federal debt at the lowest cost over time. Any delay
in raising the debt limit that affects Treasury's regular and
predictable schedule of auctions can create uncertainty in the
Treasury market. So too some of the actions Treasury takes to manage
the amount of debt near the debt limit, such as reducing the size of
auctions, can compromise the certainty of supply that Treasury relies
on to achieve the lowest borrowing cost over time. This uncertainty
can, in turn, raise the cost of financing the federal debt. In
addition, managing debt near the debt limit diverts Treasury's limited
resources away from other cash and debt management issues at a time
when Treasury already faces significant challenges in lengthening the
average maturity of its debt portfolio, which reduces rollover risk
and uncertainty about future interest payments.
Recently--and for the foreseeable future--Treasury's actions take
place in the context of rapidly growing federal debt. Treasury's past
success at managing cash and debt when near or at the debt limit is no
guarantee that it can continue to manage successfully in the future
and may be misleading. Given the size of current and projected
borrowing needs, the extraordinary actions Treasury uses to manage
debt near or at the debt limit will be more limited in coming years.
As a result, once debt is at the debt limit, Congress will likely have
less time to debate raising the debt limit before there are
disruptions to government programs and services and to the Treasury
market.
Observers and participants with whom we spoke suggested that improving
the link between the spending and revenue decisions that increase the
need to borrow and changes in the debt limit would improve the
situation. Better alignment could be possible if decisions about the
debt level occur in conjunction with spending and revenue decisions as
opposed to the after-the-fact approach now used. This would help avoid
the uncertainty and disruptions that occur during debates on the debt
limit today. It might also facilitate efforts to change the fiscal
path by highlighting the implications of these spending and revenue
decisions on debt. This will be particularly important in coming years
as the federal government addresses the challenge of unsustainable
increases in federal debt.
Matter for Congressional Consideration:
The projections of a growing debt burden have raised concerns both in
Congress and in the public. Well-designed budget processes and metrics
can help as Congress and the President seek to address the federal
government's long-term fiscal challenge. The current design of the
debt limit does not engender or facilitate debate over specific tax or
spending proposals and their effect on debt. In addition, the
uncertainty it creates can lead to disruptions in the Treasury market
and in turn to higher borrowing costs. To avoid these potential
disruptions to the Treasury market and to help inform the fiscal
policy debate in a timely way, Congress should consider ways to better
link decisions about the debt limit with decisions about spending and
revenue. Such a process would build on the approach used in 2008 and
2009 when Congress passed and the President signed three laws that
were expected to increase borrowing with a corresponding increase in
the debt limit.[Footnote 28] This report presents a number of
approaches that could serve as a basis for better linking decisions
about spending and revenue with decisions about the debt limit.
Agency Comments:
We requested comments on a draft of this report from the Secretary of
the Treasury. Treasury officials told us they appreciated the in-depth
and careful analysis contained in the report. They also provided
technical comments, which we incorporated as appropriate.
We will send copies of this report to interested congressional
committees, the Secretary of the Treasury, and other interested
parties. We will also make copies available at no charge on the GAO
Web site at [hyperlink, http://www.gao.gov].
If you or your staff have any questions about this report, please
contact Susan J. Irving at (202) 512-6806 or irvings@gao.gov or Gary
T. Engel at (202) 512-3406 or engelg@gao.gov. Contact points for our
Offices of Congressional Relations and Public Affairs may be found on
the last page of this report. GAO staff who made major contributions
to this report are listed in appendix III.
Signed by:
Susan J. Irving:
Director for Federal Budget Analysis:
Strategic Issues:
Signed by:
Gary T. Engel:
Director Financial Management and Assurance:
[End of section]
Appendix I: Objectives, Scope, and Methodology:
Our objectives were to (1) describe the actions that the Department of
the Treasury (Treasury) has taken to manage debt near the debt limit
and challenges that arise, (2) analyze the effects that approaching
the debt limit had on the market for Treasury securities, including
Treasury's borrowing costs, and (3) in light of the disconnect between
the debt limit and the policy decisions that have an effect on the
size of federal debt, describe alternative triggers or mechanisms that
would permit consideration of the link between policy decisions and
the effect on debt when or before decisions are made.
To answer our first objective, we reviewed Treasury documents and
prior GAO reports describing the actions that Treasury has taken
during debt limit debates since 1995. We used publicly available data
including Treasury's Monthly Statement of Public Debt and federal
investment account statements to estimate the amount of potential
borrowing capacity these actions could provide in fiscal year 2010 in
comparison to previous years since 2002 in which debt approached the
debt limit. To identify challenges that might arise when managing debt
near the debt limit, we examined publicly available data from the
Daily Treasury Statement to identify trends in federal receipts and
expenditures, issuance and redemption of federal debt, and changes in
Treasury's operating cash balance. To understand how managing debt
near the debt limit affected agency operations, we reviewed documents
provided by Treasury, interviewed Treasury officials involved in both
the decision-making process and implementation of extraordinary
actions, and obtained estimates when possible of the time and staff
involved in planning for when the debt limit will be reached and
implementing the extraordinary actions. To assess the reasonableness
of Treasury estimates, we reviewed calendar appointments and other
supporting documents. However, we did not obtain sufficient supporting
documentation to independently verify Treasury's estimates. We were
also unable to independently verify the foregone opportunities that
Treasury identified, such as less time to analyze short-term financing
needs, in part because it is difficult to prove what would happen in
the absence of the debt limit event.
To identify the potential effects of approaching the debt limit on the
market for Treasury securities, we reviewed publicly available
Treasury documents such as minutes from meetings of the Treasury
Borrowing Advisory Committee and academic literature, and interviewed
Treasury officials. Our review covered the last 16 years (1995-2010)
in order to include a particularly disruptive debt limit debate in
1995-1996 that required Treasury to take a number of extraordinary
actions, as well as the most recent debt limit increase. In February
2010, we asked eight market experts including six primary dealers for
their general views on the effects of delays in raising the debt limit
on the market for Treasury securities and Treasury operations.
On the basis of this initial analysis, we performed the following: (1)
We used press releases, auction announcements, and historical auction
data since 1995 to identify instances when auctions or auction
announcements or both, were delayed as a result of the debt limit. We
compared the yields at postponed auctions with yields on Treasury
securities of the same maturity being sold in the secondary market to
estimate the effect of delaying auctions or auction announcements on
Treasury's borrowing costs. We discussed our methodology with Treasury
officials and staff at the Federal Reserve Bank of New York and
incorporated their suggestions and feedback when appropriate. (2) We
used data provided by Treasury to analyze changes in the amount of
Treasury bills outstanding during debt limit debates since 2002. (3)
We performed a multivariate regression analysis to estimate the effect
of the debt limit on Treasury's borrowing costs. See appendix II for
more details on our methodology used for estimating these costs and
limitations to our analysis. (4) In August 2010, we received written
or oral responses to a standard questionnaire from four primary
dealers and managers of a large mutual fund asking for their views on
postponed auctions, reductions in bills outstanding, Treasury's
extraordinary actions, and the general uncertainty related to the
timing of debt limit increases. To obtain a broader perspective on the
effects of approaching the debt limit on financial markets, we also
spoke with representatives from two of the three major rating
agencies, a major trade organization representing securities firms and
other financial institutions, a research and consulting firm for the
municipal bond market, and a State and Local Government Series
securities subscriber.
To examine the disconnect between the debt limit and the policy
decisions that have an effect on the size of federal debt, we
conducted a literature review and reviewed the legislative history of
laws increasing the debt limit. We began our review with the debt
limit increase enacted on December 12, 1985, because (1) this was the
first in a series of debt limit increases enacted in legislation
containing budget process reforms, and (2) after 1985, Congress
provided Treasury with its current authorities related to the G-Fund
and Civil Service Retirement and Disability Fund (CSRDF) thereby
changing the way in which Treasury manages debt near or at the debt
limit.
To identify and describe alternative triggers or mechanisms that would
permit consideration of the link between policy decisions and the
effect on debt when or before decisions are made, we reviewed articles
in academic journals and reports and other information published by
credit rating agencies and policy research organizations. We conducted
semistructured interviews with budget and legislative experts. To
ensure that we captured a broad range of perspectives, we sought to
include a minimum number of representatives from the following
categories: former congressional staff with experience working for one
or more of relevant committees (i.e., the House Budget Committee, the
House Committee on Ways and Means, the Senate Budget Committee, and
the Senate Finance Committee) or senior party leadership in the House
or Senate; former Congressional Budget Office or Office of Management
and Budget Directors; representatives from a broad range of policy
research organizations that focus on issues related to the federal
debt; and academics and other experts on the legislative process. We
interviewed a total of 17 budget and legislative experts representing
one or more of these different categories. Five of those interviewed
also had experience working at Treasury.
To put U.S. practices into perspective, we examined triggers and
mechanisms used by members of the Organisation for Economic Co-
operation and Development (OECD)--an international organization
comprised of countries committed to democracy and market-based
economies. To identify countries for review, we analyzed countries'
responses to the 2007 OECD International Database of Budget Practices
and Procedures Survey, particularly questions asking about debt,
deficit, and other fiscal rules used when developing a budget.
[Footnote 29] We selected countries for further review based on their
responses to the survey as well as our review of reports by the OECD
and the International Monetary Fund. We sought to include countries
that provided a range of different mechanisms for (1) monitoring or
controlling debt and (2) delegating borrowing authority from their
legislature to debt managers. Our selection was limited to those
countries with information on fiscal rules and borrowing authority
available in English. We chose seven countries for further review:
Canada, Denmark, Germany, New Zealand, Switzerland, Sweden, and the
United Kingdom. We contacted representatives from budget offices, debt
management offices, or supreme audit institutions in each of these
countries for additional information on their respective country's
fiscal rules and borrowing authority. While selected countries offer
illustrative examples, their experiences are not always applicable to
the United States given differences in political systems and economies.
In order to assess the reliability of the data used in this report,
including proprietary data from Datastream and IHS Global Insight and
publicly available data from Treasury and the Federal Reserve, we
examined the data to look for outliers and anomalies and, when
possible, compared data from multiple sources for consistency. In
general, we chose databases that were commonly used by Treasury and
researchers to monitor changes in federal debt and related
transactions. Where possible and appropriate, we corroborated the
results of our data analysis with other sources. On the basis of our
assessment, we believe the data are reliable for the purpose of this
review.
We conducted our work from December 2009 to January 2011 in accordance
with generally accepted government auditing standards. Those standards
require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings
and conclusions based on our audit objectives. We believe that the
evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
[End of section]
Appendix II: Detailed Methodology and Findings of Statistical Analysis
of Treasury Borrowing Costs near the Debt Limit:
To measure changes in the Department of the Treasury's (Treasury)
borrowing costs when debt is approaching the debt limit, we performed
a multivariate regression analysis of yields paid on 13-week (i.e., 3-
month) Treasury bills issued during the five debt limit events
beginning in fiscal year 2002. For our purposes, a debt limit event
begins when the Secretary notifies Congress that the debt limit needs
to be raised and ends when legislation increasing the debt limit is
signed into law. The dependent variable in our analysis is the spread,
or difference, between yields on 3-month Treasury bills and yields on
3-month commercial paper. We used yield spreads during the preevent
period 3 months prior to the Secretary's letter as a benchmark against
which yields during the event can be compared. A narrowing of the
spread indicates that the market perceives the relative risk of
Treasury bills to be closer to that of commercial paper, increasing
their cost to Treasury. Conversely, a widening of the spread indicates
that the market perceives the relative risk of Treasury bills to be
less than that of commercial paper, making them less costly to
Treasury. We regressed the yield spread on key variables affecting
risk and liquidity of the financial market. Our results suggest that
Treasury paid a premium ranging from 1 to 4 basis points on 3-month
Treasury bills issued during debt limit events in 2001-2002, 2002-
2003, and most recently in 2009-2010.[Footnote 30] However, we did not
observe premiums in 2004-2005 and 2005-2006.
Variables and Model Specifications:
The existing literature on the effect of the debt limit on Treasury's
borrowing costs is limited. Our analysis was based in part on a prior
study of the effect of debt limit events on Treasury interest rates by
Liu, Shao, and Yeager.[Footnote 31] Similar to the results of our
analysis, Liu et al. (2009) found that during debt limit events in
2001-2002 and 2002-2003, the spread between 3-month Treasury bill
yields and 3-month commercial paper yields narrowed, implying that
Treasury bills were relatively more costly during this period;
however, this relationship was not observed in either the 2004-2005 or
2005-2006 debt limit events. The authors hypothesized that, during
these latter two debt limit events, investors may have assumed based
on past experience that members of Congress would resolve their
differences before there were any serious disruptions in the Treasury
market and therefore did not charge a premium on securities issued
near the debt limit. We also reviewed an earlier study by Nippani,
Liu, and Schulman which found that Treasury paid a premium on 3-month
and 6-month Treasury bills issued during the debt limit event in 1995-
1996.[Footnote 32]
On the basis of discussions with Treasury officials, staff at the
Federal Reserve Bank of New York, and market experts such as primary
dealers and larger investment funds, we determined that the model
developed by Liu et al. (2009) provided a reasonable starting point
for our analysis. We made modifications to reflect the Federal
Reserve's purchases of commercial paper through its Commercial Paper
Funding Facility (CPFF) in 2008 to enhance the liquidity of the
commercial paper market. We included a variable to control for the
volume of commercial paper held by the Federal Reserve as a percentage
of total commercial paper outstanding. We also included the Chicago
Board Options Exchange's Volatility Index (VIX) to control for
volatility and uncertainty in financial markets. The equation we used
is:
Yield Spread = B0 + B1*EVENT + B2*POST + B3*SPRET + B4*VIX +
B5*LOG(CPISSUE) + B6*LOG(TBILLTRANS) + B7*CPFFSHARE + Error:
Table 5 below describes each of the variables in the equation and
indicates the expected sign of the coefficient. Treasury officials and
market experts said that our modifications to their equation were
reasonable.
Table 5: Variables Used in Multivariate Regression:
Variable name: EVENT;
Variable description: This variable is equal to 1 during the event
period and is 0 otherwise. Consistent with the academic literature, we
defined the debt limit event period as beginning when Treasury
notified Congress that the debt limit needed to be raised and as
ending when the legislation is passed to raise the debt limit. We
expect the variable's coefficient to be negative because debt limit
events may raise the perceived risk of Treasury securities and reduce
the yield spread compared to the preevent period.
Variable name: POST;
Variable description: This variable is equal to 1 in the 3-month
postevent period and is otherwise 0. A negative coefficient indicates
that the perceived increased risk of Treasury securities persisted
beyond the end of the debt limit event period.
Variable name: SPRET;
Variable description: This variable is the daily return of the
Standard & Poor's 500 index and is an indicator of the market's
assessment of economic activity. We expect the variable's coefficient
to be negative because a stronger economy should reduce the default
risk of commercial paper, lower its yield, and reduce the yield
differential. The data for this variable were downloaded from Thomson
Reuters' proprietary statistical database Datastream.
Variable name: VIX;
Variable description: This variable represents the market expectations
of volatility over the next 30-day period and is calculated by the
Chicago Board Options Exchange using Standard & Poor's 500 stock index
option bid/ask quotes. The variable is intended to control for
volatility and uncertainty in financial markets. We expect the
coefficient to be positive because increased financial market
uncertainty should cause investors to move from private-sector
securities into Treasury securities and reduce Treasury yields
relative to other securities. The data for this variable were
downloaded from proprietary data provider IHS Global Insight.
Variable name: LOG(CPISSUE);
Variable description: This is the natural log of weekly AA commercial
paper issues with a maturity greater than 80 days and is intended to
control for the liquidity of commercial paper. Weekly data for this
variable were downloaded from the Board of Governors of the Federal
Reserve System Web site and then back-filled to obtain daily
values.[A] We expect the variable's coefficient to be negative because
an increase in liquidity should raise commercial paper prices, lower
their yields, and thus decrease the yield spread.
Variable name: LOG(TBILLTRANS);
Variable description: This is the natural log of weekly transactions
in Treasury bills traded among primary dealers and is intended to
control for the liquidity of Treasury bills. Weekly data for this
variable were downloaded from the Federal Reserve Bank of New York's
Web site and then back-filled to obtain daily values.[B] We expect the
variable's coefficient to be positive because an increase in the
liquidity of Treasury bills should reduce their yield and increase the
yield spread.
Variable name: CPFFSHARE;
Variable description: This variable is the volume of commercial paper
held by Federal Reserve through its Commercial Paper Funding Facility
(CPFF) as a percentage of total commercial paper outstanding. The
variable is intended to control for temporary increases in liquidity
from CPFF purchases. The CPFF was initiated in October 2008 to enhance
the liquidity of commercial paper during the financial crisis. Because
this facility did not exist during previous debt limit events, it is
included only in the 2009-2010 debt limit regression. Weekly data for
this variable were downloaded from the Board of Governors of the
Federal Reserve System Web site and then back-filled to obtain daily
values.[C] We expect the variable's coefficient to be negative because
increases in liquidity should raise commercial paper prices and reduce
yields, thus decreasing the yield spread.
Source: GAO analysis.
[A] Downloaded from [hyperlink,
http://www.federalreserve.gov/DataDownload/Choose.aspx?rel=CP] on May
17, 2010.
[B] Downloaded from [hyperlink,
http://www.newyorkfed.org/markets/statrel.html] on May 18, 2010.
[C] Downloaded from [hyperlink,
http://www.federalreserve.gov/datadownload/Choose.aspx?rel=H41] on May
13, 2010.
[End of table]
The regression results based on the variables listed above are
presented in table 6. Negative EVENT coefficients suggest that debt
limit events reduced the spread between commercial paper and Treasury
yields compared to the preevent period. Consistent with the Liu et al.
(2009) study, the EVENT coefficients for the debt limit events in 2001-
2002 and 2002-2003 had the expected negative sign and were
statistically significant.[Footnote 33] In addition, the coefficient
of the EVENT variable had the expected negative sign and was
statistically significant for the most recent debt limit event in 2009-
2010.
Table 6: Regression Results for Debt Limit Event Periods from Fiscal
Year 2002 to 2010:
CONSTANT;
Expected sign: n.a.;
2001-2002: Event 1 coefficients: -0.540;
2002-2003: Event 2 coefficients: 0.246;
2004-2005: Event 3 coefficients: 1.395;
2005-2006: Event 4 coefficients: -0.595;
2009-2010: Event 5 coefficients: -0.642.
EVENT;
Expected sign: -;
2001-2002: Event 1 coefficients: -0.011 [gray] [bold];
2002-2003: Event 2 coefficients: -0.038 [gray] [bold];
2004-2005: Event 3 coefficients: 0.025 [gray];
2005-2006: Event 4 coefficients: -0.006 [bold];
2009-2010: Event 5 coefficients: -0.040 [gray] [bold].
POST;
Expected sign: -/+;
2001-2002: Event 1 coefficients: -0.011 [bold];
2002-2003: Event 2 coefficients: -0.041 [gray] [bold];
2004-2005: Event 3 coefficients: 0.034 [gray] [bold];
2005-2006: Event 4 coefficients: 0.034 [gray] [bold];
2009-2010: Event 5 coefficients: -0.030 [bold].
SPRET;
Expected sign: -;
2001-2002: Event 1 coefficients: -0.268 [gray] [bold];
2002-2003: Event 2 coefficients: -0.433 [gray] [bold];
2004-2005: Event 3 coefficients: -1.092 [gray] [bold];
2005-2006: Event 4 coefficients: 0.579;
2009-2010: Event 5 coefficients: 0.083.
VIX;
Expected sign: +;
2001-2002: Event 1 coefficients: 0.000;
2002-2003: Event 2 coefficients: -0.002 [gray];
2004-2005: Event 3 coefficients: -0.008 [gray];
2005-2006: Event 4 coefficients: 0.005 [gray] [bold];
2009-2010: Event 5 coefficients: 0.005 [gray] [bold].
LOG(CPISSUE);
Expected sign: -;
2001-2002: Event 1 coefficients: 0.014 [gray];
2002-2003: Event 2 coefficients: -0.001 [bold];
2004-2005: Event 3 coefficients: 0.006;
2005-2006: Event 4 coefficients: -0.010 [gray] [bold];
2009-2010: Event 5 coefficients: 0.007 [gray].
LOG(TBILLTRANS);
Expected sign: +;
2001-2002: Event 1 coefficients: 0.052 [gray] [bold];
2002-2003: Event 2 coefficients: -0.004;
2004-2005: Event 3 coefficients: -0.107 [gray];
2005-2006: Event 4 coefficients: 0.084 [bold];
2009-2010: Event 5 coefficients: 0.060 [gray] [bold].
CPFFSHARE;
Expected sign: -;
2001-2002: Event 1 coefficients: n.a.;
2002-2003: Event 2 coefficients: n.a.;
2004-2005: Event 3 coefficients: n.a.;
2005-2006: Event 4 coefficients: n.a.;
2009-2010: Event 5 coefficients: -0.515 [gray] [bold].
Number of variables with correct sign and are statistically
significant;
2001-2002: Event 1 coefficients: 3;
2002-2003: Event 2 coefficients: 3;
2004-2005: Event 3 coefficients: 2;
2005-2006: Event 4 coefficients: 3;
2009-2010: Event 5 coefficients: 4.
Adjusted R-squared;
2001-2002: Event 1 coefficients: 0.102;
2002-2003: Event 2 coefficients: 0.241;
2004-2005: Event 3 coefficients: 0.228;
2005-2006: Event 4 coefficients: 0.029;
2009-2010: Event 5 coefficients: 0.187.
Source: GAO analysis.
Note: The bolded coefficients have the correct signs while the cells
highlighted in gray are the estimated coefficients that are
statistically significant at least at the 10 percent level.
n.a.= not applicable.
[End of table]
We explored a variety of alternative specifications to see whether our
model could be improved but found that no specification proved
particularly robust across all the events studied. On the basis of
discussions with Treasury staff and market experts, we added,
replaced, and removed variables and defined the event period
differently. For example, we explored several alternative controls for
credit risk, including the spread between London interbank offer rate
(LIBOR) and the Overnight Indexed Swap that measures risk and
liquidity in the money market, and the spread between Baa corporate
bond yields and 10-year Treasury note yields. We also redefined the
event period to begin when Treasury took its first extraordinary
actions or, in the case of 2009-2010, reduced the amount in the
Supplementary Financing Program (SFP). None of the alternative
variables or specifications produced statistically significant
results. Using the Liu et al. (2009) specifications did not result in
statistically significant results for the most recent event in 2009-
2010.
Effect of Debt Limit on Borrowing Costs:
One the basis of our analysis, we estimate that Treasury paid $78
million in additional interest costs for newly issued 3-month
securities issued during the 2009-2010 debt limit event period. We
arrived at this estimate by translating the coefficient of the EVENT
in 2009-2010 (-.040) to basis points (4) and multiplying by the amount
of 3-month Treasury bills issued during the event period. We selected
3-month bills for our analysis because it is a commonly used benchmark
in economic indicators such as the TED spread--a key indicator of
credit risk.[Footnote 34] We did not estimate the effects of the debt
limit on other Treasury securities with longer terms to maturity in
part because of a lack of reliable data on yields for private-sector
fixed-income assets with maturity dates comparable to medium-term
Treasury securities such as a 2-year note. Therefore, we do not know
whether the same 4-basis-point premium would apply to other Treasury
securities with longer or shorter terms to maturity issued during the
debt limit event period. Nippani et al. (2001) found that the effect
of the debate over the debt limit in 1995-1996 was greater on 3-month
Treasury bills than on 6-month Treasury bills, indicating the
investors may have believed that debate over the debt limit would be
resolved over time and that longer-dated securities would therefore be
less affected. However, even a smaller premium when applied to the
large amount of Treasury securities offered by Treasury would result
in a notable increase in borrowing costs. For instance, for each
additional basis point paid on bills issued during the 2009-2010 debt
limit event period, Treasury's borrowing cost would increase by
roughly $92 million.
Limitations of the Analysis:
There are a number of limitations to using a multivariate regression
to measure changes in Treasury's borrowing costs. First, the results
of our analysis explain only a small portion of the variation in the
yield spread, as indicated by the relatively low R-squared statistics.
Any equation attempting to explain the yield spread would have limited
explanatory power given inherent randomness in daily time series data
such as Treasury bill and commercial paper yields. Furthermore, the
estimates are subject to omitted variable bias. Second, there was
substantial variation in the sign, size, and significance of the
estimated coefficients across debt limit events. However, the EVENT
variable's coefficient, which is the central focus of our analysis,
had the expected negative sign in four of the five debt limit periods
included in our analysis and was significant in three of these
periods. We discussed these and other limitations with Treasury
officials, staff at the Federal Reserve Bank of New York, and other
market experts such as primary dealers and incorporated their
suggestions and feedback when appropriate. Despite these limitations,
the estimates do suggest that a debt limit event may result in a
premium.
[End of section]
Appendix III: GAO Contacts and Staff Acknowledgments:
GAO Contacts:
Susan J. Irving, (202) 512-6806 or irvings@gao.gov:
Gary T. Engel, (202) 512-3406 or engelg@gao.gov.
Staff Acknowledgments:
In addition to the contacts named above, Melissa Wolf, Assistant
Director; Dawn Simpson, Assistant Director; Thomas J. McCabe, analyst-
in-charge; Richard Krashevski, Claire Li, Inna Livits, and Nicole
McGuire made key contributions to this report.
[End of section]
Footnotes:
[1] A very small amount of total federal debt is not subject to the
debt limit. This amount is primarily comprised of unamortized
discounts on Treasury bills and Zero Coupon Treasury bonds; debt
securities issued by agencies other than Treasury, such as the
Tennessee Valley Authority; and debt securities issued by the Federal
Financing Bank. As of September 30, 2010, 99.5 percent of federal debt
was subject to the debt limit.
[2] Budget resolutions are concurrent resolutions, which are not
presented to the President for his signature and do not become law.
Therefore, debt limit increases must be passed as part of separate
legislation such as a bill or joint resolution.
[3] This was House Rule XXVIII in the 111th Congress but was not
included in the House Rules for the 112thCongress.
[4] Pub. L. No. 111-5. The other acts were the Housing and Economic
Recovery Act of 2008 (Pub. L. No. 110-289) and the Emergency Economic
Stabilization Act of 2008 (Pub. L. No. 110-343).
[5] We have issued a number of prior reports on Treasury's efforts to
achieve this goal including more recently Debt Management: Treasury
Was Able to Fund Economic Stabilization and Recovery Expenditures in a
Short Period of Time, but Debt Management Challenges Remain,
[hyperlink, http://www.gao.gov/products/GAO-10-498] (Washington, D.C.:
May 18, 2010), Debt Management: Treasury Inflation Protected
Securities Should Play a Heightened Role in Addressing Debt Management
Challenges, [hyperlink, http://www.gao.gov/products/GAO-09-932]
(Washington, D.C.: Sept. 29, 2009), and Debt Management: Treasury Has
Refined Its Use of Cash Management Bills but Should Explore Options
That May Reduce Cost Further, [hyperlink,
http://www.gao.gov/products/GAO-06-269] (Washington, D.C.: Mar. 30,
2006).
[6] See [hyperlink, http://www.gao.gov/special.pubs/longterm/debt/],
which updates information in Federal Debt: Answers to Frequently Asked
Questions: An Update, [hyperlink,
http://www.gao.gov/products/GAO-04-485SP] (Washington, D.C.: Aug. 12,
2004).
[7] Treasury generally uses CM bills to finance intramonth funding
gaps due to timing differences of large cash inflows and outflows but
has also used them in recent years to raise funds for the
Supplementary Financing Program--a temporary program begun in 2008 to
assist the Federal Reserve with its monetary policy.
[8] For information on the costs associated with issuing CM bills, see
[hyperlink, http://www.gao.gov/products/GAO-06-269] and [hyperlink,
http://www.gao.gov/products/GAO-10-498].
[9] Pub. L. No. 104-103 (Feb. 8, 1996).
[10] For additional information on the costs associated with past
transactions, see GAO, Debt Ceiling: Analysis of Actions Taken during
the 2003 Debt Issuance Suspension Period, [hyperlink,
http://www.gao.gov/products/GAO-04-526] (Washington, D.C.: May 20,
2004).
[11] Treasury also called back compensating balances after a
cancellation of a 4-week bill auction the week of September 11, 2001,
to help meet its obligations on time.
[12] Interest payments on debt held in government accounts are
credited to government accounts and do not require additional
borrowing from the public, but the investment of these interest
payments is subject to the debt limit.
[13] The increase in nonmarketable debt on December 31, 2009, includes
roughly $18 billion in interest payments on debt held by the CSRDF
trust fund that Treasury could choose to suspend during a DISP.
[14] Cash held in the Treasury's Supplementary Financing Program
account is excluded because it has not been used to finance federal
expenditures.
[15] The amount of disinvestment of securities held by CSRDF is
determined based on the length of the declared DISP, which affects the
total amount of the available extraordinary actions.
[16] [hyperlink, http://www.gao.gov/products/GAO-06-269].
[17] Our methodology was based on a prior academic study: Pu Liu,
Yingying Shao, and Timothy J. Yeager, "Did the repeated debt ceiling
controversies embed default risk in U.S. Treasury securities?" Journal
of Banking & Finance, vol. 33 (2009): 1464-1471. The regression
specification we used for the recent debt limit was different from the
Liu et al. regression specification because the financial markets were
operating in unique economic conditions (e.g., the Federal Reserve
began purchasing commercial paper in 2008 to enhance the liquidity of
the commercial paper market). We also reviewed a similar study:
Srinivas Nippani, Pu Liu, and Craig T. Schulman, "Are Treasury
Securities Free of Default?" Journal of Financial and Quantitative
Analysis, vol. 36, no. 2 (2001): 251-265. This study also found that
the market charged a premium on Treasury securities issued during the
debt limit debate in 1995-1996. See appendix II for more information.
[18] For the purposes of this study, a debt limit event period begins
when Treasury first warns of the need to raise the debt limit and ends
when legislation to raise the limit is passed. For the first four debt
limit events, we use the same dates used by Liu et al. (2009). For the
fifth debt limit event in 2009-2010, the event period began when
Treasury notified Congress that the debt limit needed to be raised and
ended when legislation increasing the debt limit was signed into law.
[19] This is consistent with the Liu et al. (2009). The authors of the
study hypothesized that during these latter two debt limit events,
investors may have assumed based on past experience that members of
Congress would resolve their differences before there were any serious
disruptions in the Treasury market and therefore did not charge a
premium on securities issued near the debt limit.
[20] The 95 percent confidence interval of the premium estimate ranges
from 1.0 to 7.1 basis points.
[21] Standard & Poor's, The U.S. Debt Ceiling: As Headroom Shrinks,
It's Time to Raise the Room Beams (Oct. 9, 2009).
[22] Moody's, U.S. Statutory Debt Limit to be Raised; Longer-Term
Fiscal Strategy the Real Question (September 2009).
[23] Automatic stabilizers are provisions built into the structure of
the federal budget that alter tax or spending levels based on economic
fluctuations without any explicit government action.
[24] Under 31 U.S.C. § 1106, the President is required to submit an
update of the federal budget, often referred to as a mid-session
review, before July 16 of each year.
[25] Congress reached agreement on budget procedures in the Balanced
Budget and Emergency Deficit Control Act of 1985 (Pub. L. No. 99-177),
Balanced Budget and Emergency Deficit Control Reaffirmation Act of
1987 (Pub. L. No. 100-119), the Budget Enforcement Act of 1990 (Pub.
L. No. 101-508), the extended budget process provisions of the Budget
Enforcement Act of 1990 (Pub. L. No. 103-66), and the Budget
Enforcement Act of 1997 (Pub. L. No. 105-33).
[26] PAYGO is a procedure requiring that the aggregate effect of
increases in mandatory spending or reductions in revenue generally be
offset (Pub. L. No. 111-139).
[27] These are the National Commission on Fiscal Responsibility and
Reform, The Moment of Truth: Report of the National Commission on
Fiscal Responsibility and Reform (Washington, D.C.: December 2010);
Bipartisan Policy Center, Restoring America's Future: Reviving the
Economy, Cutting Spending and Debt, and Creating a Simple Pro-Growth
Tax System (Washington, D.C.: November 2010); National Research
Council and National Academy of Public Administration, Choosing the
Nation's Fiscal Future (Washington, D.C.: 2010); and Peterson-Pew
Commission on Budget Reform, Red Ink Rising: A Call to Action to Stem
Mounting Federal Debt (Washington, D.C.: December 2009).
[28] The American Recovery and Reinvestment Act of 2009 (Pub. L. No.
111-5), the Housing and Economic Recovery Act of 2008 (Pub. L. No. 110-
289) and the Emergency Economic Stabilization Act of 2008 (Pub. L. No.
110-343).
[29] See [hyperlink, http://www.oecd.org/gov/budget/database] for the
results of the survey.
[30] One basis point is equal to 1/100th of 1 percent. Thus, 4 basis
points is 0.04 percent.
[31] Pu Liu, Yingying Shao, and Timothy J. Yeager, "Did the repeated
debt ceiling controversies embed default risk in U.S. Treasury
securities?" Journal of Banking and Finance, vol. 33 (2009): 1464-1471.
[32] Srinivas Nippani, Pu Liu, and Craig T. Schulman, "Are Treasury
Securities Free of Default?" Journal of Financial and Quantitative
Analysis, vol. 36, no. 2 (2001): 251-265.
[33] Because these are estimates, a test of statistical significance
attempts to rule out an effect purely attributable to chance. Our
criterion for statistical significance is that there is less than a 10
percent probability of rejecting the null hypothesis that the
coefficient is zero when the null hypothesis is true.
[34] The TED spread is the difference between the 3-month LIBOR rate
and 3-month Treasury yield.
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Phone:
The price of each GAO publication reflects GAO‘s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO‘s Web site,
[hyperlink, http://www.gao.gov/ordering.htm].
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional
information.
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548: