BANK CAPITAL, THE ANNOUNCEMENT OF THE FEDERAL RESERVE'S TERM AUCTION FACILITY, AND BANK BORROWING DURING THE FINANCIAL CRISIS OF 2007–2008

• Published date: 01 December 2018
• Date: 01 December 2018
• DOI: http://doi.org/10.1111/jfir.12160

BANK CAPITAL, THE ANNOUNCEMENT OF THE FEDERAL RESERVE’S

TERM AUCTION FACILITY, AND BANK BORROWING DURING

THE FINANCIAL CRISIS OF 2007–2008

Mujtaba Zia

Southern Arkansas University

Imre Karafiath and Niranjan Tripathy

University of North Texas

Abstract

During the financial crisis of 2007–2008, the Federal Reserve System created

emergency lending facilities to address the liquidity problem in financial markets and

alleviate the stigma attached to borrowing from its traditional lending facility, the

discount window. The most important and commonly used facility was the Term

Auction Facility (TAF) announced December 12, 2007. Using event-study methodol-

ogy, we find that stock prices of TAF participating banks experienced a significant

2.84% average cumulative abnormal return around the announcement date. The

negative abnormal returns correlated with the peak loan balances reached on average

416 days after the announcement.

JEL Classification: G14, G21, G35

I. Introduction

The Federal Reserve System (Fed) is expected to play a central role in averting systemic

liquidity crisis, by providing member banks access to its emergency lending facility

(Flannery 1996). Member banks, however, often hesitate to use the Fed’s traditional

overnight lending facility, the discount window (DW), primarily because of the stigma

and costs associated with borrowing from the “lender of the last resort.”

1

Following the

onset of the financial crisis and credit crunch in the third quarter of 2007, and in response

to the perceived stigma associated with the DW, the Fed expanded the availability of

credit and funds by establishing unprecedented lending facilities in late 2007 and

provided unconventional funds to financial institutions in 2008 and 2009. The first

and most popular of these lending facilities was the Term Auction Facility (TAF),

announced December 12, 2007. However, subsequent studies (Cyree, Griffiths, and

Winters 2013, 2016; Blau, Hein, and Whitby 2016) have shown that banks that chose to

We thank Hanzhi Xu and Yi Zheng for invaluable research assistance. We are grateful to two anonymous

referees for their excellent comments and suggestions.

1

Borrowing from the Fed DW has traditionally been considered a sign of poorly managedliquidity risk and in

some cases triggered bank audit and scrutiny. See also Section II.

The Journal of Financial Research Vol. XLI, No. 4 Pages 485–506 Winter 2018

DOI: 10.1111/jfir.12160

485

© 2018 The Southern Finance Association and the Southwestern Finance Association

participate in the TAF were not able to avoid the stigma—in terms of negative stock price

response—associated with emergency borrowing from the Fed.

For the purpose of this study, it is important to note that the data on amounts of

funds and details about borrowers were not just undisclosed by the Fed; they did not exist

at the time of the TAF announcement. As the program was being implemented—shortly

after the crisis intensified in the early part of 2008—the magnitude and conditions of

credit extended to banking firms through these facilities came under the spotlight of news

agencies. On May 2, 2008, Bloomberg LP, the parent company of Bloomberg News,

requested information from the Fed on the credit amounts extended to financial firms.

The Fed refused to provide the information, claiming that the data were sensitive and

would negatively affect the financial industry. On November 7, 2008, Bloomberg LP

filed a lawsuit against the Fed and Clearing House Association LLC (a group of largest

banks) based on the Freedom of Information Act. Bloomberg won the case in lower

courts, but the Fed and Clearing House Association LLC fought Bloomberg’s lawsuit up

to the U.S. Supreme Court. The Supreme Court declined to hear the Fed’s appeal in

March 2011. Consequently, in late 2011, the Fed had to release to Bloomberg the details

of loans extended to financial institutions that tapped one or more of the lending facilities

during the financial crisis. This data set, which we use in our study, contains daily

balances for each bank borrowing from various lending facilities of the Fed, including the

TAF, during the financial crisis.

Although these data did not exist at the time of the TAF announcement, the focus

of this study is to determine whether the market could anticipate and price the stigma

associated with TAF borrowing based on information contained in the data set. In this

sense, it is a joint test of the stigma of TAF lending and of the efficient market hypothesis

in the strong form as defined by Fama (1970). In the context of evidence provided by

other studies, this article attempts to shed light on the sequence in which new information

was reflected in the stock prices of banks participating in the TAF. The price effect could

appear during three chronologically ordered events: (1) announcement date, when the

program was first announced by the Fed but the data set did not exist; (2) borrowing date,

when banks actually borrowed through the program and when the data set was available

but not disclosed; and (3) disclosure date, when information about the banks and their

borrowing were eventually disclosed to the public.

Blau, Hein, and Whitby (2016) examine stock price reactions around borrowing

and disclosure dates, and their results show that although banks participating in the TAF

experienced a stock price decline during the six-day interval including and following the

borrowing dates, there was no additional stock price reaction during the disclosure date.

Based on standard event-study methodology, our article reports a significant 2.84%

abnormal return during the three-day interval centered on the TAF announcement date,

which predates the peak TAF borrowing of the participating banks by an average of

416 days. The long interval between borrowing and announcement dates and the negative

announcement-period price reaction are consistent with information-based trading and

the existence of a stigma for banks participating in the TAF program.

Using a cross-sectional regression designed to control for bank risk exposure

during the financial crisis, we also examine three competing and mutually exclusive

hypotheses by focusing on the relation between the observed announcement-period

486 The Journal of Financial Research