Bank net interest margins, the yield curve, and the 2007–2009 financial crisis

• Date: 01 January 2018
• Author: Peter V. Egly,André Varella Mollick,David W. Johnk
• DOI: http://doi.org/10.1002/rfe.1016
• Published date: 01 January 2018

ORIGINAL ARTICLE

Bank net interest margins, the yield curve, and the 2007–2009

financial crisis

Peter V. Egly

1

|

David W. Johnk

2

|

Andr

e Varella Mollick

3

1

Department of Economics and Finance,

University of Texas Rio Grande Valley,

Brownsville, TX, USA

2

Department of Business, Rogers State

University, Claremore, OK, USA

3

Department of Economics and Finance,

University of Texas Rio Grande Valley,

Edinburg, TX, USA

Correspondence

David W. Johnk, Department of Business,

Rogers State University, Claremore, OK,

USA.

Email: djohnk@rsu.edu

Abstract

Using quarterly call report data from 2000 to 2016, we reexamine the relationship

between net interest margins (NIM) and the yield curve for more than 5,500 U.S.

commercial banks. In the full sample, yield curve and RGDP growth have posi-

tive effects on NIM, while inflation and deposit-to-loan ratios (D/L) have negative

effects. Splitting the sample around the 2007–2009 crisis, we show the impact of

yield curve and RGDP growth on NIM increasing during the “recovery”(2009Q3

to 2016Q4), and inflation and D/L changing signs. Positive effects of yield curve

on profits vary with bank size and change over time.

JEL CLASSIFICATION

E44, G21

KEYWORDS

bank profitability, financial crisis, macroeconomics, panel data, yield curve

1

|

INTRODUCTION

By the summer of 2007, the fallout of the U.S. housing market led banks to report considerable losses and substantial write

downs of their real estate portfolios due to rising delinquencies and reductions in home values. Factors that contributed to

the softening of the housing market included increasing interest rates between 2003 and 2006 and a greatly reduced pool of

qualified homeowners.

1

Furthermore, during this time period, market capitalization of the major banks became significantly

depleted, contributing to the banking turmoil. The ensuing U.S. financial crisis led to unprecedented government interven-

tion by the U.S. Federal Reserve and the U.S. Treasury in an attempt to revive the financial system. During the financial

crisis period, bank lending was modest as banks focused on rebuilding liquidity on their balance sheets. Bank lending to

some degree was curtailed by their exposure in real estate portfolios. The effects of the 2007–2009 financial crisis have

subsided and bank profitability, for the most part, has been restored.

There are, of course, many factors that influence bank profitability. Not only business cycles but also institutional guide-

lines regarding lending and competition in the industry could be particularly important. The academic literature reviewed

below has evidence on bank profits responding to interest rates (and other macroeconomic factors), as well as to bank-

related measures, such as: assets, deposits to loan ratios, mortgage portfolios, etc. Perhaps one of the most relevant terms

behind bank profitability is the yield curve. The mechanism is well-known in the financial press, such as during the recent

U.S. presidential election of November 2016, which led to a substantial market upward response in bond yields: U.S. 10-

year benchmark, for example, moved from 1.7%–1.8% to about 2.3% in a couple of weeks, amid a sell-off in bonds and a

boom in all stock markets. Here is one very recent example summarizing the conventional wisdom on the yield curve and

bank profit margins: “A move higher in interest rates and a steeper yield curve ... suggest bank profits should improve.

More relief could be on the way if the Federal Reserve again increases short-term rates in December. And any move to

undo or lessen regulation could allow banks to return more capital to investors, which in turn would boost returns on

Received: 15 March 2017

|

Revised: 18 September 2017

|

Accepted: 29 November 2017

DOI: 10.1002/rfe.1016

12

|

©2018 The University of New Orleans wileyonlinelibrary.com/journal/rfe Rev Financ Econ. 2018;36:12–32.

equity. That would lead to higher profits and could argue for higher valuation multiples, some analysts believe”(Ensign,

2016). While often used in the financial press, there is not to–our knowledge–an effort of quantifying the yield curve on

bank profits.

The literature surrounding the financial crisis has taken various paths focusing on government intervention (e.g., Cec-

chetti, 2009; Taylor & Williams, 2009), balance sheet adjustments of financial intermediaries (e.g., Adrian & Sung, 2010;

He, Khang, & Krishnamurthy, 2010), bank lending and credit availability (e.g., Apergis & Christou, 2015; Cornett,

McNutt, Strahan, & Tehranian, 2011; Egly & Mollick, 2013; Kosak, Li, Loncarski, & Marinc, 2015), and consequences of

expansionary monetary policy for equity markets (e.g., Huang, Mollick, & Nguyen, 2016). Another stream of research re-

examines theories of financial intermediation (e.g., Acharya, Padersen, Philipon, & Richardson, 2010; Sh leifer & Vishny,

2010) and investigates equity markets’reaction to government regulation enacted in response to the financial crisis (e.g.,

Peristian, Morgan, & Savino, 2010; Subrahmanyam, Pennathur, & Smith, 2011). To acquire an overall economic perspec-

tive of the government intervention, Veronesi and Zingales (2010) calculate the costs (i.e., cost to tax-payers) and benefits

(i.e., increased value of banks’financial claims) of the government bailout and conclude it was an overall success. They

contend that from an economic viewpoint, the government intervention created value by preventing a run on banks and by

providing capital that reduced banks’inefficiencies related to excessive leverage.

An interesting area of research which has not received much attention deals with how economic forces impact bank

profitability after the 2007–2009 financial crisis. This paper examines the role of the yield curve on bank net interest mar-

gins (a measure of bank profitability) while controlling for other macro forces and bank-level characteristics. This study

explores the following research questions: (1) Does bank balance sheet structure and/or bank size play a role in banks’net

interest margin behavior in response to changes in the yield curve?, (2) Has bank profitability, captured through net interest

margins, been restored to precrisis levels?, and (3) What has been the impact of the yield curve on bank net margins during

a period of unprecedented government intervention and U.S. expansionary monetary policy? Examining the impact of eco-

nomic forces on bank profitability is particularly appealing given the significant transformation in financial intermediation,

the gradual shift in sources of bank revenue (i.e., interest income vs. fee income), and the path of the yield curve over the

time frame of our study which covers periods of economic expansion and contraction. The banking turmoil and the subse-

quent 2007–2009 financial crisis motivate the importance of understanding the main elements of bank profitability. The

motivation for this research is also driven by the recent Basel III Accord. Using data from 1990 to 2007 for several indus-

trial countries, Bolt, De Haan, Hoeberichts, Van Oordt, and Swank (2012) claim that the Accord urged banks to retai n

additional profits and payout fewer dividends when Tier 1 capital buffers are below required levels. Revisiting how the

yield curve impacts bank profitability is important since the overall functioning and wellbeing of the U.S. economy hinges

on a stable financial system with well-capitalized profit-generating financial institutions operating in an econom y with his-

torically low interest rates since the 2007–2009 financial crisis.

This paper contributes to the literature as follows. First, it complements the body of research in the financial press that

offers viewpoints on the impact of the yield curve on bank profits largely based on circumstantial evidence. Our paper stud-

ies the profit behavior of U.S. banks during a period of significant transformation in the banking system, in which the key

driving force of profits remains the price of lending long and borrowing short.

Second, in contrast to previous studies, this research covers a sample period from 2000Q1 to 2016Q4, which includes

the mild recession of 2001, economic expansion (including the peak and subsequent bust of the housing bubble), the 2007–

2009 global financial crisis, and economic recovery. During the time frame of this study, we witnessed major swings in the

yield curve. This invites a re-assessment of the relationship between the key variable of interest and bank net interest mar-

gins especially in light of shifts in bank’s sources of revenues over the years and swings in the U.S. economy captured by

real GDP growth, as documented by Stiroh (2004). Using annual call report data from 1978 to 2000, he finds that greater

reliance on non-interest income, particularly trading revenue, is associated with lower risk-adjusted profits and higher risk.

Third, the work by Bolt et al. (2012) combines theory and evidence of bank behavior. This paper borrows their empiri-

cal framework but moves focus to the U.S. banking system examining individual bank quarterly data, while their work

examines both individual and aggregate bank annual data for 17 countries. The sample period in our work extends beyond

theirs (from 1990 to 2007) to capture the recession of 2007–2009 that led short-term interest rates to zero, looks at four dif-

ferent asset classes of banks, and combines macro factors along with bank controls to explain net interest margins. In this

way, we apply panel data methods to a very large number of U.S. banks following a four-size classification scheme based

on asset size that yields 4,981 small banks, 475 medium banks, 34 large banks, and 14 money-center banks to examine

sensitivity of bank profitability to fluctuations in the yield curve. This represents an important departure from existing stud-

ies at quarterly frequency that typically considers a small number of largest banks. For the U.S., for example, at quarterly

frequency and under a similar time span than ours it is possible to mention the 10 commercial banks studied by Apergis

EGLY ET AL.

|

13