J. Finan. Intermediation: Rebecca Zarutskie

J. Finan.
Intermediation 22 (2013) 373–396
Contents lists available at SciVerse ScienceDirect
J. Finan. Intermediation
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j fi
Competition, financial innovation and commercial

bank loan portfolios
Rebecca Zarutskie
Board of Governors of the Federal Reserve System, Division of Research and Statistics, Mailstop 97, 20th and C Streets NW,
Washington, DC 20551, USA
a r t i c l e i n f o a b s t r a c t
Article history: I examine how US commercial bank loan portfolios change in

Received 7 October 2010 response to the rise of securitization markets and banking market
Available online 16 February 2013 deregulations over 1976–2003. Banks increasingly tilt their portfo-
lios toward real-estate-backed loans. However, there are signifi-
cant differences across banks. Larger banks and younger banks
disproportionately shift their lending toward real-estate-backed
loans, particularly commercial real-estate-backed loans, whereas
smaller banks and older banks maintain greater shares of their loan
portfolios in commercial and personal loans. When larger banks
make more real-estate-backed loans, they charge lower interest
rates, consistent with these banks lowering the costs of lending
and expanding credit for borrowers. In contrast, smaller banks
charge higher interest rates, consistent with these banks restricting
lending to a select group of borrowers.
Published by Elsevier Inc.
1. Introduction
Tremendous changes in US commercial banking markets occurred during the three decades since
1976. A wave of state-wide deregulations beginning in 1976 and ending with national deregulation in
1994 made banking markets more competitive. However, the number of commercial banks in the US
fell dramatically as deregulation allowed banks to merge and consolidate into larger ‘‘mega’’ banks. At
the same time, securitization markets expanded rapidly, first with the market for real-estate-backed
loans in the 1970s and subsequently with the markets for commercial and personal loans in the 1990s.
How did the larger consolidated banks and the remaining smaller banks co-exist and compete for
depositors and borrowers?
E-mail address: rebecca.zarutskie@frb.gov
1042-9573/$ - see front matter Published by Elsevier Inc.

http://dx.doi.org/10.1016/j.jfi.2013.02.001
374 R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396
Motivated by this question, I examine how banks’ loan portfolios change in response to dereg-
ulation and the concurrent innovation and growth in securitization markets over the period 1976–
2003 and whether the response varies by bank size and age. My findings suggest larger banks and
younger banks focus more on hard information loans (such as real estate backed loans) relative to
smaller banks and older banks. I focus on bank size because bank size changed dramatically during
this time and theory predicts size should matter in lending behavior. Recent studies suggest that
smaller banks may be better at processing and making loans based on soft information, or
information that cannot be easily quantified.1 Larger banks may be more likely to achieve economies
of scale and a more diversified loan portfolio from their presence across many markets and borrow-
ers.2 Bank age may also matter in lending decisions. Older banks may have a comparative advantage
in processing soft information and engaging in relationship lending since they have been in the mar-
ket longer and may be able to better source and process these kinds of loans relative to younger
banks.
I find that all banks engage in more real-estate-backed lending over the period 1976–2003 and the
increased lending is proportional to the level of securitization in the loan market. Over this period the
percentage of real-estate-backed loans in banks’ loan portfolios rose from 35% to 65% and the ratio of
the value to outstanding securitized real-estate-backed loans to the value of real-estate-backed loans
held on banks’ balance sheets rose from 0.19 to 1.89. Larger banks hold greater shares of real-estate-
backed loans in their portfolios and they lend more to commercial real-estate-backed loans compared
to smaller banks over the sample period.
I examine how bank loan portfolios change during two periods of deregulation in US commercial
banking markets – the period of state-wide deregulation between 1976 and 1994 in which states al-
lowed entry by commercial banks into new markets and the period of national deregulation in which
states adopted the provisions of the Riegle-Neal Interstate Banking and Branching Efficiency Act of
1994. Both periods yield similar findings. Larger banks and younger banks disproportionately shift
their lending toward real-estate-backed loans, particularly commercial real-estate-backed loans,
whereas smaller banks and older banks maintain greater shares of their loan portfolios in unsecured
commercial and personal loans. Although a large part of the lending shift is explained by the rise in
securitization, deregulation plays a significant role.
Examining how the pricing of loans changes for smaller versus large and young versus older banks
in response to deregulation enforces this view. Larger banks and younger banks charge lower interest
rates in response to an increase in securitization and deregulation. This evidence is consistent with
these banks lowering the costs of making loans and thus expanding credit to borrowers. In contrast,
I find that smaller and older banks charge higher interest rates, suggesting that they are able to charge
borrowers a premium for their greater screening ability relative to larger banks, or that competition in
these loan categories is reduced as larger banks are unable or unwilling to judge credit quality and
make loans to these borrowers.
These results suggest that larger banks and younger banks shift their lending towards hard infor-
mation loans in response to deregulation and the increasing ability to take advantage of loan securi-
tization markets, and that smaller banks and older banks keep more of their loan portfolios in soft
information loans. While the analysis cannot strictly identify whether innovation in securitization
markets encourages consolidation and larger bank size after deregulation, or vice versa, it points to
a clear association between larger bank size, securitization, deregulation, and greater lending in
real-estate-backed loans, especially by larger banks.
This paper contributes to the literature on the effects of credit market competition on credit supply.
Empirical studies test competing hypotheses on the effects of credit market competition on equilib-
rium lending and other economic outcomes (e.g., Petersen and Rajan, 1995; Jayaratne and Strahan,
1998). Although conclusions are mixed, they generally support the view that an increase in credit
market competition has first-order positive effects on credit supply. The results in this paper suggest
that competition or financial innovation encourages banks to better assess the quality of borrowers or
1
See for example, Stein (2002), Berger et al. (2005), Liberti and Mian (2009), Alessandrini et al. (2010), and Presbitero and
Zazzaro (2011).
2
See for example Farrell and Shapiro (1990), Jayaratne and Strahan (1998), and DeYoung et al. (1998).
R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396 375
structure lower cost contracts with them in reaction to competitive forces. Over the long run, this
might mitigate the initial negative effects of competition on loan supply (e.g., Petersen and Rajan,
1995) but may also raise systemic risk.
This paper also contributes to the literature on securitization in loan markets by exploring the rela-
tion between bank size and age and the degree banks respond to an increased ability to securitize
loans. Recent studies (e.g., Loutskina and Strahan, 2009; Mian and Sufi, 2009; Demyanyk and Van
Hermert, 2011; Keys et al., 2010) examine the impact of securitization on the cost of funds at banks
and the types and terms of loans granted. They show that securitization tends to lower the costs of
lending, or to expand debt capacity for borrowers whose loans are sold or who borrow from banks that
securitize loans. The results in this paper show that larger and younger banks charge lower interest
rates on their loans when they make more loans that can be securitized.
Finally, this paper provides a new test of theories of organizational form on information processing
and transmission. Several studies document that soft information loans are more likely to be made by
smaller banks with less hierarchical organizational structures (e.g., Berger et al., 2005; Liberti and
Mian, 2009; Alessandrini et al., 2010; Presbitero and Zazzaro, 2011). Results here show this is also true
for older banks, which suggests that the ability to process and act on soft information act on it is re-
lated to the age of a bank, in addition to its size. The analysis in this paper also suggests that the inno-
vation in securitization markets and deregulation-driven competition may encourage smaller banks
and older banks to focus more on soft information loans.
The rest of the paper proceeds as follows. Section 2 discusses the related literature. Section 3 dis-
cusses the data and empirical framework. Section 4 presents the results. Section 5 concludes.
2. Related literature
Studies on the impact of competition on commercial bank lending primarily focuses on relation-
ship lending and lending to informationally opaque borrowers, whose loans are often characterized
as soft information loans. Boot and Thakor (2000) argue that increased interbank competition will lead
to more relationship lending. Dell’Ariccia and Marquez (2004) demonstrate that banks may make
more loans to informationally opaque borrowers when competition increases to the extent that some
banks have better screening ability than others and when uninformed banks cannot free ride on the
information gathered by the informed banks. In a similar vein, Dinc (2000) shows that competition
may actually increase a bank’s incentive to engage in relationship lending. In contrast, Petersen and
Rajan (1995) argue that an increase in competition reduces relationship lending and lending to infor-
mationally opaque borrowers by banks. In a more recent model, Hauswald and Marquez (2006) dem-
onstrate that competition may reduce lending to informationally opaque borrowers, despite
competitive incentives to invest in becoming better at screening opaque borrowers to develop a cap-
tive market.
Several empirical studies use borrower-level data to explore competition on loan usage by small
and young firms (e.g., Petersen and Rajan, 1995; Black and Strahan, 2002; Zarutskie, 2006; Presbitero
and Zazzaro, 2011). Other studies examine bank-level data, but typically focus on one bank or merging
banks. Using data from a Belgian bank, Degryse and Ongena (2007) find that the bank makes more
relationship loans in branches where competition is greater. Elsas (2005) examines the empirical
determinants of relationship lending in a random sample of German banks and also finds that rela-
tionship lending increases in more competitive markets. Berger et al. (1998) examine the impact of
bank mergers on lending to informationally opaque small businesses and find that merged banks re-
duce their lending to small businesses, but that some of this effect is offset by non-merging banks in
the same market. In a more recent study, Canales and Nanda (2012) find that decentralized banks,
while making more relationship loans, may use their superior screening ability to cherry pick borrow-
ers and restrict credit.
A different set of studies suggests that size and age may influence how banks react to competitive
forces. Prior work on merger synergies and economies of scale and scope (e.g., Farrell and Shapiro
(1990)) suggest that larger firms may be able to lower loan costs relative to smaller firms. Moreover,
the types of loans that are amenable to cost reduction via technological improvements are those that
may be more amenable to arm’s length transactions, often characterized as hard information loans
(e.g., Petersen and Rajan, 2002; DeYoung et al., 2008). Theoretical work suggests that smaller banks
may have an advantage in assessing borrower quality, especially in more opaque lending markets.
The ability to assess individual borrower quality is a task which often relies on soft information
collected by loan officers through personal interactions with prospective borrowers. Older banks
may be better able to process soft information due to their accumulated knowledge and experience
(e.g., Hauswald and Marquez, 2006). Smaller banks may be better able to credibly transmit soft
information collected by their loan officers through the flatter hierarchies that characterize these
banks (e.g., Stein, 2002; Liberti and Mian, 2009; Alessandrini et al., 2010; Presbitero and Zazzaro,
2011).
Jayaratne and Strahan (1998) find that bank operating costs and loan losses decrease in periods of
increased competition in the US, especially in larger banks. Likewise, DeYoung et al. (1998) and Eva-
noff and Ors (2008) find that incumbent banks increase their cost efficiency when new banks enter
their markets. Although they do not examine the role of competition in fostering greater distance be-
tween borrower and lender, Petersen and Rajan (2002) note that distance between lenders and bor-
rowers in the US has been increasing over time, consistent with innovation in loan-screening
technology and productivity in the later, more competitive period of US banking markets. More re-
cently, Brevoort and Wolken (2009) and Brevoort et al. (2010) document a reversal in the increase
in the bank-borrowers distance.
Fewer studies directly examine the relation between bank loan portfolios and the use of secondary
loan sales and securitization markets. However recent studies by Loutskina and Strahan (2009), Mian
and Sufi (2009), Demyanyk and Van Hemert (2011), and Keys et al. (2010) examine the impact of secu-
ritization on the cost of funds at banks and the terms of loans granted. They show that securitization
tends to lower the costs of lending for borrowers whose loans are sold or who borrow from banks that
securitize loans.
3. Empirical framework and data
In this section I discuss the empirical framework used to discern differences in bank portfolio re-
sponses by bank size and age to increases in securitization and deregulation. I then describe the data.
3.1. Empirical framework
The aim of the analysis is to estimate how bank loan portfolios change in response to trends in
securitization and deregulation over the time period 1976–2003. I first examine the relation between
bank loan portfolios and trends in securitization over the entire sample period by bank size and age. As
such, the first regressions take the form of the following equation:
LoanPortfolioi;t ¼ a0 þ a1 LogBankAssetsi;t þ a2 LogBankAgei;t þ a3 Securitizationt LogBankAssetsi;t

þ a4 Securitizationt LogBankAgei;t þ a5 X i;t þ ut þ ki þ ei;t ð1Þ
LoanPortfolio is the fraction of a bank’s loan portfolio represented by a particular loan category, e.g.,
real-estate-backed loans. Securitization is the value of securitized loans in a loan category divided
by the value of loans in this category held by commercial banks. The indices i and t in Eq. (1) cor-
respond to bank and year, respectively. Matrix X includes controls for other bank-level and market-
level characteristics that may influence banks’ loan portfolios. I control for the bank’s capital ratio,
defined as the ratio of book equity to book assets, for possible changes in risk-taking by banks over
time (e.g., Keeley, 1990; Hellmann et al., 2000). I also control for state-level bank asset concentra-
tion, measured as the Herfindahl index of bank assets for banks headquartered in a state, the nat-
ural logarithm of real state-level GDP, and the natural logarithm of the inflation adjusted Federal
Housing Financing Agency (FHFA) state-level house price index to control for economic factors
that may be correlated with the types of loans demanded in a given state. Eq. (1) also includes
bank- and year fixed effects.3
Given the year fixed effects in Eq. (1), it is not possible to include Securitization on its own in this
regression. However, it is possible to include the interaction of Securitization with bank size and bank
age to test the extent differences in post-deregulation lending behavior by different banks is driven by
the rise in securitization markets. The coefficients a3 and a4 on the interaction terms in Eq. (1) tell us
whether larger banks and younger banks shift more of their lending to real-estate-backed loans as
securitization rises over the sample period.
In addition to examining the response of bank loan portfolios to securitization trends over the sam-
ple period, I also examine how loan pricing responds to these same trends. The next regressions take
the form of the following equation:
LoanPRicing i;t ¼ b0 þ b1 LogBankAssetsi;t þ b2 LogBankAgei;t þ b3 Securitizationt LogBankAssetsi;t
þ b4 Securitizationt LogBankAgei;t þ b5 X i;t þ xt þ wi þ li;t ð2Þ
The securitization literature suggests banks that securitize more of their loans may lower the cost
of lending and expand access to credit. The estimates of the coefficients b3 and b4 on the interaction
terms will us whether larger and younger banks exhibit different pricing and performance on the
loans they make in response to securitization trends.
After estimating the response of bank loan portfolios to securitization over the entire sample per-
iod, I separately examine two periods of deregulation to estimate the response of bank loan portfolios
to both deregulatory shocks and trends in securitization. As such, I estimate regressions as in the fol-
lowing equations:
LoanPortfolioi;t ¼ c0 þ c1 LogBankAssetsi;t þ c2 LogBankAgei;t þ c3 Deregulationj;t
þ c4 Deregulationj;t LogBankAssetsi;t þ c5 Deregulationj;t LogBankAgei;t
þ c6 Securitizationt LogBankAssetsi;t þ c7 Securitizationt LogBankAgei;t
þ c8 X i;t þ tt þ ji þ qi;t ð3Þ
LoanPRicing i;t ¼ d0 þ d1 LogBankAssetsi;t þ d2 LogBankAgei;t þ d3 Deregulationj;t

þ d4 Deregulationj;t LogBankAssetsi;t þ d5 Deregulationj;t LogBankAgei;t
þ d6 Securitizationt LogBankAssetsi;t þ d7 Securitizationt LogBankAgei;t
þ d8 X i;t þ st þ pi þ ri;t ð4Þ
Eq. (3) estimates the response of bank loan portfolio composition to both deregulation and securitiza-
tion. Eq. (4) estimates the response of loan pricing to deregulation and securitization. Deregulation is
an indicator variable(s) which is (are) zero in the years in which the relevant deregulation has not yet
been enacted in the bank’s headquarter state and is one in the years in which the deregulation has
been enacted. The coefficients on the interactions between bank size and age and the Deregulation
and Securitization variables allow us to discern differential responses to these trends by bank size
and age.
I estimate Eqs. (3) and (4) separately over two deregulatory periods. The first period occurs from
1976 to 1994. During this period, individual states relaxed restrictions on the ability of commercial
banks to expand within and across state borders. States previously were quite restrictive in allowing
banks located within their borders to expand within the state and in allowing banks located in other
3
Because the dependent variables in Eq. (1) range between zero and one, OLS estimations may generate predicted values that lie
outside the [0, 1] interval. Tobit regressions, which use maximum likelihood to account for right and left censoring, yield very
similar estimates in my data sample to OLS estimates, when I include state, rather than bank, fixed effects in the specifications.
Including bank fixed effects in a Tobit specification is computationally intractable and thus, I report the OLS estimates. The
predicted values in the estimation sample are greater than zero and less than one 99% of the time. I also run OLS regressions in
which I take the log odds ratio of the dependent variables and find similar results. I report OLS estimates for straightforward
interpretation of economic magnitudes. The OLS estimates using the log odds ratio transformation of the dependent variables are
available upon request.
states to enter the state.4 The state-wide deregulations took two forms – intrastate branching deregu-
lations and interstate branching deregulations. Intrastate deregulation allowed already existing banks
to expand within the state either by acquiring existing branches of another bank or by establishing de
novo branches within the state. Interstate deregulation allowed banks based in other states to enter
the state by acquiring existing branches of another bank. Having entered a state through acquisition,
an out-of-state bank could then potentially establish more de novo branches. Interstate branching dereg-
ulation typically occurred on a bi-lateral or multi-lateral basis between two or more states. The years in
which states allowed intrastate and interstate branching are listed in the appendix.
The second period of deregulation in US commercial banking markets involved the passage of the
Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994. It was passed by the US Congress
and was national in scope. The Riegle-Neal Act struck down the final barriers to interstate branching
by mandating that banks in all states could enter any other state. However, it had a differential impact
across states since states had discretion to adopt the provisions of the Act over the 1994–1997 period.5
Alaska adopted first, in 1994. Texas and Montana were the latest adopters, in 1999 and 2001, respec-
tively, after they initially opted out of the legislation. The appendix lists the years in which each state
adopted the provisions of the provisions of the Riegle-Neal Act.
3.2. Data and descriptive statistics
The two main data sources used in the analysis are the Call Reports of Condition and Income (‘‘Call
Reports’’) and the Flow of Funds Accounts. The Call Reports provide balance sheet and income state-
ment data for commercial banks. The Call Reports provide information on loans that are secured by
real estate, including both personal and commercial loans, non-real-estate-backed loans used for com-
mercial and industrial use (C&I Loans) and personal loans, agricultural loans, and other loans, such as
interbank loans. While these loan categories are broad, they allow an examination of the degree banks
concentrate their lending in loans collateralized by real estate versus loans that are not collateralized.
Collateralization is an important dimension of loans, which can plausibly distinguish hard information
loans from soft information loans. I take bank-level data from the Call Reports in June of each year
from 1976 to 2003. I use the annual version of the Flow of Funds Accounts to compute measures of
the rate of securitization in loan markets.
Table 1 presents descriptive statistics for the key variables used in the analysis. Averages are re-
ported first, followed by standard deviations in parentheses. Observations are aggregated to the bank
holding company level, when applicable. Table 1 first summarizes the shares of bank loan portfolios
comprised of the five major loan categories. On average, 19.2% of bank loan portfolios are commercial
and industrial loans and 20.9% are personal loans.6 Small C&I loans (principal amounts $250,000 or less)
make up around half of all C&I loans on bank balance sheets. Small C&I loans are recorded in the Call
Reports starting in 1993. On average, 45.3% of bank loan portfolios are real-estate-backed loans. Over half
of these loans are backed by residential real estate, representing 26.3% of bank loan portfolios, on aver-
age. Around one-quarter of real-estate-backed loans, or 11% in total, are backed by non-residential real-
estate. The remaining real-estate-backed loans are secured by new construction and farmland. On aver-
age, 13.4% of bank loan portfolios are agriculture loans.
Table 1 also presents descriptive statistics for measures of loan pricing. The loan pricing measure is
total interest and fee income in a loan category divided by total loans in that category. This measure
gives an average measure of the interest and fees collected on loans in a given category and is a proxy
for the average interest rate on loans in a given category. The Call Reports do not report interest by
subcategory for real-estate-backed loans except for a handful of the largest banks and only begin to
track interest and fee income by loan category in 1984. Real-estate-backed loans have the lowest
4
For a more detailed history of this period of state-wide commercial banking market deregulations and a discussion of the
political economy of the deregulation see Kroszner and Strahan (1999).
5
See Dick (2006) and Rice and Strahan (2010) for a more detailed discussion of the impact across states of the Riegle-Neal Act.
6
The personal loan category consists of personal installment loans and credit card loans. Personal installment loans comprise
99.5% of the personal loan portfolios held by commercial banks over the sample period.
Table 1
Descriptive statistics for key variables.
Sample period: 1976–2003 (Unique banks or bank holding

companies = 26,836)
Mean (Std.dev.) Obs.
Bank loan portfolio shares
Commercial and industrial (C&I) loans/total loans 0.192 (0.136) 273,950
Small C&I loans/total loans 0.108 (0.083) 82,104
Personal loans/total loans 0.209 (0.146) 273,950
Real-estate-backed (RE) loans/total loans 0.453 (0.222) 273,950
Residential loans/total loans 0.263 (0.190) 273,950
Non-residential loans/total loans 0.110 (0.103) 273,950
Construction loans/total loans 0.031 (0.055) 273,950
Farmland loans/total loans 0.049 (0.070) 273,950
Agriculture loans/total loans 0.134 (0.188) 273,950
Other loans/total loans 0.012 (0.067) 273,950
Loan interest and fees
C&I loan interest and fees/C&I loans 0.106 (0.124) 144,488
Personal loan interest and fees/personal loans 0.052 (0.063) 147,186
RE loan interest and fees/RE backed loans 0.042 (0.048) 147,617
Bank and state characteristics
Bank assets 382,137 (4,882,350) 273,950
Log(Bank assets) 11.0 (1.28) 273,950
Bank age 61.6 (36.5) 273,950
Log(Bank age) 3.76 (1.09) 273,950
Capital ratio 0.070 (0.039) 273,950
State bank concentration 0.093 (0.102) 273,950
State GDP 208,947 (204,303) 273,950
Log(State GDP) 11.8 (0.936) 273,950
FHFA house price index 186.7 (41.5) 273,950
Log(FHFA house price index) 5.21 (0.191) 273,950
Securitization variables
Securitization 1.01 (0.591) 273,950
Securitization residential 1.68 (0.947) 273,950
Securitization non-residential 0.09 (0.139) 273,950
Data are taken from the June Call Reports of Income and Condition between 1976 and 2003. Variables are computed at the bank
holding company level, when applicable. Sample means are reported first, followed by standard deviations in parentheses. For
each variable, the number of observations corresponds to bank-years or bank holding company-years. C&I loans are non-real-
estate-backed commercial and industrial loans. Small C&I loans are C&I loans with principal amounts of less than $250,000.
Personal loans are non-real-estate-backed personal installment and credit card loans. Bank assets are reported in thousands of
year 2000 dollars. Capital ratio is the ratio of bank book equity to book assets. State bank concentration is the Herfindahl index,
i.e., sum of squared shares of bank assets in a given state and year. State GDP is taken from the Bureau of Economic Analysis and
is reported in millions of year 2000 dollars. The FHFA house price index is taken from the Federal Housing Finance Agency by
state and year and is adjusted for inflation. Securitization is the value of real-estate-backed loans outstanding in a given year
that have been securitized as reported by the Flow of Funds accounts divided by the value of real-estate-backed loans on
commercial banks’ balance sheets.
average interest and fees at 4.2%, suggesting that greater securitization in this loan category has re-
duced the cost of borrowing.
Table 1 also summarizes key bank characteristics. The average bank size measured in year 2000 US
dollars, is $382 million in assets. The average bank age is 61.6 years. Bank capital ratio, the ratio of
book equity to book assets, and state bank concentration are also summarized. I use these variables
as additional controls for bank characteristics that may influence bank loan portfolios. State GDP is
taken from the Bureau of Economic Analysis and is reported in millions of year 2000 dollars and is
meant to capture differences in economic activity and loan demand. The house price index is taken
from the FHFA by state and year and is adjusted for inflation; it is meant to control for house prices
in a state that may influence demand for real-estate-backed loans. State bank concentration is the Her-
findahl index of bank assets in a given state and year; it is meant to control for differences in banking
market structure that may affect loan portfolios.
Finally, Table 1 summarizes the securitization variables. Securitization is the value of loans out-
standing in a given year that have been securitized as reported by the Flow of Funds accounts divided
by the value of loans on commercial banks’ balance sheets. A higher ratio indicates that more loans
have been securitized in a given loan category. The Flow of Funds accounts report the value of secu-
rities backed by assets by government sponsored entities and private asset-backed securities issuers.
The only kind of assets-backed securities based on loans are real-estate-backed loans until the late
1990s. Therefore, I summarize only the securitization ratios for real-estate-backed loans. The ratio
of all securitized real-estate-backed loans to real-estate-backed loans held by commercial banks is
1.01 over the sample period. The ratio of all securitized residential real-estate-backed loans to residen-
tial real-estate-backed loans held by commercial banks is 1.68 over the sample period, which reflects
this category of loans was the first to be securitized by government sponsored entities. The ratio of
securitized non-residential real-estate-backed loans to non-residential real-estate-backed loans held
by commercial banks is 0.09 over the sample period, reflecting a more limited securitization market
for non-residential loans.
3.3. Trends in key variables, 1976–2003
I next show how the averages of the key variables summarized in Table 1 vary over the sample per-
iod. Fig. 1 plots by year the average share of bank loan portfolios that are comprised by the main loan
categories – C&I, personal, real-estate-backed, and agriculture loans. Fig. 2 plots by year the average
share of bank loan portfolios that are comprised by the different sub-categories of real-estate-backed
loans. Both figures show bank portfolios become increasingly dominated by real-estate-backed loans
over time. In the mid-1980s, real-estate-backed loans begin to steadily rise, from 35% of in 1985 to 65%
by 2003. Likewise, we see a steady decline in commercial, personal, and agriculture loans over the
sample period as real-estate-backed loans surge. Fig. 2 shows that the rise in real-estate-backed loans
is driven by an increase in both residential real-estate-backed loans and non-residential real-estate-
backed loans. Starting in 1985, residential real-estate-backed loans increase rapidly starting in
1985, but growth levels off and declines slightly by 1998. In contrast, non-residential real-estate-
backed loans begin to increase in 1985 and continue to increase over the sample period. Real-es-
tate-backed loans backed by construction and farmland increase slightly over the sample period.
Fig. 3 shows the increase in securitization of real-estate-backed loans over the same time period
1976–2003. Residential real-estate-backed loan securitization takes off in the mid-1980s, around
Fig. 1. Bank loan portfolio shares, 1976–2003. This figure plots by year the average share of banks’ loan portfolios comprised by
C&I, personal, real-estate-backed and agriculture loans. Data are taken from the June Call Reports of Income and Condition.
Variables are computed at the bank holding company level, when applicable.
Fig. 2. Bank real-estate-backed loan portfolio shares, 1976–2003. This figure plots by year the average share of banks’ loan
portfolios comprised by all real-estate-backed, residential real-estate-backed, non-residential real-estate-backed, construction,
and farmland loans. Data are taken from the June Call Reports of Income and Condition. Variables are computed at the bank
holding company level, when applicable.
Fig. 3. Securitization, 1976–2003. This figure plots by year securitization of real-estate-backed loans from 1976 to 2003. The
measure of securitization is the value of loans outstanding in a given year that have been securitized as reported by the Flow of
Funds accounts divided by the value of loans in a given category on commercial banks’ balance sheets. Data are taken from the
June Call Reports of Income and Condition.
the same time banks shift their loan portfolios toward residential real-estate-backed loans. The in-
crease is consistent with arguments that securitization expands credit in the loan category that is
being securitized. The ratio of the value of securitized residential real-estate-backed loans to the value
of residential real-estate-backed loans held on banks’ balance sheets rises from 0.27 in 1976 to 2.85 in
2003. Securitization of non-residential real-estate-backed loans does not begin in earnest until the
1990s, rising from 0.04 in 1990 to 0.47 in 2003. Interestingly, the shift in bank loan portfolios towards
Fig. 4. Bank size, 1976–2003. This figure plots by year average bank size in thousands of year 2000 dollars. Data are taken from
the June Call Reports of Income and Condition. Banks size is computed at the bank holding company level, when applicable.
Fig. 5. Number of banks, 1976–2003. This figure plots by year the number of unique banks or bank holding companies. Data are
taken from the June Call Reports of Income and Condition.
residential and nonresidential real-estate-backed loans corresponds in time with the rise in securiti-
zation in these two loan categories.
Figs. 4 and 5 show the increase in average bank size and the decrease in the number of banks over
the sample period. In constant year 2000 dollars, average bank size increases from $213 million to
$924 million. As with the increase in real-estate-backed lending by banks and the increase in securi-
tization, the increase in bank size starts in the mid-1980s. Finally, Fig. 5 shows the steady decline in
the number of banks in the sample. There are 12,687 banks in 1976 and only 6,745 in 2003.7 The
trends in Figs. 4 and 5 reflect the wave of consolidation and rapid bank growth that occurred when bar-
riers to cross-state entry and mergers were lifted by the deregulation described in section 3.2 (e.g., Berger
7
These counts account for the fact that bank holding companies may own several individual banks. The counts are aggregated to
count only one bank holding company as a single observation.
et al., 1999). Overall the trends depicted in Figs. 1–5 show that bank size, securitization, and lending to
real-estate-backed loans are positively correlated over time. In the next section, I explicitly estimate the
degree bank size and bank age is correlated with the shift to real-estate-backed lending given the rise in
securitization.
4. Estimation results
In this section, I present the empirical results using the estimation framework detailed in Sec-
tion 3.1. I first present the results for how bank loan portfolios changed over the entire sample period
in response to trends in securitization. I then examine the response separately by the two periods of
state-wide and national deregulation.
4.1. Bank loan portfolios and securitization, 1976–2003
Table 2 presents regression estimates for Eq. (1). Columns (1) and (2) present estimates when the
dependent variables are the share of C&I loans in bank loan portfolios and the share of personal loans
in bank loan portfolios, respectively. Columns (3) and (4) present estimates when the dependent
variables are the share of non-residential real-estate-backed loans in bank loan portfolios and the
Table 2
Bank loan portfolios by bank size and age, 1976–2003.
C&I loans/ Personal loans/ Non-residential RE loans/

Total loans Total loans RE loans/Total loans Total loans
(1) (2) (3) (4)
Log(Bank assets) 0.028*** 0.013** 0.004 0.021***
(5.83) (2.12) (1.11) (2.51)
Log(Bank age) 0.046*** 0.030*** 0.023*** 0.038***
(10.58) (6.42) (3.40) (4.38)
Securitization Log(Bank assets) 0.014*** 0.004** 0.036*** 0.003
(7.91) (2.43) (5.06) (0.85)
Securitization Log(Bank age) 0.017*** 0.018*** 0.089*** 0.030***
(6.59) (13.75) (3.43) (5.75)
Capital ratio 0.076*** 0.052* 0.075*** 0.198***
(3.30) (1.77) (3.81) (4.47)
State bank concentration 0.035** 0.021 0.020 0.037
(2.49) (1.30) (0.72) (1.58)
Log(State GDP) 0.012 0.020 0.016* 0.011
(1.37) (1.31) (1.78) (0.71)
Log(FHFA house price index) 0.028* 0.0005 0.003 0.007
(1.79) (0.03) (0.32) (0.27)
Bank fixed effects? Yes Yes Yes Yes
Year fixed effects? Yes Yes Yes Yes
N 273,950 273,950 273,950 273,950
Adjusted R2 0.747 0.785 0.747 0.868
The data are taken from the June Call Reports of Income and Condition between 1976 and 2003. Variables are constructed at the
bank holding company level when applicable. Bank assets are reported in thousands of year 2000 dollars. Bank age is reported
in years. Securitization is the value of real-estate backed loans outstanding in a given year that have been securitized as
reported by the Flow of Funds accounts divided by the value of real-estate-backed loans on commercial banks’ balance sheets.
Capital ratio is the ratio of bank book equity to book assets. State bank concentration is the Herfindahl index, i.e., sum of squared
shares, of bank assets in a given state and year. State GDP is taken from the Bureau of Economic Analysis and is reported in
millions of year 2000 dollars. The FHFA house price index is taken from the Federal Housing Finance Agency by state and year
and is adjusted for inflation. All regressions are estimated using OLS. Coefficients are reported followed by t-statistics adjusted
for clustering at the state level.
*
Indicates two-tailed statistical significance at the 10% level.
**
***
share of all real-estate-backed loans in banks’ loan portfolios, respectively. The variable Securitization
measures total securitization for all real-estate-backed loans, since in the early part of the sample per-
iod other securitization measures, such as commercial securitization, are zero.
Focusing on the coefficients on bank size and the interaction between securitization and bank size,
we see that larger banks make more C&I loans from the positive and significant coefficient of 0.024 on
the LogBankAssets in column (1). However, once the variable Securitization takes values greater than 2,
larger banks make fewer C&I loans compared to smaller banks, as indicated by the negative and sig-
nificant coefficient on the interaction between Securitization and LogBankAssets of 0.014. Turning to
the estimates for personal loan shares in column (2), we see that larger banks make fewer personal
from the significant coefficient of 0.013 on LogBankAssets. As securitization increases, larger banks
make even fewer personal loans, reflected by the significant and negative coefficient of 0.004 on
the interaction between Securitization and LogBankAssets. Thus, we see from the estimates in columns
(1) and (2) that larger banks make relatively fewer commercial and personal loans in response to
greater securitization compared to smaller banks.
Turning to the estimates in Table 2, column (3), we see that non-residential real-estate-backed
loans do not comprise a significantly larger share of larger banks’ loan portfolios from the statistically
insignificant coefficient on LogBankAssets. However, the positive and significant coefficient (0.036) on
the interaction between Securitization and LogBankAssets shows that as soon as the ratio of securitized
loans to loans held by banks is positive, larger banks make more non-residential real-estate-backed
loans compared to smaller banks. Finally, Table 2, column (4) shows that larger banks make more
real-estate-backed loans overall from the significant coefficient (0.021) on LogBankAssets. However,
we do not see that larger banks make more real-estate-backed loans overall as securitization increases
over the full sample period, as shown by the statistically insignificant coefficient on the interaction
between Securitization and LogBankAssets. Both smaller banks and larger banks increased their lending
to residential real-estate-backed loans in the 1980s when government sponsored entities securitized
the largest share of these loans. This neutralizes the greater lending by larger banks to non-residential
real-estate-backed loans when we add both types together in this measure of overall real-estate-
backed lending.
Overall, the estimated coefficients reported in Table 2 show that larger banks make more real-es-
tate-backed loans relative to smaller banks in response to a greater ability to securitize these types of
loans. We also see that older banks make more C&I loans and personal loans as securitization increase
from the positive and significant coefficients of 0.017 and 0.018 on the interaction between Securiti-
zation and LogBankAge in columns (1) and (2). Older banks reduce their real-estate-backed lending,
particularly non-residential real-estate-backed lending, as securitization increases, as indicated by
the negative and significant coefficients of 0.089 and 0.030 on the interaction of Securitization
and LogBankAge in columns (3) and (4).
I report estimates for the loan pricing regression specified by Eq. (2) in Table 3. We see that larger
banks charge higher interest and fees in each loan category from the positive, significant coefficients
on LogBankAssets, but as securitization increases larger banks increasingly lower their fees relative to
smaller banks, indicated by the negative, significant coefficients on the interaction between Securiti-
zation and LogBankAssets. Likewise, older banks charge higher fees as securitization increases, as
shown by the positive and significant coefficients on the interaction between Securitization and Log-
BankAge. Overall, the estimates reported in Table 3 support the view that the rise in real-estate-backed
loan securitization over the period 1976–2003 led to an expansion of credit to borrowers and reduced
lending costs, especially among the largest and youngest banks.
4.2. State-wide and national deregulation
I next examine the relation between bank loan portfolios and securitization, accounting for dereg-
ulatory events to estimate the extent deregulation and securitization trends separately explain
changes in bank loan portfolios. Table 4 reports estimates of Eq. (3) during the period of state-wide
deregulation from 1976 to 1994. Over this period, the relevant deregulation indicator variables are
Interstate and Intrastate, which take the value one when a state enacts interstate and intrastate bank
Table 3
Loan pricing by bank size and age, 1984–2003.
C&I loan interest and Personal loan interest and fees/ RE backed loan interest and
fees/C&I loans Personal loans fees/RE loans
(1) (2) (3)
Log(Bank assets) 0.105*** 0.059*** 0.054***
(8.91) (10.41) (8.88)
Log(Bank age) 0.013*** 0.002 0.001
(2.89) (1.25) (0.32)
Securitization Log(Bank 0.037*** 0.021*** 0.019***
assets)
(7.83) (7.57) (6.50)
Securitization Log(Bank 0.018*** 0.012*** 0.010***
age)
(5.02) (6.74) (7.52)
Capital ratio 0.007 0.098*** 0.100***
(0.21) (4.57) (4.55)
State bank concentration 0.007 0.004 0.004
(0.38) (0.62) (0.86)
Log(State GDP) 0.006 0.001 0.004
(0.62) (0.26) (0.89)
Log(FHFA house price 0.005 0.007** 0.010**
index)
(0.48) (1.04) (2.13)
Bank fixed effects? Yes Yes Yes
Year fixed effects? Yes Yes Yes
N 144,488 147,186 147,617
Adjusted R2 0.579 0.588 0.668
bank holding company level, when applicable. Bank age is reported in years. Securitization is the value of real-estate-backed
loans outstanding in a given year that have been securitized as reported by the Flow of Funds accounts divided by the value of
real-estate-backed loans on commercial banks’ balance sheets. Capital ratio is the ratio of bank book equity to book assets. State
bank concentration is the Herfindahl index, i.e., sum of squared shares of bank assets in a given state and year. State GDP is
taken from the Bureau of Economic Analysis and is reported in millions of year 2000 dollars. The FHFA house price index is
taken from the Federal Housing Finance Agency by state and year and is adjusted for inflation. All regressions are estimated
using OLS. Coefficients are reported followed by t-statistics adjusted for clustering at the state level.

**
***
branching deregulations, respectively. The second specification for each loan category includes a time
trend to test the robustness of the deregulatory and securitization variables to a blunt time trend.
Examining the coefficients on the interaction terms of the Interstate and Intrastate indicator vari-
ables with LogBankAssets, we see that larger banks make fewer C&I loans after state-wide interstate
banking regulations are lifted from the negative and significant coefficient (0.003) on the interaction
between Interstate and LogBankAssets. They make more non-residential real-estate-backed loans, as
shown by the positive, significant coefficient (0.003) on the interaction between Interstate and LogBan-
kAssets. Deregulation by bank size does not have a significant impact on lending to personal loans or
all real-estate-backed loans; the coefficients on the interaction between Interstate and LogBankAssets
and the interaction between Intrastate and LogBankSize are insignificant. The insignificant coefficients
on the interactions between Interstate and LogBankAge and between Intrastate and LogBankAge across
all columns indicate there is no significant impact of deregulation on lending by bank age across all
loan categories.
Turning to the coefficients on the interactions between Securitization and LogBankAssets and be-
tween Securitization and LogBankAge in Table 4, we see that an increase in securitization leads to fewer
C&I and personal loans made by larger banks and more non-residential real-estate-backed loans made
Table 4
386
Bank loan portfolios during state-wide deregulation, 1976–1994.
C&I loans/Total loans Personal loans/Total loans Non-residential RE loans/Total loans RE loans/Total loans
(1) (2) (3) (4) (5) (6) (7) (8)
Log(Bank assets) 0.029*** 0.029*** 0.014** 0.013** 0.008* 0.007* 0.011 0.010
(6.52) (6.51) (2.32) (2.22) (1.86) (1.83) (1.32) (1.24)
Log(Bank age) 0.044*** 0.043 0.024*** 0.023*** 0.025*** 0.025*** 0.041*** 0.040***
(11.74) (12.18) (6.12) (5.92) (4.99) (5.05) (6.22) (6.31)
Interstate 0.016 0.016 0.029* 0.032* 0.018* 0.017* 0.005 0.003
(1.22) (1.30) (1.68) (1.82) (1.87) (1.85) (0.20) (0.12)
Intrastate 0.024 0.023 0.028 0.033 0.039** 0.039* 0.058* 0.058*
(1.41) (1.30) (0.79) (0.89) (2.03) (1.91) (1.84) (1.77)
Interstate Log(Bank assets) 0.003** 0.003** 0.002 0.002 0.003** 0.003** 0.001 0.001
(2.25) (2.33) (1.21) (1.25) (2.54) (2.49) (0.37) (0.39)
R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396

Intrastate Log(Bank assets) 0.001 0.001 0.001 0.001 0.002 0.002 0.0001 0.001
(0.76) (0.54) (0.36) (0.29) (1.02) (1.07) (0.03) (0.18)
Interstate Log(Bank age) 0.003 0.003 0.003 0.003 0.003 0.003 0.003 0.003
(0.93) (0.94) (1.03) (1.19) (1.60) (1.66) (0.86) (0.77)
Intrastate Log(Bank age) 0.005 0.005 0.002 0.004 0.003 0.003 0.013*** 0.015***
(1.54) (1.49) (0.68) (1.21) (0.92) (0.75) (3.03) (3.18)
Securitization Log(Bank assets) 0.010*** 0.005 0.006** 0.005*** 0.007*** 0.010*** 0.001 0.012**
(4.80) (1.09) (2.18) (2.69) (3.69) (3.40) (0.18) (2.07)
Securitization Log(Bank age) 0.016*** 0.008 0.021*** 0.010** 0.022*** 0.028*** 0.039*** 0.014
(2.86) (0.81) (8.23) (2.58) (4.71) (7.56) (5.59) (1.21)
Trend Log(Bank assets) 0.0005* 0.0001 0.0003 0.0010**
(1.68) (0.33) (1.19) (2.13)
Trend Log(Bank age) 0.0008 0.0021*** 0.0006 0.0024**
(0.84) (3.88) (1.53) (2.41)
Capital ratio 0.089*** 0.089*** 0.063* 0.059* 0.112*** 0.111*** 0.231*** 0.234***
(2.74) (2.72) (1.73) (1.66) (4.21) (4.14) (4.90) (5.10)
State bank concentration 0.038** 0.038** 0.038** 0.038** 0.016 0.016 0.008 0.008
(2.23) (2.21) (2.28) (2.26) (0.76) (0.76) (0.43) (0.41)
Log(State GDP) 0.011 0.011 0.003 0.002 0.015 0.015 0.026 0.026
(1.14) (1.14) (0.18) (0.11) (1.18) (1.16) (1.46) (1.48)
Log(FHFA house price index) 0.027* 0.028* 0.001 0.001 0.006 0.005 0.011 0.014
(1.77) (1.93) (0.08) (0.04) (0.50) (0.44) (0.51) (0.65)
Bank fixed effects? Yes Yes Yes Yes Yes Yes Yes Yes
Year fixed effects? Yes Yes Yes Yes Yes Yes Yes Yes
N 208,508 208,508 208,508 208,508 208,508 208,508 208,508 208,508
Adjusted R2 0.765 0.765 0.783 0.783 0.717 0.717 0.867 0.867
The data are taken from the June Call Reports of Income and Condition between 1976 and 1994. Variables are constructed at the bank holding company level when applicable. Interstate is
a dummy variable which takes the value one in the year in which a state allows out-of-state banks to open branches within the state and in all subsequent years. Intrastate is a dummy
variable which takes the value one in the year in which a state allows banks to branch within state and in all subsequent years. Bank assets are reported in thousands of year 2000 dollars.
Bank age is reported in years. Securitization is the value of real-estate-backed loans outstanding in a given year that have been securitized as reported by the Flow of Funds accounts
divided by the value of real-estate-backed loans on commercial banks’ balance sheets. Trend is a time trend variable which takes the value one in 1976 and increases by one in each
successive year. Capital ratio is the ratio of bank book equity to book assets. State bank concentration is the Herfindahl index, i.e., sum of squared shares of bank assets in a given state and
year. State GDP is taken from the Bureau of Economic Analysis and is reported in millions of year 2000 dollars. The FHFA house price index is taken from the Federal Housing Finance
Agency by state and year and is adjusted for inflation. All regressions are estimated using OLS. Coefficients are reported followed by t-statistics adjusted for clustering at the state level.
*
**
***
Table 5
Bank loan portfolios during national deregulation, 1990–2003.
C&I loans/Total loans Personal loans/Total loans Non-residential RE loans/Total loans RE loans/Total loans
(1) (2) (3) (4) (5) (6) (7) (8)
Log(Bank assets) 0.023*** 0.017*** 0.002 0.001 0.008 0.001 0.015* 0.001
(3.61) (2.98) (0.21) (0.09) (1.03) (0.14) (1.88) (0.15)
Log(Bank age) 0.048 0.026*** 0.014** 0.003 0.057*** 0.039*** 0.034*** 0.024
(10.63) (4.67) (2.20) (0.65) (4.62) (4.48) (4.93) (3.88)
Riegle-Neal 0.005 0.005 0.012 0.027*** 0.044*** 0.049*** 0.044*** 0.004
(0.72) (0.57) (1.01) (3.85) (2.89) (3.29) (4.39) (0.32)
Riegle-Neal Log(Bank assets) 0.001 0.001 0.001 0.002** 0.002*** 0.004*** 0.005*** 0.002*
R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396

(1.48) (0.69) (1.39) (2.18) (2.98) (4.58) (5.03) (1.88)
Riegle-Neal Log(Bank age) 0.001 0.004** 0.0004 0.003*** 0.003 0.0004 0.006*** 0.008***
(0.63) (2.65) (0.30) (2.88) (1.58) (0.24) (3.42) (4.29)
Securitization Log(Bank assets) 0.012*** 0.006** 0.003* 0.003*** 0.015*** 0.007*** 0.002 0.016***
(6.18) (2.55) (1.98) (4.50) (5.15) (3.64) (0.61) (4.16)
Securitization Log(Bank age) 0.017*** 0.00001 0.006*** 0.001 0.031*** 0.017*** 0.021*** 0.014***
(5.52) (0.00) (3.09) (0.94) (4.56) (3.93) (5.98) (3.11)
Trend Log(Bank assets) 0.001*** 0.00005 0.0007*** 0.001***
(2.96) (0.22) (2.87) (3.20)
Trend Log(Bank age) 0.002*** 0.001*** 0.002*** 0.001
(5.47) (3.39) (3.37) (1.55)
Capital ratio 0.051* 0.056* 0.031 0.030 0.003 0.001 0.095 0.093**
(1.83) (1.98) (0.77) (0.74) (0.14) (0.04) (2.40) (2.36)
State bank concentration 0.008 0.008 0.010 0.010 0.018 0.018 0.028 0.028
(0.68) (0.66) (0.67) (0.68) (0.60) (0.62) (1.17) (1.18)
Log(State GDP) 0.023*** 0.025*** 0.024** 0.023** 0.005 0.006 0.009 0.009
(3.27) (3.48) (2.20) (2.11) (0.44) (0.53) (0.69) (0.75)
Log(FHFA house price index) 0.028** 0.027** 0.013 0.012 0.050 0.051 0.050** 0.052**
(2.34) (2.45) (1.40) (1.27) (1.25) (1.23) (2.43) (2.58)
Bank fixed effects? Yes Yes Yes Yes Yes Yes Yes Yes
Year fixed effects? Yes Yes Yes Yes Yes Yes Yes Yes
N 111,251 111,251 111,251 111,251 111,251 111,251 111,251 111,251
Adjusted R2 0.789 0.789 0.849 0.849 0.828 0.829 0.890 0.890
The data are taken from the June Call Reports of Income and Condition between 1990 and 2003. In a given year, the data are aggregated and variables are computed at the bank holding
company level when applicable. Riegle-Neal deregulation is an indicator variable equal to one in the year a bank’s headquarters state enacts the provisions of the Riegle-Neal Act and one in
all subsequent years. Bank assets are reported in thousands of year 2000 dollars. Bank age is reported in years. Securitization is the value of real-estate-backed loans outstanding in a given
year that have been securitized as reported by the Flow of Funds accounts divided by the value of real-estate-backed loans on commercial banks’ balance sheets. Trend is a time trend
variable which takes the value one in 1990 and increases by one in each successive year. Capital ratio is the ratio of bank book equity to book assets. State bank concentration is the
Herfindahl index, i.e., sum of squared shares of bank assets in a given state and year. State GDP is taken from the Bureau of Economic Analysis and is reported in millions of year 2000
dollars. The FHFA house price index is taken from the Federal Housing Finance Agency by state and year and is adjusted for inflation. All regressions are estimated using OLS. Coefficients
are reported followed by t-statistics adjusted for clustering at the state level.
*
**
***
387
by larger banks.8 As in Table 2, which considered the full sample period of 1976–2003, we see that larger
banks do not make significantly more real-estate-backed loans in response to greater securitization. This
is becausse many smaller banks made lots of residential real-estate-backed loans in the 1980s when
securitization activity by government sponsored entities increased. Older banks make more C&I and per-
sonal loans and fewer real-estate-backed loans, both non-residential real-estate-backed loans and across
the board, in response to greater securitization, as shown by the positive and significant coefficients on
the interaction between Securitization and LogBankAge in columns (1)–(4) and by the significant, negative
coefficients on the interaction between Securitization and LogBankAge in columns (5)–(8).
Overall, the estimates in Table 4 show trends in both securitization and deregulation are associated
with shifts to greater non-residential real-estate-backed lending by larger and younger banks and to
greater unsecured C&I and personal lending by smaller and older banks. Table 5 reports estimates of
Eq. (3) during the period of national deregulation, 1990–2003, with similar results. In this set of
regressions, the deregulation variable is the indicator variable, RiegleNeal, which equals one in the per-
iod after a state enacts the provisions of the Riegle Neal Interstate Banking and Branching Efficiency
Act of 1994.
Table 5 shows with securitization, larger banks and younger banks make more real-estate-backed
loans relative to smaller banks and older banks, which keep more of their portfolios in unsecured C&I
and personal loans. The coefficients on the interaction between Securitzation and LogBankAssets are
significantly negative in the specifications for C&I and personal loans, and significantly positive in
the specifications for non-residential and all real-estate-backed loans. Likewise, the coefficients on
the interaction between Securitization and LogBankAge are significantly positive in the specifications
for C&I and personal loans, and significantly negative in the specifications for non-residential and
all real-estate-backed loans. There is little response to deregulation over this time period by bank size
and age, as shown by the small and mostly insignificant coefficients on the interaction of RiegleNeal
with bank size and age. Low response may reflect that most deregulation had occurred by the time
the Riegle-Neal Act was passed, so it had a smaller impact than deregulation in the earlier sample
period.
Tables 6 and 7 report estimates of Eq. (4) in the periods of state-wide and national deregulation,
respectively. In Table 6, the significant, negative coefficients on the interaction between Securitization
and LogBankAssets in all specifications indicate an increase in securitization is associated with lower
interest and fees by larger banks. The generally insignificant coefficients on the interactions between
Interstate and Intrastate and bank size and age suggest that securitization is the main driver of lower
interest and fees. In Table 7 we again see negative and significant coefficients on the interactions be-
tween Securitization and LogBankAssets. The correlation between bank age and loan interest and fees as
securitization increases, although in the first specifications for each loan category there is a positive
association between bank age and loan costs as securitization increases. In contrast to Tables 6 and
7 shows an impact of deregulation on loan costs; younger and larger banks charger lower rates after
deregulation, though the effect is dampened by the inclusion of a blunt time trend.
4.3. Economic magnitudes
I next discuss the economic magnitudes of the regression estimates. Table 8 summarizes the eco-
nomic magnitudes of the response of bank’s loan portfolios to both securitization and deregulation.
Panel A summarizes the magnitudes for the period state-wide deregulation presented in the first
regression specification for each loan category in Table 4. Panel B summarizes magnitudes for the per-
iod state-wide deregulation presented in the first regression specification for each loan category in Ta-
ble 5. To calculate the magnitudes for securitization changes I multiply the coefficient estimate for the
8
Including the time trend in the second specification for each loan category causes the coefficient on the interaction between
Securitization and LogBankAssets to become insignificant for C&I loans. Here the time trend picks up the tendency of larger banks to
lend less to C&I loans over this time period, rather than the securitization variable. When I use the ratio of securitized non-
residential real-estate-backed loans as the measure of securitization in the specification in column (2) of Table 4, I find that the
interaction between Securitization and LogBankAssets is negative and significant. This measure of securitization only becomes non-
zero in the late 1980s.
Table 6
Loan pricing during state-wide deregulation.
C&I loan interest and fees/ Personal loan interest and RE loan interest and
C&I loans fees/Personal loans fees/RE loans
(1) (2) (3) (4) (5) (6)
Log(Bank assets) 0.100*** 0.099*** 0.051*** 0.051 0.051*** 0.052***
(8.50) (7.78) (8.49) (7.55) (7.09) (7.99)
Log(Bank age) 0.014*** 0.036*** 0.012*** 0.021 0.006*** 0.012***
(2.63) (4.65) (4.52) (4.53) (3.51) (4.69)
Interstate 0.046 0.037 0.045** 0.041** 0.029** 0.026**
(1.30) (1.06) (2.55) (2.44) (2.29) (2.11)
Intrastate 0.005 0.005 0.015 0.010 0.024 0.028
(0.16) (0.13) (0.51) (0.35) (0.90) (1.10)
Interstate Log(Bank assets) 0.004 0.004 0.004** 0.004** 0.003** 0.003**
(1.26) (1.29) (2.37) (2.50) (2.56) (2.55)
Intrastate Log(Bank assets) 0.0004 0.0001 0.001 0.001 0.002 0.002
(0.12) (0.02) (0.30) (0.32) (0.98) (1.04)
Interstate Log(Bank age) 0.002 0.0001 0.001 0.00004 0.0002 0.001
(1.04) (0.07) (0.72) (0.04) (0.24) (1.22)
Intrastate Log(Bank age) 0.004* 0.001 0.002 0.0004 0.001 0.0001
(1.82) (0.39) (1.40) (0.27) (1.23) (0.14)
Securitization Log(Bank assets) 0.030*** 0.028*** 0.017*** 0.018*** 0.017*** 0.018***
(6.85) (3.73) (6.15) (2.86) (6.50) (5.17)
Securitization Log(Bank age) 0.020*** 0.009 0.010*** 0.002 0.010*** 0.002
(5.47) (1.35) (3.67) (0.50) (4.55) (0.99)
Trend Log(Bank assets) 0.0002 0.00002 0.0001
(0.25) (0.03) (0.19)
Trend Log(Bank age) 0.003*** 0.001 0.001***
(4.09) (2.35) (2.99)
Capital ratio 0.104** 0.099** 0.096*** 0.093*** 0.133** 0.132**
(2.32) (2.19) (3.60) (3.50) (2.52) (2.50)
State bank concentration 0.025 0.023 0.001 0.001 0.001 0.001
(1.21) (1.15) (0.12) (0.07) (0.28) (0.20)
Log(State GDP) 0.014 0.014 0.001 0.001 0.006 0.006
(1.43) (1.44) (0.15) (0.15) (1.18) (1.19)
Log(FHFA house price index) 0.023* 0.020 0.001 0.002 0.020*** 0.020***
(1.93) (1.67) (0.08) (0.20) (3.79) (3.69)
Bank fixed effects? Yes Yes Yes Yes Yes Yes
Year fixed effects? Yes Yes Yes Yes Yes Yes
N 85,325 85,325 86,846 86,846 87,129 87,129
Adjusted R2 0.669 0.669 0.699 0.699 0.764 0.764
bank holding company level, when applicable. Interstate is a dummy variable which takes the value one in the year in which a
state allows out-of-state banks to open branches within the state and in all subsequent years. Intrastate is a dummy variable
which takes the value one in the year in which a state allows banks to branch within state and in all subsequent years. Bank
assets are reported in thousands of year 2000 dollars. Bank age is reported in years. Securitization is the value of real-estate-
backed loans outstanding in a given year that have been securitized as reported by the Flow of Funds accounts divided by the
value of real-estate-backed loans on commercial banks’ balance sheets. Trend is a time trend variable which takes the value one
in 1984 and increases by one in each successive year. Capital ratio is the ratio of bank book equity to book assets. State bank
concentration is the Herfindahl index, i.e., sum of squared shares of bank assets in a given state and year. State GDP is taken
from the Bureau of Economic Analysis and is reported in millions of year 2000 dollars. The FHFA house price index is taken from
the Federal Housing Finance Agency by state and year and is adjusted for inflation. All regressions are estimated using OLS.
Coefficients are reported followed by t-statistics adjusted for clustering at the state level.
*
**
***
interaction of Securitization and LogBankAssets by value of the total change in securitization over the
sample period by the change in bank assets from the 25th to 75th percentile over the sample period.
I perform a similar calculation which replaces LogBankAssets with LogBankAge to calculate the
Table 7
Loan pricing during national deregulation.
C&I loan interest and fees/ Personal loan interest and RE loan interest and
C&I loans fees/Personal loans fees/RE loans
(1) (2) (3) (4) (5) (6)
Log(Bank assets) 0.089*** 0.078*** 0.056*** 0.048*** 0.046*** 0.045***
(5.10) (8.20) (7.48) (8.66) (7.35) (10.80)
Log(Bank age) 0.036*** 0.048*** 0.001 0.029*** 0.002 0.020***
(5.20) (6.02) (0.28) (7.61) (0.66) (8.12)
Riegle-Neal 0.074*** 0.056** 0.048*** 0.050 0.028*** 0.039***
(4.87) (2.14) (3.47) (2.71) (2.96) (3.03)
Riegle-Neal Log(Bank assets) 0.008*** 0.005** 0.005*** 0.003 0.003*** 0.003***
(5.80) (2.45) (4.31) (2.37) (4.03) (2.80)
Riegle-Neal Log(Bank age) 0.003 0.0005 0.003*** 0.003 0.002*** 0.001***
(1.48) (0.22) (3.79) (2.76) (5.47) (3.58)
Securitization Log(Bank assets) 0.025*** 0.014*** 0.016*** 0.008* 0.013*** 0.012***
(3.35) (4.49) (5.49) (1.96) (5.33) (4.01)
Securitization Log(Bank age) 0.004** 0.006 0.011*** 0.012*** 0.009*** 0.006***
(0.68) (1.46) (4.72) (5.56) (5.77) (5.23)
Trend Log(Bank assets) 0.001 0.001*** 0.0001
(1.30) (1.61) (0.35)
Trend Log(Bank age) 0.001* 0.002*** 0.002***
(1.98) (7.65) (9.15)
Capital ratio 0.025 0.027 0.106*** 0.108*** 0.118*** 0.118***
(1.03) (1.09) (5.11) (5.15) (4.33) (4.40)
State bank concentration 0.022 0.021 0.005 0.004 0.006 0.006
(1.49) (1.44) (0.67) (0.62) (1.44) (1.43)
Log(State GDP) 0.001 0.0001 0.001 0.002 0.00004 0.001
(0.08) (0.01) (0.12) 0.4500 (0.01) (0.25)
Log(FHFA house price index) 0.047*** 0.047*** 0.002 0.002 0.004 0.004
(2.68) (2.73) (0.28) (0.25) (0.76) (0.76)
Bank fixed effects? Yes Yes Yes Yes Yes Yes
Year fixed effects? Yes Yes Yes Yes Yes Yes
N 97,051 97,051 99,222 99,222 99,478 99,478
Adjusted R2 0.609 0.609 0.624 0.626 0.712 0.714
bank holding company level, when applicable. Riegle-Neal deregulation is an indicator variable equal to one in the year a bank’s
headquarters state enacts the provisions of the Riegle-Neal Act and in all subsequent years. Bank assets are reported in
thousands of year 2000 dollars. Bank age is reported in years. Securitization is the value of real-estate-backed loans outstanding
in a given year that have been securitized as reported by the Flow of Funds accounts divided by the value of real-estate-backed
loans on commercial banks’ balance sheets. Trend is a time trend variable which takes the value one in 1990 and increases by
one in each successive year. Capital ratio is the ratio of bank book equity to book assets. State bank concentration is the
Herfindahl index, i.e., sum of squared shares of bank assets in a given state and year. State GDP is taken from the Bureau of
Economic Analysis and is reported in millions of year 2000 dollars. The FHFA house price index is taken from the Federal
Housing Finance Agency by state and year and is adjusted for inflation. Coefficients are reported followed by t-statistics
adjusted for clustering by bank or bank holding company.
*
**
***
magnitudes for differences between younger and older banks. To calculate the magnitudes for dereg-
ulation I multiply the coefficients on the deregulator variables interactions with bank size or bank age
by a the sample increase from the 25th to 75th percentiles of bank size and age. The total change in
bank loan portfolios is reported first, followed by the percentage change in parentheses.
Focusing first on the magnitudes of loan portfolio changes in response to securitization in Table 8,
given the increase in securitization over the period 1976–1994, banks at the 75th size percentile de-
creased their lending to C&I loans by 0.021 (10.4%) relative to smaller banks at the 25th size percentile
and also decreased their lending to personal loans by 0.012 (5.09%). On the other hand, banks at the
75th size percentile increased their lending to non-residential real-estate-backed loans by 0.022
Table 8
Magnitudes of loan portfolio changes.
C&I loans/ Personal Non-residential RE backed

Total loans/Total RE loans/Total loans/Total
loans loans loans loans
Panel A – State-wide deregulation, 1976–1994
Change given increase in securitization over sample 0.021 0.012 0.022 (23.7%) 0.002
period and increase in bank size from 25th to 75th (10.4%) (5.09%) (0.39%)
percentile
Change post-deregulation given increase in bank size 0.005 0.002 0.006 (6.92%) 0.001
from 25th to 75th percentile (2.70%) (0.71%) (0.28%)
period and increase in bank age from 25th to 75th (12.8%) (23.3%) (15.4%)
percentile
Change post-deregulation given increase in bank age 0.002 0.001 0.0001 (0.12%) 0.011 (2.59%)
from 25th to 75th percentile (1.07%) (0.46%)
Panel B – National deregulation
Change given increase in securitization over sample 0.013 0.003 0.016 (10.6%) 0.002 (0.42%)
period and increase in bank size from 25th to 75th (8.20%) (2.03%)
percentile
Change post-deregulation given increase in bank size 0.001 0.002 0.003 (2.08%) 0.007
from 25th to 75th percentile (0.81%) (1.42%) (1.26%)
period and increase in bank age from 25th to 75th (9.80%) (3.49%) (3.41%)
percentile
Change post-deregulation given increase in bank age 0.001 0.0005 0.004 (2.57%) 0.007 (1.28%)
from 25th to 75th percentile (0.53%) (0.30%)
The table reports changes in loan portfolio composition given the increase in securitization and post deregulation given an
increase from the 25th to 75th percentile in log bank size and log bank age. Panel A reports the changes during the period of
state-wide deregulation, corresponding to the estimates in the first specification for each dependent variable in Table 4. Post-
deregulation refers to the increase after both intrastate and interstate branching deregulations have been enacted in a state.
Panel B reports the changes during the period of national deregulation, corresponding to the estimates in the first specification
for each dependent variable in Table 5. The changes are calculated by mulitplying the estimated coefficients on the interaction
of the securitization variable or deregulation indicator variable(s) and log bank size and log bank age by the sample difference
between the 75th and 25th percentiles of log bank size and log bank age. The percentage changes are reported in parentheses.
(23.7%) compared to smaller banks at the 25th size percentile.9 We also see sizable magnitudes by bank
age during the period of state-wide deregulation. Given the increase in securitization over the period
1976–1994, banks at the 75th age percentile increased their lending to C&I loans by 0.026 (12.8%) rela-
tive to younger banks at the 25th age percentile and also increased their lending to personal loans by
0.054 (23.3%). On the flip side, banks at the 75th age percentile decreased their lending to non-residential
real-estate-backed loans by 0.036 (38.9%) and also decreased their lending to all real-estate-backed loans
by 0.063 (15.4%) compared to younger banks at the 25th age percentile.
We see similar patterns for the magnitudes of changes in bank loan portfolios in response to the
rise in securitization in the period of national deregulation in Panel B of Table 8. Given the increase
in securitization over the period 1990–2003, banks at the 75th size percentile decreased their lending
to C&I loans by 0.013 (8.2%) relative to smaller banks at the 25th size percentile and also decreased
their lending to personal loans by 0.003 (2.03%). On the other hand, banks at the 75th size percentile
increased their lending to non-residential real-estate-backed loans by 0.016 (10.6%) compared to
smaller banks at the 25th size percentile. Given the increase in securitization over the period 1990–
2003, banks at the 75th age percentile increased their lending to C&I loans by 0.016 (9.8%) relative
to younger banks at the 25th age percentile and also increased their lending to personal loans by
9
Note that the decreases in C&I loans and personal loans for larger banks are not completely offset by increases in real-estate-
backed loans. Agriculture loans make up the difference. Smaller banks decreased their lending to agriculture loans relative to larger
banks in response to securitization. Agriculture loans are arguably a type of hard information loan relative to other types of C&I and
personal loans since the underlying assets of farms, such as crops and demand for crops, are more easily quantifiable because they
are commodities.
0.005 (3.49%) In constrast, banks at the 75th age percentile decreased their lending to non-residential
real-estate-backed loans by 0.029 (19.1%) and also decreased their lending to all real-estate-backed
loans by 0.019 (3.41%) compared to younger banks at the 25th age percentile. The smaller magnitudes
for changes in securitization in Panel B of Table 8 in part reflect that the increase in securitization over
the period 1990–2003 was smaller relative to the increase seen over the earlier period 1994–1976.
Table 8 indicates that deregulatory events have much smaller impact on in terms of economic mag-
nitudes on bank loan portfolios, although after deregulation larger banks make more real-estate-
backed loans and older banks make more C&I and personal loans in both periods.
Table 9 presents the economic magnitudes of the estimated response of loan interest and fees and
loan nonperformance to securitization and deregulation. Panel A of Table 9 presents the magnitudes
during the period of state-wide deregulation using the coefficients from the first regression specifica-
tions for each loan category in Table 6. Panel B of Table 9 presents the magnitudes during the period of
national deregulation using the coefficients from the first regression specifications for each loan cat-
egory in Table 7. Given the increase in securitization over the period of state-wide deregulation, loan
interest and fees fall by an additional 0.060 (51.4%) for C&I loans, 0.035 (69.7%) for personal loans, and
0.035 (81.6%) for real-estate-backed loans for larger banks at the 75th size percentile compared to
smaller banks at the 25th size percentile. During the period of national deregulation, we see that loan
interest and fees fall by an additional 0.028 (28.8%) for C&I loans, 0.018 (40.0%) for personal loans, and
0.014 (42.3%) for real-estate-backed loans for larger banks at the 75th size percentile compared to
smaller banks at the 25th size percentile given the increase in securitization over this period. Older
banks charge more for their loans compared to younger banks as securitization increases. During
the period of state-wide deregulation, loan interest and fees increase by an additional 0.033 (27.9%)
Table 9
Magnitudes of loan pricing changes.
C&I loan interest Personal loan interest RE loan interest

and fees/C&I and fees/Personal and fees/RE
loans loans loans
Panel A – State-wide deregulation
Change given increase in securitization over sample 0.060 (51.4%) 0.035 (69.7%) 0.035 (81.6%)
period and increase in bank size from 25th to 75th
percentile
Change post-deregulation given increase in bank size 0.005 (4.00%) 0.006 (12.6%) 0.0007 (1.53%)
from 25th to 75th percentile
period and increase in bank age from 25th to 75th
percentile
Change post-deregulation given increase in bank age 0.007 (6.35%) 0.003 (6.05%) 0.0007 (1.64%)
Panel B – National deregulation
period and increase in bank size from 25th to 75th
percentile
Change post-deregulation given increase in bank size 0.011 (11.6%) 0.008 (17.0%) 0.005 (14.3%)
period and increase in bank age from 25th to 75th
percentile
Change post-deregulation given increase in bank age 0.004 (3.84%) 0.004 (8.06%) 0.003 (8.03%)
The table reports the changes in loan interest and fees given the increase in securitization and post deregulation given an
increase from the 25th to 75th percentile in log bank size and log bank age. Panel A reports the changes during the period of
state-wide deregulation, corresponding to the estimates in the first specification for each dependent variable in Table 6. Panel B
reports the changes during the period of national deregulation, corresponding to the estimates in the first specification for each
dependent variable in Table 7. The changes are calculated by mulitplying the estimated coefficients on the interaction of the
securitization variable or the deregulation indicator variable(s) and log bank size and log bank age by the increase from the 25th
to 75th sample percentiles for log bank size and log bank age. The percentage changes are reported in parentheses.
for C&I loans, 0.016 (32.3%) for personal loans, and 0.016 (37.2%) for real-estate-backed loans for older
banks at the 75th age percentile compared to younger banks at the 25th age percentile given the in-
crease in securitization over this period. During the period of national deregulation, loan interest and
fees increase by an additional 0.003 (3.43%) for C&I loans, 0.010 (22.4%) for personal loans, and 0.008
(24.1%) for real-estate-backed loans for older banks at the 75th age percentile compared to younger
banks at the 25th age percentile given the increase in securitization over this period.
The magnitudes of the changes in loan interest and fees associated with the deregulation variables
in Table 9 are much smaller than those associated with securitization. Overall the results reported on
the changes in bank loan portfolios associated with the deregulatory events over the period 1976–
2003 and the concurrent rise in securitization of real-estate-backed loans show that larger and youn-
ger banks shift their lending towards real-estate-backed loans in response to an increase in competi-
tion and the increasing ability to take advantage of loan securitization markets. Smaller and older
banks keep more of the loan portfolio in unsecured commercial and personal loans. Securitization en-
ters more strongly as an explanatory factor in the shifts in commercial banks’ loan portfolios over this
time period relative to the deregulatory variables. However, the estimation techniques employed can-
not discern a potentially more complex mechanism through which deregulation encouraged consoli-
dation and larger banks, which may in turn have encouraged innovation in securitization markets.
Even if deregulation encouraged innovation in securitization markets, it is clear from the regression
estimates that trends in securitization are better explanatory factors for shifts in bank loan portfolios
than trends in deregulation.
5. Conclusion
I examine how commercial banks’ loan portfolios respond to the rise of securitization markets and
to deregulation over the period 1976–2003. All banks engage in more real-estate-backed loans and
this increase in lending is proportional to the level of securitization in the loan market. Larger banks
make more to non-residential real-estate-backed loans compared to smaller banks in response to both
securitization and deregulation, as do younger banks. Smaller banks and older banks make more unse-
cured commercial and personal loans. Larger banks and younger banks charge lower interest and fees
in response to the rise in securitization. These findings are consistent with larger and younger banks
making lower cost loans in response to the rise in securitization markets and expanding access to
credit for the borrowers they serve. Smaller banks charge higher interest and fees, suggesting that they
charge borrowers a premium for their greater screening ability relative to larger banks, or that com-
petition in these loan categories is reduced as larger banks are not able to judge credit quality and
make loans to these borrowers as extensively as smaller banks.
These results further suggest that larger banks and younger banks shift their lending towards hard
information loans in response to deregulation and the increasing ability to take advantage of loan
securitization markets. While the analysis cannot strictly identify whether innovation in securitization
markets encourages consolidation and larger bank size after deregulation, or vice versa, it points to a
clear association between larger bank size, securitization, deregulation, and greater lending in real-es-
tate-backed loans, especially by larger banks.
The analysis raises several future research questions. By shifting their lending behavior in reaction
to competition do banks take on extra portfolio risk or are they able to diversify their portfolios prop-
erly within their larger portfolios? What are the systemic risk implications of this shift in bank size
and loan portfolio composition? What are the welfare implications for borrowers who are charged
higher interest rates by smaller banks after deregulation?
Acknowledgments
I thank Philip Strahan (the editor), two anonymous referees, Laurence Ball, Nicola Cetorelli, Giov-
anni Dell’Ariccia, Bob DeYoung, Manju Puri, Evren Ors, Robert Marquez, Hamid Mehran, Philipp Sch-
nabl, Lucy White and Andrew Winton and seminar and conference participants at Duke University,
Johns Hopkins University, the San Francisco Fed, Virginia Tech, the 2008 Annual NYU-New York Fed
Conference on Financial Intermediation, the 2009 Financial Intermediation Research Society Annual
Meeting, the 2009 Chicago Fed Bank Structure and Competition Annual Conference, and the 2010 An-
nual Meeting of the American Finance Association for helpful comments. I thank Jingrui Xie for re-
search assistance. The analysis and conclusions in this paper and do not necessarily indicate
concurrence by the Federal Reserve Board or other members of the Federal Reserve System. All errors
are mine.
Appendix A. Timing of US commercial bank deregulation
This appendix lists the years in which states enacted the three types of deregulation considered in
the empirical analysis. The second column shows the year in which states first allowed interstate bank
branching. The third column shows the year in which states first allowed intrastate bank branching.
The fourth column shows the year in which a state enacted the provisions of the Riegle-Neal Interstate
Banking and Branching Efficiency Act of 1994. The data in the second and third columns are taken
from Kroszner and Strahan (1999). The data in the fourth column are taken from Dick (2006).
State Year allowed interstate Year allowed intrastate Year enacted Riegle-Neal
branching branching provisions
Alabama 1987 1981 1997
Alaska 1982 <1970 1994
Arizona 1986 <1970 1996
Arkansas 1989 1994 1997
California 1987 <1970 1995
Colorado 1988 1991 1997
Connecticut 1983 1980 1995
Delaware 1988 <1970 1995
District of 1985 <1970 1996
Columbia
Florida 1985 1988 1997
Georgia 1985 1983 1997
Hawaii n.a. 1986 1997
Idaho 1985 <1970 1995
Illinois 1986 1988 1997
Indiana 1986 1989 1996
Iowa 1991 n.a. 1996
Kansas 1992 1987 1997
Kentucky 1984 1990 1997
Louisiana 1987 1988 1997
Maine 1978 1975 1997
Maryland 1985 <1970 1995
Massachusetts 1983 1984 1996
Michigan 1986 1987 1995
Minnesota 1986 1993 1997
Mississippi 1988 1986 1997
Missouri 1986 1990 1997
Montana 1993 1990 1997
Nebraska 1990 1985 1997
Nevada 1985 <1970 1995
New Hampshire 1987 1987 1997
New Jersey 1986 1977 1996
New Mexico 1989 1991 1996
Appendix A. Timing of US commercial bank deregulation (continued)

State Year allowed interstate Year allowed intrastate Year enacted Riegle-Neal
branching branching provisions
New York 1982 1976 1996
North Carolina 1985 <1970 1995
North Dakota 1991 1987 1997
Ohio 1985 1979 1997
Oklahoma 1987 1988 1997
Oregon 1986 1985 1995
Pennsylvania 1986 1982 1995
Rhode Island 1984 <1970 1995
South Carolina 1986 <1970 1996
South Dakota 1988 <1970 1996
Tennessee 1985 1985 1997
Texas 1987 1988 1995
Utah 1984 1981 1995
Vermont 1988 1970 1995
Virginia 1985 1978 1995
Washington 1987 1985 1996
West Virginia 1988 1987 1997
Wisconsin 1987 1990 1997
Wyoming 1987 1988 1997
References
Alessandrini, P., Presbitero, A.F., Zazzaro, A., 2010. Bank size or distance: what hampers innovation adoption by SMEs? J. Econ.
Geogr. 10 (6), 845–881.
Berger, A., Saunders, A., Scalise, J., Udell, G., 1998. The effects of bank mergers and acquisitions on small business lending. J.
Finan. Econ. 50, 187–229.
Berger, A., Demsetz, R., Strahan, P., 1999. The consolidation of the financial services industry: causes, consequences, and
implications for the future. J. Banking Finance 23, 135–194.
Berger, A., Miller, N., Petersen, M., Rajan, R., Stein, J., 2005. Does function follow organizational form? Evidence from the lending
practices of large and small banks. J. Finan. Econ. 76, 237–269.
Black, S., Strahan, P., 2002. Entrepreneurship and bank credit availability. J. Finance 57, 2807–2833.
Boot, A., Thakor, A., 2000. Can relationship banking survive competition? J. Finance 55, 679–713.
Brevoort, K.P., Wolken, J.D., 2009. Does distance matter in banking? In: Alessandrini, P., Fratianni, M., Zazzaro, A. (Eds.), The
Changing Geography of Banking and Finance. Springer, Berlin, Heidelberg.
Brevoort, K.P., Holmes J.A., Wolken, J.D., 2010. Distance still matters: The information revolution in small business lending and
the persistent role of location, 1993–2003. Finance and Economics Discussion Series 2010–59, Board of Governors of the
Federal Reserve System.
Canales, R., Nanda, R., 2012. A darker side of decentralized banks: market power and credit rationing in SME lending. J. Finan.
Econ. 105, 353–366.
Degryse, H., Ongena, S., 2007. The impact of competition on bank orientation. J. Finan. Intermediation 16, 399–424.
Dell’Ariccia, G., Marquez, R., 2004. Information and bank credit allocation. J. Finan. Econ. 72, 185–214.
Demyanyk, Y., Van Hermert, O., 2011. Understanding the subprime mortgage crisis. Rev. Finan. Stud. 24, 1848–1880.
DeYoung, R., Hasan, I., Kirchoff, B., 1998. The impact of out-of-state entry on cost efficiency of local commercial banks. J. Econ.
Bus. 50, 191–203.
DeYoung, R., Glennon, D., Nigro, P., 2008. Borrower-lender distance, credit scoring, and loan performance: evidence from
informational-opaque small business borrowers. J. Finan. Intermediation 17, 113–143.
Dick, A., 2006. Nationwide branching and its impact on market structure, quality and bank performance. J. Bus. 79, 567–592.
Dinc, S., 2000. Bank reputation, bank commitment, and the effects of competition in credit markets. Rev. Finan. Stud. 13, 781–
812.
Elsas, R., 2005. Empirical determinants of relationship banking. J. Finan. Intermediation 14, 32–57.
Evanoff, D., Ors, E., 2008. The competitive dynamics of geographic deregulation in banking: implications for productive
efficiency. J. Money, Credit, Banking 40, 897–930.
Farrell, J., Shapiro, C., 1990. Horizontal mergers: an equilibrium analysis. Amer. Econ. Rev. 80, 107–126.
Hauswald, R., Marquez, R., 2006. Competition and strategic information acquisition in credit markets. Rev. Finan. Stud. 19, 967–
1000.
Hellmann, T.F., Murdock, K.C., Stiglitz, J.E., 2000. Moral hazard in banking and prudential regulation: are capital requirements
enough? Amer. Econ. Rev. 90, 147–165.
Jayaratne, J., Strahan, P., 1998. Entry restrictions, industry evolution and dynamic efficiency: evidence from commercial
banking. J. Law Econ. 41, 239–273.
Keeley, M., 1990. Deposit insurance, risk and market power in banking. Amer. Econ. Rev. 80, 1183–1200.
Keys, B.J., Mukerjee, T.K., Seru, A., Vig, V., 2010. Did securitization lead to lax screening? Evidence from sub-prime loans. Quart. J.
Econ. 125, 307–362.
Kroszner, R., Strahan, P., 1999. What drives deregulation? Economics and politics of the relaxation of bank branching
restrictions. Quart. J. Econ. 114, 1437–1467.
Liberti, J., Mian, A., 2009. Estimating the effect of hierarchies on information use. Rev. Finan. Stud. 22, 4057–4090.
Loutskina, E., Strahan, P., 2009. Securitization and the declining impact of bank finance on loan supply: evidence from mortgage
originations. J. Finance 65, 861–889.
Mian, A., Sufi, A., 2009. The consequences of mortgage credit expansion: evidence from the 2007 mortgage default crisis. Quart.
J. Econ. 124, 1149–1496.
Petersen, M., Rajan, R., 1995. The effect of credit market competition on lending relationships. Quart. J. Econ. 110, 407–443.
Petersen, M., Rajan, R., 2002. Does distance still matter? The information revolution in small business lending. J. Finance 57,
2533–2569.
Presbitero, A.F., Zazzaro, A., 2011. Competition and relationship lending: friends or foes? J. Finan. Intermediation 20, 387–413.
Rice, T., Strahan, P., 2010. Does credit competition affect small firm finance. J. Finance 65, 861–889.
Stein, J., 2002. Information production and capital allocation: decentralized vs. hierarchical firms. J. Finance 57, 1891–1921.
Zarutskie, R., 2006. Evidence on the effects of bank competition on firm borrowing and investment. J. Finan. Econ. 81, 503–537.

J. Finan. Intermediation: Rebecca Zarutskie

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

J. Finan. Intermediation: Rebecca Zarutskie

Uploaded by

Copyright:

Available Formats

J. Finan.

Intermediation 22 (2013) 373–396

Contents lists available at SciVerse ScienceDirect

Competition, ﬁnancial innovation and commercial

Article history: I examine how US commercial bank loan portfolios change in

E-mail address: rebecca.zarutskie@frb.gov

1042-9573/$ - see front matter Published by Elsevier Inc.

3. Empirical framework and data

3.1. Empirical framework

LoanPortfolioi;t ¼ a0 þ a1 LogBankAssetsi;t þ a2 LogBankAgei;t þ a3 Securitizationt LogBankAssetsi;t

LoanPRicing i;t ¼ d0 þ d1 LogBankAssetsi;t þ d2 LogBankAgei;t þ d3 Deregulationj;t

3.2. Data and descriptive statistics

Sample period: 1976–2003 (Unique banks or bank holding

3.3. Trends in key variables, 1976–2003

4.1. Bank loan portfolios and securitization, 1976–2003

C&I loans/ Personal loans/ Non-residential RE loans/

4.2. State-wide and national deregulation

(1) (2) (3) (4) (5) (6) (7) (8)

R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396

(1) (2) (3) (4) (5) (6) (7) (8)

R. Zarutskie / J. Finan. Intermediation 22 (2013) 373–396

4.3. Economic magnitudes

C&I loans/ Personal Non-residential RE backed

C&I loan interest Personal loan interest RE loan interest

Appendix A. Timing of US commercial bank deregulation

Appendix A. Timing of US commercial bank deregulation (continued)

You might also like