Determinants of bank loan charge-off rates: evidence from the USA

Author:Amit Ghosh
Position:Department of Economics, Illinois Wesleyan University, Bloomington, Illinois, USA
Pages:526-542
SUMMARY

Purpose Using data on 5,176 commercial banks in the USA for the period 1999Q1-2016Q3, the present study aims to examine the underlying determinants of loan charge-off rates. Design/methodology/approach The study uses panel data fixed-effects estimation methodology. Findings Greater regulatory capital, more diversification, higher profits and cost efficiency reduce charge-... (see full summary)

 
FREE EXCERPT
Determinants of bank loan charge-
orates: evidence from the USA
Amit Ghosh
Department of Economics, Illinois Wesleyan University, Bloomington, Illinois, USA
Abstract
Purpose Using data on 5,176 commercial banks in the USA for the period 1999Q1-2016Q3, the present
study aimsto examine the underlying determinants of loan charge-offrates.
Design/methodology/approach The study usespanel data xed-effects estimation methodology.
Findings Greater regulatorycapital, more diversication, higher prots and cost efciencyreduce charge-
off rates. On the contrary, a higher share of loans in banksasset portfolio and a higher share of real estate
loans have a detrimental impact on loan performance. Moreover, strong US macroeconomic fundamentals
reduce loancharge-offs. Finally, real estate loancharge-offs are most sensitive to balancesheet conditions.
Practical Implications Consistent with Basel-III regulation, the resultsunderscore the importance of
banks to remain well capitalized. Greater tier-1 capital refrains banks from riskylending practices, thereby
improving their loan performance.It is also important that banks maintain a diversied income streamand
earn higher protability.Finally, managerial inefciencies leading to higher non-interestexpense needs to be
reduced toimprove loan performance.
Originality/value Although a burgeoning body of literature has examined the underlying factorsthat
affect poor quality loans in both the USA and elsewhere, fewerstudies have focused on loan performance.
From the perspective of banking regulation and fostering banking stability, determining the factors that
affect loan charge-offs is extremely crucial to identify channels through which loan performance is either
worsened or improved.If we understand poor loan performance, we can use thatknowledge to anticipate the
possibilityof bankruptcy.
Keywords Panel data, GrammleachBliley act, US commercial banking, Net loan charge-off rates,
Tier-1 capital ratio
Paper type Research paper
1. Introduction
A key feature of the recent nancial crisis was deterioration in the performance of loans
given by banks. Lax lending standards by bank managers,as well as poor loan monitoring
mechanisms, resulted in sharp rise loan charge-off rates, as well as escalations in non-
performing loans(henceforth, NPLs) of banks in the USA (Ghosh,2015, 2017).
In the USA, banks currently operate under what is called a incurred loss model.Inthis
model, loan losses are not recognized in the nancial statements until a loss has become
probable based on information available as of the date of the nancial statement. So, when
banks underwrite new loans some portions of these loansare allocated to reserves for loan
losses. Additions to these reserves come from an income statement item, loan loss
provisions, which are set aside as new loans are made and as potential problem loans are
identied. Often the available information will indicate that there have been losses to the
loan portfolio but those losses cannot yet be attributed to specic assets. When the losses
can be attributed to individual assets (or the losses are conrmed), the bank is required to
charge-off the loan (OCCs Handbook,2012)[1].
Once interest becomes overdue on a loan, it is designated past duebut still accruing
interest, as if interestwas still being paid. If a loan is past due, 90 days or more and accruing
JFRC
26,4
526
Journalof Financial Regulation
andCompliance
Vol.26 No. 4, 2018
pp. 526-542
© Emerald Publishing Limited
1358-1988
DOI 10.1108/JFRC-02-2018-0021
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1358-1988.htm
interest, then it must be reported to the regulatory authorities. At some stage, the loan is
given a non-accrual status, indicating that the bank no longer continuesto count the unpaid
interest. The nal stage places the loaninto the default category. Then the loan is written off
through the accounting entry charge-off (see discussions in Miller and Noulas, 1994). In
some cases, a bank may have recoveredon a loan than it had anticipated when it charged-off
part of the loan. The difference between charge-offs and recoveries, that is, net charge-offs,
enters as a reconciliation item in thebalance sheet. Hence, high rates of net loan charge-off
indicate a poorly performingloan.
Using an extensive dataset of all 5,176 banks that were operational from 1999Q1 till
2016Q3, the present study examines the inuence of different balance-sheet conditions and
US macroeconomic fundamentals on banks loan performance. From the perspective of
banking regulation, determining the underlying factors that affect loan charge-offs is
extremely crucialto identify channels through which loan performanceis either worsened or
improved. This enables banks to mitigate factors that deteriorate loan performance. Any
policy response bybanking regulatory authorities in the wake of risingloan charge-off rates
requires an understanding of their determinants. If we understand poor loan performance
and the factors that inuence it, we can use that knowledgeto anticipate bank failures, and
hence promote banking sectorstability.
The study also examinessector-specic charge-offrates in real estate and individual loan
categories. Aggregate loans can mask important differences between the heterogeneous
inuences of different balance-sheet variables on the performance of different specic
categoriesof loans. This study uncovers such relationships by examiningsector-specicloan
charge-offs. Thisinformation could be useful for banks thatspecialize in a specic category
of loan.
The rest of the paper proceeds as follows: Section 2 succinctly surveys the relevant
literature. Section 3 shows trends and patterns in banks loan charge-off rates in the USA.
Section 4 discusses the theoretical foundations, empirical model and explains the results.
Section 5 performs severalsensitivity analyses. Finally, Section 6 concludes with thepolicy
implications.
2. Extant literature
Miller and Noulas (1994), is the one of the very few studies that has examined net loan
charge-offs of US commercial banks in the late 1980s.The authors nd a rise in the share of
total loans to total assets to lower net charge offs while an increase in the ratio of total
deposits to total assets to reduce loan charge offs. Keeton and Morris (1987) examine net
loan charge-off rates in 1984-1985 for a sampleof about 2,500 banks in the Federal Reserve
District of Kansas City. They nd substantial variation in loan performance because of
differences in both local economic conditions and also because of the poor performance of
certain industries. In a somewhat different but related context, Shaffer (1998) examine the
role of bank age, structureand local demographic factors in inuencing loan charge-offrates
using MSA-level data from 1986-1995. A higher number of banks increase loan charge off
rates. Likewise, charge-off rates are higher for newer banks. Rajan (1994) examine loan
charge-offs for banks in the Federal Reserve Bank of Boston district for 1986-1992. The
share of real estate loans in totalloans signicantly increase loans charge off to total assets.
However, an absence is such studies is the role of key balance sheet factors like bank
capitalization, protability or incorporating the underlying macroeconomic environment
in inuencing loan performance. The present study distinguishes itself from the
aforementioned one by using a wider range of balance sheet and macroeconomic
determinants, as well an extensiveempirical treatment covering the most recent time period.
Determinants
of bank loan
charge-o
rates
527

To continue reading

REQUEST YOUR TRIAL