Risk Management in Financial Institutions

• Date: 01 April 2020
• DOI: http://doi.org/10.1111/jofi.12868
• Author: S. VISWANATHAN,ADRIANO A. RAMPINI,GUILLAUME VUILLEMEY
• Published date: 01 April 2020

THE JOURNAL OF FINANCE •VOL. LXXV, NO. 2 •APRIL 2020

Risk Management in Financial Institutions

ADRIANO A. RAMPINI, S. VISWANATHAN, and GUILLAUME VUILLEMEY∗

ABSTRACT

We study risk management in financial institutions using data on hedging of interest

rate and foreign exchange risk. We find strong evidence that institutions with higher

net worth hedge more, controlling for risk exposures, across institutions and within

institutions over time. For identification, we exploit net worth shocks resulting from

loan losses due to declines in house prices. Institutions that sustain such shocks

reduce hedging significantly relative to otherwise-similar institutions. The reduction

in hedging is differentially larger among institutions with high real estate exposure.

The evidence is consistent with the theory that financial constraints impede both

financing and hedging.

DESPITE MUCH DEBATE ABOUT BANK risk management and its purported failure

during the financial crisis, the basic patterns of risk management in financial

institutions are not known and the main determinants of banks’ risk manage-

ment are not well understood. Since financial institutions play a key role in

the macroeconomy and in the transmission of monetary policy, understanding

their exposure to shocks is essential for monetary and macro-prudential policy.1

Financial institutions can manage the risk exposures arising from lending and

∗Adriano A. Rampini and S. Viswanathan are with Duke University. Guillaume Vuillemey

is with HEC Paris. We thank Manuel Adelino; Juliane Begenau; Markus Brunnermeier; Mark

Flannery; Dalida Kadyrzhanova; Divya Kirti; Kebin Ma; Justin Murfin; Dimitris Papanikolaou;

Roberta Romano; Farzad Saidi; Jo˜

ao Santos; David Scharfstein; Jeremy Stein; Jason Sturgess;

Amir Sufi; Adi Sunderam; Yuri Tserlukevich;James Vickery; seminar participants at Duke, HEC

Paris, Princeton, Georgia State, the Federal Reserve Bank of New York, the NBER Insurance

and Corporate Finance Joint Meeting, the FIRS Conference, the NYU-NY Fed-RFS Conference on

Financial Intermediation, the ETH-NYU Law & Banking/Finance Conference, the Paul Woolley

Centre Conference at LSE, the LBS Summer Finance Symposium, the Barcelona GSE Summer

Forum, INSEAD, Frankfurt School of Finance & Management, the CEPR ESSFM in Gerzensee,

the EFA Annual Meeting, EIEF, the Wharton Conference on Liquidity and Financial Crises, the

AFA Annual Meeting, the Jackson Hole Finance Conference, the NY Fed Conference on OTC

Derivatives, National University of Singapore, Paris-Dauphine, Mannheim, the Conference of

Swiss Economists Abroad, the Bank of Italy, and Zurich; as well as Amit Seru (the Editor); an

anonymous associate editor; and two anonymous referees for helpful comments. Rampini is an

NBER Research Associate and a CEPR Research Fellow. Viswanathan is an NBER Research

Associate. Vuillemey is a CEPR Research Affiliate. The authors have no conflicts of interest to

disclose as identified in The Journal of Finance disclosure policy.

Correspondence: Adriano A. Rampini; e-mail: rampini@duke.edu.

1Following Bernanke and Gertler (1989), Holmstr¨

om and Tirole (1997), Gertler and Kiyotaki

(2010), Brunnermeier and Sannikov (2014), and Rampini and Viswanathan (2019), among others,

analyze the effects of financial institutions’ net worth on the availability of intermediated finance

DOI: 10.1111/jofi.12868

C2019 the American Finance Association

591

592 The Journal of Finance R

deposit-taking activities using financial derivatives. Indeed, financial institu-

tions are the largest users of derivatives, measured in terms of gross notional

exposures. In this paper, we study risk management in U.S. financial institu-

tions using panel data on interest rate and foreign exchange rate derivatives

positions, which represent on average 94% and 5% of the notional value of all

derivatives used for hedging, that is, almost all of the derivatives that financial

institutions use for hedging purposes.2

We show that the net worth of financial institutions is a principal deter-

minant of their risk management both across institutions and within insti-

tutions over time: institutions with higher net worth hedge more, and insti-

tutions whose net worth declines reduce hedging. To study the causal effect

of net worth on hedging, we propose a novel identification strategy using net

worth shocks resulting from loan losses that are due in turn to house price

declines. Using difference-in-differences specifications, we find that institu-

tions that sustain such shocks reduce hedging of both interest rate and foreign

exchange risk substantially relative to otherwise-similar institutions. Using

triple-differences specifications to control for differences in lending opportuni-

ties, we show that the reduction in both types of hedging is differentially larger

for institutions with higher real estate exposure, which sustain larger shocks to

their net worth. We conclude that the financing needs associated with hedging

are a major barrier to risk management.

We use theory to inform our measurement. A leading theory of risk man-

agement argues that firms subject to financial constraints are effectively risk

averse, which gives them an incentive to hedge (see Froot, Scharfstein, and

Stein (1993)). Based on this rationale, Rampini and Viswanathan (2010,2013)

show that when financing and risk management are subject to the same fi-

nancial constraints, that is, when promises to both financiers and hedging

counterparties need to be collateralized, both require net worth and thus risk

management has an opportunity cost that is higher for more constrained firms.

The same risk management concerns arise in the context of financial institu-

tions (see Froot and Stein (1998), Rampini and Viswanathan (2019)). Financial

institutions face a trade-off between lending and risk management, which im-

plies that financially constrained institutions must allocate their limited net

worth between the two. Hedging has an opportunity cost in terms of forgone

lending. The main prediction is that more financially constrained financial in-

stitutions, that is, institutions with lower net worth, hedge less, as the cost of

and real activity.Empirically, Peek and Rosengren (1997,2000) document the effects of banks’ net

worth on lending and real activity and Chodorow-Reich (2014) studies the effects on employment.

Gertler and Gilchrist (1994), Bernanke and Gertler (1995), Kashyap and Stein (2000), and Jim´

enez

et al. (2012) examine financial institutions’ central role in the transmission of monetary policy.

2According to the Bank for International Settlements Derivative Statistics (December 2014),

financial institutions account for more than 97% of all gross derivatives exposures. Financial

institutions’ derivatives positions for hedging include, in addition to interest rate and foreign

exchange derivatives, equity derivatives (0.7%) and commodity derivatives (0.1%). Not included in

these calculations are credit derivatives, as no breakdown between uses for hedging and trading

is available.

Risk Management in Financial Institutions 593

forgoing lending or cutting credit lines is higher at the margin for such insti-

tutions.

We use panel data on U.S. financial institutions, focusing on bank hold-

ing companies (BHCs). We first establish a new basic stylized fact about risk

management in financial institutions: across institutions, financial institutions

with higher net worth hedge interest rate and foreign exchange rate risk to a

greater extent, while within institutions, those institutions whose net worth

declines reduce hedging. We control for risk exposures throughout.

We next test the main prediction of the theory using a novel identifica-

tion strategy by focusing on decreases in financial institutions’ net worth

due to loan losses attributable to house price declines. We use 2009 as the

treatment year and define treated institutions as those with a below-median

mortgage-weighted-average local house price change in the previous two years

or below-median mortgage-weighted-average local housing supply elasticity.3

In difference-in-differences specifications using either definition of treatment,

we find that treated institutions reduce hedging both economically and statis-

tically significantly relative to otherwise-similar control institutions. Indeed,

treated institutions cut hedging by as much as one-half. Similar results obtain

when treatment and control groups are propensity-score matched. We note that

treatment affects institutions’ net worth significantly, but not their risk expo-

sures. This evidence supports the hypothesis that financial constraints are a

major determinant of hedging.

We use a triple-differences specification to control for differences in lending

opportunities. This specification compares the effect of treatment across ter-

ciles of institutions based on their exposure to real estate. We find that treated

institutions with higher real estate exposure reduce both interest rate and

foreign exchange hedging differentially more, by about one-half. Moreover, for

such treated institutions net worth also decreases differentially more, whereas

exposures do not change differentially. These results suggest that the mecha-

nism works through net worth as theory predicts, rather than through lending

opportunities or exposures.

Our results are not likely due to alternative explanations for several rea-

sons. First, since our hedging variables are scaled by total assets, that is, our

results indicate that treated institutions hedge less per unit of asset, a reduc-

tion in lending or assets alone cannot explain our findings. Second, changes

in interest rate risk exposures due to changes in the lending environment are

not likely to drive our results, as we control for exposures throughout and

we show that treatment does not affect exposures. Moreover, our results are

qualitatively similar for foreign exchange hedging, and arguably institutions’

foreign exchange exposures are not directly affected by the domestic lending

3A growing literature uses house prices to instrument for the collateral value of firms (see, e.g.,

Chaney, Sraer, and Thesmar (2012)) and entrepreneurs (see, e.g., Adelino, Schoar, and Severino

(2015)). For financial institutions, several recent studies of the determinants of the supply of

bank lending use a measure of local house prices in a similar spirit to ours, albeit at a more

aggregated level (see Kleiner (2015), Bord, Ivashina, and Taliaferro (2018), Cu ˜

nat, Cvijanovi´

c, and

Yua n (2018)).