Monetary Policy and Global Banking

• Author: FALK BRÄUNING,VICTORIA IVASHINA
• DOI: http://doi.org/10.1111/jofi.12959
• Published date: 01 December 2020
• Date: 01 December 2020

THE JOURNAL OF FINANCE •VOL. LXXV, NO. 6 •DECEMBER 2020

Monetary Policy and Global Banking

FALK BRÄUNING and VICTORIA IVASHINA∗

ABSTRACT

When central banks adjust interest rates, the opportunity cost of lending in local

currency changes, but—absent frictions—there is no spillover effect to lending in

other currencies. However, when equity capital is limited, global banks must bench-

mark domestic and foreign lending opportunities. We show that, in equilibrium, the

marginal return on foreign lending is affected by the interest rate differential, with

lower domestic rates leading to an increase in local lending, at the expense of a re-

duction in foreign lending. We test our prediction in the context of changes in interest

rates in six major currency areas.

FOREIGN (“GLOBAL”) BANKS PLAY AN important role in many countries. Ac-

cording to the Bank for International Settlements (BIS), as of June 2015, Eu-

ropean and Japanese banks’ claims on U.S. nonbank firms were USD 1.61 and

0.72 trillion, respectively. DealScan data indicate that foreign banks help orig-

inate close to a quarter of all syndicated corporate loans in the United States.

Similarly,U.S. banks are important lenders abroad: as of June 2015, U.S. banks

held the equivalent of USD 0.74 and 0.11 trillion in claims on European and

Japanese nonbank companies, respectively. More generally, it is estimated that

∗Bräuning is with the Federal Reserve Bank of Boston and Ivashina is with Harvard Business

School, NBER, and CEPR. We thank seminar participants at the AFA 2017, Bank of Canada,

Banque de France, Copenhagen Business School, Duke University, European Central Bank, FRIC

Annual Conference, Federal Reserve Bank of Boston, Federal Reserve Bank of Chicago, Fed-

eral Reserve Bank of Cleveland, Federal Reserve Board, Global Research Forum on Interna-

tional Macroeconomics and Finance 2016, IBEFA Summer Meeting 2016, IMF Spillover Work-

shop, NBER Summer Institute 2016, Nova School of Business and Economics, UIUC, University

of Miami, University of Porto, and WCWIF 2016 for helpful comments. Weare particularly grate-

ful to François Gourio; Stefan Nagel (Editor); Ali Ozdagli; Joe Peek; Jeremy Stein; Jenny Tang;

Christina Wang;two anonymous referees; and our discussants Morten Bech, Nicola Cetorelli, Stijn

Claessens, Ricardo Correa, Linda Goldberg, Michael Hutchison, Jose Lopez, Teodora Paligorova,

and Skander Van den Heuvel for detailed feedback. Botao Wu,Kovid Puria, and Sam Tugendhaft

provided remarkable research assistance. The views expressed in this paper are those of the au-

thors and do not necessarily represent the views of the Federal Reserve Bank of Boston or the

Federal Reserve System. We have read The Journal of Finance’s disclosure policy and have no

conflict of interest related to this research.

Correspondence: Victoria Ivashina, Harvard Business School, Baker Library,Bloomberg Center

233, Boston, MA 02163; e-mail: vivashina@hbs.edu.

DOI: 10.1111/jofi.12959

© 2020 the American Finance Association

3055

3056 The Journal of Finance®

foreign banks account for about 10% of the assets of the French and Italian

banking sectors (World Bank (2008)).

Given the economic significance of global banks, questions have been raised

about their role in the propagation of economic shocks from one country to an-

other. This is the first paper that provides a framework that shows how mon-

etary policy in one country affects loan supply in different currencies through

the balance sheet of global banks. A standard effect of monetary policy is that,

by setting the interest rate on reserve assets, it affects the opportunity cost

of lending: A higher domestic policy rate increases the required marginal rate

of return on domestic lending. We show that if a global bank is capital con-

strained, that is, if it needs to consider the allocation of a fixed amount of equity

to back lending in multiple currencies—monetary policy in one country will af-

fect its loan supply in all currencies. This is because, in equilibrium, marginal

returns (expressed in the same currency) on domestic and foreign lending have

to be equal. We present a model that formalizes this point and derives testable

predictions that guide the empirical analysis. Overall, the central insight of

the paper is that global banks’ lending decisions in different currencies are

interlinked and respond to both foreign and domestic monetary policy shocks.

In addition to lending, banks have a global approach to their liquidity man-

agement: They prefer to hold liquidity in countries in which reserve assets,

such as central bank reserves or government bonds, carry a higher risk-

adjusted yield. Using U.S. Call Report and BIS data, we find a strong positive

relationship between foreign reserve holdings and the difference in the interest

on excess reserves (IOER) rate between foreign and bank’s domestic market.

While global liquidity management is not essential to derive cross-currency ef-

fects of monetary policy on lending, given that these liquidity flows are sizable

and respond to monetary policy, we incorporate global banks’ liquidity man-

agement in our model.

Our empirical results can be divided into two parts: (i) aggregate macro ev-

idence and (ii) firm- and loan-level micro evidence. In line with the central

mechanism of the model, we show that foreign banks in the United States that

are headquartered in countries where the interest rate has been lowered de-

crease their lending to U.S. firms by up to 10% per 25-basis-point increase in

the IOER rate differential. The results are robust to using U.S. monetary pol-

icy shocks and alternative measures of interest rate differentials. Using BIS

data, we show that similar patterns emerge in a cross-country setting: We find

a decrease in foreign-currency cross-border claims on foreign firms of 1.2% per

25-basis-point increase in the IOER rate differential between the foreign cur-

rency and the currency of a bank’s home country. We estimate this effect after

controlling for concurrent movements in the spot exchange rate. Consistent

with our model predictions and our findings for the United States, the reduc-

tion in foreign lending is stronger after we take into account the appreciation

of the foreign currency.

Our second set of results is based on loan-level data, which are crucial for

a tight empirical identification of the mechanism at work. For this purpose,

we employ loan-level data on syndicated loan originations by global banks in

Monetary Policy and Global Banking 3057

six major currencies from 2000 to 2016. At the bank level, we find that, in a

given quarter, there is a larger contraction in credit in currencies that carry a

higher IOER differential with respect to the bank’s home country. With regard

to lending volume, a 25-basis-point increase in the IOER rate difference is

associated with a roughly 1-percentage-point reduction in the share of foreign

currency lending (relative to the bank’s lending in both foreign and domestic

currencies), which corresponds to a reduction of 2% relative to the mean share

of foreign currency lending. The impact on the lending share based on the

number of loans is equally sizable. In line with our mechanism, the effect of

the IOER rate differential on foreign currency lending is particularly strong

for banks with a low equity-to-assets ratio and is not confined to the U.S. dollar

but rather holds for all major currencies.

A similar picture arises when we conduct bank-firm-level regressions. Here

we find that, among foreign banks, the propensity to lend (extensive mar-

gin) and the size of the loan commitment (intensive margin) are negatively

related to the difference in interest rates set by the monetary authorities in

the host country and a bank’s home country. In particular, after controlling

for borrower-quarter and lender-quarter fixed effects, we find a decrease of

roughly 1.6% in the probability of lending (relative to the mean lending proba-

bility) and 3.6% in lending volume for a 25-basis-point differential in IOER. Be-

cause the loans in our sample are syndicated, each loan has several banks with

large commitments. This setup allows us to include both borrower-quarter and

lender-quarter fixed effects, which account for factors related to time-varying

loan demand and bank behavior, such as a bank’s response to changes in eco-

nomic conditions in the bank’s home country to focus on the differential sup-

ply of credit in different currencies. For example, the shock to the Japanese

banks analyzed in Peek and Rosengren (1997,2000) would lead to an over-

all contraction in credit by Japanese banks, which would be subsumed by the

lender-quarter fixed effects.

We also estimate changes in aggregate credit supply at the domestic firm

level following foreign monetary policy shocks to provide evidence on whether

the reduction in credit is binding for individual firms. We find that firms that,

in the past, had a larger share of foreign global banks in the syndicate and

that subsequently experienced a monetary easing policy shock in their home

country face a stronger contraction in credit than other firms. Economically, we

find that a one-standard-deviation increase in the past share of foreign global

banks is associated with a 12% lower probability of obtaining a loan and with

a 5% reduction in the volume of loans granted after an expansionary monetary

policy shock in the foreign bank’s home country.

Our paper contributes to several strands of the literature. First, we extend

a growing body of research on global banks and the role they play in trans-

mitting shocks across borders. In particular, our work is closely related to em-

pirical work by Cetorelli and Goldberg (2012) and Morais et al. (2019).1These

1In addition to the seminal contributions of Peek and Rosengren (1997,2000), other research in

this area includes Acharya and Schnabl (2010), Chava and Purnanandam (2011), Schnabl (2012),