Estimating the direct impact of bank liquidity shocks on the real economy: Evidence from letter‐of‐credit import transactions in Colombia

• Date: 01 November 2019
• Published date: 01 November 2019
• DOI: http://doi.org/10.1111/roie.12433

1510

|

wileyonlinelibrary.com/journal/roie Rev Int Econ. 2019;27:1510–1536.

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

Money is fungible. So are most types of loans. There is no guarantee that a borrower will always abide

by the purpose of a loan as stated on the loan request. This makes it an almost elusive quest to fully

identify the direct link between loans and economic activity. A rare exception is a letter of credit. Its

sole purpose—to pay for an import transaction specified therein—connects the financial sector with

Received: 19 September 2017

|

Revised: 2 July 2019

|

Accepted: 8 July 2019

DOI: 10.1111/roie.12433

ORIGINAL ARTICLE

Estimating the direct impact of bank liquidity

shocks on the real economy: Evidence from

letter‐of‐credit import transactions in Colombia

JaeBinAhn1

|

MiguelSarmiento2,3

1Graduate School of International

Studies,Seoul National University, Seoul,

Korea

2Financial Stability Department,Banco de

la República, Bogotá, Colombia

3European Banking Center,Tilburg

University, Tilburg, Netherlands

Correspondence

JaeBin Ahn, Graduate School of

International Studies, Seoul National

University, 1 Gwanak‐ro, Gwanak‐gu,

Seoul, Korea.

Email: jaebin.ahn@snu.ac.kr

Abstract

This study identifies and provides an estimate of the im-

pact of bank liquidity shocks on real economic activity by

exploring letter‐of‐credit import transactions in Colombia

during the 2008 to 2009 global financial crisis. The detailed

dataset on letter‐of‐credit transactions allows for exploiting

within‐importer–exporter variation across issuing banks.

The study finds substantial effects of bank liquidity shocks

on letter‐of‐credit import transactions: banks that were more

vulnerable to adverse liquidity shocks—proxied by the

ex ante reliance on wholesale funding or borrowings from

foreign banks—reduced letter‐of‐credit issuances more in

both intensive and extensive margins. The study also con-

firms that it had real effects: importer–exporter pairs that

relied more on letter‐of‐credit transactions experienced a

greater reduction in their total imports in response to ad-

verse bank liquidity shocks.

JEL CLASSIFICATION

F1; F4; G1; G2

|

1511

AHN et Al.

the real sector, and thus provides a unique prism through which to view the impact of financial shocks

on real economic activity. The analysis in this paper draws on that information.1

Since the global financial crisis, academic research has centered on two main transmission chan-

nels through which the financial crisis led to what has been called the Great Recession. One channel

is the global transmission of financial shocks across countries, and the other is the subsequent trans-

mission of financial shocks to the real economy. This paper aims to identify the second transmission

channel, taking bank liquidity shocks transmitted by the first channel as a given.2

Banks can facilitate import transactions by issuing a letter of credit on behalf of importers par-

ticularly when they have greater capacity to mitigate contracting and financial frictions. However,

adverse liquidity shocks might affect banks’ ability to extend credit, thereby cutting back on the sup-

ply of letters of credit.3

This would directly affect importers’ ability to buy abroad, resulting in real

economic impacts—that is, import reductions. Based on this notion, this study identifies and provides

an estimate of the direct impact of bank liquidity shocks on real economic activity by exploring letter‐

of‐credit import transactions in Colombia during the 2008 to 2009 global financial crisis.4

The identification strategy exploits the extremely detailed nature of Colombian import transaction

data, which provide the name of importers, exporters, and banks for each letter‐of‐credit transaction.

Together with the fact that there are many importer–exporter pairs that use multiple issuing banks for

letter‐of‐credit transactions, the data allow for controlling for importer–exporter specific shocks, and

for exploring within‐importer–exporter variation in letter‐of‐credit transactions across issuing banks.

It finds a substantial impact of bank liquidity shocks on the supply of letters of credit: banks that

were more vulnerable to adverse liquidity shocks—proxied by the reliance on wholesale funding—re-

duced letter‐of‐credit issuances by around 80 percent in intensive margins, and were about 13 times

more likely to discontinue the issuance in extensive margins between the pre‐ and post‐crisis periods.5

One potential concern about this estimation strategy is that it reflects only relative changes in the

supply of letters of credit across banks for a given importer–exporter pair. It is plausible that the results

were mainly driven by switching across issuing banks within an importer–exporter pair, without any

significant effect on total imports via letters of credit. Similarly, it is conceivable that importer–ex-

porter pairs switched to alternative payment methods in response to a reduced supply in letters of

credit, without resulting in overall reductions in import transactions. The study thus undertakes fur-

ther analysis to check these possibilities by investigating within‐importer variation across exporters

for total import (including but not limited to letter‐of‐credit) transactions. It confirms that there is no

evidence of such switching across banks or payment methods, and that the estimated impact of bank

liquidity shocks on total import transactions is indeed stronger for those who relied more on letters of

creidt in the pre‐crisis period.

All these findings are robust to an alternative measure of bank liquidity shocks based on ex ante

exposure to borrowings from foreign banks as well as to an alternative time frame addressing season-

ality issues. Notwithstanding the robust and clear identification results, a strong caution is warranted

against deriving a conclusion that the “letter‐of‐credit” channel identified in this paper was mainly

responsible for the import collapse in Colombia. This is because letters of credit accounted for only

about 4% of total import transactions in Colombia during the period.6

In this regard, it needs to be

reemphasized that the main contribution of this paper should be interpreted as identifying one specific

channel through which trade finance shocks affect import transactions. This could still serve as the

benchmark to deliver potentially more meaningful aggregate implications to those countries that rely

more on letters of credit for their international transactions.

This paper is related to two strands of literature. First, it builds on the literature that studies the

bank lending channel through which financial shocks affect the real economy.7

Identifying the real

transmission of financial shocks has often been quite challenging not least because of the difficulty in