Funding Liquidity without Banks: Evidence from a Shock to the Cost of Very Short‐Term Debt

• Author: LINA CARDONA‐SOSA,PHILIP E. STRAHAN,FELIPE RESTREPO
• Date: 01 December 2019
• Published date: 01 December 2019
• DOI: http://doi.org/10.1111/jofi.12832

THE JOURNAL OF FINANCE •VOL. LXXIV, NO. 6 •DECEMBER 2019

Funding Liquidity without Banks: Evidence from

a Shock to the Cost of Very Short-Term Debt

FELIPE RESTREPO, LINA CARDONA-SOSA, and PHILIP E. STRAHAN∗

ABSTRACT

In 2011, Colombia instituted a tax on repayment of bank loans, which increased the

cost of short-term bank credit more than long-term credit. Firms responded by cutting

short-term loans for liquidity management purposes and increasing the use of cash

and trade credit. In industries in which trade credit is more accessible (based on U.S.

Compustat firms), we find substitution into accounts payable and little effect on cash

and investment. Where trade credit is less available, firms increase cash and cut

investment. Thus, trade credit provides an alternative source of liquidity that can

insulate some firms from bank liquidity shocks.

HOW DO FIRMS MANAGE LIQUIDITY? Most of the economics and finance literature

focuses on the banking system, with banks acting as the primary supplier

of liquidity and payment mechanisms between counterparties. Cash, demand

deposits, and credit lines—or equivalently, very short-term bank loans—are

fundamental payment products supplied almost uniquely by banks (Kashyap,

Rajan, and Stein (2002), Gatev and Strahan (2009)). Many firms use credit

lines to smooth out liquidity needs over time, drawing funds when payments

need to be made (e.g., to make payroll or to pay for supplies) and repaying those

funds when payments are received (e.g., after realizing sales receipts). Demand

deposits and credit lines differ mainly in that the latter include a credit aspect

(they are not prefunded by the customer), whereas the former do not. Firms

with limited access to lines of credit use deposits (out of their buffer stock of

cash) to make payments.1

∗Restrepo is with the Ivey Business School at Western University. Cardona-Sosa is with Banco

de la Rep´ublica. Strahan is with Boston College and NBER. [The views expressed in this paper are

those of the authors alone and do not necessarily represent those of the institutions with which

they are affiliated.] We thank our discussants (Jean-Noel Barrot, Thomas Mosk, and Mitchell Pe-

tersen); Amit Seru (the Editor); an anonymous Associate Editor; two anonymous referees; seminar

participants at the Bank of Canada, University of Pittsburg, Federal Reserve Bank of New York,

SFS Cavalcade 2017, FIRS Conference 2017, WEAI 13th International Conference, AFA Annual

Meeting 2018, Colombia’s Central Bank, and EAFIT University for helpful comments. We would

also like to thank Lina Maria Montoya, Maria Luisa Mu˜

noz, and Ana Maria Rios for helpful in-

sights on the institutional background. The authors do not have any potential conflicts of interest

to disclose, as identified in the Journal of Finance Disclosure Policy. The contents of the article do

not bind Banco de la Rep ´ublica or its Board of Directors.

[Correction added on 02 August 2019, after first online publication: in the second footnote, “Whether

the observed change in cash observed leads to permanently lower investment” was corrected to

“Whether the observed change in cash leads to permanently lower investment”.]

1Banks also supply term loans as a source of credit to firms, usually structuring them with a

short maturity as a key contracting tool to help solve information and monitoring problems. Indeed,

DOI: 10.1111/jofi.12832

2875

2876 The Journal of Finance R

In this paper, we analyze how firms’ liquidity management changes following

an exogenous increase in the relative price of very short-term bank debt. We

find that some firms substitute into cash, whereas others substitute into trade

credit as an alternative source of liquidity. In industries in which trade credit

is less available, the increase in cash is large. As a result, these firms cut both

long-term and short-term investment. In contrast, in industries with greater

access to trade credit, firms increase net accounts payable, but neither their

cash balances nor their investment change after the shock. Taken together,

these results lead to two implications. First, using cash can be a costly substi-

tute for bank liquidity facilities. When external funds are costlier than internal

sources of financing, firms that need to increase their stock of cash (because

short-term bank credit has become expensive) must do so by drawing funds

away from investment.2Second, for firms in some industries, trade credit pro-

vides liquidity services that can substitute for bank-provided liquidity without

distorting real decisions.

Toconduct our analysis, we consider the empirical setting of Colombia, whose

two unique institutional features allow us to identify how an increase in the

cost—or equivalently, a decrease in the supply—of bank liquidity affects firms.

First, in Colombia, “preferential clients,” that is, clients with annual sales above

a specific sales cutoff set by banks, have greater access to bank credit than do

smaller firms. In particular, these preferential clients have access to Treasury

Facilities—very short-term bank loans—which they can use to manage their

liquidity needs but not to finance real investment due to the facilities’ very

short maturity.3Second, in 2011 Colombia adopted a new law that taxes each

payment on a bank loan. This regulatory change made short-maturity Treasury

outside of real estate and other projects collateralizable by hard assets, most bank loans are short

term, which implies that borrowers must roll them over frequently.This practice provides banks the

option to restrict or deny credit, thereby improving borrowers’ ex ante incentives. Diamond (1991)

offers the seminal theoretical treatment of debt maturity and argues that small and information-

intensive firms are often only able to borrow short term. Stohs and Mauer (1996) provide empirical

support for this prediction.

2Whether the observed change in cash leads to permanently lower investment is hard to assess

from our results, but holding more cash can increase firms’ agency costs (Jensen (1986), Yun

(2009)). Sufi (2009) argues that many firms are forced to use cash to self-insure against liquidity

shocks despite such costs because their access to credit lines is limited by low cash flow, a key

contractual control mechanism that banks use to monitor borrowers. Credit lines expose banks to

substantial credit as well as liquidity risk because firms may want to draw funds when cash flow

is low due to poor fundamentals. Cash-flow-based covenants help alleviate this risk to lenders, but

imply that credit lines do not provide firms insurance against low cash flow in bad states. In line

with this argument, Lins, Servaes, and Tufano(2010) survey CFOs from the largest firms across 29

countries and conclude that credit lines are a more important source of liquidity than cash among

their sampled firms, but at the same time firms hold cash buffers to self-insure against bad states.

Thus, firms often use credit lines to take advantage of future uncertain investment opportunities.

3Smaller firms also have access to bank credit lines for liquidity management, but these come

with much more restricted access and tighter covenants. Small firms can also access short-term

debt (say a 30-day loan), but supply is more constrained and conditions (rates, collateral, and

covenants) less favorable. Operationally, firms with sales below the given sales threshold would

bank out of a branch office, where firms with higher sales are assigned to a loan officer/account

manager who gives this customer preferential treatment.

Funding Liquidity without Banks 2877

Facilities (and other very short-term debt instruments) prohibitively expensive,

but had little effect on the cost of long-term bank loans (due to the low tax

rate of 0.4%). Given these two features, we focus our empirical analysis on

large firms with access to Treasury Facilities, and construct a difference-in-

differences identification strategy to exploit variation in this access, as heavy

users of Treasury Facilities are expected to be most affected by the tax change.

Firms with access to Treasury Facilities (as proxied by preregulatory change

use of short-term bank debt issuance with original maturity less than 60 days)

constitute our treatment sample; firms without such access are taken to be the

control sample.

We report three core results. First, the use of short-term bank borrowing for

liquidity falls sharply after 2011 for firms exposed to the tax shock. Second, the

decline in short-term bank borrowing leads to a substitution into both trade

credit and cash. In particular,both net accounts payable and cash increase after

2011, with the increases being larger for firms exposed to the tax shock. Third,

the decline in bank-supplied liquidity leads to a decrease in both long-term

and short-term investment. Specifically, capital expenditures and a proxy for

investment in research and development (R&D) both decline for some treated

firms. In addition, short-term inventory investment in raw materials and un-

finished goods declines. Across industries, however, those with high access to

trade credit substitute into accounts payable and experience no investment

declines. Moreover, the substitution to trade credit is complete in the sense

that there is no change in cash for firms in these sectors. In contrast, firms

in industries with greater frictions in their ability to borrow from suppliers

experience no change in accounts payable but a large increase in cash and a

large decrease in investment (in both long-term and short-term assets). Thus,

in these segments the decline in bank liquidity spills over to real decisions.

A decline in long-term investment likely occurs because firms need to divert

funds to increase their buffer stock of cash to manage liquidity. This increase

in cash displaces investment. Whether this decline is permanent is difficult to

determine, given our short time window of just four years following treatment.

The effect on inventories, however, is more direct because firms use short-term

bank debt (along with trade credit) as a key source of funds for investments in

working capital.4

As we describe in Section I, the tax regime in Colombia distorts the rela-

tive after-tax prices of tools that firms use to manage liquidity—bank credit

facilities, trade credit, and cash—in the pre- and post-2011 periods. Our pa-

per thus compares two equilibria. In the first regime (pre-2011), many firms

minimize taxes by using relatively tax-advantaged very short-term bank debt,

rather than cash or trade credit, to smooth payments. In this regime, efforts to

avoid the tax distort firm behavior by encouraging firms to use short-term debt.

In the second regime (after 2011), the tax change reduces this distortion by in-

creasing the relative price of short-term bank debt (so that its tax treatment

becomes the same as that on the other payment mechanisms). This increase

4For completeness, we also check for effects on investment in acquisitions but find no changes.