Consumer Default, Credit Reporting, and Borrowing Constraints

• Author: GABRIEL NATIVIDAD,MARK J. GARMAISE
• Published date: 01 October 2017
• Date: 01 October 2017
• DOI: http://doi.org/10.1111/jofi.12522

THE JOURNAL OF FINANCE •VOL. LXXII, NO. 5 •OCTOBER 2017

Consumer Default, Credit Reporting,

and Borrowing Constraints

MARK J. GARMAISE and GABRIEL NATIVIDAD∗

ABSTRACT

Why do negative credit events lead to long-term borrowing constraints? Exploiting

banking regulations in Peru and utilizing currency movements, we show that con-

sumers who face a credit rating downgrade due to bad luck experience a three-year

reduction in financing. Consumers respond to the shock by paying down their most

troubled loans, but nonetheless end up more likely to exit the credit market. For

a set of borrowers who experience severe delinquency, we find that the associated

credit reporting downgrade itself accounts for 25% to 65% of their observed decline in

borrowing at various horizons over the following several years.

THE RECOVERY FROM THE RECESSION OF 2008 has been anemic. An influen-

tial stream of research has attributed this sustained period of lackluster

growth to financial constraints that bind heavily indebted consumers and limit

their participation in the economy (Mian and Sufi (2010,2011), Hall (2011),

Eggertsson and Krugman (2012), Mian, Rao, and Sufi (2013), and Guerrieri

and Lorenzoni (2016)). What explains the power and duration of these finan-

cial constraints? While it is clear that many households experienced negative

credit events during the recession, what is less apparent is precisely why these

events might have had such long-lasting consequences for access to loans. In

this paper, we provide an empirical analysis of the impact of unfavorable credit

events on future financing for consumers, with a particular focus on the role

played by formal credit reporting systems.

Our empirical setting is a broad panel of consumer loans in Peru. We begin

by showing that in Peru, as in other countries, negative credit events are

associated with serious medium-term restrictions to credit access. Our main

interest, however, is in analyzing the determinants of this relationship. To

that end, we exploit features of local banking regulations to identify exogenous

shocks to the risk classifications of some borrowers. These shocks have no

information content, yet we show that they lead to a three-year reduction in

financing for the affected consumers. That is, consumers who experience a

∗Mark J. Garmaise is at UCLA Anderson. Gabriel Natividad is at Universidad de Piura. This

document has been officially screened by Superintendencia de Banca, Seguros, y AFP del Peru

(SBS) to ensure that no confidential information is revealed. We are grateful to SBS for access

to the banking data, and to the Editor (Kenneth Singleton), the Associate Editor, two anonymous

referees, and numerous seminar audiences for useful comments. Guillermo Ramirez-Chiang and

Renzo Severino provided excellent research assistance. The authors do not have any potential

conflicts of interest to disclose, as identified in the JF Disclosure Policy.

DOI: 10.1111/jofi.12522

2331

2332 The Journal of Finance R

credit downgrade due simply to bad luck are subjected to an extended period

of reduced financial access.

We next apply our methodology to a set of borrowers who have had their

loans downgraded to the lowest level by all of their lenders. We label this event

“complete default.” For some of these borrowers, complete default arose due

to the exogenous rating shocks, while others were not subject to these shocks.

Contrasting the outcomes for these two classes of borrowers, we disentangle

the extent to which future lending restrictions are driven simply by negative

credit reporting, as opposed to the persistent long-run real shocks that often

cause default. At various horizons up to three years, we estimate that 25% to

65% of the observed credit decline after complete default arises solely due to the

sustained negative impact of the borrower’s poor credit rating. These findings

offer some of the first evidence that credit reporting documenting the defaults,

foreclosures, and bankruptcies of consumers may itself be a key mechanism

substantially reducing future borrowing, irrespective of expansionary central

bank policies or other macroeconomic stimuli such as those implemented in the

United States after the 2008 crisis.

Assessing the impact of exogenous credit rating shocks on consumers may be

challenging, as credit performance is determined endogenously by the actions

of the consumer and the evaluations of banks and other credit raters. The con-

sumer borrowing market in Peru has two features that allow for an empirical

examination of the central questions outlined above.1First, Peruvian banking

regulations require that banks provide to a central credit registry a quantita-

tive risk assessment of each client, which is available for anyone to see. For

borrowers with more than one bank, the regulations further require that these

ratings display a degree of alignment. In particular, a poor risk rating given

by any bank with a share of 20% or more of a given borrower’s total lending

should be reflected in the ratings of all other lenders. Second, during our 2001

to 2011 sample period, Peruvian consumers routinely borrowed in a mix of local

currency (sol) and U.S. dollar debt.

The strict 20% cutoff for the alignment requirement and the combination of

sol and dollar borrowing create the possibility that one borrower may have a

poorly performing loan pushed across the 20% threshold purely by exchange

rate movements, while another borrower with a similar loan profile but a some-

what different currency exposure may remain below the threshold. We imple-

ment a regression discontinuity design that compares borrowers with banking

relationships whose exchange-rate-adjusted balances (i.e., previous month’s

balances adjusted by changes in the current month’s exchange rate) are just

above 20% to borrowers with relationships whose exchange-rate-adjusted bal-

ances are just below 20%. A borrower with a delinquent loan that crosses

the 20% threshold will experience a downgrade that is imposed by regula-

tion, leading to an overall rating record that appears very weak. Another bor-

rower with a delinquent loan just below 20% of her total loan portfolio will not

1We focus exclusively on pure consumers, not businesses; the individuals in this study do not

have a personal tax ID for business purposes and have never received a business loan.

Consumer Default, Credit Reporting, and Borrowing Constraints 2333

experience these consequences, due simply to specific movements in the cur-

rency market.

We show that our exchange-rate-adjusted balances clearly predict whether

a borrower’s actual loan balance will shift to over 20% of her overall balance,

despite the fact that the former ignores any changes made in the current

month (to avoid endogeneity concerns). We also show that, in terms of ob-

servable characteristics, borrowers with exchange-rate-adjusted balances just

over 20% look very similar to those with exchange-rate-adjusted balances just

below this threshold, which is not surprising given that currency movements

are exogenous for any given consumer. We further show that borrowers with

low-rated loans pushed above the threshold by exchange rate movements expe-

rience a negative rating shock of moderate duration (the effect is statistically

significant for no longer than five months, though the estimated coefficients

do not fall much over the first year). These effects are confined, as expected,

to borrowers with highly heterogeneous loan ratings; borrowers whose loan

ratings are all somewhat similar are unaffected by the alignment mandate

and do not experience a significant rating change when a loan passes the

threshold.

We consider the impact of this negative rating shock on the borrower’s bank-

ing relationships. We find that above-threshold borrowers experience a reduc-

tion in their consumer loan balances and receive less new consumer financing

over the next three years, relative to below-threshold borrowers. Moreover,

these consumers are less likely to initiate new banking relationships and are

subject to reductions in their unused credit line balance.

Banks have access to all of the information necessary for unraveling the

quasi-random source of the downgrade, but it nonetheless has a meaningful

negative impact on lending to the borrower. This relates to findings in other

settings that rating events with no information content can have an impact

on financial outcomes when regulations are linked to credit ratings (Kisgen

and Strahan (2010), Ellul, Jotikasthira, and Lundblad (2011), and Bongaerts,

Cremers, and Goetzmann (2012)). We provide new evidence that the impact of

these purely regulatory effects may not be immediate but nonetheless may be

sustained over a long period of time, as we document a decrease in financing

that manifests two years after the rating shock and remains present during

the third year.

We further consider the effect of the shock on the consumer’s actions. First,

we show that above-threshold consumers are more likely to pay down their most

delinquent loans (loans that are subject to judicial collection) in the three years

after the shock. Further, we find that borrowers who receive a negative shock

are more likely to achieve a zero balance on their credit card accounts in the

year following the shock. These results suggest that the rating shock serves as

a wake-up call for the consumer, inducing her to improve her financial profile.

Despite these corrective actions taken by above-threshold consumers, however,

we find that the medium-term impact of the shock is quite negative. Shocked

clients are more likely to completely exit the consumer loan market in the

subsequent two and three years. The rating shock appears to lead consumers