Strengthening Local Credit Markets Through Lender‐Level Index Insurance

• Published date: 01 June 2020
• Date: 01 June 2020
• DOI: http://doi.org/10.1111/jori.12277
• Author: Benjamin L. Collier

©2019 The Journal of Risk and Insurance. Vol.XX, No. XX, 1–31 (2019).

DOI: 10.1111/jori.12277

Strengthening Local Credit Markets Through

Lender-Level Index Insurance

Benjamin L. Collier

Abstract

This article considers lender-level index insurance as a means of expanding

access to credit in disaster-prone communities. In this approach, the lender

transfers the disaster risk of loans in its portfolio by contracting on an observ-

able measure of the catastrophe.I develop and calibrate a dynamic, stochastic

model using data from a community lender in Peru that is vulnerable to El

Ni˜

no–related flooding. The modeled lender can insure against El Ni˜

no using

an index-based product that is available for purchase by financial interme-

diaries in Peru. I examine how premium rates, basis risk, and background

risk may influence the lender’s insurance decision and credit supply. Over-

all, the results suggest that lender-level index insurance holds promise for

reducing disaster-related credit supply shocks and expanding credit access

in vulnerable communities.

Introduction

Natural disaster exposures are ubiquitously underinsured.For example, about 70 per-

cent of natural disaster losses in the last decade were uninsured (2008–2017, Swiss Re,

2018). This insurance gap affects not only disaster-exposed firms and households, but

third parties connected to them. Recent research examines the uninsured disaster ex-

posures of local credit markets, noting that disasters can cause spatially concentrated

loan defaults. Laux, Lenciauskaite, and Muermann (2017) consider the low take-up of

earthquake insurance and the exposure of California mortgage lenders. Collier and

Babich (2017) examine lending in community credit markets of developing and emerg-

ing economies. They find that disaster-related losses on existing loans erode lenders’

Benjamin L. Collier is at the Fox School of Business, Temple University. Collier can be con-

tacted via e-mail: collier@temple.edu. This research is the result of collaboration with excellent

professionals at Banco de Cr´

edito del Per´

u, BBVA Banco Continental, Edyficar, Caja Nuestra

Gente, Caja Piura, Caja Paita, Caja Sullana, Caja Trujillo,La Positiva Seguros, Pacifico Seguros,

Rimac Seguros, Willis Peru, and the banking regulator in Peru, Superintendencia de Banca,

Seguros, y AFP. I want to especially thank Caja Trujillo for its guidance, including feedback on

this manuscript. I thank Jerry Skees for his mentoring, sage guidance, thoughtful leadership on

the topic, and conceptual contributions to this article. I thank Mario Miranda for his mentoring

and early contributions to this article. Thanks also to VolodymyrBabich, Grant Cavanaugh, Ja-

son Hartell, Thorsten Moenig, Ruben Lobel, Richard Peter,Marc Ragin, Paul Shea, and George

Zanjani for their help and insightful comments.

1

319

319

Vol. 87, No. 2, 319–349 (2020).

2The Journal of Risk and Insurance

equity capital, causing them to reduce the creditsupply after the disaster. In turn, these

credit market frictions seem to exacerbate the economic consequences of the disas-

ter by delaying the recovery of affected firms and households (Del Ninno, Dorosh,

and Smith, 2003; Noy,2009). These findings highlight the importance of reducing the

catastrophe insurance gap. However, in developed and developing countries alike,

getting households and businesses to insure remains a formidable challenge and so

creates a need for lenders and other third parties to manage their disaster risks.

This article considers an approach in which lenders transfer their portfolio concentra-

tions of disaster risk—the lender, rather than the borrowing household or business,

insures its risk. The article’s primary research question is how lender-level risk trans-

fer may influence access to credit for disaster-prone communities in developing and

emerging economies. I consider both ex ante and ex post access to credit. Regarding

ex post access, I assess how transferring disaster risk is likely to affect credit supply

shocks created by the disaster. Regardingex ante access, I assess whether transferring

disaster risk may influence the self-insuring strategies of the lender during nondisas-

ter conditions, for example, whether it more fully leverages its capital, lending more

per dollar of equity.

The focus and modeled application is on financial intermediaries (FIs) lending to

households and small and medium enterprises (SMEs) in developing and emerging

economies. I refer to these FIs as “SME lenders” though “community lenders” would

also be appropriate.

I develop a dynamic, partial equilibrium model to examine the behavior of a repre-

sentative lender under risk. The credit portfolio of the FI is exposed to a systemic

catastrophe and uninsurable background risks. The FI manages a stock of equity and

maximizes its expected, discounted stream of returns through its lending decisions.

The lender must meet minimum capital requirements, incurring compliance costs if

its capital falls too low.I calibrate the model for an SME lender in Peru that is vulnera-

ble to severe El Ni˜

no–related flooding. The calibration data are from the lender, which

conducted a risk assessment survey among its field office and credit risk managers,

and its banking regulator, which provided monthly income and balance sheet data.

The model’s mechanics are consistent with previous research (e.g., Van den Heuvel,

2009; Collier and Babich, 2017): in response to losses on existing loans from a catastro-

phe, the lender contracts credit, reducing loan allocations to bring them in line with

a smaller equity capital base. The risk of these shocks motivates the lender ex ante

to maintain a capital buffer above minimum requirements, which has the effect of

reducing the credit supply in nondisaster conditions.

Next, I consider the development of a market to transfer the lenders’ catastrophe risk

through index insurance.1Index insurance makes payments using an objective mea-

sure of an adverse event (e.g., deficit rainfall at local weather stations as a measure

1While I refer to financial risk transfer as “insurance,” Cummins and Weiss (2009) describe a

variety of structures that could be used in this setting such as derivatives (e.g., see their Figure

10 and accompanying text).

2The Journal of Risk and Insurance

320