Paying for expertise: The effect of experience on insurance demand

• Published date: 01 September 2021
• Author: Vaibhav Anand,James Tyler Leverty,Kenny Wunder
• Date: 01 September 2021
• DOI: http://doi.org/10.1111/jori.12331

Received: 17 July 2019

|

Revised: 1 August 2020

|

Accepted: 18 October 2020

DOI: 10.1111/jori.12331

ORIGINAL ARTICLE

Paying for expertise: The effect of experience

on insurance demand

Vaibhav Anand

1

|James Tyler Leverty

2

|Kenny Wunder

3

1

Wisconsin School of Business,

University of Wisconsin, Madison,

Wisconsin, USA

2

Gerald D. Stephens CPCU Distinguished

Chair in Risk Management and

Insurance, Wisconsin School of Business,

University of Wisconsin, Madison,

Wisconsin, USA

3

Dai‐ichi Life Endowed Teaching Chair

in Actuarial Science and Risk

Management, Culverhouse College of

Business, University of Alabama,

Tuscaloosa, Alabama, USA

Correspondence

James Tyler Leverty, Gerald D. Stephens

CPCU Distinguished Chair in Risk

Management and Insurance, Wisconsin

School of Business, University of

Wisconsin, 5284A Grainger Hall, 975

University Ave, Madison, WI 53706,

USA.

Email: ty.leverty@wisc.edu

Abstract

This paper investigates whether corporations purchase

insurance for the real services that insurers provide.

We examine the real‐service efficiency hypothesis in

the insurance industry by exploiting within firm‐year

variation in reinsurance usage and experience at the

line of business level. Our results show that insurers

purchase reinsurance to access reinsurers’expertise

and specialized knowledge. We find that, within the

same year, the average multi‐line insurer purchases

2.4%–3.3% points more reinsurance for new lines

relative to existing lines. The demand for reinsurance

declines by 0.55%–0.80% points with each additional

year of experience. The size of these effects differs by

the line of business, indicating that the development of

internal expertise and specialized knowledge varies

by line.

KEYWORDS

expertise, insurance, reinsurance, risk management

1|INTRODUCTION

In frictionless capital markets, firm value does not depend on risk management (Modigliani &

Miller, 1958). Yet frictions do exist, and several—such as financial distress, credit‐rationing,

taxes, and agency conflicts—have been shown to be theoretically relevant to risk management

increasing firm value (Froot et al., 1993; Holmstrom & Tirole, 2000; Mayers & Smith, 1982;

Smith & Stulz, 1985). Empirical confirmation of these theories, using derivatives to

measure risk management, has been mixed (e.g., Géczy et al., 1997; Graham & Rogers, 2002;

J Risk Insur. 2021;88:727–756. wileyonlinelibrary.com/journal/JORI

|

727

© 2020 American Risk and Insurance Association

Haushalter, 2000; Nance et al., 1993; Tufano, 1996). Even without frictions, however, managing

risk with insurance may increase firm value because insurers have a comparative advantage in

the provision of real services (e.g., risk identification, analysis, and evaluation; claims admin-

istration; loss monitoring and prevention; and pricing; Mayers & Smith, 1982). By exploiting

within‐firm variation in insurance usage at the line of business level, we identify a strong

demand for the real services that insurers provide.

We investigate the real‐service efficiency hypothesis in the insurance industry. In contrast

to noninsurance corporations that do not systematically disclose their risk management ac-

tivities, regulatory requirements mandate that insurers disclose their purchases of reinsurance,

making the insurance industry an excellent laboratory to test the real‐service efficiency hy-

pothesis. Reinsurance is insurance purchased by one insurance company, the ceding (or pri-

mary) company, from another, the reinsurer. Within the insurance industry, reinsurance

purchases are similar to traditional insurance purchases by noninsurance corporations. Since

insurers already have expertise in the provision of insurance services (i.e., they employ actu-

aries, claims adjusters, underwriters, etc.), evidence that insurers purchase insurance for the

real services provided by other insurers offers strong support for the real‐service efficiency

hypothesis.

Using a panel data set of reinsurance purchases by group and unaffiliated single insurers at

the firm‐line‐year level, we exploit within firm‐year variation in experience and reinsurance

usage at the line of business level to identify the demand for real services. Many insurers

operate in multiple lines of insurance in a year (e.g., workers’compensation and auto liability).

These multi‐line insurers purchase different amounts of reinsurance in each line and have

varying amounts of experience in each line (e.g., it could be the first year that an insurer

operates in workers’compensation insurance, but its 10th year in auto liability insurance).

Whether a reinsurer has a comparative advantage in providing real services depends on the

primary insurer's experience in a line: In general, inexperienced insurers will demand the

specialized knowledge and expertise of reinsurers to a greater degree than experienced insurers.

Accordingly, we estimate the demand for real services using firm‐year and line fixed effects

regressions that compare the demand for reinsurance in new lines (the treatment group) with

the demand in existing lines (the control group).

1

The firm‐year fixed effects remove the

variation between firms in a year, making the comparison between the treatment and control

groups within the same firm and year (which only occurs when an insurer operates in two or

more lines in a year).

2

The inclusion of the line fixed effects, which control for constant line‐

level unobserved heterogeneity over time, means that the comparison between the treatment

and control groups also holds fixed any differences in reinsurance usage by line of business.

In sum, we compare how experience affects reinsurance usage between different lines of

business within the same firm‐year, controlling for fixed differences in reinsurance usage at the

line of business level.

We find that in the same year, the average multi‐line insurer purchases 2.4%–3.3% points

more reinsurance for new lines than existing lines. Our estimates suggest that insurers cede

9.9%–13.6% more of their premiums to reinsurers in new lines relative to existing lines. The size

1

When an insurer first enters a line, it has not yet accumulated loss data, acquired specialized knowledge, or developed

internal expertise for the marketing, underwriting, and claims handling of the line.

2

Since within firm‐year differences in experience at the line of business level do not affect firm‐level motivations for

reinsurance (e.g., cost of financial distress, underinvestment incentives, tax structure, diversification, and organiza-

tional structure), the demand for real‐service is separated from firm‐level motivations for reinsurance.

728

|

ANAND ET AL.