Systemic Risk in the Insurance Sector: A Review and Directions for Future Research

• Published date: 01 September 2016
• Date: 01 September 2016
• DOI: http://doi.org/10.1111/rmir.12062
• Author: David Antonius Pankoke,Martin Eling

Risk Management and Insurance Review

C

Risk Management and Insurance Review, 2016, Vol.19, No. 2, 249-284

DOI: 10.1111/rmir.12062

SYSTEMIC RISK IN THE INSURANCE SECTOR:AREVIEW

AND DIRECTIONS FOR FUTURE RESEARCH

Martin Eling

David Antonius Pankoke

ABSTRACT

This article reviews the extant research on systemic risk in the insurance sector

and outlines new areas of research in this field. We summarize and classify 48

theoretical and empirical research papers from both academia and practitioner

organizations. The survey reveals that traditional insurance activity in the life,

nonlife, and reinsurance sectors neither contributes to systemic risk nor in-

creases insurers’ vulnerability to impairments of the financial system. However,

nontraditional activities (e.g., credit default swap underwriting) might increase

vulnerability, and life insurers might be more vulnerable than nonlife insur-

ers due to higher leverage. Whether nontraditional activities also contribute to

systemic risk is not entirely clear; however, the activities with the potential to

contribute to systemic risk include underwriting financial derivatives and pro-

viding financial guarantees. This article is not only likely of interest to academics

but also highly relevant for the industry,regulators, and policymakers.

INTRODUCTION

In the wake of the financial crisis and the collapse of Lehman Brothers and AIG, systemic

risk has been widely discussed in the financial services sector. A number of research

papers on the subject have been published, and regulators and industry think tanks

have issued reports. Recently, the Financial Stability Board (FSB) published a list of nine

global systemically important insurers and it intends to implement several special policy

measures for these institutions by January 2019.1

A more detailed review of existing work on this topic and an examination of what

remains to be investigated is worthwhile for at least two reasons. First, researchers doc-

ument that certain business activities might contribute to systemic risk in the insurance

sector (see, e.g., Acharya et al., 2011; Besar et al., 2011, for securities lending). A struc-

tured review can identify those activities and the situations in which the activities may

be cause for alarm. Second, the literature reports mixed results regarding systemic risk,

Martin Eling and David Pankoke are both with the Institute of Insurance Economics, University

of St. Gallen, Rosenbergstr.22, 9000 St. Gallen, Switzerland. The authors can be reached by e-mail

at martin.eling@unisg.ch and dpankoke@hotmail.de, respectively.

1See FSB (2013a). The FSB is an international organization that was established by the G-20 in

April 2009. Its purpose is to monitor the finance industry and to make recommendations for

addressing systemic risk.

249

250 RISK MANAGEMENT AND INSURANCE REVIEW

for example, in the case of credit default swaps (CDSs),2which calls for a structured

review of what has been studied to date and what remains to be done to settle this issue.

This article makes three contributions to the ongoing discussion. First, we discuss how

systemic risk can be understood conceptually and how it can (or cannot) be measured.

Second, based on this framework, we review the literature on systemic risk in the

insurance sector. Third, we highlight areas in need of further research. The survey is

intended to enhance our understanding of systemic risk in the insurance sector and to

motivate additional research in this field. The literature indicates that policymakers and

regulators need to closely analyze systemic risk, especially with respect to nontraditional

insurance activities.

The remainder of the article is organizedas follows. We begin by discussing the definition

and classification of systemic risk and systemic risk measures. Subsequently, we review

the extant insurance literature on systemic risk and summarize the main results for

different lines and activities of insurance companies. Then, we investigate whether

banking regulation should be extended to insurance companies. We close the article

with a conclusion and a discussion of future research directions.

CLASSIFICATION OF SYSTEMIC RISK

The financial system can be thought of as a network with two types of nodes (financial

institutions and nonfinancial actors that have business relationships with financial insti-

tutions) and edges (business activities).3Toidentify the origins of a possible impairment

and the contributors to systemic risk, one can focus on the nodes, the edges, or both.

Furthermore, it is important not only to identify which parts of the financial system can

originate impairment and contribute to systemic risk but also to discover which parts

are most vulnerable to impairment.

In the remainder of the article, we use the term “contribution to systemic risk” for an

institution or business activity that increases systemic risk. Weuse “vulnerability” when

describing those parts of the financial system (institution or business activity) that are

most vulnerable to impairment.4

Systemic Risk Definitions

There is no generally agreed-upon definition of “systemic risk” other than that it involves

uncertainty about the occurrence of a specific event. We reviewed 26 definitions of

systemic risk and identified three important elements:

2See Trichet (2005), Baluch et al. (2011), Klein (2013), Grace (2011), Baranoff (2012), Chen et al.

(2013), and Cummins and Weiss (2013), in addition to the industry study by the Geneva Asso-

ciation (2010a) versus Radice (2010); Wallison reported by Harrington(2009, 2013).

3See, for example, Anand et al. (2013).

4Our classification is similar to the “contribution approach”and “participation approach” defined

by Tarashevet al. (2010) and used by Drehmann and Tarashev (2011) and Jobst (2012). Wedo not

use that terminology here because, for our purposes, it is too narrow.Both approaches consider

only institutions and bankruptcies and ignore, for example, institutions that are in distress but

not insolvent, in addition to business activities. Our approach is also employed by others, for

example, Klein (2013) and Schwarcz and Schwarcz (2014).

SYSTEMIC RISK IN THE INSURANCE SECTOR 251

1. Risk of an event: For each risk, there must be an associated event that can occur.

The associated event is the dysfunction of financial services, default of financial

institutions, or a shock to the economy.5

2. Impact of the event: Most definitions specify the consequences if the event occurs,

which are typically that the real economy is negatively affected.

3. Causation of the event: Some definitions require the risk to have a certaincausation

before it is labeled systemic. These causations can be general in nature and/or

specific6and are mostly related to the financial services sector.

The variety of definitions makes obvious the enormous difficulty involved in differ-

entiating among cause, impact, and events when discussing systemic risk.7Thus, it is

not surprising that to date, no agreement on the definition of systemic risk has been

reached.8However, most of the definitions considered in this article relate to the 2008

financial crisis;9thus, it might be useful to examine the events of 2008 as a means of

differentiating systemic risk from other risks:

1. Event: Certain financial services became unavailable (e.g., interbank lending) or

had virtually no market (e.g., credit).

2. Impact of the event: As a consequence, there was a substantial negative effect on

the economy.

3. Causation of the event: In 2008, an external shock (falling prices in the U.S. sub-

prime mortgage market) impaired several financial institutions. Due to contagion

and interdependence, other financial institutions and services also became im-

paired.

An appropriate definition should encompass all risks that can lead to the reoccurrence

of the 2008 crisis and exclude all others. Studies often ignore this last point. An excep-

tion is the concept of systemic risk proposed by De Bandt and Hartmann (2000), who

distinguish between risks of shocks based on their second-round effects (the focus is not

on institutions affected by the shock but on the consequences of these institutions’ being

5The first two factors involve financial services; the latter factor involves the general economy.

We use “financial instability” as a synonym for “dysfunction of financial services.” It must be

kept in mind that neither financial “stability” nor “instability” have clear-cut definitions. See,

for example, Allen and Wood(2006), the European Central Bank (2013), and the Federal Reserve

Bank of Cleveland (2013).

6Klein (2013) writes that idiosyncratic events (e.g., the failure of a single entity or cluster of

entities) or general conditions in financial intermediaries might cause systemic risk. The general

conditions arerelated to the linkages between financial institutions, which can lead to a cascading

effect of bankruptcies, especially in the case of excessive risk taking.

7For example, it is not clear whether the shock to the real economy is the event or the consequence

of the event. Similarly, it is far from clear whether the default of institutions is the event or the

causation of an event.

8See, for example, Liedtke (2010) and Dwyer (2009) for a critical discussion of several definitions

of systemic risk.

9See, for example, Bach and Nguyen (2012), Billio et al. (2012), and Rodr´

ıguez-Moreno and Pe˜

na

(2013).