Tail Risk Networks of Insurers Around the Globe: An Empirical Examination of Systemic Risk for G‐SIIs vs Non‐G‐SIIs

• Published date: 01 June 2020
• Date: 01 June 2020
• Author: Hua Chen,Tao Sun
• DOI: http://doi.org/10.1111/jori.12296

© 2019 The Journal of Risk and Insurance. Vol. 87, No. 2, 285–318 (2020).

DOI: 10.1111/jori.12296

TAIL RISK NETWORKS OF INSURERS AROUND THE GLOBE:

ANEMPIRICAL EXAMINATION OF SYSTEMIC RISK FOR G‐SIIS

VS NON‐G‐SIIS

Hua Chen

Tao Sun

ABSTRACT

In this article, we investigate systemic risk of 157 insurers around the

globe. We construct tail risk networks among these insurers using a

single‐index model for quantile regressions with a variable selection

technique. We develop a new network‐based systemic risk indices,

taking into account expected tail losses of insurers, direct and indirect

contagion effects, and the time‐varying strength of tail risk spillover.

Our systemic risk indices successfully recognize global systemically

important insurers (G‐SIIs).Wefind that on average G‐SIIs are more

systemically relevant than non‐G‐SIIs, particularly during the recent

U.S. financial crisis. We also find a small group of non‐G‐SIIs that are

more important than G‐SIIs. Our results have significant implications for

systemic risk regulation.

INTRODUCTION

The recent global financial crisis has clearly demonstrated the need for re-

searchers, practitioners, and regulators to develop a better understanding of

systemic risk and identify “Systemically Important Financial Institutions”

(SIFIs).

1

Specific to the insurance industry, the FSB, in consultation with the

International Association of Insurance Supervisors (IAIS),identified nine global

systemically important insurers (G‐SIIs)based on the assessment methodology

Hua Chen is at the University of Hawai’i at Manoa. Chen can be contacted via e‐mail:

huachen@hawaii.edu. Tao Sun is at the Department of Finance and Insurance, Lingnan

University, Hong Kong. Sun can be contacted via e‐mail: taosun@ln.edu.hk.

1

The Financial Stability Board (FSB, 2011)defines SIFIs as “financial institutions whose dis-

tress or disorderly failure, because of their size, complexity and systemic interconnectedness,

would cause a significant disruption to the wider financial system and economic activity”.

285

developed by the IAIS.

2,3

Though the FSB’sG‐SII designation is advisory only

and nonbinding, in the United States the Financial Stability Oversight Council

(FSOC)has designated all three U.S.‐based G‐SIIs (AIG, MetLife, and Prudential

Financial)as systemically important (National Association of Insurance Com-

missioners [NAIC], 2016). Ever since the release of the IAIS assessment metho-

dology and the G‐SII list, questions exist regarding the appropriateness and

effectiveness of the proposed framework from both the insurance industry and

academia. MetLife has been constantly fought to remove its “systemically im-

portant”label and received a favorable ruling in the U.S. District Court in March

2016 (Wall Street Journal, 2016).

4

Following MetLife’s success, AIG’sSIFIdes-

ignation was rescinded by the FSOC in September 2017, and Prudential Financial

shed its SIFI label in October 2018.

5

Prior studies (e.g., Weiß and Mühlnickel,

2014; Bierth, Irresberger, and Weiß, 2015)also indicate that some of the indicators

2

The methodology was first published in 2013 (IAIS, 2013)and updated in 2016 (IAIS, 2016).

Specifically, a composite score is calculated for insurers based on a wide variety of indicators

and assigned weights. In the 2013 methodology, there were 19 indicators under five cate-

gories size, global activities, interconnectedness, nontraditional insurance and noninsurance

activities, and substitutability. In the updated methodology, there are 17 indicators under

five categories size, global activity, interconnectedness, asset liquidation, and substitut-

ability. Particularly, the overall weights assigned to the category of “interconnectedness”

increased from 40 percent in IAIS (2013)to 49.3 percent in IAIS (2016).

3

On the basis of the IAIS developed methodology, the FSB identified an initial list of nine

G‐SIIs (AIG, Allianz, Assicurazioni Generali, Aviva, AXA, MetLife, Ping An, Prudential, and

Prudential Financial)in 2013 (FSB, 2013). This G‐SII list has been updated annually: there

was no change in 2014 (FSB, 2014); in 2015, one insurer (Aegon)was added and one insurer

(Assicurazioni Generali)was removed (FSB, 2015); there was no change in 2016 and 2017.

G‐SIIs will be subject to a framework of policy measures, including higher loss absorbency,

enhanced group‐wide supervision, and group‐wide recovery and resolution planning and

regular resolvability assessments, with the objective of reducing the negative externalities

stemming from the potential disorderly failure posed by a G‐SII (FSB, 2016).

4

According to the ruling by the federal judge, the FSOC made an “arbitrary and capricious”

decision when it designated MetLife as systemically important. The Department of Justice on

behalf of the FSOC appealed that decision in the United States Court of Appeals for the D.C.

Circuit. Even though the appellate court had not rendered a decision, on January 18, 2018,

MetLife and the FSOC settled and agreed to end the appeal, effectively ending the fight.

5

In September 2017, AIG was de‐designated from the SIFI list as it convinced the FSOC that it

had dramatically changed its business by significantly de‐risking its businesses, substantially

reducing its leverage, debt, derivative positions, securities lending and repurchase agree-

ments, and drastically shrinking in size. FSOC said that “AIG is notably different from the

company as it existed leading up to the financial crisis.”—quoted from “Insurance Sector

Trends: 2017 Year End Review and Forecast for 2018.”Prudential Financial also has worked

closely through the FSOC’s rigorous review process, which resulted in the Council’s con-

clusion that Prudential Financial does not pose systemic risk in October 2018.

286 THE JOURNAL OF RISK AND INSURANCE

in the IAIS assessment methodology maynotbeabletoexplaininsurers’

exposure and/or contribution to systemic risk.

6

That being said, little is known about the role of identified G‐SIIs in transmit-

ting systemic risk at the global level, that is, are G‐SIIs more likely to be victims

or propagators of systemic risk than other insurers? This calls for more detailed

analysis of the interconnectedness in the international insurance market and the

assessment of systemic risk for G‐SIIs and other insurers. In this article, we

construct tail risk networks for 157 insurers around the globe during the period

of 2006–2015. To do this, we use a single‐index model for quantile regressions

with a variable selection technique, that is, the least absolute shrinkage and

selection operator (LASSO), to identify key risk drivers for each insurer. We

develop a set of network‐based systemic risk indices, which take into account

expected tail losses of insurers, the direct and indirect effects of tail risk spil-

lover, and the time‐varying strength of tail risk spillover. Using these systemic

risk indices, we find that on average G‐SIIs are more systemically relevant than

non‐G‐SIIs, particularly during the market turmoil periods. However, we also

find a small group of non‐G‐SIIs that are as important as, if not more important

than, G‐SIIs.

This article is among the very few international studies on systemic risk

in the insurance sector (e.g., Bierth, Irresberger, and Weiß, 2015; Mühlnickel

and Weiß, 2015; Kanno, 2016). An international perspective is crucial to un-

derstand the interconnectedness and systemic risk in the insurance industry.

The recent U.S. financial crisis and ensuing European Sovereign debt crisis have

demonstrated that, given a more integrated global economy, the impact of a

negative shock originated from an industry or a country could spread to other

industries or countries quickly (Minoiu and Reyes, 2013; Diebold and Yilmaz,

2014). Particularly to insurers, their business models heavily rely on risk

transfer and risk diversification through reinsurance, which is essentially a

global market. Additionally, this international approach is also consistent with

the financial regulatory reforms advanced by the FSB to create guidelines

for regulatory coordination and supervise systemic risk in the international

financial system.

Our article is related to a large strand of systemic risk literature focusing on

deriving quantitative systemic risk measures, which could serve regulatory

purposes to identify SIFIs or provide early warning signals regarding a potential

future crisis. Among them, four mostly cited market‐based systemic risk mea-

sures are marginal expected shortfall (MES)and systemic expected shortfall

(SES)of Acharya et al. (2010), SRISK of Brownlees and Engle (2017),andΔCoVaR

6

Weiß and Mühlnickel (2014)show that size, nonpolicyholder liabilities and investment

income can be used to identify insurers that are most vulnerable to systemic risk, but the

contribution of insurers to systemic risk is only driven by insurer size. They find that

substitutability, global activity, and the exposure to the reinsurance industry do not

predict the systemic relevance of an insurer and thus should not be penalized by the IAIS

assessment methodology.

287TAIL RISK NETWORKS OF INSURERS AROUND THE GLOBE