Correlated Trading by Life Insurers and Its Impact on Bond Prices

• Author: Greg Niehaus,Chia‐Chun Chiang
• DOI: http://doi.org/10.1111/jori.12276
• Published date: 01 September 2020
• Date: 01 September 2020

CORRELATED TRADING BY LIFE INSURERS AND ITS IMPACT

ON BOND PRICES

Chia-Chun Chiang

Greg Niehaus

ABSTRACT

Our evidence indicates that U.S. life insurers’ decisions to buy and sell

individual corporate bonds are correlated across companies within the life

insurance industry. On average, the correlation in sell decisions is greater in

smaller bonds, bonds with lower ratings, bonds that have been downgraded,

and bonds that have recently experienced relatively large abnormal returns.

Correlated trading was also elevated during the financial crisis. In addition,

correlated buying and selling are greater when insurers designated as

systemically important financial institutions are actively trading. We also

find that the bonds that insurers sell in a correlated manner exhibit negative

average abnormal returns during the quarter in which insurers are selling. One

explanation is that insurers’ correlated selling is temporarily pushing bond

prices below their fundamental value. In this case, we would expect prices to

bounce back in the subsequent quarter. However, we do not find a rebound in

prices and therefore our evidence supports the alternative explanation that

insurers’ correlated selling is impounding information into bond prices.

INTRODUCTION

Academics, insurance professionals, and regulators continue to debate whether

traditional insurance activities of insurers are a source of systemic risk.

1

For example,

Chia-Chun Chiang is at the University of Texas at El Paso, El Paso, Texas. She can be contacted

via e-mail: cchiang@utep.edu. Greg Niehaus is at the University of South Carolina, Columbia,

South Carolina. He can be contacted via e-mail: gregn@moore.sc.edu. The authors appreciate

the helpful comments from two anonymous reviewers, the co-editor, Anna Agapova, Allen

Berger, Yongqiang Chu, David Cummins, Marco Eugster, Shingo Goto, Song Han, Jean

Helwege, Ozgur Ince, Pierre Joos, Philipp Kiergassner, Liang Ma, Andreas Milidonis, Ashleigh

Poindexter, Eric Powers, Herman Saheruddin, Daniel Schwarcz, Mary Weiss, and DH Zhang,

and seminar participants at the University of South Carolina, Temple University, the 2016Fixed

Income and Financial Institutions Conference, the University of North Carolina Charlotte, and

Ludwig-Maximilian University.

1

Many commentators acknowledge that nontraditional activities, such as trading credit default

swaps, could cause insurance groups to be systemically important (see Harrington, 2009;

Cummins and Weiss, 2014).

©2019 The Journal of Risk and Insurance. Vol. 9999, No. 9999, 1–29 (2018).

DOI: 10.1111/jori.12276

1

597

597

Vol. 87, No. 3, 597–625 (2020).

the Financial Stability Oversight Council (FSOC) in December of 2014 designated

MetLife as a systemically important financial institution (SIFI) despite objections from

Metlife and other commentators (see, e.g., Wallison, 2014). Metlife challenged the

ruling, and in March 2016, a judge rescinded the SIFI designation. The Department of

Justice on behalf of FSOC appealed the decision, but in January 2018, the government

agreed not to pursue the case further (Schroeder, 2018).

One channel through which life insurers could contribute to systemic risk is via their

correlated trading of corporate bonds.

2

As of 2016, life insurers held over $2.3 trillion

in bonds, which represents about 26 percent of the value of all U.S. corporate bonds

outstanding.

3

If insurers’ trading of securities is correlated across insurers, then they

could potentially disrupt financial markets by causing security prices to move

temporarily away from fundamental values. Schwarcz and Schwarcz (2014)

forcefully make this argument and call for greater regulation. The FSOC also refers

to this mechanism when justifying its designation of Prudential as an SIFI (FSOC,

2013).

4

On the other side of the spectrum from those who argue that life insurer correlated

trading contributes to systemic risk, Vaughn (2012) argues that the life insurance

industry provides a stabilizing force in financial markets during times of crisis. This

would occur, for example, if during liquidity shocks that induce sales from other

institutions, insurers maintain their positions and/or even step in on the buy side. A

related point of view is that insurer investment decisions are unlikely to influence

security prices because even though life insurers’ asset portfolios are large, they are

typically buy-and-hold investors. Paulson and Rosen (2016) report that the annual

turnover rate of corporate bonds held by life insurers is about one-fifth of the turnover

rate of corporate bonds in general. On the other hand, trading in corporate bonds is

relatively thin, and so even a relatively small amount of trading can potentially

impact prices. Also, recent evidence by Ellul, Jotikasthira, and Lundblad (2011, 2014)

and Merrill, Nadauld, and Strahan (2014, 2017) is consistent with life insurer

investment behavior temporarily impacting security prices.

The purpose of this article is to examine (1) the extent to which life insurers’

investment decisions in corporate bonds are correlated with each other, (2) insurer

characteristics that are associated with life insurer correlated trading, and (3) whether

life insurer correlated trading in corporate bonds cause bond prices to move

2

There are other arguments for how insurers could contribute to systemic risk, including

concerns about an insolvency of one insurer reducing confidence in the ability of otherinsurers

to make good on their promises, which in turn could cause policyholder runs and cause

insurers to liquidate assets quickly and at fire sale prices. See (Foley-Fisher, Narajabad, and

Verani, 2015). Cummins and Weiss (2014) focus on whether reinsurance activities contribute to

systemic risk. Also see Acharya, Biggs, and Richardson (2014), Billio et al. (2012), Chen et al.

(2014); Manconi, Massa, and Yasuda (2012), Neale, et al. (2012), Weiss and Muhlnickel (2014),

and Weiss, Bierth, and Irresberger (2015).

3

See American Council of Life Insurers (2016) and the Securities Industry and Financial Markets

Association (2018).

4

Also, see Getmansky et al. (2016) and Koijen and Yogo (2016).

2THE JOURNAL OF RISK AND INSURANCE

2The Journal of Risk and Insurance

598