Measuring the Performance of the Secondary Market for Life Insurance Policies

• Author: Joseph Golec,Carmelo Giaccotto,Bryan P. Schmutz
• DOI: http://doi.org/10.1111/jori.12078
• Date: 01 March 2017
• Published date: 01 March 2017

MEASURING THE PERFORMANCE OF THE SECONDARY

MARKET FOR LIFE INSURANCE POLICIES

Carmelo Giaccotto

Joseph Golec

Bryan P. Schmutz

ABSTRACT

We construct an index of life insurance policies purchased in the secondary

market by viatical and life settlement companies. Using the repeat sales

method to measure returns over our 1993–2009 sample period, we find that

policy returns average about 8 percent annually compared to 5.5 percent for

the S&P 500 and 7 percent for corporate bonds, but they are twice as volatile

as the S&P and four times as volatile as bonds. Nevertheless, because the

index return is relatively uncorrelated with stock or bond returns, life

insurance policies make attractive additions to well-diversified portfolios.

INTRODUCTION

The secondary market for individual life insurance policies in the United States has

grown from about $200 million in 1993 to $44 billion in 2010.

1

The market began with

policies on individuals with less than 2 years of life expectancy (called viaticals) and

grew to include those with more than 2 years of life expectancy (life settlements). Our

data source includes both; hence, we refer to these life insurance investments

collectively as viatical life settlements investments (VLSI). Except for recent work by

Braun, Gatzert, and Schmeiser (2012) (hereafter BGS) and Davo, Resco, and Barroso

(2013) (hereafter DRB), there is little research that systematically analyzes the return

characteristics of these investments. We add to this line of research.

Carmelo Giaccotto and Joseph Golec are at the Department of Finance, School of Business,

University of Connecticut. Bryan P. Schmutz is at the Department of Accounting and Finance,

College of Business, Western New England University. The authors can be contacted via e-mail:

CGiaccotto@Business.Uconn.edu, Jgolec@Business.Uconn.edu, and Bryan.Schmutz@wne.edu.

For their helpful comments and suggestions, we would like to thank session participants at the

2009 Eastern Finance Association meetings and the 2012 Swiss Society for Financial Market

Research meetings. We thank the Center for International Business and Business Education and

Research at the University of Connecticut for support for data gathering.

1

See the Life Insurance Settlements Association, http://www.lisa.org/content/3/Industry.

aspx (accessed June 30, 2013).

© 2015 The Journal of Risk and Insurance. Vol. 84, No. 1, 127–151 (2017).

DOI: 10.1111/jori.12078

127

BGS provide a detailed analysis of the performance of an index of 17 open-end life

settlements mutual funds, covering December 2003–June 2010. They also provide an

excellent overview of the life settlements market and the life settlements mutual fund

industry. DRB combine two life settlements mutual funds with fixed income and

equity index funds over September 2006–February 2010 to form efficient portfolios.

To our knowledge, our article is the first to examine the risk and return characteristics

of an index composed of direct life insurance policies, including 1,724 policies with a

face value of about $300 million. Specifically, we develop a quarterly index of VLSI

returns starting from the fourth quarter of 1993 through the fourth quarter of 2009.

We providethree advances compared to BGS and DRB. First, we usedata on individual

VLSI contracts instead of mutual fund portfolios. This allows us to show that VLSI

returns differ across disease types; for example, VLSI from AIDS patients have

relatively low returns. Second, we cover a longer sample period; thus we are able to

illustrate how major VLSI-related events lead to VLSI return volatility. Third, we use

the repeat sales method to compute a VLSI return index, because the method was

createdby Bailey, Muth, and Nourse (1963) to handle infrequentlytraded assets such as

VLSI. It is used to compute the widely quoted S&P Case–Shillerreal estate price index.

Our results are in line with those of BGS and DRB in many respects, except we find

that the returns of VLSI computed using the repeat sales method are much more

volatile than those self-reported by mutual funds. Indeed, BGS anticipate this

possibility, noting that because life settlements investments are illiquid, fund

managers can value their investments with a mark-to-model approach instead of a

mark-to-market approach, allowing more leeway to smooth their reported returns.

They find little correlation between fund returns and the returns of other assets, but

this could be due to artificially smoothed fund returns.

Although our VLSI returns are quite volatile, we confirm that they are still little

correlated with the returns of more traditional investments, such as stocks and

corporate bonds. Indeed, we illustrate some of the reasons why. For example, during

the early years of the VLSI market from 1993 through 1996, our VLSI return series is

comparable to that of corporate bonds; however, breakthroughs in AIDS drug

treatments extended the lives of many AIDS patients, driving down VLSI returns

from 1997 to 1999. Changes in institutional investor demand and supply of VLSI

could also affect return volatility. The newness and general illiquidity of the VLSI

market could partly explain the high volatility.

Our results show that VLSI earn about 8 percent annually compared to 7 percent for

long-term corporate bonds, and 5.5 percent for the Standard & Poor’s 500 stock index

(S&P 500). If we exclude the market crash between 2008 and 2009 from our sample

period, we find that VLSI earn about 7.3 percent compared to 8.5 percent for the S&P

500 and about 7.1 percent for bonds. But the VLSI return volatility is about twice that

of the S&P 500, and about four times that of bonds.

2

2

Our data source for VLSI does not extend beyond 2009 because there is a considerable lag in

their availability from the data source. Extended further to include recent stock market

appreciation, the superior performance of bonds and VLSI would likely be much less.

128 THE JOURNAL OF RISK AND INSURANCE