The Impact of Private Equity on a Life Insurer's Capital Charges Under Solvency II and the Swiss Solvency Test

• Author: Hato Schmeiser,Caroline Siegel,Alexander Braun
• Published date: 01 March 2014
• Date: 01 March 2014
• DOI: http://doi.org/10.1111/j.1539-6975.2012.01500.x

© The Journal of Risk and Insurance, 2014, Vol. 81, No. 1, 113–158

DOI: 10.1111/j.1539-6975.2012.01500.x

113

THE IMPACT OF PRIVATE EQUITY ON A LIFE INSURER’S

CAPITAL CHARGES UNDER SOLVENCY II AND THE SWISS

SOLVENCY TEST

Alexander Braun

Hato Schmeiser

Caroline Siegel

ABSTRACT

In this article, we conduct an in-depth analysis of the impact of private eq-

uity investments on the capital requirements faced by a representative life

insurance company under Solvency II as well as the Swiss Solvency Test.Our

discussion begins with an empirical performance measurement of the asset

class over the period from 2001 to 2010, suggesting that limited partnership

private equity funds may be suited for the purpose of portfolio enhance-

ment. Subsequently,we review the market risk standard approaches set out

by both regulatory regimes and outline a potential framework for an in-

ternal model. Based on an implementation of these solvency models, it is

possible to demonstrate that private equity is overly penalized by the stan-

dard approaches.Hence, life insurers aiming to exploit the asset class’s return

potential may expect significantly lower capital charges when applying an

economically sound internal model. Finally, we show that, from a regula-

tory capital perspective, it can even be less costly to increase the exposure to

private rather than public equity.

INTRODUCTION

Within the last decade, the regulation of the European insurance sector was subject

to fundamental reforms aimed at the introduction of risk-based solvency standards.

One of the first of these initiatives to revise solvency surveillance came into effect in

2004 in the United Kingdom (see, e.g., Cummins and Phillips, 2009). Switzerland fol-

lowed with its Swiss Solvency Test (SST) in 2006. Beyond these regulatory reformsof

individual countries, Solvency II, the EU’s flagship project to modernize and harmo-

nize European insurance supervision, has entered its final development phase and is

expected to come into force in 2013. Solvency II and the SST in particular are viewed

Alexander Braun, Hato Schmeiser, and Caroline Siegel are from the Institute of Insur-

ance Economics, University of St. Gallen, Kirchlistrasse 2, CH-9010 St. Gallen. The authors

can be contacted via e-mail: alexander.braun@unisg.ch, hato.schmeiser@unisg.ch, and caro-

line.siegel@unisg.ch, respectively.We are grateful to two anonymous referees as well as Karin

Jans, Andy Hawson, and Joerg Allenspach of Swiss Re for their helpful comments and sugges-

tions.

114 THE JOURNAL OF RISK AND INSURANCE

to be the most innovative frameworks currently available and should thus have a

major impact on insurance regulation in the near future. Despite this fact, design

and calibration of their standard approaches for market risk clearly promote bond

holdings. The prescribed treatment of alternative investments such as private equity,

on the contrary, is at least questionable. Since the attractiveness of an asset class for

insurance companies does not only depend on its performance characteristics but also

on the associated capital charges, an inappropriate assessment from a solvency per-

spective may cause an underrepresentation of asset classes with favorable risk-return

profiles, which would otherwise be well suited for portfolio diversification.

In the article at hand, we address this issue for the specific case of private equity.

Our contribution is twofold. First, identifying its risk-return profile within the critical

period from 2001 to 2010, we evaluate the attractiveness of the asset class from a per-

formance perspective and compare it with various investment alternatives. Second,

we shift our focus to the associated market risk capital requirements under Solvency

II and the SST. Our discussion begins with a review of the standard approaches for

market risk set out by both regulatory regimes as well as an outline of a potential

framework for an internal model. Subsequently, we compute and compare the re-

spective capital charges for a stylized life insurance company based on an empirical

implementation of these three solvency models. The results are shown to be robust

with regard to the benchmark index employed for parameter estimation, the percent-

age of private equity in the portfolio, and the chosen calibration period (including

and excluding the recent financial crisis). Finally, we assess how costly it is, from a

regulatory capital perspective, to increase the portfolio weight of private equity in

comparison with public equity and hedge funds.

The rest of the article is organized as follows. The second section contains an overview

of the literature on private equity and insurance regulation. In the third section,

we briefly discuss the characteristics of private equity investments and conduct an

extensive empirical performance analysis, including a comparison with other asset

classes. The market risk standard approaches of Solvency II and the SST as well as the

outline of an internal model are presented in the fourth section. In the fifth section, we

implement the market risk models and provide an in-depth discussion of the resulting

capital charges. Finally, in the sixth section we discuss economic implications and state

our conclusion.

LITERATURE REVIEW

Due to the fact that there is a substantial body of extant literature on both private

equity and the regulation of financial institutions, we will only review recent research

in those areas that are most relevant to our study. Webegin with the latest work on the

risk-return characteristics as well as the historical performance of the asset class, par-

ticularly in comparison with public equity markets. Although industry professionals

and general partners regularly stress the attractiveness of private equity investments,

existing empirical evidence conveys a rather ambiguous picture. Moskowitz and

Vissing Jørgensen (2002) control for distorted market values and estimate returns for

the entire U.S. private equity market, concluding that it did not outperform pub-

lic equity between 1989 and 1998. Ljungqvist and Richardson (2003), in contrast,

conduct a performance analysis of private equity funds over the two decades from

THE IMPACT OF PRIVATE EQUITY ON CAPITAL CHARGES 115

1981 to 2001 and show that their returns exceeded those of the aggregate public equity

market by at least 5 percent per year. Yet, Zhu et al. (2004) again challenge the al-

legedly superior risk-return profile of the asset class. They argue that, due to the long

investment horizon, severe liquidity constraints, and high default probabilities of

the portfolio companies, private equity funds hide latent risks, which investors need

to take into account. In another empirical paper, Kaserer and Diller (2004) analyze

European private equity funds based on cash flow data, documenting an underper-

formance relative to public equity from 1980 to 2003 but an outperformance during

the shorter period from 1989 to 2003. The work of Kaplan and Schoar (2005) is proba-

bly one of the most influential studies of private equity performance. Examining the

capital flows of more than 1,000 limited partnerships for the period between 1980

and 2001, they demonstrate that the cross-sectional mean return net of fees did vir-

tually not differ from that of the S&P 500. Furthermore, Driessen et al. (2008) devise

a statistical methodology based on the generalized method of moments to estimate

the risk-return characteristics of nontraded assets from cash flow data. Applying it

to a sample of venture capital and buyout funds between 1980 and 2003, they get a

mixed impression of the investments’ performance. Phalippou and Gottschalg (2009)

consider an updated version of the Kaplan and Schoar data set, covering the period

from 1980 to 2003. From their results they conclude that the performance figures as

disseminated by industry representatives and earlier research are overstated due to

sample selection issues and artificially inflated net asset values. When corrected for

these biases, the average fund return net of fees falls short of the S&P 500 by as much

as three percent per year.Finally, Franzoni et al. (Forthcoming) provide evidence that

private equity returns include a sizable liquidity risk premium. As a consequence, the

diversification benefits of the asset class may be lower than traditionally assumed.

Owing to the absence of objective market values, analyses of private equity perfor-

mance need to rely on appraisal-based figures reported by the limited partnership

funds themselves. This, however, causes a variety of problems that are the subject

matter of another extensive strand of the private equity literature. Kaplan et al. (2002)

aim to assist academics and practitioners with the interpretation of their empirical

results by cross-checking the contract specifications of over 140 venture capital fi-

nancings and pointing out major biases in the two leading private equity databases.

Similarly,Woodward (2004) argues that the returns, as disclosed by general partners,

are inaccurate measures of the actual changes in value. Introducing an approach that

allows her to control for this issue, she is able to show that the risk of private equity

is higher than commonly assumed, both in terms of return volatility and beta. An-

other arguably seminal article has been contributed by Cochrane (2005), who applies

maximum likelihood estimation to correct the mean returns, volatilities, alphas, and

betas of venture capital investments between 1987 and mid-2000 for selection bias.

His results indicate that private equity behaves quite similar to traded securities.

Conroy and Harris (2007) find understated risks and exaggerated returns based on

data for the period from 1989 to 2005, which they attribute to the prevailing practices

of appraisal-based portfolio valuation and information disclosure by private equity

funds. Systematic valuation biases are also documented by Cumming and Walz(2010),

who analyze a data set of more than 5,000 portfolio companies as well as 221 funds

between 1971 and 2003. In addition, Cumming et al. (2010) criticize appraisal-based

indices for being subject to return smoothing and stale pricing, which results in an