Managing Capital via Internal Capital Market Transactions: The Case of Life Insurers

• Author: Greg Niehaus
• Published date: 01 March 2018
• DOI: http://doi.org/10.1111/jori.12143
• Date: 01 March 2018

© 2016 The Journal of Risk and Insurance. Vol. 85, No. 1, 69–106 (2018).

DOI: 10.1111/jori.12143

69

MANAGING CAPITAL VIA INTERNAL CAPITAL MARKET

TRANSACTIONS:THE CASE OF LIFE INSURERS

Greg Niehaus

ABSTRACT

The movement of capital within insurance groups is important for

understanding insolvency risk management, as well as regulatory policies

regarding capital standards and group supervision. Panel data estimates

indicate that, on average, a dollar decrease in performance (net income plus

unrealized capital gains) when performance is negative is associated with a

$0.26 increase in capital contributions to life insurers from other entities in

the group, and that a dollar increase in performance when performance is

positive is associated with a $0.56 increase in the amount of internal

shareholder dividends paid by life insurers to other entities in the group.

Moreover, the sensitivity of internal dividends to performance is higher

during the financial crisis than the noncrisis period. Also, insurers with

low (high) risk-based capital ratios receive more (less) internal capital

contributions than other insurers, holding other factors constant.

INTRODUCTION

Despite a large literature on the determinants and regulation of capital held by

insurance companies, there is limited research on how these institutions manage

capital internally, that is, how they move capital among companies under common

Greg Niehaus a Professor of Insurance and Finance, Moore School of Business, University of

South Carolina, Columbia, SC 29208. Niehaus can be contacted via e-mail: gregn@moore.sc.

edu. The author appreciates the research assistance of Justin Choma (who was supported by the

Moore School Research Grant Program) and Chia-Chun Chiang. The author also recognizes the

comments and suggestions from an anonymous reviewer, D. H. Zhang, Motohiro Yogo, Cindy

Soo, Daniel Schwarcz, Richard Rosen, Tom Rietz, Richard Paulsen, Greg Nini, Tanakorn

Makaew, Tyler Leverty, Robert Kasinow, Chuck Fang, Cameron Ellis, Randy Dumm, Michele

Barnes, seminar participants at the Universities of Georgia, Iowa, and South Carolina, and

conference participants at the 2014 Annual meetings for the Financial Intermediation Research

Society, Western Economic Association, Insurance Risk and Finance Research Conference, and

American Risk and Insurance Association.

70 THE JOURNAL OF RISK AND INSURANCE

ownership.

1

Understanding the allocation of internal capital in financial institutions

is important for insolvency risk assessment purposes, whether this is done by

consumers, regulatory authorities, rating agencies, or internal risk managers.

Understanding internal capital flows is also important for the formulation of policies

regarding capital regulation and group versus company level supervision, topics that

have received considerable attention since the financial crisis.

2

The purpose of this article is to provide a peek inside insurance holding company

systems to see how capital flows among the internal entities. The variables examined

are motivated by a simple framework that views a holding company structure as an

arrangement for allocating capital where it is most productive. The framework

predicts that the flow of capital within a group, all else equal, will be inversely related

to an insurer’s performance. That is, when an insurer performs poorly (well) for

exogenous reasons, capital will flow to (away from) the insurer. The empirical

evidence, from both aggregate data and panel data, support this prediction. The

negative relation between performance and internal capital flows may appear

contradictory to intuition, and studies in the internal capital markets literature that

indicate that capital should flow to entities that add the most value, which under

plausible conditions means to entities that have performed well. The apparent

discrepancy is easily resolved by the institutional context: in the life insurance

industry, deterioration in a firm’s capital, which follows from poor performance, is

likely to be associated with reduced demand for the firm’s products and a negative

impact on franchise value.

3

As a consequence, for insurers whose capital is

exogenously depleted (augmented), the marginal impact of additional capital on the

1

Examples of insurer capital structure studies are Baronoff and Sager (2002), Cummins and

Nini (2002), Klein, Phillips, and Shiu (2002), Harrington and Niehaus (2003), and Fier et al.

(2013). The literature related to insurance groups will be reviewed below. Papers on internal

capital markets in banks include Houston, James, and Marcus (1997), Houston and James

(1998), Campello (2002), Holod and Peek (2010), de Haas and van Lelyveld (2010), and

Cremers et al. (2011).

2

AIG’s problems during the financial crisis have led to calls for greater group-level supervision,

federal regulation of insurance, and greater coordination between regulatory authorities in

other countries. See Harrington (2009) for a discussion of AIG and the financial crisis; also see

Committee on Financial Services (2010). The National Association of Insurance Commis-

sioners (NAIC) identified group regulatory issues as one of the focus areas of its Solvency

Modernization Initiative (see Solvency Modernization Initiative Task Force, 2013). In 2010 the

NAIC modified model laws to enhance the ability of insurance regulators to “look in the

window” of noninsurance affiliates. See Insurance Holding Company System Regulatory Act

(Model #440) and the Insurance Holding Company System Model Regulation with Reporting

Forms and Instructions (Model #450). Also, the task force has stated that it will require a group

capital assessment under the Own Risk Solvency Assessment (ORSA). The Federal Insurance

Office’s (2013) report recommends movement toward more uniform regulation of national

and global insurers and group supervision.

3

See, for example, Epermanis and Harrington (2006) for empirical evidence on the response of

premiums to ratings, and see Eling (2012) for a review of the literature on market discipline in

insurance.

MANAGING CAPITAL VIA INTERNAL CAPITAL MARKET TRANSACTIONS 71

insurer’s franchise value is high (low), and so internal capital is transferred to (away

from) these insurers.

There are of course potential costs of shifting additional capital to an insurance

company in a group. For example, adding capital will diminish the value of the option

to default on that entity’s claims. In addition, given regulatory restrictions on the

ability to remove capital from insurance companies, placing capital in an insurance

entity potentially diminishes the ability to reallocate that capital to better uses in

future periods. Given these trade-offs, insurance groups have an incentive to allocate

sufficient capital to their insurance entities to provide assurance to policyholders/

rating agencies that the insurers will meet claim payments, but not put excess capital

in the insurers. Instead, the group will hold excess capital in noninsurance entities so

that it can be allocated to other projects. This reasoning implies that groups will

transfer capital out of insurers when they are performing well and transfer capital to

insurers following poor performance. This negative relationship between perfor-

mance and internal capital transfers is the focus of this article.

The data are from life insurance companies from 2006 to 2011. This time period was

selected so that internal capital movements during the financial crisis could be

examined. As discussed by Berry-Stolzle, Nini, and Wende (2014) and Niehaus

(2013), the life insurance industry, especially annuity providers, experienced lower

operating income and large losses on their investments during 2008 and 2009. As a

consequence, the capital cushion of many life insurers and annuity providers would

have declined substantially during the crisis if not for actions that replenished capital

(discussed further below), including internal capital market transactions.

I examine insurance company internal capital market transactions in the aggregate

over time and using a panel data set. The aggregate data indicate substantial variation

in internal capital market activity over the 6-year sample period. For example, in the

2 years before (after) the 2008–2009 financial crisis, life insurers contributed

$2.7 billion ($3.0 billion) in capital to nonlife insurance entities in their group, but

during the financial crisis these life insurers received $22.3 billion in capital

contributions from nonlife insurance entities in their group. This evidence, along with

other evidence using the aggregate time series data, suggests that internal capital

markets act as a risk sharing arrangement for life insurers, where internal capital

flows out of the industry during good times and into the industry during bad times. It

also indicates that the capital of noninsurance entities is used in insurer capital

management.

The panel data analysis indicates that internal capital market transfers (capital

contributions and shareholder dividends) are related to the performance and

capitalization of individual insurance companies. When life insurer performance (net

income plus unrealized capital gains) is negative, a dollar decrease in performance is

associated with a $0.26 increase on average in capital contributions received. When

life insurer performance is positive, a dollar increase in performance is associated

with a $0.56 increase in internal shareholder dividends paid, on average. Stated

differently, if one just considers internal capital market transactions, life insurers

augment their capital by about $0.44 for each dollar that they earn and their capital is

reduced by about $0.74 for each dollar that they lose, on average. Since performance