The Relations among Organizational and Distribution Forms and Capital and Asset Risk Structures in the Life Insurance Industry

Baranoff, Etti, Sager, Thomas, Journal of Risk and Insurance

ABSTRACT

This article is the first step toward integrating in a single framework two previously separate lines of research on major structural decisions of life insurers. The literature has previously studied the relation between capital structure and asset risk on the one hand, and the relation between organizational form and distribution system on the other hand, without integrating them. Using life insurer data for 1993-1999, we model the four key insurer decisions of capital structure, asset risk, organizational form, and distribution system as endogenous choices in a single interrelated set of simultaneous equations. The model assesses the nature of the interactions among these decisions. The model also assesses the impact of insurers" fundamental business strategy (treated as predetermined) on these choices. The business--strategy hypothesis views other key decisions as jointly determined and driven by the fundamental business strategy, once the latter is set in motion. Confirming previous studies, we find a positive relation between capital ratios and asset risk. We also find an association in the simultaneous context between stock ownership and brokerage distribution, which was not found in prior studies. Stock ownership is related to greater financial and asset risk taking, whereas brokerage distribution is associated with lower risk taking. These and other results are interpreted in light of several theories, including transaction-cost economics (TCE), agency theory, and regulatory and bankruptcy cost avoidance. Deriving from these theories, the finite risk paradigm emerges as the most comprehensive interpretation of the results, as opposed to the risk-subsidy hypothesis of the impact of guarantee funds. We also find support for the notion that the business strategy drives the capital and distribution decisions, as predicted by TCE.

INTRODUCTION

Motivation

This article is the first step toward integrating two separate strands of research. The relation between capital structure and asset risk has been examined previously in a simultaneous framework (Cummins and Sommer, 1996; Baranoff and Sager, 2002). Similarly, the relation between organizational form and distribution system has received attention (Regan and Tzeng, 1999). (1) But the four factors of capital structure, asset risk, organizational form, and distribution system have not previously been analyzed as a single integrated set of simultaneously determined decisions. It has also been proposed that the firm's selection of a business product logically precedes and strongly influences some of these factors (Regan and Tzeng, 1999; Williamson, 1985, 1988). In this study we undertake an analysis of the life insurance industry for the period 1993-1999 using a model that treats the major structural decisions as an interrelated process once the product has been determined. We use the phrase business-strategy hypothesis for the idea that the product decision underlies and affects the four strategic decisions mentioned above.

This study is an empirical expedition. To be sure, we interpret our results in light of various applicable theories. But we do not attempt to validate any theory. The theories used for this purpose include transaction-cost economics (TCE), agency theory, and bankruptcy and regulatory costs hypotheses, which underlie the finite risk paradigm. The idea that firms limit total risk lies at the heart of the finite risk framework. Both TCE and agency theory provide a foundation for finite risk: TCE through its core idea that firms exist to minimize transactional costs, and agency theory through its idea that owners, managers, and customers monitor each other's behaviors. We find that finite risk provides the most comprehensive framework for our results. An opposing theory is the risk-subsidy hypothesis, which maintains that the operation of guarantee funds provides consumer safeguards. These safeguards reduce incentive to limit risk taking, especially by marginally solvent insurers.

Our objectives are threefold: (1) to interpret the relations among capital structure, asset risk, organizational form, and distribution system in an integrated decision making context, (2) to investigate the empirical foundation for the business-strategy hypothesis in this simultaneous framework, and (3) to examine the effects of various canonical controls, such as firm size and regulatory pressure.

Prior Literature

Integrated treatment of the capital structure/asset risk decision has received extensive attention in the banking industry. (2) Such studies are relatively new to insurance. We mention Cummins and Sommer (1996) for property/casualty insurers and Baranoff and Sager (2002) for life insurers. Cummins and Sommer (1996) found a positive relation between capital ratio and market measure for a portfolio of a firm's asset and product risks. (3) Baranoff and Sager (2002) found a positive relation between a measure of regulated asset risk (4) and the capital-to-asset ratio for life insurers, but a negative relation between a measure of product risk (5) and the capital ratio. They also found a positive relation between product risk and regulated asset risk.

The relation between organizational form and distribution structure has received much attention in studies in the property/casualty area, but not as simultaneously integrated determined decisions prior to Regan and Tzeng (1999). (6) The latter authors thoroughly reviewed the literature on the reciprocal effects of organizational form and insurer distribution system. They also contributed a new model that analyzes organizational form and distribution system as simultaneously determined decisions. Using this model, Regan and Tzeng (1999) found that organizational form and distribution system were not significantly related in the simultaneous framework for property/casualty insurers. However, in a nonsimultaneous mode, the stock form of organization was associated with the independent agency distribution system.

The Model

To study the capital/risk/organization/distribution relations, we use a model that combines categorical and continuous endogenous variables in a single framework. Four structural equations that describe the capital (continuous variable), asset risk behavior (continuous variable), organizational form (categorical variable), and distribution system (categorical variable) of lire insurers are estimated simultaneously in a limited-information approach based on the two-stage least squares paradigm. Capital and asset risk are modeled autoregressively; organization and distribution are modeled via probit regression.

To represent the business-focus driver of these relations, we used the proportions of an insurer's premiums in various product lines. Results show that a focus on annuities is associated with a low capital-to-asset ratio, low asset risk, and broker distribution; and a focus on group health business is associated with brokerage distribution and high capital-to-asset ratios, in line with theoretical expectations based on TCE as will be discussed in the next section.

The article is structured as follows. In the next section we develop the theoretical background for the strategic decision of the lire insurance industry during the 1990s. Next, we explain the data and the variables. This is followed by the development of the model and the analytical procedures. The "Empirical Analyses and Results" section provides the basic statistics and the results of the model followed by the discussion of the results. The article concludes with a summary.

THEORETICAL BACKGROUND

The Business-Strategy Aspect

Regan and Tzeng (1999) discussed the notion that an insurer first selects its business strategy (or product focus) and then decides jointly on the organizational form (stock or mutual) and distribution system (captive or independent agency). From this perspective, the organizational form/distribution system decisions are viewed as flowing from the basic business strategy. As noted above, we call this view the business-strategy hypothesis. In this article, we expand the scope of the business-strategy hypothesis to include the possibility that the capital/risk structure decisions are also consequences of the basic business strategy. To model the business-strategy hypothesis, we treat the business strategy/product focus as a predetermined (7) variable, and the four factors of capital structure, asset risk, organizational structure, and distribution system as endogenous.

We now present a brief discussion of the theoretical foundation for the business-strategy hypothesis. Several theories have a hand in this, including transaction-cost economics (TCE), agency theory (monitoring hypothesis), bankruptcy cost avoidance, and regulatory cost. All of these are consistent with a finite risk framework since they predict constraints on overall risk taking. TCE provides an especially natural explanation for the impact of the business strategy on capital structure and distribution system. The other theories collaborate to explain the impact of business strategy on asset risk and organizational form.

TCE theory was first introduced by Coase (1937) and further expanded by Williamson (1975, 1985, 1988, 1990) and Klein, Crawford, and Alchian (1978), among others. (8) Under TCE, the level of transaction costs generated by the characteristics of the products largely determines the degree of vertical integration (Williamson, 1985) and the form of capital structure (Williamson, 1988). These transaction costs derive chiefly from the uncertainties of contractual relationships for producing and distributing the products. Products that involve large contractual uncertainties are subject to opportunistic behavior and conflicts among the stakeholders, thus increasing the transaction costs. According to Williamson, firms seek to reduce or offset such costs by vertical integration (1985). Williamson (1988) notes that a firm with assets of greater specificity, complexity, and idiosyncrasies would have a higher proportion of equity than a firm specializing in assets with less specificity. Kochhar (1996) elaborates on Williamson's hypothesis and explains that in TCE, equity is considered the more powerful governance instrument. Kochhar indicates that since "information asymmetry cannot be reduced in the transaction cost logic, the result is failure of the market form of exchange (debt) for firms with assets of high specificity level." Thus, TCE speaks directly to the effect of the product on capital and distribution structures. TCE, however, does not speak directly to the influence of product on the organizational structure or asset risk. It should also be noted that agency theory can provide a complementary interpretation. When contracts create conflicts among the stakeholders, owners develop monitoring techniques to control management (Mayers and Smith, 1981).

To apply TCE and the other theories to the life insurance industry we need a clear appreciation of the risk characteristics of life insurance products and an evaluation of which life insurance products are riskiest. Life insurers market a wide variety of products, not just life insurance policies. From the standpoint of TCE (Williamson, 1985), the riskiest products involve relational contracts that are complex, incomplete, and uncertain. The least risky products involve classical contracts that are simpler, more explicit, and less uncertain. Based on this idea, Baranoff and Sager (2002) view health products as the riskiest line for life insurers. Carr, Cummins, and Regan (1999) consider annuities the least risky and view group insurance as less risky than individual contracts. Group business is the commercial line of the life industry. As in the property/casualty industry, commercial business is more complex and less uniform. Group contracts are customized through negotiation with employers and so are less standardized than policies sold to individuals. They are more incomplete and subject to opportunistic behavior. Thus we view group business as riskier than individual contracts in the life industry. Following Williamson (1988) and Kochhar (1996), it would be expected that an emphasis on riskier products such as health, life, and group business would be associated with capital financing rather than with debt. Costs of regulation and bankruptcy avoidance (the finite risk paradigm) add support to this expectation, since an orientation toward risky products could be offset by a less risky capital structure. The impact of risky products on the distribution system is more complex in its application to the life industry. Group products, especially group health, represent the greatest risk, both to insurer and insureds since the contracts are incomplete and relational. It would be expected that the contracting parties would seek to protect their interests by integrating the group products. Consistent with Regan (1997), commercial insureds seek to do this by selecting brokers, who legally represent them (9) and who can monitor insurer behavior on their behalf. An insurer may also prefer to have a broker as a third party to buffer conflicts. Thus, we may expect that …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: The Relations among Organizational and Distribution Forms and Capital and Asset Risk Structures in the Life Insurance Industry. Contributors: Baranoff, Etti - Author, Sager, Thomas - Author. Journal title: Journal of Risk and Insurance. Volume: 70. Issue: 3 Publication date: September 2003. Page number: 375+. © 2009 American Risk and Insurance Association, Inc. COPYRIGHT 2003 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.