Determinants of Cash Holdings by Property-Liability Insurers

Abstract

ABSTRACT This study investigates the differences in cash holdings across property-liability insurers. We conclude that relative cash holdings are less for insurers with better access to cash through capital markets and/or other group members. We also conclude that larger insurers, higher quality insurers, insurers that write longer tail lines of business, and firms with higher degrees of leverage hold less cash. Also, we find that insurers with a higher variance of cash flows tend to hold more cash. Another interesting finding is that, contrary to what managerial discretion arguments might suggest, stock insurers tend to hold more cash than do mutuals. INTRODUCTION This article examines the variation in cash holdings across property-liability insurers during the three-year period from 1993 to 1995. For the property-liability insurer, virtually all business transactions occur in cash. Premiums are received in cash, and claims are paid in cash. As a result, the insurer's decision regarding the amount of cash to hold is critical to its operations and therefore to its overall financial stability. In making that decision, an insurer may choose to hold a large amount of cash; such a strategy affords much flexibility to the insurer, but the flexibility gains are offset by the opportunity costs resulting from the lower returns generated by cash compared to less liquid assets (Amihud and Mendelson, 1986). On the other hand, an insurer may choose to hold a small amount of cash; such a strategy maximizes returns by investing the cash in higher yielding assets but exposes the insurer to transaction costs--and potentially unfavorable economic conditions--when assets must be liquidat ed to meet obligations. Because there are competing costs and benefits of holding cash, the insurer must compare the marginal costs of holding cash to the marginal benefits of holding cash when determining its optimal level of cash (Opler, Pinkowitz, Stulz and Williamson, 1999). Cash holdings vary considerably across insurers. [1] Such differences can be expected for a number of reasons, including the degree of agency conflict among the insurer's management, owners, and policyholders; the ability of the insurer to generate cash from alternative sources; the nature of the insurer's operations; and the composition of the insurer's portfolio of non-cash assets. In order to determine which insurer characteristics affect the extent of cash holdings, we employ regression analysis on a large sample of property-liability insurers over a three-year period. Our examination of cash holdings by property-liability insurers serves several purposes. First, it contributes to the large body of literature investigating corporate cash holdings (Chudson, 1945; Baumol, 1952; Meltzer, 1963; Frazer, 1964; Vogel and Maddala, 1967; Gertler and Gilchrist, 1994; Kim, Mauer and Sherman, 1998; Opler, Pinkowitz, Stulz and Williamson, 1999). Because prior literature has excluded the analysis of financial firms, investigation of insurer cash holdings provides the opportunity to evaluate the extent to which past findings hold for a totally different class of firms. Perhaps more importantly, the unique aspects of the insurance industry, such as the different organizational forms available to insurers and the phenomenon of insurer groups, provide the opportunity to generate additional insights into cash holdings. Additionally, the examination complements previous research investigating other components or aspects of insurer investment portfolios (Colquitt and Hoyt, 1997; Cummins, Philli ps and Smith, 1997; Lee, Mayers and Smith, 1997; Cox, Gayer and Wells, 1998). Finally, the examination demonstrates that insurers choose cash balances systematically based on their organizational and operational characteristics. This information should benefit stakeholders, regulators, and academics in their attempts to understand and predict insurer behavior. …