...distressed. They are generally small and have low return on assets and cession ratios. Relative to holdings of liquid assets they have high levels of property and reinsurance assets, and low levels of equity holdings. They also write more overseas business, and less motor insurance and long-tailed insurance lines, relative to fire and household insurance.

INTRODUCTION

The rapid pace of financial innovation, globalization, and deregulation of the financial system over the past decade has made the operations of financial intermediaries (FIs) more complex and potentially more risky. These developments, in turn, have led prudential supervisors to design new processes for monitoring and identifying FIs experiencing deteriorating financial conditions (Sahajwala and Van den Berg, 2000). These processes include on-site examination systems, statistical early warning systems, and market based measures of FI risk. (1)

Although these techniques have their broadest application in banking, a number of insurer failures in the United States in the mid 1980s and early 1990s led to a questioning of the effectiveness of insurance regulation and solvency monitoring and the ability of regulators to identify insurers at risk of insolvency at an early stage of their financial deterioration. The subsequent regulatory reforms led to the adoption of the FAST (Financial Analysis and Surveillance Tracking) solvency monitoring system for insurers and the introduction of risk-based capital requirements. These developments were associated with a number of empirical models that examined the ability of accounting based statistical models to predict the insolvency of life and health insurers and property-liability insurers. (2)

Outside the United States there has, until recently, been little interest in developing and applying solvency prediction models to the insurance industry. In the Australian context, Black (2004) provides an overview of the formation of the Australian Prudential Regulation Authority (APRA) in 1998 and its move toward a risk-based prudential approach. The failure and appointment of provisional liquidators in March 2001 to HIH, a large insurance company, led to concerns relating to the adequacy of APRA's approach to prudential regulation and supervision and to the appointment of a Royal Commission to investigate the collapse of HIH. The Commission was critical of APRA, reporting that "the manner in which APRA exercised its powers and discharged its responsibilities under the Insurance Act fell short of that which the community was entitled to expect from the prudential regulator" (HIH Royal Commission, 2003). In April 2001, APRA's board concluded that it should have had a mechanism for identifying institutions at risk, and that it should require more explicit and timely information about those institutions (Black, 2004, p. 33).

Drawing on the Canadian and U.K. approaches, APRA's response was to develop a risk-based approach within the Probability and Impact Rating System (PAIRS) and Supervisory Oversight and Response System (SOARS) frameworks. Although numerically scored, the approach is largely qualitative relying on assessments provided by APRA's supervisory staff. There has been little effort within APRA or the Australian academic or professional community to develop or assess quantitatively based systems to identify financial institutions in financial distress. (3)

We address this gap in the literature by developing a model to identify Australian general insurance companies (GIs) in financial distress. As general insurance includes any insurance that is not life insurance, it is also referred to as nonlife insurance. GIs provide cover against loss from a particular financial event including workers compensation, public liability, product liability, automobile and homeowners' policies, pet insurance, mortgage and creditor insurance, and others. In the United States, GIs are known as property and casualty insurers or property-liability insurers.

There is an extensive body of U.S. literature that develops and tests static and/or dynamic insolvency prediction models for property-liability and life-health insurers (Chen and Wong, 2004). The dynamic approach involves deterministic or stochastic cash flow modeling (Ceccarelli, 2003; Cummins, Grace, and Phillips, 1999). It is assumed that the insurer collects premiums for the next 1 or 2 years, after which time it goes into "run-off." The principal cash flows for each insurance line are then modeled for a horizon of 20+ years, adopting a set of assumptions for investment returns, loss (claims) development factors for each line, expenses, etc. The net worth of the insurer is then determined at the 20+ year time horizon and the insurer is classified as insolvent if its projected net worth is zero or negative.

Data limitations currently preclude an application of the cash flow simulation approach to Australian GIs. Moreover, because detailed data on an insurer's claims experience are typically not in the public domain, (4) it can be argued that good dynamic modeling is best undertaken by insurance companies as part of APRA's dynamic simulation modeling and stress testing of large financial institutions. The static approach may then be viewed as complementary to the dynamic internal stress testing approach. In this respect, Cummins, Grace, and Phillips (1999) find that the cash flow simulation model adds significant discriminatory power to the risk-based capital and FAST early warning models used in the United States.

The static approach, which we adopt in this article, utilizes financial ratios and a variety of statistical techniques to identify factors that discriminate between samples of solvent and insolvent insurers. The techniques include multiple discriminant analysis, recursive partitioning, neural networks, and logit and probit analysis.

There is, however, a difficulty in applying this approach outside the United States because most countries have limited experience with failing insurers and lack the number of insurer insolvencies to be able to estimate the models. Hence, our focus is on financial distress, rather than insolvency, of insurers. The costs of financial distress borne by stakeholders of a troubled insurer, although less than the costs of insolvency or regulatory takeover, are significant. They include reductions in the market value of the insurer's debt and equity, additional supervisory costs incurred by the prudential regulator, and disruption to policyholders and insurance markets with increased uncertainty as to the value of the insurer's policy reserves.

In the following section, we critically review two approaches to defining financial distress that have been used in studies of Asian and Dutch insurers where the limited number of insurer insolvencies precludes a study of insolvencies. The approaches suffer from a number of problems including subjectivity, a focus on supervisor's assessments, and/or explaining "financially unstable" insurers using financial ratio regressors similar to those used in the classification system to determine whether the insurer was financially unstable. We avoid these problems by developing an alternative regulatory based measure of financial distress derived from the Solvency Condition in the Insurance Act applying to Australian GIs.

After defining financial distress, we specify a model of the determinants of financial distress, describe the data used in estimating the model, summarize the results, and conclude with a summary and suggestions for further research. From a regulatory perspective, our objective is to identify GIs likely to experience financial distress so that prompt and appropriate corrective action can be implemented. (5)

DEFINING FINANCIAL DISTRESS

In the absence of a significant number of insurer insolvencies in Australia, we focus on predicting "financial instability or weakness" as an indicator of financial distress. There are two approaches in the literature for handling this problem. Kramer (1996) evaluates the financial solidity of Dutch nonlife insurance companies using a threeway subjective classification system of whether the company is "strong," "moderate," or "weak" as the dependent variable. In the Netherlands, supervisors provide a written assessment of the insurer following an examination process. "The major weak and strong points of a company are described in that text as value statements, criticisms, or indications of negative or positive trends" (Kramer, 1996, p. 81). The researcher's classification is then based on the number and severity of the weak and strong points in the assessment.

This approach raises several issues. First, there is considerable subjectivity involved in applying the classification from the written assessments. Second, the approach is really a study of how supervisors assess the financial solidity of insurers and thus is more appropriate for studies of determinants of examiner ratings when a precise rating is not available.

The second approach is used by Chen and Wong (2004) in a study of the financial health of Asian insurance companies. They adopt different classification approaches for general and life insurers. For general insurers they use a variant of the Insurance Regulatory Information System (IRIS) as implemented by the National Association of Insurance Commissioners (NAIC) in the United States in the 1970s and 1980s. They examine 14 financial ratios under three categories: liquidity, profitability, and capacity. Each of the ratios for an insurer is compared to an industry norm and the insurer is classified as "financially unstable" if it fails to meet the industry standard in five or more of the 14 ratios. They then run a logit regression to explain "financially unstable" insurers using similar financial ratios to those in the classification system.

For life insurers Chen and Wong use the HHM model (see Hollman, Hayes, and Murrey, 1993). This model computes an index of financial solidity based on the relative change in a set of financial ratios over the period. The implicit assumption is that companies in financial distress make larger changes in key financial statement ratios than those that are not distressed. The solidity index for individual insurers is averaged and insurers with an index value worse than the average are classified as financially unstable. The dichotomous stability variable is then used as the dependent variable in a logit regression.

A weakness of the Chen and Wong approach lies in the regressions which, in effect, explain dichotomous dependent variables derived from financial ratios by independent variables which are almost identical to the financial ratios used in the classification system. However, the problem may not be quite as severe for the application of the HHM model where the classification criteria are based...

NOTE: All illustrations and photos have been removed from this article.




Looking for additional articles?
Search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.