Robert E. Hoyt and Kathleen A. McCullough discuss managerial discretion and property/casualty reserving practices.

Financial statements are a major source of information to a property/casualty insurer's many stakeholders: investors, policyholders, managers, regulators, tax authorities, rating agencies, and reinsurers. For the typical property/casualty insurer, over two-thirds of its liabilities are represented by reserves (liabilities) for losses and related expenses that are not yet paid. These loss reserves typically fall into three categories: reserves for losses that have been reported and settled, but not yet paid; reserves for losses that have been reported, but are still in the process of settlement; and, finally, reserves for losses that have been incurred but are not yet reported, so called IBNR reserves. Since a number of factors can influence the ultimate magnitude of the actual loss payments an insurer makes, the statement reserves may be over or understated relative to these ultimate payments to claimants.

Most insurance professionals realize that these loss reserve calculations require a significant degree of estimation. In fact, many would suggest that loss reserving is as much an art as it is a science. This means that a figure on insurer balance sheets that has an aggregate value of well over $300 billion in the US property/casualty insurance industry is subject to a significant level of accounting discretion. Relatively modest adjustments in an insurer's reported reserves can have a significant impact on the magnitude of other financial measures, including reported earnings, tax liabilities, and surplus or equity.

Given the financial magnitude of loss reserves, it is perhaps not surprising that academics have taken a great interest in the level of accuracy of management's estimates of these liabilities. An extensive literature has developed in the academic press regarding the following questions: (1) Do property/casualty insurers manage reserves?; (2) Why do property/casualty insurers engage in intentional management of reserves?; and, (3) Do interested parties recognize that intentional reserve manipulation goes on? In this article we review what the academics have had to say on these questions.

Do property/casualty insurers manage reserves?
Recognizing that loss reserves are estimates of future obligations that are subject to varying degrees of uncertainty, most of the initial academic studies sought to isolate “unintentional” reserving errors from those that were “deliberate.” Unanticipated changes in factors that affect the loss distribution, such as inflation and an evolving legal environment, can contribute to reserving errors that are unintentional. However, once these unintentional factors are controlled for, is there any evidence that insurers are engaging in a systematic pattern of earnings management which is consistent with the exercise of managerial discretion over the accounting statement of loss reserves? Several studies have provided relatively convincing evidence that insurers do in fact engage in some level of deliberate reserve manipulation. (See, for example, Weiss, Journal of Risk and Insurance, 1985; and Grace, Journal of Risk and Insurance, 1990.) Further, it seems clear that since IBNR reserves are the most subjective, they also would provide the greatest opportunity for the exercise of managerial discretion. Evidence of earnings management in this context was provided by Aiuppa and Trieschmann (Journal of Risk and Insurance, 1987) in their study of IBNR reserving practices.

Why do property/casualty insurers engage in intentional management of reserves?
With a relatively convincing case built in the literature for the existence of intentional reserve management by insurers, academics turned their attention to evaluating the various motives for such manipulation. The major motives for earnings management generally relate to efforts directed at: making the firm appear more financially sound than it really is; reducing, or at least smoothing, reported earnings to reduce the present value of taxes owed; bolstering a claim to the regulator for a rate increase due to higher losses, or justifying a rate decrease due to declining losses.

As participants in a highly regulated industry, insurers are under scrutiny from regulators for a number of reasons. Clearly, a key focus of regulatory oversight is assuring the financial solvency of insurers operating in the market. Regulatory oversight tools like IRIS, FAST, and, more recently, risk-based capital (RBC), are evidence of this focus. Since these oversight tools rely, at least in part, on financial ratios or formulas containing financial data, it is interesting to consider what impact these tools have on insurers' incentives to manage reserves.

Several studies, including Petroni (Journal of Accounting and Economics, 1992) and Gaver and Paterson (Contemporary Accounting Research, 1999), find evidence that insurers that have been flagged as financially troubled under IRIS, or that are near the financially-troubled range, are more likely to engage in reserve manipulation to “enhance” their financial condition. Since understating reserves simply postpones the recognition of loss payments that will eventually have to be made, this activity can exacerbate the financial fallout that results if an insurer ultimately becomes insolvent. Also, in a recently completed study of the effect of risk-based capital requirements on reserving practices, we find strong evidence that insurers exercised accounting discretion in an attempt to avoid the unwanted regulatory attention that might accompany a level of capital that falls below the regulatory action or control levels proscribed by the NAIC's risk-based capital guidelines. Our results suggest that firms that had capital below the company action level (200% of the final result of the RBC formula) systematically reduced their estimates for IBNR reserves relative to firms that had capital levels well outside of the range of regulatory scrutiny (we defined this as above 400% of the RBC formula value).

Although it has not been the primary focus of academic studies in this area, it also seems clear that management's discretion over the appropriate level of loss reserve estimates provides a vehicle for supporting a desired level of pricing. In the context of regulated rates, this discretion could be used to justify a rate increase by overstating losses, or a rate decrease by understating losses.

Tax planning affects a host of financial decisions made by any firm. A number of academic studies have considered the potential for insurers to minimize their income tax liability by managing loss reserves to smooth earnings. The results provide evidence of reserving practices targeted at minimizing income taxes. Of course, overstated reserves must be adjusted downward at some point. However, given the asymmetric nature of the income tax system resulting from progressive tax rates and the unequal tax treatment of profits versus losses, a firm can reduce its present value of expected taxes by smoothing income. Reserve manipulation is a powerful tool in accomplishing this income smoothing. A recent study by Petroni and Shackelford (NBER, 1998) also provides evidence that multi-state insurers exercise accounting discretion to allocate premiums and losses across states with differing premium tax rates in an effort to minimize their premium tax liabilities.The power of this discretion is perhaps most apparent in the findings of a study by Nelson (Stanford University, 1997). Under statutory accounting practices in the US loss reserves are not explicitly discounted for the time value of money. Nelson found evidence consistent with insurers exercising discretion over reserves to build in an implicit discount for the time value money in their statement of loss reserves. This implicit discount is greater the longer the payout tail of the insurer's liabilities. That is, relative to the undiscounted accounting liability for losses, reserves are understated by a greater amount in a line like products liability line than in a shorter tailed line like auto property damage.Do interested parties recognize that intentional reserve manipulation goes on?Given the evidence that insurer managers have significant discretion over the levels of these accounting numbers, it is interesting to consider whether the various stakeholders recognize this. A question that is considered by several studies is: Do investors and others recognize these incentives for earnings management and adjust accordingly?Not surprisingly, investors seem to view earnings announcements from insurers with noisier earnings (due to claim expense estimation errors) as being less informative (Anthony and Petroni, Journal of Accounting Auditing and Finance, 1997). However, more importantly, investors seem to be able to anticipate the relative incentives that managers have to manage earnings. Christensen, Hoyt and Paterson (Journal of Business Finance and Accounting, 1999) found that earnings announcements from insurers whose managers were expected to have greater incentives to manage earnings through the exercise of accounting discretion produced less of a reaction in the market. In effect, investors adjust their assessment of the quality of earnings reports based on their belief regarding managers' incentives to manage these numbers through manipulation of loss reserves. This finding suggests that investors are not fooled by managers' efforts at shining up the balance sheet and income statement. However, of the over 3,000 property/casualty insurers in the US, perhaps fewer than 200 are actively traded public companies. It is much less clear whether regulators and policyholders are capable of identifying the nature and extent of managers' incentives to manage earnings. The fact that so many academic studies find evidence of intentional reserve manipulation suggests that at least some of the stakeholders are unaware of the full magnitude of this accounting discretion.So are insurers the only corporations that engage in discretionary reporting to manage earnings? Certainly not. The academic accounting literature is full of examples of firms that exercise managerial discretion over accounting accruals to achieve various goals. However, few industries have two-thirds of their liabilities in items that are subject to such significant estimation uncertainty. As a result, the opportunity for management of a firm's financial results through exercise of managerial accounting discretion is significant in the property/casualty insurance industry. It should be clear from this that insurer financial statement consumers, such as reinsurers, regulators, investors, policyholders, and others, must carefully look behind the numbers if they expect to develop a reliable picture of the financial performance and condition of the insurers with which they do business.Robert E. Hoyt, Ph.D., is Associate Professor of Risk Management and Insurance in the Terry College of Business at the University of Georgia. Kathleen A. McCullough is a Ph.D. candidate in risk management and insurance at the University of Georgia.