Insurance Meeting

May 21, 2010
Ken Froot of Harvard Business School and NBER and Howard Kunreuther of the University of Pennsylvania and NBER, Organizers

Martin F. Grace, Georgia State University, and J. Tyler Leverty, University of Iowa
How Tort Reform Affects Insurance Markets

One critique of tort reform is that promised reductions in insurance rates do not follow reform enactment. However, a number of reforms are subsequently declared unconstitutional or repealed legislatively. Accordingly, Grace and Leverty investigate the duration of tort reforms enacted between 1985 and 2005. They then use the estimated survival probability of tort reform to examine its impact on state liability insurance markets. Unlike previous studies that rely on a binary measure of tort reform-that is, whether a state has a reform or not-theirs finds that tort reforms reduce premiums. Their results suggest that examining the effect of a current law without accounting for its expected future treatment may produce misleading results.

Brian Cheyne and Greg Nini, University of Pennsylvania
Creditor Mandated Purchases of Corporate Insurance

Cheyne and Nini present novel empirical evidence on the contractually mandated purchase of insurance by corporate creditors. Analyzing a large sample of private credit agreements of publicly-traded firms, they find that nearly all of them contain at least a boilerplate provision requiring the borrower to purchase insurance. In about 80 percent of the agreements, the insurance covenant is more explicit. The researchers focus on four additional features of the insurance covenant: explicit permission for the borrower to self-insure; requirements of coverage for specific risks; naming the lender as a loss payee; and mandating that any insurance proceeds be used to repay the loan. They find that credit agreements contain more stringent insurance requirements for borrowers who are smaller and pose higher credit risk, measured in a variety of ways. They also find that insurance requirements are highly correlated with many other terms of the loan and are very strongly positively correlated with the loan being secured by collateral and the loan size being limited by a borrowing base. This latter evidence suggests that insurance creates value by protecting lenders from unexpected changes in seniority that might occur after the destruction of collateral or a large liability suit. Mandatory insurance requirements appear to be an important ingredient of credit agreements, designed to encourage monitoring by senior, secured lenders.

Francis Ghesquiere and Olivier Mahul, World Bank
Financial Protection of the State against Natural Disasters

Ghesquiere and Mahul address the establishment or strengthening of financial strategies to increase the financial-response capacity of governments of developing countries in the aftermath of natural disasters, while still protecting their long-term fiscal balance. They analyze various aspects of emergency financing, including the type of instruments available, their relative costs and disbursement speed, and how these can be combined to provide cost-effective financing for the different phases that follow a disaster. Their paper shows in a didactic manner why governments are usually better served by retaining most of their natural disaster risk through risk retention, using risk-transfer mechanisms to manage the excess volatility on their budget, or to access immediate liquidity in the aftermath of a disaster. Finally, they discuss innovative approaches to disaster risk financing and provide examples of strategies implemented by developing countries in recent years.

Anastasia Kartasheva, University of Pennsylvania, and Sojung Park, California State University, Fullerton
Rating Standards for Catastrophic Risks and the Insurers' Capital Structure

In the aftermath of Hurricane Katrina in 2005, the major rating agencies increased the amount of capital that an insurer has to hold in order to maintain the current rating standard. The changes in the rating methodology and in the catastrophic models had a significant effect on the amount and composition of capital and reinsurance needed to achieve a particular rating. Kartasheva and Park analyze the impact of the new standards on the capital structure of insurance companies and show that new standards have a heterogeneous effect. While some insurers increase the amount of capital and improve the credit quality, others find the new standard too costly. In the latter case, the insurers either reduce the exposure to catastrophic risk by selling fewer policies or admit the rating downgrade, reduce the capital, and become more risky.

John A. Major, Guy Carpenter & Company, Inc.
Information Asymmetry in the M-Curve Model

Major adapts two assymetric information methodologies to the Major-Froot M-curve model in order to calculate the probability of success, and effective cost, of raising capital via seasoned equity offering. His core assumption is that while both management and investors know the relationship between wealth and market value, only management knows current wealth. In the context of insurance companies, this can be interpreted as asymmetric information about reserve adequacy. Based on a plausible M-curve and parameters representing the U.S. property and casualty insurance industry as a whole as if it were a single firm, effective costs peak in the range of one hundred-to-four-hundred percent. In states of high surplus, information effects drive the probability of successful issue to nearly zero. In states of covert insolvency, however, there is a significant probability -- thirty to fifty percent -- of a successful issue that transfers wealth from new to existing shareholders.

Daniel Schwarcz, University of Minnesota
Regulating Consumer Demand in Insurance Markets

In recent years, it has become increasingly clear that Expected Utility Theory (EUT) is remarkably poor at explaining how and why individuals purchase insurance, but the normative implications of this conclusion have remained largely unexplored. Schwarz takes up this issue, arguing that many observed deviations from EUT are likely the result of mistakes, in the sense that consumers would act differently than they do if they possessed perfect information and cognitive resources. From this perspective, regulatory interventions designed to improve consumer decisionmaking about insurance are potentially desirable. At the same time, he argues that some deviations from EUT may actually reflect sophisticated consumer behavior. In some cases, seemingly puzzling insurance decisions may help consumers to manage emotions such as anxiety, regret, and loss aversion, while in other cases they may represent valuable commitment strategies. Because consumers' insurance decisions may reflect sophisticated, rather than mistaken, decisionmaking, regulatory interventions that limit consumer choice are normatively troubling. Given these conflicting explanations for EUT's failure as a descriptive theory of consumer demand in insurance markets, Schwarz explores a spectrum of "Libertarian Paternalistic" regulatory interventions. He argues that regulatory strategies that aim to encourage presumptively welfare-maximizing insurance decisions without restricting individual choice represent a promising and normatively defensible opportunity for improving consumer behavior in insurance markets.

Liran Einav, Stanford University and NBER; Amy Finkelstein, MIT and NBER; Iuliana Pascu, MIT; and Mark R. Cullen, Yale University School of Medicine
How General are Risk Preferences? Choices under Uncertainty in Different Domains (NBER Working Paper No. 15686)

Einav, Finkelstein, Pascu, and Cullen examine the extent to which individuals' actual insurance and investment choices display a stable ranking in willingness to bear risk, relative to their peers, across different contexts. The researchers examine the same individuals' decisions regarding their 401(k) asset allocations and their choices in five different employer-provided insurance domains, including health and disability insurance. Among the five insurance domains, the magnitude of the domain-general component of preferences appears substantial; for example, one's choices in other insurance domains are substantially more predictive of one's choice in a given insurance domain than either one's detailed demographic characteristics or one's claims experience in that domain. However, the authors find considerably less predictive power between one's insurance choices and the riskiness of one's 401(k) asset allocations, suggesting that the common element of an individual's preferences may be stronger among domains that are closer in context. They also find that the relationship between insurance and investment choices appears considerably larger for employees who may be associated with better financial sophistication. Overall, they view these findings as largely consistent with an important domain-general component of risk preferences.