Study compares the accuracy of valuation methods of insurance companies

A study by Columbia Business School Professor Doron Nissim, Ernst & Young Professor of Accounting & Finance, reveals a better understanding of how investors value insurance companies. Two alternative approaches are typically used when estimating a company's equity value: fundamental valuation and relative valuation. Academic research and teaching tends to emphasize fundamental valuation models, although relative valuation models – which typically involve price multiples, or ratios used to compare a company to a group of similar companies – are much more common in practice. Unlike most prior studies, this research, forthcoming in the Review of Accounting Studies, examines the impact of industry-specific adjustments on the precision of estimated value relative to stock price.

Although book value multiples tend not to be an accurate measure when valuing non-financial firms, these multiples perform relatively well when valuing insurance companies, the study shows. In fact, over the last decade, book value multiples have performed significantly better than earnings multiples. Contributing factors include the financial nature of the majority of insurance company assets and liabilities, the relatively small size of unrecognized intangibles, and the role of capital-related regulation.

Analysts often exclude Accumulated Other Comprehensive Income (AOCI) from book value, a practice that is unique to the insurance industry. This is usually seen as a way to reduce the volatility of book value and mitigate accounting distortions. However, excluding AOCI tends to worsen, rather than improve, the accuracy of valuation, the study shows. AOCI measures unrecognized economic gains and losses that increase net invested assets, and should therefore be included in book value.

In another surprising finding, the study also showed that excluding realized investment gains and losses from earnings does not improve valuation accuracy. An exception occurred during the recent financial crisis, most likely caused by an increase in "gains trading," or the selective realization of gains.

Consistent with common industry practice, the study finds that conditioning the price-to-book ratio on return on equity significantly improves the valuation accuracy of book value multiples. In contrast, incorporating proxies for growth, earnings quality and risk does not consistently improve out-of-sample predictions.

Two methodological issues relevant for the use of price multiples are the definition of industry group and the measurement of shares. The study finds that, for insurance companies, limiting peers to the same sub-industry (as opposed to using all insurance companies) improves valuation accuracy, and that adjusting with respect to potentially dilutive shares improves earnings-based valuations but not book value-based valuations.

The sample used in the study includes information from all insurance companies available in the intersection of three databases: IBES, CRSP, and COMPUSTAT. Market-related data, such as price, stock returns, shares, and adjustment factors, were extracted from CRSP and Yahoo! Finance (for recent data).