How are P&C Insurance Loss reserves computed?

Analysis of Loss reserves plays an integral part of valuation of an Insurance business. This blog post looks at basics of the Loss reserve methods prevalent in the industry. Standard actuarial reserving methods include, but are not limited to, expected loss ratio, paid and reported loss development, and Bornhuetter-Ferguson methods. A general description of these methods is provided below. Reference Ace Insurance 2014 10-K report.

The expected loss ratio for any particular origin year is selected after consideration of a number of factors, including historical loss ratios adjusted for rate changes, premium and loss trends, industry benchmarks, the results of policy level loss modeling at the time of underwriting, and/or other more subjective considerations for the product line (e.g., terms and conditions) and external environment as noted above. The expected loss ratio for a given origin year is initially established at the start of the origin year as part of the planning process. This analysis is performed in conjunction with underwriters and management. The expected loss ratio method arrives at an ultimate loss estimate by multiplying the expected ultimate loss ratio by the corresponding premium base. This method is most commonly used as the basis for the actuarial central estimate for immature origin periods on product lines where the actual paid or reported loss experience is not yet deemed sufficiently credible to serve as the principal basis for the selection of ultimate losses. The expected loss ratio for a given origin year may be modified over time if the underlying assumptions differ from the original assumptions (for example, the assessment of prior year loss ratios, loss trend, rate changes, actual claims, or other information). Selected paid and reported development patterns provide a benchmark against which the actual emerging loss experience can be monitored. Where possible, development patterns are selected based on historical loss emergence by origin year with appropriate allowance for changes in business mix, claims handling process, and/or ceded reinsurance that are likely to lead to a discernible difference between the rate of historical and future loss emergence. For product lines where the historical data is viewed to have low statistical credibility, the selected development patterns also reflect relevant industry benchmarks and/or experience from similar product lines written elsewhere within ACE. This most commonly occurs for relatively new product lines that have limited historical data or for high severity/low frequency portfolios where historical experience exhibits considerable volatility and/or lacks credibility. The paid and reported loss development methods convert the selected loss emergence pattern to a set of multiplicative factors which are then applied to actual paid or reported losses to arrive at an estimate of ultimate losses for each period. Due to their multiplicative nature, the paid and reported loss development methods will leverage differences between actual and expected loss emergence. These methods tend to be utilized for more mature origin periods and for those portfolios where the loss emergence has been relatively consistent over time. The Bornhuetter-Ferguson method is essentially a combination of the expected loss ratio method and the loss development method, where the loss development method is given more weight as the origin year matures. This approach allows a logical transition between the expected loss ratio method which is generally utilized at earlier maturities and the loss development methods which are typically utilized at later maturities. This method is applied using reported loss data although paid data may also be used. The applicability of actuarial methods will also be impacted by the attachment point of the policy or contract with the insured or ceding company. In the case of low attachment points typical of primary insurance or working layer reinsurance, the experience tends to be more frequency driven. For these product types, standard actuarial methods are generally applicable in determining loss reserve levels given sufficient history and credible loss experience (although still subject to the same limitations and uncertainties described elsewhere in this section, for example, changing inflationary or legal environments). In the case of high attachment points typical of excess insurance or excess of loss reinsurance, the experience tends to be severity driven, as only a loss of significant size will enter the layer. For these product lines, it typically takes longer for loss experience to gain credibility, which adds uncertainty to the estimates derived from standard actuarial methods. For products such as Ace’s assumed reinsurance business, they typically supplement the standard actuarial methods with an analysis of each contract's terms, original pricing information, subsequent internal and external analyses of the ongoing contracts, market exposures and history, and qualitative input from claims managers. This approach is also used for structured or unique contracts. RESERVING CONSIDERATION FOR SHORT TAIL AND LONG TAIL INSURANCE : The time period between the date of loss occurrence and the final payment date of the ensuing claim(s) is referred to as the "claim-tail." The following is a discussion of specific reserving considerations for both short-tail and long-tail product lines. In this section, we reference the nature of recent prior period development to give a high-level understanding of how these considerations translate through the reserving process into financial decisions. SHORT TAIL BUSINESSES: Short-tail business generally describes product lines for which losses are typically known and paid shortly after the loss actually occurs. This would include, for example, most property, personal accident, aviation hull, and automobile physical damage policies that we write. There are some exceptions on certain product lines or events (e.g., major hurricanes or aviation crashes) where the event has occurred, but the final settlement amount is highly uncertain (e.g., coverage disputes or liability-related claims) and not known with certainty for a potentially lengthy period. Due to the short reporting and development pattern for these product lines, the uncertainty associated with estimate of ultimate losses for any particular accident period diminishes relatively quickly as actual loss experience emerges. The company typically assigns credibility to methods that incorporate actual loss emergence, such as the paid and reported loss development and Bornhuetter-Ferguson methods, sooner than would be the case for long-tail lines at a similar stage of development for a given origin year. The reserving process for short-tail losses arising from catastrophic events typically involves an assessment by the claims department, in conjunction with underwriters and actuaries, of our exposure and estimated losses immediately following an event and then subsequent revisions of the estimated losses as our insureds provide updated actual loss information. LONG TAIL BUSINESS : For those long-tail product lines that are less claim frequency and more claim severity oriented, such as high excess professional and casualty lines, more reliance is placed upon expert legal and claims review of the specific circumstances underlying reported cases rather than loss development patterns. Where appropriate, this is supplemented with loss development and Bornhuetter-Ferguson approaches to provide for claims that have been reported but are too immature to develop individual claims estimates and also to provide for claims that have occurred but have not been reported. The assumptions used for these lines of business are updated over time to reflect new claim and legal advice judged to be of significance.