Progressive 2012 Annual Report Download - page 73

V. CRITICAL ACCOUNTING POLICIES
Progressive is required to make certain estimates and assumptions when preparing its financial statements and
accompanying notes in conformity with GAAP. Actual results could differ from those estimates in a variety of areas. The two
areas that we view as most critical with respect to the application of estimates and assumptions are the establishment of our
loss reserves and the method of determining impairments in our investment portfolio.
A. Loss and LAE Reserves
Loss and loss adjustment expense (LAE) reserves represent our best estimate of our ultimate liability for losses and LAE
relating to events that occurred prior to the end of any given accounting period but have not yet been paid. At December 31,
2012, we had $7.0 billion of net loss and LAE reserves, which included $5.5 billion of case reserves and $1.5 billion of
incurred but not recorded (IBNR) reserves.
Progressive’s actuarial staff reviews over 400 subsets of business data, which are at a combined state, product, and line
coverage level (the “products”), to calculate the needed loss and LAE reserves. We begin our review of a set of data by
producing multiple estimates of needed reserves, using both paid and incurred data, to determine if a reserve change is
required. In the event of a wide variation among results generated by the different projections, our actuarial group will
further analyze the data using additional techniques. Each review develops a point estimate for a relatively small subset of
the business, which allows us to establish meaningful reserve levels for that subset. In addition, the actuarial staff
completes separate projections of needed case and IBNR reserves.
We do not review loss reserves on a macro level and, therefore, do not derive a companywide range of reserves to
compare to a standard deviation. Instead, we review a large majority of our reserves by product/state combination on a
quarterly time frame, with the remaining reserves generally reviewed on a semiannual basis. A change in our scheduled
reviews of a particular subset of the business depends on the size of the subset or emerging issues relating to the product
or state. By reviewing the reserves at such a detailed level, we have the ability to identify and measure variances in the
trends by state, product, and line coverage that otherwise would not be seen on a consolidated basis. Our intricate process
of reviewing the aforementioned subsets makes compiling a companywide roll up to generate a range of needed loss
reserves not meaningful.
In analyzing the ultimate accident year loss experience, our actuarial staff reviews in detail, at the subset level, frequency
(number of losses per earned car year), severity (dollars of loss per each claim), and average premium (dollars of premium
per earned car year). The loss ratio, a primary measure of loss experience, is equal to the product of frequency times
severity divided by the average premium. The average premium for personal and commercial auto businesses is not
estimated. The actual frequency experienced will vary depending on the change in mix of class of drivers insured by
Progressive, but the frequency projections for these lines of business is generally stable in the short term, because a large
majority of the parties involved in an accident report their claims within a short time period after the occurrence. The severity
experienced by Progressive is much more difficult to estimate, especially for injury claims, since severity is affected by
changes in underlying costs, such as medical costs, jury verdicts, and regulatory changes. In addition, severity will vary
relative to the change in our mix of business by limit.
Assumptions regarding needed reserve levels made by the actuarial staff take into consideration influences on available
historical data that reduce the predictiveness of our projected future loss costs. Internal considerations that are process-
related, which generally result from changes in our claims organization’s activities, include claim closure rates, the number
of claims that are closed without payment, and the level of the claims representatives’ estimates of the needed case reserve
for each claim. These changes and their effect on the historical data are studied at the state level versus on a larger, less
indicative, countrywide basis.
App.-A-73