Determining NU Profit Leaders & Champs

In some cases, combined filers do not include every company we may recognize as a member of an insurance group. For example, Zurich Group consists of two combined filers–Farmers Insurance Group (ranked 148) and Zurich Insurance Company Group (ranked 191).

o Must have reported non-negative combined ratios on its regulatory filings in each of the six years from 2001-2006 to be a candidate for the "Profit Leaders" list (shown on page 14). This requirement eliminated 43 of 264 individual insurers or combined filers that did not report combined ratios for each of the last six full years.

Among the insurers eliminated by this requirement were the U.S. operations of "Class of 2001″ Bermuda companies such as Arch Capital, which recorded an average combined ratio of 87.4 over the last five years.

Also eliminated were two four-year-old specialty insurers, James River Insurance Group in Richmond, Va., which had a four-year average combined ratio of 87.5, and Birmingham, Ala.-based nonstandard auto insurer Infinity Property & Casualty, with a four-year average combined ratio of 91.0.

In an effort to make our listing as complete as possible, NU editors reviewed data available for each of the 43 organizations that did not make it past this hurdle. In cases where a combined filer had only one combined ratio missing in the six years, and where it was possible for NU to estimate the missing figure by adding together available results for individual members of the combined filer, the estimate was used to put the organization back into the ranking.

This was possible for three combined filers–Amerisure Mutual Insurance, Medical Liability Mutual Group and United Automobile Insurance Group.

To qualify for the "Profit Champions" list (page 13), an organization must have reported non-negative combined ratios on its regulatory filings in each of the 12 years from 1995 to 2006. Again, NU editors made an effort to estimate missing combined ratios for nearly 50 additional organizations that did not report combined ratios for each of the early years (1995 to 2000).

Estimates of missing combined ratios were possible for the following 10 groups: Balboa Casualty Group; Navigators Insurance Group; Penn American Group; Wawanesa Insurance Group; Germania Insurance Group; Arbella Insurance Group; Doctors Company Group; Promutual Group; United Automobile Insurance Group; Amerisure Mutual Insurance.

o Must have reported net written premiums greater than $100 million in the latest full year–2006.

With this requirement, we attempted to eliminate insurers with data that is not credible enough to reliably produce combined ratios indicative of their actual performance. When premiums levels are extremely small, minimal losses or sizable loss levels in a single year can skew six-year average ratios.

Consistent with our prior analyses, we applied this size requirement to the most recent year only. Although we considered applying this rule to each of the six years analyzed to generate even more credible results, we ultimately rejected this approach after a review of the organizations that would have been eliminated by this more stringent size requirement revealed that many profitable small insurers with stable combined ratios would have been cut.

The inclusion of organizations with less than $100 million in net premiums for prior years, however, penalizes companies in start-up mode early in the six-year period, like Seattle-based specialty workers' compensation insurer Seabright Insurance.

Seabright, with written premiums over $100 million for each of the last three years (2004, 2005 and 2006) and an average combined ratio of 83.4 for those same three years, wrote only $8 million, on average, in each of the three prior years and reported an average combined ratio of 159.7.

To eliminate such distortions, we reviewed the volatility of combined ratios over the six-year timeframe for seven insurance organizations whose premiums jumped by more than a factor of 10 over the period–a time when industry premiums rose just 37 percent, or 6.5 percent per year on average. Based on this analysis, we judgmentally removed five of the seven from our analysis.

Along with Seabright, with this fine-tuning of our criteria we also booted:

o FFG Insurance Company, a Chicago-based affiliate of Aon that once insured the auto dealer performance of extended service agreements.

o Nau Country Insurance Company, a Ramsey, Minn.-based crop insurer.

o Universal Property & Casualty Insurance, a Fort Lauderdale, Fla.-based homeowners insurer.

o Stonebridge Casualty, an affiliate of Dutch life insurer, Aegon.

It's worth noting that Stonebridge was ranked first among our "Profit Leaders" in a similar report published late last year. With less than $10 million in net p-c premiums in 2001, Stonebridge recorded a combined ratio of 46.1, pushing the overall six-year average down to 90.2. In contrast, the average combined ratio for the most recent three years was 109.1, when premiums neared or topped the $100 million mark.

o Must not specialize in lines not traditionally considered property-casualty.

Specifically, our criteria excluded filers with 75 percent of their 2006 premiums in financial or mortgage guaranty, accident and health, surety, fidelity, credit or "other write-in" lines from analysis, dropping 17 insurance organizations which would have otherwise ranked among the top 25 profit leaders (mainly mortgage and financial guaranty insurers).

The limitation, however, was not enough to eliminate some significant writers of accident and credit lines–Courtesy Insurance or IDS Property Casualty. (See related article on page 12 for details.)

Taken together, our criteria reduced the universe of p-c organizations to 219. Among these, we have classified 116 as commercial writers (shown on page 17), which recorded 55 percent or more of their 2006 net written premiums in commercial lines, while 79 others recorded at least 55 percent of their premiums in personal lines (see page 19). The remainder had mixed books of business.

The starting point of our analysis is the "U.S. Insurance" product of Highline Data Services, an NU-affiliated, monthly CD-ROM-based product which compiles annual statement data reported to U.S. insurance regulators.

The database contains variables to allow users to immediately capture six full years of combined ratios for reporting insurance entities, leading us to choose the six-year average combined ratio as the principal measure of profit for our analysis here and for two prior reports published by NU (Jan. 24, 2005 and Dec. 4, 2006).

Recognizing that a longer-term average would provide a better picture of long-term profit trends, we have supplemented this year's analysis with combined ratio data for six prior years available from Highline Data's online product, Analyst PRO.

The "Profit Leaders" and "Laggards" tables published in this issue (and in our two prior reports) rank the entities based on their six-year average combined ratios from lowest to highest.

The newly added "Profit Champions" table ranks the entities based on their 12-year average combined ratios from lowest to highest.

Even 12-year averages, however, are not entirely sufficient to isolate insurance organizations that are consistently profitable. A company reporting six years at a combined ratio of 75 and six at a dismal 125 has a breakeven average, for example.

With that in mind, we display the latest six-year and earliest six-year average combined ratios on our "Profit Champions" chart, along with a volatility measure. These metrics can help readers separate organizations that have been profitable across the entire 12-year period from those that have suffered some relatively unprofitable years (Franklin Mutual and Amerisafe in recent years, or North Star Companies in prior years, for example).

Beyond the problems with averages, for some insurers combined ratios are a particularly poor measure of profitability.

One such statistical anomaly, as we indicated in past years, is Delaware-based Nuclear Electric Insurance. The company, as in past reports, still had the worst six-year average combined ratio among the groups we analyzed, along with the second-worst 12-year average combined ratio, but it has actually been quite profitable in some years.

Insuring nuclear utilities for the costs of interruptions, decontaminations and other risks, in fact, was such a profitable business in 2003, for example, that Nuclear Electric reported the best loss ratio among all the insurers we analyzed for that year.

Distributing high dividends from underwriting and investment earnings, however, landed the combined ratio at the bottom of the list.

Combined ratios are only one measure of profitability–and an imperfect measure of financial strength. Rating agencies, for example, review underwriting profit together with capitalization, parental support, risk management and other factors before opining on financial strength. And because of the size of the entities included in our list, many of the companies are not rated by the major rating firms.

The editors who compiled our rankings have not attempted to analyze financial statements or audit data. As such, we have accepted the data as reported, making no attempt to determine whether loss reserves are sufficient, reinsurance is recoverable, or other items impacting underwriting income are accurately reported.

We note that had we performed a similar analysis in the 1990s, Reliance Group, with combined ratios consistently below 100, would likely have been listed among the profit leaders before its eventual insolvency.