The auto line showed a five-point loss ratio dropby no means the largest among the insurance lines. However, with auto premiums representing over 40 percent of overall industry premiums, the lines improvement shaved roughly two points off the overall industry combined ratio, equaling nearly all of overall improvement reported from 2003.

Higher premium levels resulted in improved nine-month loss ratios for most lines, but only the auto insurance and general liability lines also had lower dollars of losses in 2004.

While the impact of auto improvements over several years is clear, nagging questions about whats driving losses down remain a topic of speculation. Why are they falling, and can loss frequency trend lines possibly keep moving downward?

During a recent meeting with analysts, Glenn Renwick, chief executive officer of Mayfield Village, Ohio-based Progressive Insurance, discussed a graphical presentation of month-by-month accident frequencies by claims-per-vehicle. “You draw the next one in,” he said, coaxing the analysts to try their hand at predicting the frequency level that will prevail in 2005.

“We don't know where the next line is,” he said, referring to a series of lines drawn to represent accident frequencies for each year since 2000. The year 2000 was the uppermost frequency-trend line with individual points marked from left to right for each month from January to December. The line for each successive year fell below the prior one, indicating lower ratios of accident claims-per-vehicle in each of the next four years.

In this month's “Data Insights” report, NU presents its own set of line graphs. Ours, showing loss ratios, also invite readers to draw the next onein this case, the 2005 loss ratio points.

In addition to the graphswhich may also be used for benchmarking individual insurer results against the industrywe provide expert commentary on auto insurance and overall trends below.

Bottom LineGroundhogs Bullish On Profit, Not Growth:

A familiar triangle is repeated for nearly every line of insurance displayed in this editionand the all-lines graphs reflect the same pattern. Loss ratios rise to a high point in 2001, falling to their low in 2004. But with premium growth figures also falling to their lowest levels since 2002, will underwriting profits continue in 2005?

Some points to consider:

Absent hurricane losses in 2004, the nine-month combined ratio would have been roughly 94.

Thirteen analysts polled by the Insurance Information Institute in February forecast a meager 2.7 percent rise in net written premiums in 2005 and a 98.9 combined ratio, on average. Only three of the 13 predicted combined ratios over 100.

For February, a monthly barometer of average commercial price increases released by Dallas-based MarketScout.com hit zero for first time.

AutosSafe In The Garage?

With nearly $150 billion of direct premiums in the first nine months of 2004, the overall auto loss ratio fell to 57.6. Assuming levels of loss adjustment and underwriting expenses similar to 2003, the implied combined ratio is roughly 94a level nearly 10 points below that of the most favorable prior year on our graphs, 1998.

“If that was [in] a textbook, you would clearly be looking for the explanation of what drove that,” said Mr. Renwick, making a remark about a frequency graph that could easily apply to NUs loss ratio graphs also.

For personal auto, experts offer possible explanations.

The number of vehicles per household is now greater than the number of drivers per household, Mr. Renwick noted. In other words, there are more cars per registered driver, and car owners can only drive one car at a time. “Clearly, we have vehicles that are not getting the same level of exposure as they might have in the mid-1990s,” he said.

Higher gas priceswhich result in less drivingmight be another factor, according to James Eck, a rating analyst of Moodys Investors Service in New York. In a written report predicting a stable outlook for the personal lines sector, he also noted the larger proportion of newer, safer cars, adding that higher average costs per claim have been offset by continued frequency declines.

Demographics play a role, according to William Wilt, an analyst from Morgan Stanley in New York.

He noted that when accident frequencies are graphically plotted for various age groups, a U-shaped curve results indicating the highest levels of accidents for the youngest (teenaged) and oldest drivers. The “sweet spot” in the middle with the lowest accident ratesthe baby boomersare also the largest population segment.

“There is a long-term secular decline, I think, in accident frequency,” he said, noting the other factors as well. “Analysts are too quick sometimes to say, This is a string of good luck,” he added, suggesting that frequencies might not revert back to a high long-term average level as others predict.

Turning from sweet spots to hot spots, the analysts each highlighted a different area to watch with some concern.

“Its going to be interesting to see how the nonstandard auto segment plays out this year,” according to Mr. Eck. “Competition is particularly fierce,” he said, noting that capital moved into the business because recent results have been particularly good.

Mr. Wilt raised caution flags for commercial auto, which he identified as the first line to harden in 1999. Truck tonnage hauled began rising in 2002, while premium growth is falling, he saidadding that Morgan Stanley trucking analysts believe driver turnover is at its highest level, leaving less experienced truckers on the road.

For standard personal auto, however, the analysts expect underwriting profits to continue in spite of falling premiums.

“Were definitely seeing companies reduce prices,” Mr. Eck said, pointing to announcements by State Farm recently and USAA last year. However, “the major writers had a great year last year even with decent price cuts,” still coming under a 95 combined ratio, he noted.

Mr. Wilt said that while concentration of business among the top-10 writers puts pricing control in the hands of a few large competitors, the level of pricing sophistication is a “real change” from 1998 and 1999. Also, the publicly-traded auto insurers returned earnings to shareholders in 2004so capital isnt building, he noted.

Asked about potential for regulators and lawmakers to force changes in price-setting techniques as a result of recent news of a security breach for a provider of credit-related informationAlpharetta, Ga.-based ChoicePointMr. Wilt didnt rule it out. “I definitely have that on my radar screen,” he said, noting that politicians inclined to seek reasons to disallow the use of credit-scoring could potentially find one in that news.

Reproduced from National Underwriter Edition, March 25, 2005. Copyright 2005 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.