Is the Economy Driving the Sub-Standard Auto Market in Texas’

March 24, 2003 by

Have we finally arrived at the point where the economy is a critical factor in personal automobile insurance? Not quite, but it is my opinion that we are closing in. Although insurance company annual statements do not discern between the various tiers of personal automobile insurance, i.e., superior, preferred, standard, sub-standard, etc., my hunch is that the economy is beginning to influence the sub-standard automobile insurance marketplace. If my theory is correct, then policyholders in Texas may be adversely impacted.

In Texas, over the period from 1998 through 2001, the top three writers in the state—State Farm Mutual, Mid-Century Ins. Co of TX and Allstate Indemnity—wrote 37 percent of the personal automobile insurance in Texas. There was virtually no change in their aggregate market share over that period. The top 25 writers saw their market share grow from 83 percent in 1998 to 85 percent in 2001. Unfortunately over the same time period, the top three watched their calendar year loss ratios increase from 60 percent to 79 percent. Although year-end 2002 data will show premium increases were put in place, the fundamental question is whether or not loss ratios will decline.

Developed by Demotech, Inc., Columbus, Ohio from Thomson Financial Insurance Solutions data.
Inquiries or requests for additional information should be directed to Barry J. Koestler II, Senior Consultant, at (614) 761-8602, ext. 27.

Premiums & Losses in the Texas Private Passenger Auto Market
Private Passenger Auto Liability and Physical Damage (in ‘000s)
Rank Company
2000 Direct Premiums Written
2000 Direct Premiums Earned
2000
Direct Losses Incurred
1 State Farm Mutual Auto Ins. Co.
1,829,089
1,947,803
1,497,902
2 Mid-Century Insurance Co of Texas
721,139
715,895
561,181
3 Allstate Indemnity Company
1,372,759
542,126
360,416
4 Progressive County Mutual Ins Co
563,983
491,621
340,126
5 Allstate Property & Casualty Ins Co
474,751
411,142
322,608
6 Allstate County Mutual Insurance Co
424,251
311,465
224,318
7 State & County Mutual Fire Ins Co
302,758
288,944
251,333
8 United Services Auto Assoc
272,735
260,671
200,061
9 Farmers Texas Cnty Mutual Ins Co
264,643
91,736
57,730
10 Geico General Insurance Co
106,660
225,702
198,196
11 Southern Farm Bureau Cas Ins Co
232,554
202,546
157,893
12 USAA Casualty Insurance Co
203,841
193,837
136,618
13 Home State County Mutual Ins Co
200,069
213,066
150,096
14 Nationwide Mutual Insurance Co
214,306
175,561
104,564
15 Old American Cty Mutual Fire Ins Co
181,338
139,514
246,518
16 Texas Farmers Insurance Co
145,568
257,282
73,557
17 Colonial Cnty Mutual Ins Co
248,033
93,518
70,903
18 Allstate Insurance Company
102,526
110,231
36,576
19 Southern County Mutual Insurance Co
78,433
66,287
37,884
20 Charter County Mutual Insurance Co
63,284
52,370
39,603
21 USAA County Mutual Insurance Co
53,847
52,729
56,107
22 Charter Oak Fire Insurance Co
70,186
69,204
48,073
23 Government Employees Insurance Co
58,330
54,997
37,065
24 Dairyland County Mut Ins Co of TX
57,823
55,671
99,530
25 Consumers Cnty Mutual Ins Co
125,728
135,301
99,530
TOP 25 TEXAS Private Passenger Auto Liability and Physical Damage Combined 7,103,642
7,159,219
5,451,478
ALL OTHER TEXAS Private Passenger Auto Liability and Physical Damage Combined
1,310,502
1,283,149
925,792
TOTAL TEXAS Private Passenger Auto Liability and Physical Damage Combined
8,414,144
8,442,368
6,377,270
* Auto Physical Damage is not separated into Personal and Commercial in the NAIC quarterly statement.
* Rankings based upon Texas Private Passenger Auto Liability and Physical Damage Combined Direct Written Premium for 2001.
* Data from Exhibit of Premiums and Losses (page 24 in the annual statement) represents calendar year experience for P&C insurers reporting data to the National Association of Insurance Commissioners.
* Quarterly data from Part 1 and Part 2 represents year to date calendar year experience for P&C insurers reporting data to the National Association of Insurance Commissioners.

Developed by Demotech, Inc., Columbus, Ohio from Thomson Financial Insurance Solutions data.
Inquiries or requests for additional information should be directed to Barry J. Koestler II, Senior Consultant, at (614) 761-8602, ext. 27.

In an effort to combat the rising tide of losses, insurers are increasingly turning to the use of credit scoring in their underwriting. According to the Federal Trade Commission’s Web site, credit scoring is a system that creditors use to help determine whether to give individuals credit. The FTC states that information about individuals and their credit experiences, such as bill-paying history, the number and type of accounts they have, late payments, collection actions, outstanding debt and the age of their accounts, is collected from one’s credit application and the credit report. A credit scoring system awards points for each factor that helps to predict how creditworthy a person is.

As insurance companies begin to phase credit scoring into their underwriting processes, drivers may be classified as sub-standard because of their credit scores. In this situation, an automobile insurance premium can be based on the score assigned by a credit scoring model.

Although at first glance the two seem unrelated, residents of states that have below average homeownership rates, and therefore may not own a home, may be adversely impacted when it comes their auto insurance rates. Homeownership is a measure of financial stability and a measure of residence stability. For decades, these have been two critical factors evaluated by personal automobile insurance company underwriters. Consider this: the United States Census Bureau recently published its annual statistics on housing vacancies and homeownership. Texas has a 63.8 percent homeownership rate. California has a 58.0 percent homeownership rate. The comparable national average is 67.9 percent. Are the credit scores of Texas and California residents being unduly impacted because residents do not own homes?

If a credit scoring model assigns a favorable weight to homeownership, then it seems reasonable to assume that the credit scores in states with lower homeownership rates would be lower than the credit scores in states with higher homeownership rates. In contrast to Texas and California, Indiana’s homeownership rate for 2002 was 74.2; Minnesota’s was even higher, at 75.4 percent.

In an uncertain economy, the economics of unemployment can hurt individuals twice. First, they suffer a loss in income because their unemployment benefits are less than their previous income. Second, when one’s credit score is lower, higher insurance premiums cut even further into a person’s buying power.

The Texas and California Departments of Insurance have been at the forefront of consumer protection. Accordingly, they are at the forefront of the analysis of credit scoring models and the application of credit scoring models to consumers of insurance. Concurrent with lower homeownership rates seems to be a relatively high cost of living or relatively high unemployment. Further, large and geographically diverse states, such as California and Texas, tend to have population clusters. Additionally, these cities may have certain sections with above average incidences of car theft and other crimes.

While it seems reasonable to require garaging of a vehicle for the vehicle to be eligible for insurance at a preferred premium; a garage might also be considered a surrogate for homeownership. When homeownership rates are lower than average, it seems to me the likelihood of having access to a garage is lower than average.

The companies that develop credit scoring models are adamant that their models and the inner workings of their models are proprietary information. Their argument is that they have invested in the development of that model and they have verified the predictive ability of their model. Faced with this dilemma, sophisticated departments of insurance must balance their consumer protection oversight with the realities of the business community. Is it possible for a regulator to assess the objectivity of a credit scoring model without breaking open the black box?

The flip side of this discussion is that credit scoring helps some consumers save money. Consumers with better than average credit scores are eligible for preferred rates. However, where does an insurance department draw the line on the evaluation of credit scoring? It seems to me that credit scoring may become the New Millennium’s equivalent of redlining.

Independent agents should know how their markets utilize credit scoring.

Most consumers will accept the correlation between their driving record and the cost of their insurance; however, when they are quoted a higher premium, the correlation of a credit scoring model versus their potential loss experience maybe too subtle to accept. If an applicant has a relatively clean driving record yet pays a standard or non-standard premium, the applicant may be critical of the process.

With the economy at a low point and insurers’ use of credit scoring on the rise, if you have access to auto markets that do not utilize credit scoring, now may be the time to become re-acquainted with their underwriting guidelines.

Joseph Petrelli is the president and founder of Demotech Inc. Since 1985 Demotech Inc. has been a financial analysis and actuarial services firm serving the needs of the property and casualty insurance industry. Demotech Inc. was the first company to have its rating process formally reviewed and accepted by Fannie Mae, Freddie Mac and HUD.