Profitability from 1990-1997 Continued

Padded Reserves in the early 1990s Mask Auto Insurers Profitability: Some have argued that, when viewed over a longer period, Texas private passenger automobile insurers have not been overly profitable at all. Rather, they argue the better profitability of 1996 and 1997 simply averages out the inadequate profitability of the early 1990’s. Table 4, however, shows that Texas auto insurers masked their profitability in the early 1990’s by dramatically overstating the reserves included with incurred losses.

Table 4 shows pure loss ratios for Texas private passenger automobile insurance from 1990 through 1997. "Paid Loss to Written Premium" relates dollars paid out by insurers in a calendar year to premium on policies written in that calendar year. "Incurred Loss to Earned Premium" relates paid losses plus changes in loss reserves for anticipated payments to premium earned on policies written in that calendar year. Incurred losses are considered a better indicator of losses associated with policies in a particular year and are used in developing automobile insurance rates. These loss ratios are termed "pure" because they relate losses to premiums without any other adjustment and represent the percentage of the premium dollar returning to consumers in claim payments.

Several items are noteworthy in Table 4. First, the comparison of paid-to-written loss ratios with incurred-to-earned loss ratios shows that incurred-to-earned loss ratios greatly exceeded paid to written loss ratios in the early 1990s. Insurers set aside substantial reserves in the early 1990s – resulting in high incurred-to-earned loss ratios – for claims they allegedly expected to pay out in coming years. If these reserve estimates were accurate, we would expect that after a few years, the paid-to-written loss ratios would increase as the claims from, say 1990 and 1991, were paid in 1992, 1993 and 1994. The reversal of paid-to-written and incurred-to earned loss ratios never occurred. Rather, the liability and overall paid-to-written loss ratios remained remarkably and consistently low. Table 4 shows that insurers padded reserves in the early 1990s and thereby masked their profitability.

Further evidence of over-reserving for automobile insurance is provided insurance industry analysts Dowling and Partners Securities. Dowling and Partners have analyzed countrywide private passenger automobile insurance reserves for several years for their small clientele of very large investors. The 1998 Dowling and Partners analysis shows $8 billion of "redundant" automobile liability reserves at the end of 1997 – representing about 12% of total automobile liability reserves. The Dowling and Partners study shows that the amount of redundant reserves did not change from the $8 billion total at the end of 1997. For example, aggregate incurred loss plus allocated loss adjustment expense ratios initially established by insurers at 83.8% in 1991 have been restated to 75.9% by 1997. The Dowling analysis shows reductions of 10.9% and 9.8% for 1992 and 1993 accident year experience loss and loss adjustment expense ratios since inception.

Paid-to-written loss ratios are not a good indicator of automobile insurance profitability over a shorter period – one or two years. However, the long-term, consistent pattern of overall paid-to-written loss ratios under 70% in Texas shows that Texas private passenger automobile insurers achieved reasonable or better-than-reasonable profitability from 1990 through 1995. After consideration of excess reserves, Texas auto insurers were profitable even before the windfall profits of 1996 and 1997.

Finally, it should be noted that the rate setting process starts with historical incurred losses. To the extent that historical incurred losses were exaggerated by excess reserves, rate indications during the 1990’s were overstated.

Early 1998 results show continued reduction in liability losses: Data from the "Fast Track Monitoring System" show that Texas private passenger automobile bodily injury losses continue to decline in 1998. For the first quarter 1998, the average paid loss per exposure (or pure premium per earned car year, in actuarial parlance) declined to about $111 from about $123 in 1997 – a reduction of almost 10%. The first quarter 1998 figure stands fully $65 less than the $176 pure premium in the fourth quarter 1994 – a reduction of over 35% in three years.

The combination of current Fast Track data with the Table 2 calculations of excess premiums show that, even after rate decreases in 1998, Texas automobile insurance rates remain significantly excessive. Table 2 shows that 1997 premiums were 14.5% excessive. With declining losses and assuming overall rate reductions of 5% to 7% in 1998 to date, Texas automobile insurance rates remain at least excessive in the neighborhood of 10%.