The lights dim on

The fair-weather profit years which the property and casualty insurance industry has enjoyed since the mid-1990s is likely to run into a storm in 1999 with companies generating single digit returns on equity (ROE).

Projections based on third-quarter StatsCan data suggests that the p&c industry achieved a modest 3% growth in premium income for 1998, while claims rose by at least the same rate with expenses rising by about 4%. The past year also saw a dramatic turn in the investment income earnings of insurers, with final returns likely to be between 10% and 12% below that for the full 12-month period of 1997.

Overall, the bottom-line earnings of companies for 1998 will be 30% to 35% lower than the previous year, the result of which will be a sharp decline in investor returns. The industry’s ROE for 1998 will probably be in the region of 8%, almost half the return fetched in 1997. This will place the industry back at 1994 earning levels as it enters into 1999, a year unlikely to see a fundamental shift in market conditions.

The lower returns generated for 1998 and 1999 will bring shareholder pressure on companies, particularly as returns will compare unfavourably with other elements of the financial services sector which is in the early stages of integration or what some are referring to as convergence.

A ROE of 8% will surely be a disappointment to shareholders hoping that the upward trend of 1995-1997 could be sustained. However, the precipitous drop in the rate of return cannot be counted upon to trigger a broadly based hardening of the market such as was experienced during the mid ’80s “liability crisis. With the capacity utilisation rate hovering around the 37% mark there is just too much supply in relation to demand as measured by economic growth.

What we will most likely see is a highly selective tightening of pricing and underwriting scrutiny in those classes of business and territories inflicting the most pain to the industry’s underwriting account. The obvious candidate is commercial property which is likely to end 1998 with a loss ratio in the range of 80% to 85%.

1999 and ahead

The outlook for 1999 is not encouraging. The prospect of improved results for 1999 is almost non-existent as the majority of companies failed to heed the warnings signs in 1998 and any selective market corrections applied now will have little or no impact for the full year. Therefore, the likelihood is that 1999 will produce a worse industry bottom-line than that of 1998. The industry’s average ROE will most likely slip to between 6% and 7% — this is a fairly optimistic view as the financial picture could worsen if bond prices suffer a big drop. So, the earliest improved returns could be expected is the year 2000, and even then the outlook is a low probability.

Amongst all of this gloom there are, of course, some bright lights. As always, there will be a significant number of insurers occupying profitable niches and those who have been successful in achieving meaningful cost reductions.

At the expense of traditional distribution methods, direct response call centre operators will continue to gain personal lines marketshare at an annual rate of 3% to 4%. Those companies who have achieved critical mass will also display better than average results. It is also apparent that consolidation will continue if for no other reason than the fact that the industry’s “generic growth” prospects will be lower due to an expected reduction of gross domestic product (GDP) growth. All in all, 1999 will test the mettle of p&c players and will complicate their adjustment to the restructuring of the financial services sector.

Reporting dates

Readers should realise that forecasting final results for 1998 with any degree of confidence is problematic because the only reliable data is six months old. The third-quarter results released by StatsCan in mid-November were based on returns from companies representing only 43.8% of the industry.

As such, the figures reported by StatsCan are subject to significant revision. Furthermore, the StatsCan preliminary figures for the full year of 1998 — which should be released by the end of February — will have a lower level of credibility as traditionally the last quarter draws a significantly reduced number of respondents. For instance, the early results released for 1997 were based on only 20% of the industry.

The poor company response to StatsCan’s yearend filing deadline is puzzling given the speed with which today’s information systems can spit out data. Considering the importance of having reliable industry-wide data for comparative purposes, there really is no excuse for the industry’s tardy reporting.