Falling interest rates are rendering traditional rule-of-thumb obsolete when evaluating profitability in liability insurance. Performance benchmarks such as the loss ratio need to be adjusted for the fact that fixed income investments return much less than they used to.
Appearances can be deceiving.
At first glance, liability results seemed quite strong overall for the 2012 accident year. The direct loss and loss expense ratio for the industry was 66% in Other Liability, 85% in Medical Malpractice, 1 and 73% in Commercial Auto. As seen in the figures, all ratios are comfortably below long-term averages, especially so in Medical Malpractice and Other Liability. Indeed, since 1990, lower ratios have been recorded only in the aftermath of the most recent hard market—between 2003 and 2007 for Other Liability and 2004-2007 for Medical Malpractice.
Granted, these are first reports, and there are some worrying signals—especially in Workers Compensation and Commercial Auto—that industry reserving is getting a bit aggressive. But, nitpicking aside, things seemed pretty good, and the trade press has been crowing about solid underwriting results.
But there’s a problem. The traditional loss ratio is no longer a reliable guide to profitability in liability lines, and the benchmarks that many of us have in mind when we evaluate performance may not be accurate. The problem lies in interest rates. They’ve been falling since the early 1980’s and reached historic lows for a number of maturities following the financial crisis in 2007-2008.
This does not matter much in property insurance lines. But in liability and casualty lines, where payments can stretch out over years and even decades, it can make a big difference. The amount you set aside today to fund a liability that is 10 years away depends crucially on the interest rate. A higher interest rate means that you will earn more on the funds you set aside today, so you can set aside less now to pay the same obligation in the future. Because of this, higher interest rate environments permit us to tolerate higher loss ratios. Conversely, in low interest rate environments, we must tighten our belts and target lower loss ratios.
How much of a difference does this make? Probably more than you realize. Using the average Schedule P payout pattern for the industry for accident years 1994-2003 and some simplifying assumptions, 2 I found that falling interest rates have had an effect equivalent to the following premium rate decreases:
|Line of Insurance||Equivalent Premium Change 1990-2013||Equivalent Premium Change 2000-2013||Equivalent Premium Change 2007-2013|
|Medical Malpractice (Occ)||-26%||-22%||-14%|
|Other Liability (Occ)||-20%||-17%||-11%|
|Product Liability (Occ)||-27%||-23%||-15%|
Thus, the industry has been facing a strong headwind in long-tailed lines over the last few decades. Falling interest rates have had the same impact as a 20% to 30% rate decrease since 1990 in occurrence-based liability lines. Even since 2007, falling interest rates have had the same impact as a double digit rate decrease for long-tailed lines.
Hence, once we account for interest rates by calculating a “present value loss ratio”—by discounting the loss ratios based on the payout patterns and spot rates described earlier—the results of 2012 no longer appear so strong. As shown below, the results are basically no better than average.
The challenge of low interest rates is familiar to life insurers and pension funds. Life insurers in Europe and Japan have for years grappled with the problem of “negative spread,” as the crediting rates on their policies exceed the rates of interest that the earn on their investments. Pension funds have also seen falling interest rates cause the fair value of their liabilities soar, while their fixed income portfolio yields have plummeted.
Liability insurers now face similar issues and a similar dilemma concerning what to do about pricing. As with life insurance and pensions, the problem is more of a chronic disease than an acute fever: If interest rates stay low for a long period of time, a gap between what a company should be charging and what it is actually charging serves to grind down the strength of the balance sheet. For example, in Japan it took about 10 years before the balance sheet erosion started to produce failures. Today’s liability underwriters must decide whether today’s interest rate environment is a long-term or a short-term phenomenon, and adjust their pricing benchmarks accordingly.
- I have removed the nontabular discount from Medical Malpractice. ↩
- I assumed that any remaining payout after 10 years was evenly spaced over the next 5 years. (Ideally, one would of course push this out over a longer time horizon.) Payments were assumed to occur in the middle of each year. For interest rates, we used the Treasury curve for the first trading day of each calendar year to estimate spot rates, using linear interpolation for yields at maturities that were unavailable. ↩