April 8th, 2009

Manage the Cycle, Part III: A History Lesson

Posted at 1:00 AM ET

venter_gary_thumbGary Venter, Adjunct Professor, Statistics, Columbia University

Like reinsurance rates and average elapsed times between catastrophes, the insurance and reinsurance industry’s average profitability shows false signs of cyclicality. The temporal distances between peaks (and valleys) are fairly regular. It’s easy to see the pattern and expect recovery or prepare for a drop. If this were true, though, action would be of little value. A cycle of generally low profitability recurring on a 10-year basis, for example, would imply that carriers could do little but brace themselves for an imminent drop — and later wait for the rising tide of a general market recovery. Of course, this is not how insurers and reinsurers operate.

There are events that occur which erode profits and deplete capital, but they are unexpected. The external factors that influence the market also do not follow cycles. Each industry-shaping event has had different causes, warranting its own response. Some insurers and reinsurers fare better than others, because of underwriting discipline, capital structures, operational efficiency, and other controllable measures.

The profitability of the U.S. property and casualty (P&C) industry reflects the outside drivers that shape the peaks, valleys, and spaces in between. Macroeconomic conditions, capital availability, and information lags resulted in peak profitability in 1977, 1987, 1997, and 2006. Almost every 10 years, the market hits a high point. The lows are fairly regular as well, coming in 1975, 1984, 1992, and 2001. This regularity suggests cyclicality, but a look at the specific drivers for each of these cases reinforces the notion that outside factors — not elapsed time — lead to favorable or unfavorable market conditions.


Each of the low points had a fundamentally different cause. In 1975, the industry’s 2.4 percent average profitability was caused by auto and workers compensation insurance losses, as well as high inflation related to oil prices and struggling equity markets. Further, the industry’s capital dropped 25 percent. Another drop in industry capital occurred in 1984, alongside a liability crisis. Profitability, on average, reached only 1.8 percent.

The next “valley,” 1992, was fundamentally different. There was no drop in industry capital following Hurricane Andrew, but profitability fell to 4.5 percent. Due to a pricing response, profits exceeded 10 percent the following year but over time did not rise as dramatically as in earlier recoveries.

In 2001, the U.S. P&C industry experienced an actual bottom-line unprofitable year, as a result of the stock market’s collapse in the wake of the “dotcom” boom. Industry capital dropped two years in a row. The recovery peaked in 2006, with both underwriting and investment showing good returns, the latter largely as a result of the credit-fueled market conditions that culminated in the subprime mortgage crisis.

Following almost every major event since 1975, years of low profitability were followed by fairly rapid recoveries. Uniformly, strong market recoveries occurred only after drops in industry capital levels. Recapitalization post-event and broader market and macroeconomic factors sometimes drove profitability higher, and quieter years led to gradual declines, as carriers wound up with excess capital.

The insurance and reinsurance market exhibits predictable behavior given certain outside factors. When market conditions or substantial catastrophe losses occur, profitability drops sharply. The capital shortage pushes profits higher quickly, followed by an easy post-event recapitalization driven by the hard markets (historically within two to three years but more rapid recently). Between events, the accumulation of capital and a decline in pricing in both primary and reinsurance markets based on loss history leads to an abundance of cash on hand. As carriers look for ways to make that capital productive, profitability declines.

Ostensibly, history can be deceiving. But, when one looks behind the pattern, coincidence is salient. Drivers differ, even if the results appear to be similar. This is borne out by the periods between the highs and lows. Post-event recoveries were uniform in neither size nor time. Every event is different … as are the consequences.

Gary Venter was previously a Managing Director in Guy Carpenter’s Instrat® Unit

This is the third in a six-part series. To have the next installment delivered directly to your inbox, register for e-mail updates.

AddThis Feed Button
Bookmark and Share

Related Posts