A year of heavy catastrophe losses has left Canadian insurers and regulators more cognizant of the risks presented by nonmodeled risks and earthquakes, speakers said Monday at the National Insurance Conference of Canada in Gatineau, Quebec.

Julie Dickson, superintendent of the Office of the Superintendent of Financial Institutions Canada, said during a keynote address that three major catastrophic events — the floods that swamped Calgary in June and Toronto in July, as well the July derailment of a train in Lac-Mégantic, Quebec — have reminded the industry to focus on risks.

“In terms of the (property/casualty) industry, what a difference a year makes,” Ms. Dickson said. “As the Queen said in 1992, it has been an annus horribilis.”

Noting that while factors such as changing weather patterns, aging infrastructure and increased urbanization have all been cited as reasons for mounting catastrophe losses in Canada, Ms. Dickson said insufficient catastrophe models also were part of the problem.

“The catastrophe models available for Canada focus on earthquake with limited modeling for winter storms and hail,” she said. “There are no commercially available models that cover flooding, forest fires or hurricanes.”

Moreover, during a panel discussion, Gregor Robinson, senior vice president for policy and chief economist for the Insurance Bureau of Canada, noted that his organization, which represents companies that comprise 90% of the property/casualty market in Canada, recently commissioned a study by Boston-based catastrophe modeling firm AIR Worldwide Corp. to help insurers get a sense of the risks posed by earthquakes.

%%BREAK%%

“Our board of directors launched this initiative in 2011 after the events in Haiti, Chile, New Zealand and Japan,” he said. “Each event revealed more about what we didn't know about nonmodeled risks such as tsunami and liquefaction.”

The AIR report, which Mr. Robinson said is due out in October, contains loss projections of what would happen under two hypothetical scenarios: an earthquake off the shore of British Columbia and an earthquake under the St. Lawrence River near Quebec City.

Fellow panelist Peter Gordon, professor of economics at the University of Southern California, said one of the reasons the insured losses resulting from earthquakes are difficult to project is that many of the claims result from business interruption policies.

“We know more about direct costs than we do business interruption costs,” he said. “The question is whether we can model business interruption.”

Yet the modeling of damages from earthquakes is relatively straightforward compared with the challenge of predicting them, said panelist Ross Stein, a geophysicist with the U.S. Geological Survey and chair of the scientific board of the Global Earthquake Model. While recent earthquakes have revealed new insights into how separate quakes are related, large gaps remain in basic earthquake science, Mr. Stein said.

“The depths of our ignorance is truly shocking, and it comes to us all the time,” he said, adding that scientists are beginning to better understand how quakes transfer stress down fault lines and how they spur aftershocks.

“There is progress being made even though some problems are currently beyond our understanding,” Mr. Stein said.