The devastating earthquakes in Haiti and Chile have pushed the potential for catastrophe to the forefront of everyone's mind. And for people in catastrophe-prone areas—think California and Japan for earthquakes, Texas for flash floods, the Gulf Coast for hurricanes—it brings another thought: It can happen where they live at any time.

Conversely, residents of other areas—those in large swaths of the middle United States—might think they are impervious to the ground-rattling effects of an earthquake and focus efforts on building businesses there to mitigate risk. And they'd be wrong.

Unbeknownst to many, Illinois, Missouri, Tennessee and Arkansas lie directly above the New Madrid Fault, which in the early 1800s created the four strongest earthquakes in U.S. history. In fact, one (centered over what would eventually become Memphis, Tenn.) registered as a 9.0 quake—which would have meant total devastation. During the quake, buildings were rattled as far away as Washington, D.C.

Since two centuries have passed, memory of the calamities has faded. But the danger is still there. So how can a company plan locations, mitigate risk overall for existing locations, and get a sense of what may come down the pike, adjusting insurance accordingly?

For that, you need Monte Carlo and some attractive models.

Calculated risks

The combination isn't the basic ingredient of a James Bond thriller. Instead, it is part of a sophisticated process to determine an insured entity's risk of experiencing a catastrophic loss. Catastrophe modeling, or cat modeling for short, can be used to determine the overall catastrophe risk at an organization's locations and predict the losses that would be sustained if a specific catastrophe hit—thus helping companies decide how much insurance coverage to obtain.

The process uses computer-assisted calculations, impacted by a diverse set of data points. For example, historical data on natural disasters—such as hurricanes, wildfires and floods—are fed into the calculations for the models, along with seismology and meteorology reports, and information on national wind patterns and the structures potentially affected by catastrophes. The amounts of insurance a company currently maintains or is considering are included as well. Some models even take into consideration the soil content of the ground a structure sits on.

Then, the computation begins. Monte Carlo simulations—based on the Monte Carlo method of using repeated random samplings to develop a result—are run to create as many as 10,000 simulations, helping to determine a company's risk of exposure to a natural disaster. "It looks at the impact from past events in order to predict the future based on a detailed supply of input data," says Mike Andler, managing director of Aon Property.

The resulting catastrophe model then helps companies plan their insurance needs—and plan for the future. "Number one, you get a general sense of what their insurance retention, or deductible, would be," Andler says. "Second, [the model] measures the loss excess of the deductible—what you foresee the largest loss to be, based on the return period. By understanding this metric, clients can make a more informed insurance-limits buying decision to protect against that risk."

According to Andler, catastrophe models can also help address compliance in the sections of the Sarbanes-Oxley Act of 2002 focused on "Enterprise Risk." Providing better metrics around catastrophic risks allows companies to better describe how they have minimized risk.

PROACTIVE RISK MANAGEMENT

Another benefit of catastrophe modeling lies in preemption—gathering more and better data than the underwriters of insurance, so a company can more effectively negotiate premiums and deductibles. "Since 2001, there's been a lot of emphasis on catastrophe exposure because of the cost of hurricane losses," says Brian Innes, executive vice president of Aon Global. "As a result, our more proactive clients invest in cat modeling to identify the [potential] exposure to the underwriters before the underwriters do—a preemption, if you will, to get a better price to transfer the risk.