You can’t manage what you can’t measure. Catastrophe exposure is no exception. Insurers’ ability to price cat risk depends on knowing who and what is exposed—people, property, businesses and infrastructure—and to what perils.

This is why the world’s economic losses from natural hazards are 70% uninsured—more than 90% uninsured in many low-income countries.

The business of cat risk modeling has grown during the past few decades as insurers increasingly demanded predictability in return for underwriting risk in places like Florida. As a result, when a series of hurricanes struck the U.S. in August and September, insurers were able to tap into a staggering wealth of real-time information about possible losses under various scenarios.
Multiple times a day, as Tropical Storm Harvey and Hurricane Irma advanced toward Texas and Florida, the community of cat-risk analysts embedded in the leading insurers ran and reran detailed loss models based on enormous amounts of data about property values, weather behavior and flooding—all data that in the modern era help insurers price risk and decide whether to expose their capital to it.

But data about property value and exposures don’t exist for most of the world. Where data are scarce, so is insurance. And the same is true of effective interventions to reduce exposure and build resilience.

Underscoring this point is the recent devastation and loss of life from monsoon flooding that left more than 1,200 dead in Bangladesh, India and Nepal. By contrast, the massive flooding from Harvey, which hit Texas around the same time, cost some 60 lives. The death toll from Irma was north of 50. The reason the U.S. death toll was much lower than it was in Asia is preparedness informed by years of data and analysis that has led to better building standards and emergency response systems.

Modelers, and the data they collect and analyze, have become the critical foundation for building resilience in cat-prone regions rich and poor. Even in strong economies, the proportion of insured risk can be low, as Harvey demonstrated in Texas, where flood losses dominated. But at least in those markets, a mature insurance industry is able to fund investments in the practical application of data, science and analytics.

In low-income nations, by contrast, the work of cat modelers is largely philanthropic. The commercial incentive to invest is minimal. Lack of commercial motive isn’t the only obstacle to developing cat models. Numerous academic organizations build models but not generally of a type that has utility outside of research.

The issues of commercial incentives, transparency and standardization in cat modeling were the subject of a discussion with members of the Insurance Development Forum (IDF) at the Global Insurance Forum in London in July. On the dais were representatives of the largest professional cat modelers—Risk Management Solutions (RMS) and AIR Worldwide—and the nonprofit Oasis Loss Modeling Framework. The mission of Oasis, which gets most of its funding from the insurance industry, is to increase the availability of models by encouraging all manner of modelers to use its open source code and providing free access to nations most in need.

The notion of giving away all this valuable information might pose a threat to the for-profit sellers of proprietary models in the market. The founding organizers of the IDF—including more than a dozen global insurers and brokers and the United Nations and World Bank—recognized this. That’s why the IDF established a working group charged with finding a way to bring the ability to measure and price risk to vulnerable, underinsured markets in a way that is commercially viable for all stakeholders.

“The next big phase of activity for the [IDF’s] risk mapping and modeling group…is looking at interoperability,” Dickie Whitaker, Oasis CEO, says. “It’s fantastic…that RMS and AIR relish the opportunity to have that conversation, because if they’re not part of that conversation and they’re not part of the solution, there won’t actually be a solution.”

Or at least not an optimal solution.

Oasis is working with local academics and national and regional governments not only to build models—for example, for flood in the Philippines and for cyclone in Bangladesh—but to teach people in those countries how to do it themselves and to give them all the necessary tools for free, Whitaker said. When the Philippine and Bangladesh projects are completed, “We’re going to be able to see how we can replicate that around the world. It’s ‘modeling for the masses.’”

The for-profit firms, rather than resisting this effort, have in fact joined it. RMS, for example, has been funded by the U.K. and Germany to help the governments of the world’s 70 poorest countries close the so-called protection gap. And AIR has listed its models on the Oasis online marketplace, OasisHub.

RMS global managing director Daniel Stander cites his firm’s commitment to transparency. The company discloses its model methodologies to its clients. This doesn’t just help insurers refine their underwriting practices. It also helps insurers demonstrate to regulators and rating agencies that they truly understand their risk and are adequately capitalized.

Cat risk models do much more than help insurers manage their exposures, Stander says. “Models make markets,” he says. “Risk and capital cannot find each other unless there is a common language to describe what is being transferred. And the commercial models are that reference view for that trade, the standard around which a market can form and function.”

The markets that need to be made today, though, aren’t in Miami or Houston. They are in Bangladesh, the Philippines and scores of other at-risk, underinsured nations.

Daniel Kaniewski, AIR Worldwide vice president, says: “Simply understanding your risk would be a huge benefit regardless of when that mature insurance market comes along.”

Winans is principal of Chris Winans Consulting.