Printed from BusinessInsurance.com

Putting big data to work in search of next asbestos

Posted On: Oct. 25, 2015 12:00 AM CST

Putting big data to work in search of next asbestos

Casualty catastrophe modeling is increasingly a focus of insurers and reinsurers as markets seek to identify and measure the next mass tort exposure.

Lloyd's of London and Los Angeles-based modeler Praedicat Inc. earlier this month released a report outlining the challenges of casualty cat analysis. Praedicat has developed models that mine data from a variety of public sources to identify and quantify emerging mass tort risks and then apply the information to insurer portfolios.

Reinsurance units of Willis Group Holdings P.L.C. and Marsh & McLennan Cos. Inc. have also developed liability cat models using large databases of historic losses.

While the insurance industry's typical response to liability disasters like asbestos has been to limit or bar coverage, successful modeling may allow insurers to underwrite more selectively and manage risk accumulation, reducing coverage bans and limitations, according to Lloyd's.

Reinsurers, including the insurance-linked securities market, may also develop products to spread accumulation risk, Lloyd's noted in its report.

While property cat modeling has been able to rely on meteorology, seismology, historical loss patterns and geo-specific risk data, modeling emerging casualty risks “is a relatively new field without an established methodology,” Lloyd's said in its Oct. 15 report, “Emerging Liability Risks: Harnessing Big Data Analytics.”

Praedicat — which, backed by Newark, California-based catastrophe modeler Risk Management Solutions Inc., released its first liability cat model in February — starts by mining millions of peer-reviewed science journal articles and other data sources to identify substances and products that may pose threats of bodily injury or environmental damage. The model monitors scientific research on possible causal links between a product and injuries “before it matures to the point that it can support litigation,” according to the Lloyd's report.

The model is also designed to “contextualize” a given risk, comparing its potential impact to other emerging and historical risks; and to use detailed information about a policyholder company's products and operations to place the company on a “map” of those exposed to liability disasters, the report explains.

Praedicat then applies the model's data to an insurer's portfolio to quantify its exposure to mass litigation. The model can also be used to analyze an industry's mass tort exposure and to measure an insurer's clash exposure across industries in a given line of coverage, according to Praedicat.

Others have developed their own models for casualty catastrophes. Marsh & McLennan's Guy Carpenter & Co. L.L.C. unveiled its ForCas model in March, focusing first on measuring insurers' exposure to “sudden disasters,” such as oil rig explosions, that can generate multibillion-dollar cleanup costs, environmental liabilities and fines. In June, it expanded the model to cover catastrophe risks of financial institutions across directors and officers liability, errors and omissions, excess casualty and fidelity lines; and cyber exposures including privacy liability, network security liability, business interruption and data asset losses.

Willis Re, meanwhile, has developed eNTAIL, which uses an analysis of more than 30 years of casualty cat events to project a probability distribution of potential future client insurer losses, according to the broker. Separately, Willis Re earlier this year announced a $400 million reinsurance facility, dubbed PRIMO, to provide ceding insurers with coverage for liability cat accumulation exposures. The facility, backed by 20 reinsurers, covers multiple accident years across all casualty and professional liability lines.

“While the world has changed since asbestos litigation emerged, the interplay between science, technology-driven innovation and risk which can drive the accumulation of exposure has not,” Lloyd's observed in its report. Lloyd's noted possible toxicity of nanomaterials and potential environmental impacts of oil and gas hydrofracking among the emerging risks.

Modeling that makes these exposures more predictable could help insurers and policyholders, sources say.

“The results from casualty catastrophe modeling can lead to better underwriting — for example, risk selection, pricing for future events, portfolio optimization (and) exposure through policy terms,” according to a 2013 analysis by Towers Watson & Co.

Modeling can also help map realistic disaster scenarios, validate capital model assumptions and test risk mitigation strategies, Towers Watson said.