Verisk develops predictive model for civil unrest
Reprints
Verisk Inc. on Wednesday said its risk intelligence business, Verisk Maplecroft, has developed a predictive model for the strike, riots and civil commotion peril.
The model provides 12-month forecasts for 50,000 counties and districts globally on the risk of severe protests occurring that could result in insured losses. It was developed for political violence underwriters, exposure analysts, modelers and specialty reinsurers as part of the Lloyd’s Lab accelerator program, a Verisk statement said.
The machine learning model validates its predictions against actual insured losses and uses geospatial data covering the size of recent protests, concentrations of economic value, demographics, and a range of political risk, climate, and other indicators.
Access to the SRCC Predictive Model and its underlying data is available via Verisk platforms and through API.
Sam Haynes, head of risk analytics for Verisk Maplecroft, said in the statement that insured losses linked to major bouts of unrest have reached new highs in recent years, and data shows that in the last 12 months SRCC risks have risen in over 50% of countries.
Historical findings from the model shows that SRCC risks have generally worsened in all regions over the past two years except in the Middle East and North Africa and suggests that the trend for costly, major civil unrest events across the world will likely continue, the statement said.
