(Image source: Luminant Analytics homepage.)
Luminant Analytics (Hartford/Basel, Switzerland), an InsurTech focused on innovating insurance pricing, has released projections for commercial auto liability loss frequency up to 2020. The projections are predictions of commercial auto accidents that provide carriers a forward-looking view and serve as an external environment benchmark to recalibrate their internal projections. Luminant Analytics describes the model as having extremely high reliability, as exemplified by how it tracks compared to actual results over a 25+ year period.
Luminant articulates its value proposition as helping actuaries and product managers price, select and segment risks better using relevant, curated external data and predictive models. Founded by Renu Ann Joseph, Ph.D., an economist with over 15 years of global experience in industry and academics. Luminant aims to bring a fresh perspective to insurance pricing by blending insights from various disciplines, such as econometrics, statistics, finance, health economics and public policy to insurance pricing analytics.
Addressing Climbing Loss Ratios
The vendor chose the commercial auto liability line because it represents 15 percent of total auto insurance premiums in the U.S., with auto also being the single largest line in U.S. property/casualty insurance. “This line has been plagued by a decade of climbing loss ratios that rate increases have not been able to tackle effectively,” says a Luminant statement. “Some of the reasons for the negative performance lie in the industry’s archaic, pricing methods that are mostly inward and backward looking, which fail to take into account key external trend changes that impact loss development. Luminant Analytics’ projections are entirely powered by relevant external data.”
Luminant offers a free view to the predictions that includes the results of its back tested model up until 2017. Insurers interested in getting the latest forecasts until 2020 can access them by filling out a contact form at the company’s homepage. Interested parties can contact the company at this link for more granular versions of the predictive models and projections at the state level, by vehicle type or for loss severity.