Bringing together stress testing and capital models – a Bayesian approach

Dan Georgescu & Manuel Sales.

Capital requirements for financial institutions are typically calculated using a statistical model and a risk measure such as VaR, whereas stress tests designed by regulators and risk managers are often based on subjective scenarios with no associated probability level. The stress test cannot therefore be easily linked to the capital measure. Taking insurance as an example, we show how to establish the link using intuitive tools which (i) respect the stress test designer’s intuition about causal direction, (ii) can be calibrated to pre-determined parameters such as correlations between risks, and (iii) can be easily communicated to and challenged by non-technical audiences.

Continue reading “Bringing together stress testing and capital models – a Bayesian approach”

Driverless Cars: Insurers Cannot be Asleep at the Wheel

Neha Jain, James O’Reilly & Nicholas Silk

In 2020 Google plans to launch a self-driving car which has already driven nearly one million miles without causing an accident; it doesn’t get tired and irritable, swerve into lamp posts or require a driving test. The in-built chauffeur comes in the form of a rotating LIDAR laser taking 1.3 million recordings per second, and it’s a better driver than you. By eliminating the element of human blunders, driverless cars are forecast to reduce motor accidents by up to 90% in the US according to McKinsey. That might imply a substantial impact on the insurance industry, with liability potentially shifting to car manufacturers. Such developments would pose challenging questions for the PRA in regulating UK insurance firms.

Continue reading “Driverless Cars: Insurers Cannot be Asleep at the Wheel”