Why Lloyd’s Underwriters Need to Consider Casualty Disaster ScenariosBy Charles Tinworth | August 11, 2020
Lloyd’s of London (Lloyd’s) has a long history dating back to Edward Lloyd’s coffee shop in 1688. Lloyd’s is unique in the insurance industry, but this singularity does not come without challenges. Toward the end of the most recent century, three events created significant aggregations of property losses across the market: the explosion of the North Sea oil drilling platform Piper Alpha (1988), Hurricane Andrew (1992), and the Northridge Earthquake (1994). These events fundamentally changed the future of insurance. Lloyd’s responded by introducing Realistic Disaster Scenarios (RDSs) to the market in the mid-1990s.
The basic principle behind the RDSs is that they represent plausible high-loss events of low probability. An RDS acts as a mechanism to stress test a scenario against a (re)insurer’s book of business and helps Lloyd’s syndicates identify and measure aggregations of insurance risk.
Today, Lloyd’s has a sophisticated approach to property RDS returns across the market. Indeed, underwriters and catastrophe modelers will be familiar with the bi-annual deadlines to submit the latest RDS returns. AIR has been pivotal in helping syndicates complete their property RDS returns accurately and efficiently.
Casualty catastrophe modeling & RDSs
Given that the concept of calculating RDS returns for property exposure has been around since the mid-1990s, why hasn’t the market streamlined RDSs for casualty risk?
Although Piper Alpha, Andrew, and Northridge caused great financial pain to Lloyd’s, these events never threatened the very existence of Lloyd’s. Throughout the 1980s and 1990s the market was dealing with a greater challenge—the gradual emergence of a huge accumulation of exposure across the market arising from asbestos, pollution, and health liability claims from the U.S.
Even in 2019, the casualty scenarios were not part of the compulsory RDS scenarios at Lloyd’s. Perhaps this is due to the availability of casualty catastrophe models? It is common knowledge that casualty catastrophe modeling has lagged behind property catastrophe modeling, however, AIR’s Casualty solution Arium provides syndicates and Lloyd’s with a solution.
Ongoing trends, including social inflation, have resulted in industrywide deteriorations in casualty underwriting results, putting casualty exposure management at the forefront of (re)insurers’ minds in recent times. Indeed, industry heavyweight Stephen Catlin commented in February 2020 that casualty business has been “significantly underpriced” by insurers for the past 10 years and commented that the industry could “benefit from some fresh thinking” around casualty insurance business.
With increased regulatory scrutiny on casualty lines of business, as well as leaps forward in casualty modeling sophistication made by companies such as AIR, now is a great time to revisit the topic of building out a more well-defined set of highly systemic events that could generate significant losses to the Lloyd’s market.
As with property scenarios, the exact approach across the insurance market regarding casualty scenarios may vary, but some common anticipated challenges will include:
- Availability, consistency, and quality of casualty exposure data
- Building and validating processes that consume exposure data and apply a prudent methodology to estimate scenario metrics
- Understanding drivers of scenario results and managing your portfolio to optimize future results
Fortunately, Arium licensees can leverage Arium’s capabilities to enhance their casualty exposure data and put it into a consistent format for all lines of business, thus saving catastrophe modelers time. Arium also enables them to run casualty scenarios and to calculate both gross and net loss with the click of a button, saving yet more valuable time. Furthermore, Arium licensees can also use model outputs to dive deeply into the results, resulting in a better understanding of their organization’s risk drivers for each casualty scenario.
Tagged with: ,