Analysts at re/insurance risk, capital and strategic advisor TigerRisk Partners have warned that methods of indexing in catastrophe models can skew perspectives on recent loss trends and lead to “inconsistent conclusions” when it comes to the impact of climate change.
This extends to whether nat cat models are adequately capturing the full range of risks they are intended to represent, a topic which TigerRisk has sought to explore further.
Authors behind the report include: Tim Edwards, EMEA, Head of Catastrophe Analytics; Anna Neely, Head of R&D and Event Response; Consultant David Clouston; and David Flandro, Head of Analytics.
These analysts concluded that methodological challenges and indexation methods can greatly influence the trends that specific catastrophe models claim to have identified, and warned against the use of “broad-brush” approaches.
Some of the main challenges include how to accurately account for the size of underlying markets, changes in insurance penetration, and terms and conditions, while also factoring in the inflation of losses due to changes in repair costs over time.
They also pointed to issues around expressing whether insurance and reinsurance deductibles have kept pace with increased exposure trends, as loss trends can be seen to skew upwards where deductibles and attachment points have remained fixed.
Additionally, they questioned whether larger events should be discarded from analysis, due to the propensity for outlier losses to shift the view of overall trends.
“There are many underlying phenomena which need to be taken into account,” the TigerRisk analysts concluded.
“A warming climate may increase the propensity for natural catastrophe losses for some contributors to the large category of ‘secondary perils.’ However, for perils like severe convective storm there is less consensus, and any reported loss trends should be treated with caution,” the report stated, adding: “An increased focus on understanding and attempting to qualify the impact of prior exposure changes is key.”
“It should be welcomed that nat cat models increasingly look to explicitly quantify all sources of risk that may correlate with peak perils in operation, such that ‘non-modelled risk’ reduces over time. Surprise losses, or ‘unknown knowns,’ will still occur, and these will continue to drive further understanding in the mechanisms impacting loss in the (re)insurance market. This particularly – and explicitly – includes the effects of climate change.”
