SIRC 2021: Break siloes to harness Asia cat data
November 16 2021 by Andrew Tjaardstra
The insurance industry is collecting more data than ever when modelling natural catastrophes in Asia but making the best use of it remains a challenge, according to panelists addressing the 2021 Singapore International Reinsurance Conference (SIRC) on November 16.
“The challenge is not to work in siloes. We need to link global knowledge. We need underwriters to understand data science and machine learning. The future of modelling will be more accurate models,” said Amaury Dufetel, who leads Axa’s climate parametric insurance unit.
Dufetel added that the huge volume of nat cat model data can lead to a massive transformation across the industry.
“Climate change is front and center — our models have been challenged with adapting to it and the industry is tackling this head-on through innovation,” said Oriol Gaspa, Aon’s head of analytics for Japan.
This is often done through (re)insurers capturing more data to provide greater knowledge of the different locations impacted by nat cats.
“Data is difficult to gather from different sources and difficult to gather from different countries. Floods cause staggering economic losses every year and is particularly challenging to monitor,” said Irene Garonna, a property and specialty solutions manager with Swiss Re, which has developed its own modelling tool to pinpoint flood threats at specific locations globally.
Meanwhile, satellite data company Iceye have also created a flood mapping tool.
“We want a set of flood data points … We capture data with satellite imagery and then analyse the data daily in a review process. Our flood monitoring system [involves] terrain, rainfall, secondary sources, meteorologists and hundreds of bits of data,” said Nathan Uhlenbrock, a solutions architect at Iceye.
“There is also a timing component — when will it flood. However, there will still be floods that no-one saw coming,” Uhlenbrock warned.
Data usage
With this increase in data, it is important for (re)insurers to put it to good use.
Ed Gaze, a senior manager at Lloyd’s Lab, noted the need for more processing power, better algorithms and data standardisation.
“Insurers often don’t have a dedicated innovation team which is needed to lobby internally to do new things. Being able to predict floods and warn customers to protect their belongings is the future,” he said.
Adding to this theme, Gospa said that near-term climate variability data is providing a new level of capability to the industry.
“We could use the data to anticipate supply chain disruptions and help clients build new ones to avoid business interruption losses,” Gospa said.
-
FM Global | Resilience: No longer a choice
As climate disclosure becomes mandatory and new risks emerge from natural hazards, understanding the tools that are available to build resilience is more important than ever.
