By: 2 April 2024
Accelerating data maturity to better price climate risk

Simon Axon, financial services industry director, international at Teradata.

Weather and climate disasters, such as the horrific firestorm in Hawaii, the extreme heatwave and drought in Europe, and Hurricane Ian in the Caribbean and US, have dramatically reinforced the impacts of climate change. Such events are only becoming more frequent and severe, leading to a significant increase in the number and value of insurance claims.

In fact, research from Capgemini and Qorus revealed that insured losses from natural catastrophes have increased 250 percent in the last 30 years.

McKinsey predicted that the value at stake from climate-induced hazards could, conservatively, increase from about 2 percent of global GDP to more than 4 percent by 2050. The risks include not only acute events like wildfires, floods, and storms, but also more chronic issues such as droughts and rising mean temperatures, which directly impact food production, productivity, operating costs, and more. For example, figures from the Association of British Insurers (ABI) revealed that £219 million was paid out in the UK just for subsidence claims related to the record-breaking heatwave in 2022.


A crisis of confidence in prediction models

It is not only the balance sheets of insurers that are threatened by the increasing cost and frequency of climate-related claims. It is the sector’s entire business model that is at risk. There is a growing realisation among experts that climate volatility is undermining the foundations of an industry that relies on historical data to predict risk. Eric Andersen, the president of Aon, the world’s largest insurance business, told a US Senate committee that climate change is creating uncertainty in the industry and leading to “a crisis of confidence around the ability to predict loss.”

It is not surprising that industry insiders are worried. Traditionally, insurers have used climate-related risk models that predict the likelihood of future floods and fires, for example, based on historical frequency. This method has always been imperfect, but with climate patterns looking less and less like those of the past, is nearly useless. It is no longer enough to update tables with new intervals. Understanding rapidly changing and interrelated climate events requires more sophisticated models.


A new paradigm for a new reality

So the question is, how can insurers develop more complex and reliable models? The short answer is data. Insurers need more and better data, the right platforms and processes to analyse it, and the decision-making culture required to turn data into risk analysis. Currently, the insurance sector lags many others, including retail, consumer goods, technology, automotive, and telecom in using predictive analytics to help price risk, and this needs to change.

The good news is that there is no shortage of high-quality data. From granular data from IoT sensors to satellite imagery and sea-surface temperature readings, there is a wealth of real-time and longitudinal datasets that can inform more precise risk pricing. But these data streams need to be integrated and understood alongside many other inputs. The models used must also pivot from static analysis based on historic averages to dynamic scenario planning and predictive models fed with the most recent data available.


It takes an enterprise-wide, data-led strategy

To make the most of available data to drive accurate risk pricing, insurers need to develop, extend, and sustain the data foundation needed to support AI initiatives and other advanced analytics. This requires investing in an open and connected analytics and data platform that integrates data from a wide variety of sources and includes powerful AI/ML capabilities that can quickly make sense of vast datasets at scale. Potentially billions of lines of data need to be analysed and modelled to precisely understand the insurance risk linked to a single potential scenario—a task that can be quickly automated with machine learning.


A top-down approach to transformation

Becoming a truly data-driven organisation requires a leadership-led transformation. The successful insurer of the future will put data at the heart of the business and actively sustain a data culture led from the top. This will not only enable rapid, accurate risk pricing over the long term, but it will also create new lines of business and a much stronger balance sheet.

As climate change continues to impact every aspect of the insurance business, forward-thinking leaders will meet these challenges head-on and turn them into opportunities. By using AI initiatives and other advanced analytics to drive deeper understanding of risk, customers, and assets, insurers will be able to profitably insure more risks, differentiate themselves in the market, and future-proof their investments.


Image: Canva.
Emma Cockings
Emma is a content editor for Claims Media.