12 September 2022

Climate and inflation threats spark response from technology vendors

Risk modelling and management teams are being pushed to improve how they deal with climate and inflation risk – and software vendors are responding, as Ronan McCaughey explains

This year's InsuranceERM technology guide provides a breakdown of over 50 insurance technology vendors with details of more than 110 products. The guide also contains eight corporate statements from leading software solution vendors, describing the systems that insurers use for risk, capital and asset management.

Vendors with corporate statements in this year's InsuranceERM technology guide are:

The market for risk management and modelling solutions continues to develop in interesting ways. Technologies such as artificial intelligence (AI) are being introduced to help improve how insurers manage traditional risks, such as natural catastrophes, and perform regular tasks such as pricing, reserving and ALM.

There are also a host of emerging threats where sophisticated solutions are being desperately sought, including climate change, inflation and cyber. Start-up vendors with specialist expertise are stepping in to provide insurers with new products and services, while more established firms are turning to partnerships and collaborations to bring specialist knowledge into their business.

Risk and financial modeller Ortec Finance says integrating climate risk and broader environmental, social and governance (ESG) considerations into product development, risk management and investment decision-making, is a top priority for customers currently.

Hens SteehouwerOrtec Finance's head of research Hens Steehouwer says the firm has collaborated with Cambridge Econometrics, a specialist in ESG and economic modelling, to enable its scenarios to incorporate climate risks and opportunities associated with different global warming pathways.

Climate change also has the potential to radically change the frequency and intensity of weather-related risks. Understanding the impacts of climate risk is becoming vital for insurers globally, and is requiring a step-change in how catastrophe models perform.

Australian risk modelling company Reask, for example, is using AI to connect atmospheric hazards with the changing climate. The firm uses physics-driven catastrophe models that enable insurers to better understand how tropical cyclones and windstorms might evolve in various climate scenarios.

In August, Reask announced it was broadening its capabilities to cover flood risk, through a partnership with UK flood risk modeller Fathom. From 2023, the two companies also aim to bring new hazard data maps and catastrophe risk models to market, covering flood and wind.

Dynamic ALM

Besides the further integration of climate and ESG, Steehouwer expects to see an increasing importance and growing adoption by insurers of what can be called "dynamic ALM", compared to traditional more short-term static interpretations of ALM.

He says: "Traditionally, a lot of effort goes into assessing the current balance sheet and corresponding short-term capital requirements. These are very important of course. But another question is how to get insights on the potential future developments of the balance sheet under a wide range of scenarios."

Steehouwer says insurers are now picking up on the benefits of the dynamic ALM approach, and the required technology is modular and flexible for easier integration into existing technology infrastructures.

Inflation risk

Data gets bigger

Asked how risk modelling will likely change in future, Alun Marriott, chair of the technology group within Aon's strategy and technology division, says the volume of data is increasing exponentially and the insurance market needs more powerful models to handle it.

In his view, it is about trying to service that need with more cloud-based calculation servers.

Marriott comments: "As the insurance industry tackles inflation and climate change, coupled with the tightening in capacity, insurers need financial modelling tools to better understand the risks and opportunities they face."

Dealing with longer-lasting inflation has arguably become the top priority for insurers this year. And the insurance technology market is responding to the challenge.

Alun Marriott, chair of the technology group within Aon's strategy and technology division, says its ReMetrica and Tyche tools allow economic scenario projections to be brought into a model and applied to both sides of the balance sheet.

"Our tools will feed in economic scenarios to model the effect of inflation on asset returns, and also apply the same inflation projection to simulated claims based on their payment times, meaning long-tail classes are naturally highly impacted."

The spike in claims inflation has put a focus on whether reserves have been set appropriately. To help insurers understand reserve risk, insurance consultancy LCP has launched a new version of its analytics platform, LCP InsurSight.

LCP says the platform's new features include integrated artificial intelligence, as well as the ability for key reserving assumptions to be automatically highlighted to help focus judgements. Overall, the updated platform will enable insurers to accelerate their data analytics and reserving processes, says LCP.

Stress and scenario testing

In the post-Covid world, and following Russia's invasion of Ukraine earlier this year, the importance of forecasting scenarios has become more important than ever for insurers.

Patricia RenziPatricia Renzi, CEO of life technology solutions at Milliman, says the consultancy is seeing clients focusing on more regular stress and scenario testing, and focusing less on single stress evaluations other than for validation.

She says: "Firms are increasingly focused on scenarios with rich and multi-variate structures, which yield real insights into the business rather than a purely mathematical output – insight and foresight are key outcomes from these exercises. The narrative is equally as important as the numbers."

Ortec's Steehouwer agrees, saying in an increasingly uncertain world, there is a growing need and application of scenario-based stress testing and sensitivity analysis.

Geospatial data

One of the principal threats to insurer profitability and solvency remains natural catastrophes. Global estimated insured losses from natural catastrophes totalled $35bn in the first half of 2022, some 22% above the average of the past 10 years, according to Swiss Re.

The reinsurer has also noted insured losses from global flooding between 2011 and 2020 totalled $80bn, double the total of the previous decade.

Obtaining accurate and timely data for flood risk has been a long-standing challenge for insurers, but new technologies are being introduce to help. For example, Finland's ICEYE uses a constellation of space satellites to track and detect changes on the earth's surface.

This geospatial data can be fed to insurers in near real-time to understand how flood threats are developing. Among the users are innovative US flood insurer Neptune Flood, and French parametric insurer Descartes.

Cyber risk

Another area of rapid technological development is in cyber risk modelling, where established model vendors are competing against a growing crop of specialist providers.

Cyber risk experts have warned that ransomware attacks have become "weaponised" in recent years and said many clients are unprepared for cyber risk, therefore making them uninsurable. Over the last three years, premiums for cyber cover have soared and Russia's invasion of Ukraine sparked fears of a new wave of cyber-attacks.

So far this year, cyber claims have not escalated to a significant extent. But the threat is ever-present: in July, cyber risk analytics provider CyberCube reported there were more than 70 different cyber threat actors related to the war in Ukraine, double the number identified at the beginning of March.

Re/insurers holding large books of Eastern European business have been urged to stress test their portfolios against the threat of cyber-attacks sparked by the Russia-Ukraine war.

CyberCube also warned ransomware threat actors are increasingly targeting large cargo ships' onboard operational technology systems and connected infrastructure at critical port facilities.

William Altman, CyberCube's principal cyber security consultant, told InsuranceERM: "There is absolutely some cat risk from a couple of major shipping lines going down at the same time or maybe some vessels blocking key ports because their systems are disabled."

AI transparency

The development of artificial intelligence (AI) solutions for the insurance market has coincided with growing concerns about the impact AI could have on the financial inclusion of groups of protected classes, or vulnerable consumers.

Serhat GuvenRegulators have stepped in to provide guidance on best practice. For example, the European Insurance and Occupational Pensions Authority has set out six governance principles to guide insurers when implementing AI technology.

Serhat Guven, managing director and global proposition leader for personal lines pricing product claims and underwriting at WTW, tells InsuranceERM the availability of more data and the sophistication of machine learning algorithms have created increased concerns about transparency among insurers, consumers and regulators.

Guven says recent updates in WTW's Radar suite overcome this by introducing a "market-first" capability that makes it significantly easier to benefit from the full predictive power of traditional machine learning, without losing transparency.

Other tools for ethical AI implementation are emerging. US technology company FairPlay launched a tool to assist insurers in reducing algorithmic bias in their automated decision models and improve visibility for regulators in this area.

The California-based firm said its Input Intelligence product will help insurers to review their data for bias, ensuring they are not proxies for protected characteristics, such as race or gender.

A spokesperson for FairPlay tells InsuranceERM: "The insurance industry has many opportunities throughout the product cycle to use algorithms, such as marketing, pricing/underwriting, claims and fraud detection. At each of these points a consumer could be negatively impacted by an unfair algorithm, or a carrier might be in breach of a statute or regulation related to illegal discrimination."