Intro

The insurance sector makes money by taking calculated gambles on risk of policies having to pay out. Insurance policies can be made for almost any scenario as long as the risk can be accurately modelled. Competitive advantage in the insurance industry is decided by who can most effectively model risk and therefore price their policies. Overestimating risk leads to a higher premium and a loss of business as policy seekers choose competitors. Underestimating risk leads to potential ruin as claims may stack up in the event that an extreme event occurs.

Up until recently insurers have been able to model most risks effectively by looking at data from the past which has mostly correlated closely to the present and the near future. However as climate change affects how the world works it introduces new complexity into risk calculations as historical data no longer correlates well with the future due to the numerous knock on effects of a changing climate. This is leading to a widening insurance gap (the number of uninsured assets) which could have catastrophic effects in the event of a disaster as governments have to take up the slack if insurers aren’t paying out.

The insurance gap

For every dollar covered by insurance it is currently estimated that governments, businesses and communities will have to bear three-four dollars in uninsurable damages (https://kpmg.com/ca/en/home/insights/2025/09/climate-risk-modeling-for-insurers.html)

The problem manifests across two distinct categories, Physical risk and Transition risk.

Physical risk

Physical risk is the risk that a physical event will lead to a claim on an insurance policy such as a flood or wildfire. Physical risk can be either acute or chronic. Acute risks are high intensity and short duration events such as hurricanes, flash floods or hailstorms.

Chronic risks are lower intensity, drawn out events such as air quality degradation or sea level rise.

Transition risk

Transition risk refers to the financial, operational, and reputational challenges companies focus when transitioning to a lower carbon economy. Examples of this could be real estate asset managers facing high costs to retrofit buildings to a government imposed efficiency standard, or Fossil fuel companies seeing asset values decline as renewable energy sources become cheaper.

Current modelling and methodology

Traditional catastrophe modelling relies on historical data to model expected future frequency of disasters. In response to climate risk there has started to be a rise in climate risk modelling which uses scenario analysis such as what happens if we hit a 2.0C rise in global temperatures. Climate risk modelling uses physical models and simulations based on varying scenarios to work out the risks of policy claims by policy holders.

Climate risk modelling can either be done using representative concentration pathways RCPs which describe the scenarios faced given a certain greenhouse gas concentration trajectory, or they use shared socioeconomic pathways which model based on different future global developments such as economic growth or expected regulatory policy changes. The latter is more useful for analysing transition risk.

Once these models have been created they are typically used by insurers to model on 3 different timeframes:

• 0-3 years for direct underwriting & pricing
• 4-8 years for reserving capital and planning based on medium term risk accumulation
• 10-20 years+ strategic long term planning for market positioning, entry and exit of different sectors, and product development.

Challenges

Despite gradually shifting towards scenario analysis the insurance sector faces critical challenges which make effective climate risk management difficult.

The limitations of traditions Cat models

Most modelling is still using old school Cat models which don’t work well with a non-stationary climate. They struggle with:

• Non-linearity: climate impacts are not gradual, they involve tipping points, and they have cascading knock on effects. Basically climate effects are far more interlinked than traditional risks. They often underestimate the risks to a portfolio as they fail to consider that if one property is hit by a hurricane then so will the adjacent property.

Data deficit

Effective climate risk modelling relies on lots of accurate and connectable data. Currently there is a lack of useful data for multiple reasons:

• Global climate models are too coarse for property level underwriting decisions
• Lack of high resolution data linking chronic climate stressors such as extreme heat to mortality rates (mainly in health insurance but could be said for sea level rise effect on property)
• Climate risk data fragmentation and divergence. Currently different vendors will use different methodologies and assumptions which leads to different risk calculations and an erosion of trust in the industry as everyone provides different answers.

Opportunities

Strategic collaboration and data sharing

No single company is able to solve the data deficit alone as there are just too many data points to cover and different scales of data.

Working to standardise climate risk data under a risk framework would go a long way towards helping companies share data and model climate risks. This is tough as it relies on working with regulators to lead the change.

Climate risk adaption modelling

Using climate models to advise businesses and consumers on how adapting assets to climate changes will benefit them and safe money on policies.

Transition risk modelling for asset managers to analyse their investment portfolios and advise them on how to invest in the net-zero transition.

The competitive landscape

This is in one column

This is in the second column

• list
• sgsd
• sdgdsg
  Hmmmm

sdgsdgsdgdgsThis is a quote

afasfafsasfasf