The second part of the Life Climate Change Working Party’s swan blog series explores the financial and non-life risks of climate change.
In this blog series we discuss climate related risks on the spectrum of well established, known, quantifiable risks (‘white swans’) to the extreme, unknowable risks (‘black swans’). In between we find risks that are highly probable, high impact in potential, but are often ignored. These are the ‘gray rhinos’.
In the second part of our swan blog series, we take a detour to look at financial and non-life risks of climate change. The Life Climate Change Working Party focuses mainly on life underwriting risks of climate change. But it is useful to first consider risks that have already (partly) materialised. They can be quantified, measured, reported on using models and managed by quantitative limits.
Financial markets, whether efficient or inefficient, will ensure that financial risks of climate change are priced in the current market value of assets. As long-term impacts are highly uncertain, this will inevitably result in higher volatility, fluctuating asset values and abrupt changes in liquidity.
But there will at least be a best estimate value available to include on the balance sheet and be used as a starting point for projections. To ensure that projections into the future are not solely based on past trends, climate models provide a plethora of output on shocked financial variables that could be readily used in financial models.
These will then calculate capital levels (available and required) on different time horizons, risk metrics and confidence intervals and disclose results accordingly to multiple stakeholders (internal, regulator, rating agency, investors).
The impact of climate change on financial risks can therefore largely be described as white swans. That is because the risk is identified, mostly quantified, actively managed and reported on.
While the financial risks of climate change can often be quantified through asset prices, the same cannot be said for non-life underwriting risks. These risks are already manifesting in rising insured losses from floods, wildfires, storms and other extreme events. Their frequency and severity have increased to such an extent that they can no longer be viewed as unpredictable black swans. They can now be seen as recurring white swans or even looming gray rhinos.
Insurers have responded by recalibrating catastrophe models, revising zoning maps and tightening underwriting terms. But the magnitude and interdependence of these perils remain challenging. Secondary perils – such as convective storms or flash floods – are becoming primary drivers of loss, while the compounding effects of climate, inflation and social factors create new layers of uncertainty.
From a risk-management perspective, these non-life risks are identifiable, modelled, and priced, yet still subject to deep uncertainty and structural change. Reinsurance markets are increasingly reflecting this through higher costs and reduced capacity. This signals a shift from insurable to potentially uninsurable territory in certain regions. In this sense, climate-related non-life risks sit midway on the black swan – gray rhino spectrum: well recognised but accelerating faster than models or balance sheets can adapt.
By contrast, life underwriting risks remain several steps behind on this evolutionary curve. While non-life risks can at least be modelled and capitalised for, the pathways through which climate change affects mortality, morbidity and policyholder behaviour are far less quantifiable. The next section explores these challenges and asks how life actuaries can begin to bring their own grey rhinos into focus.
We argue that life underwriting risks, such as mortality, morbidity and policyholder behaviour risks, stemming from climate change, do not have any of the ‘desirable’ characteristics above. There is certainly no consensus or a best estimate that could be included on the balance sheet. Although some would argue that impacts of realised climate change are, by definition, included in published mortality tables (to the extent that mortality data is not very outdated).
Mortality tables sometimes include a temporary allowance for Covid impacts, but we are not aware of any tables that would include impacts of climate change. The same holds for SCR calculations or the ORSA where shocks to underwriting risks typically do not consider climate change impacts. This is also somewhat expected as climate models provide output as to factors like temperature levels, wind speed and storm surges, and not mortality, morbidity or lapse rates and are therefore not readily usable in liability models.
In the next parts of the series, we will address each of the life underwriting risks in turn.
