Climate risk has caused significant challenges for life insurance financial risk modelers. There is a lot of noise in the market, with emphasis on either very precise or very complex modelling. However, it is important to also look at the broader, systematic impacts on financial markets exposures; such as yield curves, inflation, credit spreads, risk premia, and asset class returns.
How do you find the balance and a sufficient level of granularity in your modeling? How do you select and assess the impact of different climate models? Are there risks we are missing? We have created a series of short articles to help those responsible for climate modeling in the actuarial, risk, and strategic asset allocation functions of a life insurance company. These articles begin to address and simplify the complexities of climate change in processes such as the ORSA, stress testing, and strategic asset allocation.
In the fourth of the series of six articles, we explore ways to broaden your modeling approach and identify missing components in your scenarios to improve interpretation.
Expanding Beyond the Network for Greening the Financial System (NGFS) – What are we Missing?
It looks as though scenario analysis is here to stay as a preferred method for understanding climate risks. Already an important part of Task Force on Climate-related Financial Disclosures (TCFD) reporting, it is increasingly becoming an own risk and solvency assessment (ORSA) requirement. Climate scenario analysis has also formed the basis of regulatory exploratory (stress) testing, and looks likely to be incorporated into non-financial accounting disclosures.
This article discusses the fact that, as with any modeling or analytical exercise, a significant part of the work to leverage climate scenarios is identifying and recognizing the limitations and omissions, and the model risks and uncertainties. How do we broaden our scenario sets and analysis, and therefore the interpretations and recommendations we produce? To do this well, risk managers are faced with a new lexicon to master: scientific uncertainty, environmental externalities, extreme outcomes like tipping points, the possibility of green growth, and broader socio-economic impacts.
The first limitation of any scenario analysis, is to recognize that any individual scenario will usually be overly precise, and precisely wrong. Scenarios are not intended to be forecasts, and it is not possible to adequately capture expectations, risks, and uncertainties without looking across a range of scenarios. In climate modeling, one parameter in particular, the ‘climate sensitivity’ is poorly constrained by science. Given the central role of this sensitivity in driving impacts and physical outcomes, all climate work is best quantified in a probabilistic framework, which can define likelihood and uncertainties. Doing so will also align the scenario analysis with broader approaches like IPCC reports, which have uncertainty and consensus at their core.
The second limitation of any scenario analysis, is what is missing; a broad range of environmental, ecological, and social impacts, known as ’externalities’. Most economic and financial scenario work focuses on the impacts that are monetized, and shorter-term—productivity impacts, damages, and the social cost of carbon. It is important to realize that these don’t capture a number of broader representative key risks; ecosystem damage, ocean acidification, deglaciation and sea level rises, and social and political upheaval. Given the significance of these impacts, it is important to recognize them in any narrative which supports the scenario analysis.
The third limitation of any scenario analysis is closely related to the first two. If climate change occurs more quickly than expected, and the broader impacts trigger extreme ‘tipping points’; then outcomes could be significantly worse than expected. Several tipping points have likely already occurred–for example, more extreme climate events, bleaching and die off of sea corals, and loss of summer sea ice in the Artic. A range of even more extreme outcomes are expected as climate change continues—such as loss of ice sheets, and slowdowns in the natural carbon cycle. We don’t know exactly when, due to uncertainties in climate sensitivities and the response times of natural systems (systems’ inertia).
The fourth limitation of any scenario analysis is to recognize not just the downside of transition scenarios (for example, stranded assets and inflation) but also the possible upsides. In climate scenario modeling it has become increasingly common to incorporate broader socio-economic trends into the climate scenarios. For example associating a socio-economic move towards sustainability with lower emissions pathways. In particular, an assumption about green growth is commonly made–for example in the shared socio-economic pathway ‘SSP 1 - pathway to sustainability’ and by the Glasgow Financial Alliance for Net Zero (GFANZ) groups. More challenging scenarios would be to look at political economic responses like a move towards de-growth politics.
Identifying the missing components in climate scenarios is a critical step when interpreting the results and relevance to businesses. For example, understanding the broader externalities and the associated ‘tragedy of the horizon’ should stop anybody from treating the exercise as a cost-benefit analysis intended to identify the ‘optimal’ pathway for their business to support. Likewise, consideration of extreme tipping points should generally caution against interpreting the scenario modeling as ‘worst case’ analysis.
Importantly, it can also help to suggest a broader set of scenarios to look at. Some of these more extreme climate sensitivities, or green growth, can potentially be explored by building on existing modeling (for example the NGFS scenario database). Others, like tipping points and broader shared socio economic pathways, will require a broader set of scenarios as their basis.
For more insights on this topic, listen to our podcast series.
Speak to our Experts about how we can help with your climate-related modeling needs.