KBS Mathematics Informatics Research Discussion on Climate Finance

Earlier this month, Teemu Pennanen, Carmine Ventre and David Aikman organised a one-hour virtual meeting to discuss potential areas of research collaboration in the field of climate finance. The meeting invitation was extended to researchers in the Departments of Mathematics, Informatics and in King's Business School and we had around 25 attendees.

We invited four external experts to give us their perspectives on the most pressing questions where research could usefully contribute. These were:

• Teresa Lober (Bank of England)
• David Carlin (UN Environment Programme's Finance Initiative)
• Caspar Siegert (JP Morgan Asset Management), and
• Jens Larsen (Wellington Management).

Each speaker gave some short introductory remarks, followed by questions from the participants.

In this note, we record the main themes that emerged in the discussion, including specific research questions posed by our guest speakers. As the meeting was held under Chatham House rules, we do not attribute these to individual speakers/organisations.

Main themes of discussion and specific research questions

1. On the nature of climate risks, nonlinearities, tipping points

What do we mean by climate risk and what links it to financial risk? We can understand this through a Value at Risk lens: what is the percentage of financial value might I lose because of the physical impacts of climate change and because of the impact of policies to support the transition to net-zero? An example of the former would be the impact of flooding on the value of real estate; an example of the latter would be the impact of an announced future tightening in climate policies which depresses the value of fossil fuel-related assets today, generating a mark to market loss. There is also the possibility of "stranded assets" in some extraction-focused industries if some fraction of outstanding reserves become unusable.

Climate risks are best thought of as low probability and high impact events, where the possibility of tipping points creates enormous uncertainty about the validity of any scenarios used for risk management purposes. They differ from other potential sources of systemic risk in several respects. One difference is that the physical risks of climate change will inevitably increase over time – they cannot be managed away in the way normal financial risks can.

Possible research questions raised:

• While there has been good modelling of the physical impacts of alternative climate pathways, including by the IPCC, much work remains in turning these estimated physical impacts into economic/financial effects. An example given was understanding the impact of heat waves on labour productivity.
• Interactions between the physical and transition risks of climate change are generally underappreciated in the modelling that has been done. We need to develop models that capture the cumulative effect of both physical and transition risks, given we are living in a world with both at play. There is good work by Lars Peter Hansen (Chicago) on the multiplicative nature of these risks.

2. The role of the financial system in the transition to net zero and the role of climate stress testing

The financial system has three roles to play on climate: (a) managing risk; (b) aligning portfolios with net zero, where the focus is on reducing exposure to brown assets; and (c) providing the enabling green finance that will allow the transition to occur, particularly in emerging economies. Stress testing is a very valuable tool for understanding the risks to financial stability that could arise under alternative climate pathways.

Research questions raised:

• Regarding the scenarios used for climate stress tests, more work is needed on the triggering mechanisms, dislocations and nonlinearities that could arise, particularly in the short term, and particularly under an incomplete or suboptimal transition.
• For risk management purposes we need to move away from excessive focus on the modal path in any scenario and consider the 99.5% version. We currently risk giving a false degree of comfort to financial institutions in their stress testing.
• What risks does climate change pose for the wider financial system (beyond banks and insurers), including asset managers, pension funds, and shadow banking institutions?
• Patent data for green innovations are a rich source of information for research. Which ideas are attracting the greatest interest? What does this tell us about the likely technological solutions of the future? What financing is required to scale these ideas up? What are the frictions to providing this finance?

3. Monitoring whether we are on track to meet Paris targets, and the unintended consequences of the transition to net zero

Research questions raised:

• What are the clear "no-regret" actions toward decarbonisation that need to happen immediately (eg investing in renewable generation capacity, moving away from coal in the OECD), and what is the required order of operation that takes account of the availability of viable substitute energy options?
• The Paris Agreement requires that public and private financial flows are aligned to 2 degrees warming relative to pre-industrial temperature levels. How can we monitor/measure whether this is being met?
• Is there space to use digital platforms to monitor different regulations and their effects? One difficulty is that regulation is still at a very early stage and evolving. It also often takes the form of guidance and recommendations, leaving lots of discretion over implementation to firms. Could there be a role for scenario analysis to assess which regulatory actions could be most important, carrying the greatest impact?
• As regulation shifts requiring banks and insurers to disclose their climate exposures and form a strategy for the transition to net zero, will risk shift to the unregulated shadow financial sector, over which there is little visibility? Will banks play the role of stewards in gradually managing down brown assets or will they divest?
• If advanced economies pursue policies to support the transition to net zero, does this risk the counterproductive effect of global capital moving out of emerging markets? Emerging market economies will be hardest hit by climate risks and require the greatest financing for mitigation policies.

4. Investors' expectations and portfolio choice

Research questions raised:

• What climate scenarios are priced into current asset prices? How can we understand the expectations that underlie current valuations? Climate risk may well be fully priced in for the "green poster children" of the transition, eg stocks like Tesla. But it probably is not for slightly more carbon-efficient cement producer. This speaks to the need for a granular analysis of the issue.
• Prices of emission rights (and forward contracts on those rights) can also be used to gauge investors' expectations about the aggressiveness of future climate policies.
• From an investor portfolio perspective, if physical and transition risks are negatively correlated, does that mean the natural owners of oil and gas stocks should be investors who also hold assets that are subject to substantial physical risk? And from a political economy perspective, if investors diversify their exposure to climate risk in this way, would that make it easier for policymakers to implement policies to support the transition to net zero?
• Optimal portfolio choice under climate uncertainty needs to move away from simple mean-variance considerations. It needs to consider tails and the prospect of shifts in distributions, as well as multiplicative risks. This is a good area of focus for an interdisciplinary research programme.

5. Climate change and the macroeconomy

Some of the most important and basic features of climate change are not covered in the current generation of integrated assessment models. These include non-stationarity, irreversibility, and deep uncertainty about the risks. We all know these factors are important, but they are hard to model, and our current toolkit essentially just ignores them. The models are designed to analyse small shocks around an equilibrium path rather than the large structural shift that will characterise the transition. This area is ripe for careful interdisciplinary research, in particular as the mathematics involved is highly complex.

Research questions raised:

• The transition to net zero will require a large shift in relative prices, and a major increase in fixed investment in new green technologies. Integrated assessment models do a good job of capturing what we might call the ordinary effects of shifting demand around and shifting prices. But a point Nick Sterne has emphasised is that these models miss scale effects and the impact of investment on innovation.
• Uncertainty vs risk. If we think the distribution of shocks is much fatter tailed but there is Knightian uncertainty about how climate risks will affect the economy, how should this affect optimal transition policy
• What impact will these effects have on monetary policy? We could be moving to a world with far greater and more frequent shifts in relative prices. What will this mean for long-term interest rates, for inflation targeting etc.
• As carbon pricing takes greater effect in the years ahead, some industries will be able to pass on this cost to consumers whereas others will not. There is a need for more research on this issue, understanding how the incidence of carbon pricing will vary by sector.

6. Data gaps

Research questions raised:

• We need to improve the quality of data available to measure the physical risks of climate change. Financial institutions need a "golden source" which provides them with a holistic picture of their risk exposures.
• There are also large gaps in our understanding of where climate risks lie within the financial system.