ISM-UCL-UCSB-MQ Workshop

Stochastic Modelling in Climate Risk: Financial Mathematics and Economics

Dates

3 days, 21st - 23rd November 2023

Venue

Auditorium, The Institute of Statistical Mathematics (ISM) in Tokyo, Japan
(https://www.ism.ac.jp/access/index_e.html)

Participation Modes

In-Person or Zoom

Overview

Climate change stands as one of the most daunting global challenges we face today. With palpable consequences on society, the economy, and our environment, its looming impacts are anticipated to intensify. Indeed, it poses a significant threat to the stability and growth of the global economy.

Statistics and Financial Mathematics play a crucial role in mitigating the effects of climate change on the public. Financial Mathematics provides tools for assessing and managing risk, while Statistics, which form the basis of machine learning and data science, offer methods to model, assess, and monitor climate processes. This workshop aims to bring together experts in mathematics, statistics, and environmental studies to explore the challenges and opportunities of climate change.

Objective

Workshop Objectives:

Workshop Format:

Over three days, attendees will engage in a mix of lectures, debates, and hands-on workshops. As the event draws to a close, a roundtable discussion will focus on future challenges related to climate change, including financial mathematics and economic implications.

Outcome

The proposed outcomes of this workshop include:

Throughout this workshop, participants will delve into the most recent findings in statistics, mathematics, and machine learning, specifically focusing on financial mathematics and its applications in addressing climate change. This exploration is crucial for understanding and managing the multifaceted concerns posed by climate change.

Organizers

Prof. Tomoko Matsui, The Institute of Statistical Mathematics (ISM)

Prof. Andrea Macrina, University College London (UCL) 

Prof. Gareth W. Peters, University of California, Santa Barbara (UCSB )

Prof. Pavel V. Shevchenko, Macquarie University (MQ)


Local arrangement

Prof. Vu Tran (ISM)

Mrs. Saori Karuki (ISM)

Program (Times shown are in Japan Standard Time)

(1) 21st November

Opening Welcome

09:45 - 10:00 Director-General Hiroe Tsubaki (The Institute of Statistical Mathematics)  

Theme: Environmental Statistics and Data Science

Topics: Statistical exploration across environmental studies and economics

Session 1: 

10:00 - 11:00 Prof. Yongyang Cai (The Ohio State University) [online]

"Climate Policy under Spatial Heat Transport and Risks: Cooperative and Noncooperative Regional Outcomes"
 We build a novel stochastic dynamic regional integrated assessment model (IAM) of the climate and economic system with elements missed in most IAMs, including spatial heat transport, sea level rise, permafrost thaw, and tipping points. We study optimal policies under cooperation or noncooperation between two regions (the North and the Tropic) in the face of risks and recursive utility. Our results suggest that when the climate elements are ignored, important policy variables such as the optimal regional carbon tax and adaptation could be seriously biased. We also find the North has higher carbon taxes than the Tropic, and climate risks significantly increase the regional social cost of carbon under either cooperation or noncooperation.

Session 2:

11:15 - 12:15 Dr. Silva Herran Diego (National Institute for Environmental Studies)  [online]

"Analysis of global scenarios of climate change mitigation with a general equilibrium integrated assessment model"
 This model introduces the analysis of global scenarios for climate change mitigation with an integrated assessment model (IAM) based on general equilibrium approach. This is a process-based IAM that represents the interactions among the energy system, the land and agricultural system, and the economy across 17 global regions. It describes the trends in energy resources and technologies, commodities and policies throughout the 21st century. Reducing the emissions of greenhouse gases (GHGs) from human activities will have important implications on the supply and demand of energy, on the agricultural activities and on the economy.  The possible development of the socio-economic drivers of emissions in the future will determine different GHG emissions pathways, which will lead to different climate change outcomes (namely global temperature change).
 In order to assess the feasibility of these future possibilities, researchers have proposed and analyzed scenarios which describe the trends and development directions of socio-economic drivers in qualitative terms.  In addition, researchers have developed process-based IAMs to quantify such scenarios, that provide knowledge on the trends in diverse indicators related to the energy and agricultural system, the economy and the environment.

Session 3:

14 :00 - 14:30 Prof. Tomoko Matsui (ISM)

"Generalised hyperbolic state space models with application to spatio-temporal heat wave prediction"
 As global warming progresses, it is increasingly important to monitor and analyze spatio-temporal patterns of heat waves and other extreme climate-related events that impact urban areas. In this work, we present a novel dynamic spatio-temporal model by combining a state space model (SSM) and a generalized hyperbolic distribution to flexibly describe a spatial–temporal profile of the tail behavior, skewness and kurtosis of the local urban temperature distribution of the greater Tokyo metropolitan area. Such a model can be used to study local dynamics of temperature effects, specifically those that characterize extreme heat or cold. The focus of the application in this paper will be heat wave events in the greater Tokyo metropolitan area which is known to be prone to some of the most severe heat wave events that have one of the largest population exposures due to high density living in Tokyo city. The advantages the proposed model offers are as follows: it accommodates skewed and fat-tail distributions for temperature profiles; the model can be expressed as a location-scale linear Gaussian SSM which allows the development of an efficient Monte Carlo mixture Kalman Filter solution for the estimation. The proposed model is compared with the Gaussian SSM through application to maximum temperature data in the Tokyo metropolitan area between 1978–2016. The result suggests that the proposed model estimates the temperature distribution more accurately than the conventional linear Gaussian SSM and that the predictive variance of our method tends to be smaller than that obtained from the conventional spate time linear Gaussian SSM benchmark model.

14:30  - 15:00 Prof. Daisuke Murakami (ISM)

"Gridded GDP projections compatible with Shared Socioeconomic Pathways"
 Historical spatially explicit gross domestic product (GDP) data are needed for the analysis of future climate risks. Our objective is to project a high-resolution and long-term GDPs under Shared Socioeconomic Pathways (SSPs) by using a wide variety of geographic auxiliary information. We projected GDPs in a 1/12-degree grid scale. The projection is done through a downscaling of the country GDP data for 1850–2010 and SSP scenario data for 2010–2100. In the downscaling, we first modeled the spatial and economic interactions among cities and projected difference in future growth patterns in each city. Subsequently, the projected patterns and other auxiliary geographic data were used to project GDP by grid. Finally, the GDP projections were visualized via three-dimensional mapping to enhance the clarity for multiple stakeholders. Our results suggest that the spatial pattern of urban and peri-urban GDP depends considerably on the SSPs; the GDP in existing major cities grew rapidly under SSP1, moderately grew under SSP 2 and SSP4, slowly grew under SSP3, and dispersed growth under SSP5.

Session 4:

15:15  - 16:30 Discussion Session - Collaboration Opportunities - Key Focus Topics

Additional Session (only for Speakers and Organizers):

17:30 - 21:00 Discussion and Dinner

(2) 22nd November

Theme: Financial Mathematics and Climate Change

Topics:  Mathematical Climate Finance Markets, pricing, and hedging

Session 1: 

09:00 - 10:00 Prof. Andrea Macrina (UCL)

"The Carbon Equivalence Principle for Financial Markets"
 The carbon equivalence principle requires that all financial products incorporate the caused and enabled carbon flows. This produces a fundamental impact on the design, pricing and hedging of carbon-linked financial instruments. This presentation shall show how financial markets can build in carbon emission and absorption, and thus contribute to the decarbonization needed to achieve net zero targets.

10:00 - 11:00 Prof. Ryoichi Suzuki (Ritsumeikan University)

"Clark-Ocone-Haussmann type formulas for additive processes and applied to finance"
 We investigate Clark-Ocone-Haussmann (COH) type formulas for additive processes by using Malliavin-Skorohod calculus. This talk gives various types of COH formulas, such as the Mallivin-Taylor-Mancino type COH formula under original and equivalent probability measures. These formulas are useful to solve financial problems in many situations, including climate finance and risk. Hence, we would like to consider potential applications of these COH type formulas. 

Session 2:

11:15 - 12:15 Prof. Jiro Akahori (Ritsumeikan University)

"Continuous-time DICE type models"
 I will discuss continuous-time DICE-type model evaluation.
                                

Session 3:

14:00 - 15:00 Dr. Tadashi Hayashi (Mitsubishi UFJ Trust and Banking Corporation)

"Carbon Emissions Pricing by Forward and Double Barrier Backward SDE approach"
 Under the circumstances of global warming caused by increasing in greenhouse gases, there are many theoretical and empirical studies in carbon emissions to control and reduce the gases. Our study is focused on the carbon emissions pricing via Forward and Double Barrier BSDE, so-called, as another pricing approach.

Session 4:

15:15  - 16:30 Discussion Session - Collaboration Opportunities - Key Focus Topics

     Additional Session (only for Speakers and Organizers):

17:30 - 21:00 Discussion and Dinner

(3) 23rd November

Theme: Climate Risk

Topics: Climate Risk Effects on Demographics, Pensions, and Long Term Portfolio Management

Session 1: 

09:00 - 10:00 Dr. Richard J. Matear (CSIRO Leader of the Future Climate and Hazard Activity in the Australian Climate Service)

"Future Climate and Hazard Modelling with a Perspective on Quantifying Climate Risk and the Loss and Damages of Anthropogenic Climate Change"
 Over the last decade, Australia has experienced frequent extreme climate events that have devastated people, communities, the economy and the environment (e.g. Black Summer of 2019-20).  To help Australia better prepare and plan for climate extremes in a changing climate and to minimise their impacts, the Australian Government established a new multi-agency partnership called the Australian Climate Service (ACS). 
This presentation will first summarise the future climate and hazard modelling work being led by ACS to elucidate how climate extremes will evolve with a changing climate.  Second, I will present an example of how ACS is using the climate information to assess future climate risk (e.g. heatwaves and health), and third, I will explore how the climate hazard information could provide a way to cost the impact of anthropogenic climate change.
 The Australian Climate Service climate modelling will provide intelligence on key climate hazards (e.g. Heatwaves and Extreme Fire weather) such as changes in their frequency, intensity and duration with climate change.  The climate modelling and insights will be used to assess the climate risks that Australia will face in the coming decades and to help guide the Australian Government's climate adaptation strategy.
The Australian Government has initiated a National Climate Risk Assessment.  Initial work by ACS explored how changes in heatwaves in a future climate would affect human health.  We used Global Warming Levels to overlay future heatwaves with demographic and health data (i.e., exposure and vulnerability data) to estimate morbidity changes with a future climate.  I will show insights gained through this heat health risk assessment.
 At the last United Nations Climate Change Conference (COP27 in Nov 2022), a breakthrough agreement was reached to provide “loss and damage” funding for vulnerable countries hit hard by climate disasters.  The research challenge for us is how we use our climate and risk knowledge to cost the impact of climate change and justify the funding to assist developing countries in responding to climate change losses and damages. In the third part of my talk, I will explore how one could use the change in the hazard with exposure and vulnerability information to estimate the loss and damage of anthropogenic climate change.

10:00 - 11:00 Prof. Pavel V. Shevchenko (MQ)

"Solving stochastic dynamic integrated climate-economy models using Least Squares Monte Carlo methods"
 The classical dynamic integrated climate-economy (DICE) model has become the iconic typical reference point for the joint modeling of economic and climate systems, where all six model state variables (including carbon concentration, temperature, and economic capital) evolve over time deterministically and are affected by two controls (carbon emission mitigation rate and consumption). We consider the DICE model with stochastic shocks in various parts of the model and solve it under several scenarios as an optimal stochastic control problem using the Least Square Monte Carlo method (LSMC) - a popular simulation method for solving optimal stochastic control problems in quantitative finance. We consider the application of various LSMC methods (including the use of neural network approximation) and discuss their pros and cons.

Session 2:

11:15 - 12:15 Dr. Ragnar Levi Gudmundarson (Heriot-Watt University) [online]

"Assessing Portfolio Diversification via Two-Sample Graph Kernel Inference"
 In this work we seek to enhance the frameworks practitioners in asset management and wealth management may adopt to assess how different screening rules may influence the diversification benefits of portfolios. The problem arises naturally in the area of Environmental, Social, and Governance (ESG) based investing practices as practitioners need to select subsets of the total available assets using screening rules of ESG ratings and to compare the subsequent risk and return profile of the portfolios created from different selective portfolios. We propose a novel method to compare the diversification relationships of assets in different portfolios based on a machine learning hypothesis testing framework called the kernel two-sample test. The objective of the test is to determine whether two samples come from the same underlying probability distribution. In the case of asset management, the samples are sequences of graph-valued data points that represent a dynamic portfolio obtained by a certain ESG screening rule and certain portfolio optimization criteria such as the global minimum variance or max Sharpe. The fact that the sample data points are graphs means that one needs graph testing frameworks to compare diversification benefits. The problem is natural for kernel two-sample testing as one can use so-called graph kernels to work with samples of graphs. The objective is then to determine if the two dynamic portfolios have the same generating mechanism. A failure to reject the null hypothesis would indicate that ESG screening does not affect diversification while rejection would indicate that ESG screening does have an effect. The article describes the graph kernel two-sample testing framework, further, it provides a brief overview of different graph kernels. We then demonstrate the power of the graph two-sample testing framework under different realistic scenarios. We finally apply the framework to demonstrate the workflow one can use in asset management to test for structural differences in the diversification of portfolios under different ESG screening rules.

Session 3:

14:00 - 15:00 Dr. Pasin Marupanthorn (Heriot-Watt University)

"Mechanisms to incentivise fossil fuel divestment and implications on portfolio risk and returns"
 Mechanisms to incentivize divestment strategies, such as divestment schedules, are an important component of carbon reduction strategies. We use dynamic asset allocation methodologies to assess this impact over time on index portfolios (S&P 500 and FTSE 100), and global exchange-traded funds (ETFs). While return profiles are unaffected, the risk profile of S&P 500 divestment portfolios is impacted by rapid divestment strategies as the divestment concentration increases. Instantaneous divestment may benefit management structure, while slower divestment provides greater stability in portfolios’ tracking errors and benefits carbon reduction, especially from reinvested capital. Divesting from both energy and utilities sectors offers a reduced carbon footprint of up to 7%, while ETFs’ divesting from highly carbon-concentrated ETFs offers further carbon footprint reductions. Investing in funds with low carbon footprints results in lower dividend returns and management fees. Although ETFs’ returns are insensitive to divestment strategies and schedules, their risk profiles are affected, proportionally to their carbon intensity, especially for rapid divestment and at the expense of higher tracking errors. Divestment strategies based on ESG rating screening of FTSE 100 portfolios provide evidence of improved diversification and impact on risk/return performance. Our study underscores the importance of considering investors’ demographics, such as dividends, management structure, and carbon reduction targets, on gradual releases of divestment strategies. 

Session 4:

15:15 - 16:30 Discussion Session - Collaboration Opportunities - Key Focus Topics

Closing Remarks:

16:30 - 16:40