Quantitative modelling of climate risk is still in its early stages. Different groups have different definitions of climate risk. The Intergovernmental Panel on Climate Change (IPCC) defined climate risk as the product of a hazard and its consequences. In finance literature, the risks associated with climate change can be categorized into either a physical risk or a transition risk. Physical risk refers to the harm, damage, or loss caused by climate or weather events. Transition risk, on the other hand, includes risks resulting from mandates and government regulations that aim to decarbonize the economy. For the United Nations Inter-Agency Task Force on Climate Change and Disaster Risk Reduction, the most basic formula of climate risk is: climate risk = (natural hazard x vulnerability) / adaptive capacity. Generally, extreme climate and weather events are considered natural hazards, vulnerability is the probability to be affected by extreme climate and weather events, and adaptive capacity refers to the economic resources of the country. From these three definitions alone, it seems there is still confusion as to what climate risk is and how to compute it. How do we compute climate risks when climate models are not reliable? How different types of climate risk could interact? There is an urgent need to measure, understand, and address climate risks as several climate events have highlighted the importance of properly estimating climate risk. After Hurricane Katrina, All State and State Farm reduced their insurance exposures to coastal states. Last January 2019, Pacific Gas & Electric, California’s largest utility, filed for bankruptcy in the face of $30 billion liabilities for a wildfire that started from their power line. The extremely hot and dry conditions of the forest set the stage for the deadliest and most destructive wildfire in California’s history. Early this year, insurance companies AIG and Chubb has backed out of insuring homes in California due to wildfire risk. All because nobody has successfully quantified climate risk. My research group in Climate Statistics will work on defining and measuring climate risk.

  1. Develop climate risk models. Most of the progress in measuring climate risk is on producing a climate risk index. Currently, there are five popular indices, namely, Global Climate Risk Index (CRI), WorldRiskIndex (WRI), Notre Dame Global Adaptation Index (ND-GAIN), Center for Global Development (CGDev), and Climate Vulnerability Monitor (DARA) that provide country-level measure of climate risk. The main issue with these indices is that they are merely tallying the number of climate and weather events, death tolls, and economic losses reported to insurance agencies. This method is retrospective and conservative and will not give the correct probabilities for extreme events that we have not yet seen. I believe the ingredients to mathematically sound measures of climate risk will involve:
    a. the correct joint distribution of extreme events,
    b. the correct joint distribution of all types of losses and damages, e.g., injury, health impacts, loss of lives, livelihoods, properties, infrastructures, ecosystems, environmental resources, etc.,
    c. and the correct joint distribution between the extreme events and losses,
    which will vary from one location to another. Developing these models will require expertise in extreme events modelling and economics.
  2. Develop pricing models with climate risk. Modelling innovations must be introduced to financial models to account for climate risk.  Market participants have been underreacting to climate risk mostly because it is a new kind of risk which nobody knows how to price. Climate change has messed up asset pricing models because the past can no longer predict the future. Some markets have started to integrate climate risk in the asset prices such as the real estate, bonds, loans, and insurance markets. However, how they measure climate risk is not based on rigorous statistical methods and their pricing models differ. Furthermore, they consider carbon emissions as the only variable that represents climate risk. Central banks and financial regulators also need to know how to price climate risk to properly change capital and reserve requirements. The production processes of corporations in the stock market are exposed to climate events. Thus, there is a need to explore how shocks stemming from never-before-seen climate events can affect asset prices and the economy. Our research areas in extreme events and climate risk modelling will put us in a position to lead the charge in understanding how to add climate risk to asset pricing models. Fundamental to modelling asset prices with climate risk are the following:
    a. the correct joint distribution of extreme events,
    b. the correct joint distribution of asset prices following extreme events, e.g., stocks, foreign currencies, commodities, interest rates, and other financial instruments,
    c. and the correct joint distribution between the extreme events and asset prices,
    which will vary from one sector or industry to another.






Leave a Reply

Your email address will not be published. Required fields are marked *