Municipality Finance Climate Risk – Pilot applies stress test methods
MuniFin is one of Finland’s largest credit institutions. In the fall of 2022, they launched a pilot project to assess and model climate risks in their lending portfolio together with FCG.
Risk management
Regulatory compliance
Modelling methodology
Climate risk assessment
The thought behind the project was to combine Municipality Finance’s data and FCG’s, now Advisense’s, unique experience from working with credit and climate risks in the finance industry.
“This pilot gives us a potential method for climate stress testing”
Joni Hirvelä, Senior Risk Manager at Municipality Finance
The Challenge
Climate stress testing methodologies are rapidly emerging partly due to regulatory and central bank demands but also because of their measurable impact on cash flows and balance sheet valuations. The approach for climate stress testing differs from traditional stress testing and there are new elements to consider. Firstly, transmission mechanisms are new and presumably long. Secondly, stress testing builds on data sets that are common in natural sciences but new and non-existent for financial models. Thirdly, the methodological basis for including climate or other ESG (Environmental, Social and Governance) attributes for that matter is thin. The challenge is how to do climate stress testing reliably and based on the best available data.
Based on the state of play, FCG’s task was to test the proposed methodology for climate risk stress testing at Municipality Finance.
“For Municipality Finance climate stress testing is becoming a must-have both from a regulatory perspective and more importantly, from our own risk management perspective. By assessing climate risks we will better manage direct and indirect transition risks and physical risks of climate change that Municipality Finance is exposed to.”
Elina Sääskilahti, ESG analyst at Municipality Finance
“We did extensive work compiling data sets for climate risk assessment. Clearly fragmented and anecdotal data exists, but a lot of effort is needed to translate that into a format fit to be used in PD and LGD models”
Konsta Lindqvist, Head of Analytics & Services, Advisense
The Solution
The Proof-of-Concept project aimed to validate that the climate stress testing model developed earlier by FCG can be deployed at Municipality Finance, and that FCG’s modelling methodology gives reasonable results and can be further developed.
The developed methodology incorporates the uncertainties relating to climate change and long forecast horizons into Municipality Finance’s stress testing calculations.
Data and translation
A number of data sets including emissions, geographic flood data, real-estate location and energy performance were collected and refined for use in the climate stress testing model. Moreover, allocation to counterparties and financial contracts was conducted. Additional work was done in translating physical world data into financial data measured in Euros. The assessment is based on a dynamic balance sheet structure and 30 years’ perspective, which has not been common in PD (Probability of Default) or LGD (Loss Given Default) models before in climate risk modelling.
“It turned out that FCG’s climate stress testing model already follows many recommendations given by the ECB (European Central Bank). However further work is needed to include even more parameters into climate risk assessment and to obtain more accurate data. Moreover, scenarios need to be better aligned with those recommended by the regulators”
Elina Sääskilahti, ESG analyst at Municipality Finance
The Results
FCG provided a calculation model, which produced capital adequacy VaR including material climate factors for Municipality Finance lending portfolio. That means modelling simulation-based outcomes for Municipality Finance under two climate risk stress scenarios. Modelling framework, calculation methods and data input were established in the applied model.
The FCG team has strong prior expertise in dynamic balance sheet and capital adequacy forecasting. Nevertheless, this project challenged the methodology development skills of the team. It is also safe to say that source data within this field is far from perfect if compared to traditional financial stress testing. More work is needed to get better data quality or even simulated data sets to be used for creating baseline scenarios. The long time horizon created interesting additional features but also added uncertainty into the modelling.
According to FCG, the pilot project with Municipality Finance provided clear evidence of the potential and suitability of the climate risk stress testing methodology. Key development areas for continuation of model development were identified and can be developed in possible next stages of co-operation.
