Financial stability in response to climate change in a northern temperate economy

Climate change poses a significant threat to global economic stability. While previous research has primarily focused on the effects of average temperature on economic damages, this study addresses critical gaps by incorporating a wider range of climate variables and analyzing sub-annual economic cycles. Existing literature lacks attention to the impacts of climate volatility on these shorter-term economic fluctuations, which are crucial for understanding financial stability. The increased frequency and severity of extreme weather events (hurricanes, floods, droughts, heatwaves) present significant physical risks to infrastructure and financial systems. Governments and financial institutions are increasingly recognizing the need for adaptation and resilience management, including targeted investments at sub-national levels to address localized impacts that exceed global trends. The study uses Canada as a case study due to its robust economy, geographical diversity, and disproportionate warming compared to global averages. This northern temperate economy provides a comprehensive example with diverse climate conditions that can be applied to other similar economies worldwide.

Existing literature on climate-related economics predominantly employs probabilistic modeling and primarily focuses on average temperature as a climate variable, neglecting other significant factors. There's limited attention to financial stability, and most studies lack the sub-annual temporal resolution necessary to understand the dynamics of economic cycles and their sensitivity to climate volatility. While estimates of overall economic damages are converging, there's a need to enhance understanding of aggregated economic output beyond simply quantifying damages. Previous research often shows a decrease in GDP due to climate change, or minimal increases, particularly when relying solely on temperature data. This study aims to improve on this by analyzing a broader spectrum of climate variables and employing sub-annual data analysis.

The researchers developed an empirically based method to project economic changes due to climate change, focusing on sub-annual economic cycle changes (period, amplitude, trough depth). Canada was selected as a case study, divided into four main regions based on climatic and economic factors: Pacific Maritime/Western Cordillera, Prairie/Northwestern Forest, Southeastern/Northeastern Forest, and Atlantic Maritime. Monthly climate data (average, maximum, minimum temperature, precipitation, snow, heating degree days, cooling degree days) were gathered from 2000-2019 (historical) and projected from 2025-2095 (future) using data from the Climate Atlas of Canada and Climate Data Canada, based on RCP 8.5. National and provincial GDP data, along with monthly wholesale trade data and interprovincial trade flows, were obtained from Statistics Canada. Data were detrended using the average annual growth rate to isolate climate impacts. Linear and non-linear models (general linear model, generalized linear mixed-effects model, least angle regression, decision trees, random forest, multivariate adaptive regression splines (MARS)) were tested. The MARS model was ultimately selected due to its superior performance, as indicated by R² and GR². The optimized MARS model was used to project future GDP, and economic cycle analysis was performed using the Savitzky–Golay filter in the TIMESAT program. Uncertainty assessments considered variations in data binning and climate model projections. Trade interconnectivity was analyzed using Supply Input-Output Tables to determine the percentage of trade at risk due to climate impacts.

Inclusion of a broad range of climate variables significantly improved model accuracy, increasing the explained variance by over 20%. Using only temperature underestimated annual GDP in 2095 by $400 billion CAD (20%) compared to the optimized MARS model. The optimized model projected an annual increase of 0.03% in GDP above the normal average growth rate (2.6%) under the RCP 8.5 scenario. In contrast, temperature-only models predicted a decrease in GDP. Regional variations were substantial: The Prairie region showed the greatest GDP changes, with projections ranging from a 3.2% increase in Alberta to a -12.6% decrease in Saskatchewan. Ontario and Quebec (Southeastern region) showed increases of 1.1% and 3.9%, respectively. The Atlantic Maritime region exhibited high variability. Economic cycle stress varied regionally, with Alberta, Saskatchewan, Prince Edward Island, and New Brunswick facing increased stress (increased period and amplitude), indicating greater volatility. Ontario and Quebec showed lower stress (decreased period and amplitude). Southeastern Canada's projected stability may positively influence national financial stability due to its strong trade connections with other regions. Saskatchewan's trade was most at risk, with 48% of imports originating from high-stress zones.

The study's findings challenge previous research by demonstrating that focusing solely on temperature as a climate variable can be misleading. A more comprehensive approach incorporating a range of climate variables is crucial for accurate projections of economic impacts. The sub-annual analysis reveals regional disparities in vulnerability to climate change, highlighting the importance of targeted adaptation strategies. The analysis of economic cycles provides valuable insights for infrastructure investment and resilience planning. Regions facing increased cycle amplitude require infrastructure capable of handling greater economic fluctuations, while regions experiencing shorter cycles need infrastructure to support higher productivity within a compressed timeframe. The differing impacts across Canadian provinces may be attributed to the unique industrial structure and economic drivers of each region. The study emphasizes the importance of considering both regional economic interdependencies and the industry-specific vulnerabilities in the development of effective climate adaptation policies.

This research presents a robust empirical method for assessing the economic impacts of climate change by incorporating multiple climate variables and sub-annual economic cycles. The study reveals significant regional variations in vulnerability and the importance of considering cold-weather variables for accurate projections. Future research could explore industry-specific impacts, regional resilience in greater depth, and the role of trade and investment networks in climate-economic interactions. Applying this methodology to other northern temperate economies is crucial for expanding our understanding of climate change implications.

The projections are based on historical relationships between climate and the economy and assume a stable global economic system, excluding factors like conflict, migration, or resource scarcity. The study uses RCP 8.5, a high-emissions scenario, which may overestimate economic growth as extreme weather events may exceed historical frequency and severity. Acute climate-related events are not explicitly modeled, potentially leading to underestimation of damages. Further research incorporating these factors would improve robustness.