Abstract
The study examined the impact of climate change risk on financial sector development,
focusing on financial stability, efficiency, depth, and access across financial institutions and
markets in selected EMEA countries. Additionally, the study examined the differential impact
of climate change risk on financial sector development across various regions, in selected
EMEA countries, each at different levels of economic development. Furthermore, machine
learning techniques were used to predict the future impact of climate change risk on financial
sector development. The study formed its theoretical and model foundations in Systemic Risk
Theory, the Efficient Market Hypothesis, the Financial Instability Hypothesis, the DICE, ESFC,
and DSK frameworks. However, among existing theories, there is an absence of a
complete theory that links climate change risk and financial sector development. The study
employed both classical econometric models and machine learning techniques, utilising the
PMG-ARDL estimator to capture short-run and long-run dynamics, alongside panel-aware
XGBoost and Random Forest models for predicting the impact of climate change risk on
financial sector development. The PMG-ARDL results reveal a positive long-run relationship
between climate change risk and financial markets development in selected EMEA countries,
particularly in regions with robust financial systems and advanced climate financing
instruments, such as Europe. Conversely, the findings indicate a negative long-run impact of
climate risk on financial institutions’ development in regions with limited regulatory depth and
institutional resilience, notably in parts of Africa and the Middle East. This highlights structural
sensitivity and institutional fragility in these contexts. The research revealed that climate
change risk negatively affects financial access and depth in Africa, while Europe and the
Middle East show positive impacts on depth and mixed outcomes on access. Random Forest
and panel-aware XGBoost, demonstrated superior predictive accuracy compared to other
machine learning models, especially in forecasting financial development indicators. While the
immediate statistical impact of climate risk on financial development may appear marginal, the
broader evidence points to a more nuanced reality: climate risk is embedded in the structural
and behavioural dynamics of financial systems. Its influence may be indirect, delayed, or
masked by conventional metrics, but it remains potent. For financial regulators and
policymakers, the muted short-term impacts of climate risk may create an illusion of stability,
potentially postponing critical reforms. However, the evidence of asymmetric data patterns and
the superior predictive capacity of machine learning models suggests that climate
vulnerabilities are quietly embedding themselves within financial systems—often in ways that...