 |
U-shaped climate-credit risk dependency at the loan portfolio level: implications from the genetic algorithm for the correlated Bernoulli trials simulation
1 : Research and Forecasting Department, Bank of Russia
The climate change agenda stimulated banks to promote "green" finance initiatives. We use a non-parametric (genetic) algorithm to simulate portfolios with low and high default correlation within "brown" and "green" sectors separately and material (while feasible) negative default correlation between the "brown" and "green" sectors. We find a U-shaped dependency of credit-climate risks. Initially, rise in "green" lending promotes diversification and reduces credit risk, but further rise of "green" lending and its domination within the loan book leads to credit risk further accumulation.
