1764. Volatility Modelling of Climate Benchmarks: A Long-Short Term Memory Network Approach
Invited abstract in session MA-9: Innovation in Insurance and Financial Risk Management, stream OR in Finance and Insurance .
Monday, 8:30-10:00Room: Clarendon SR 2.01
Authors (first author is the speaker)
| 1. | Salvatore scognamiglio
|
| Department of Management and Quantitative Sciences, Parthenope University | |
| 2. | Francesca Battaglia
|
Abstract
In recent years, the growing importance of climate benchmarks in sustainable finance has highlighted the need for robust and accurate volatility modeling of these indices. Traditional econometric methods often struggle to capture the complex, nonlinear, and dynamic behavior of climate-related financial data. This paper proposes a novel approach using Long-Short Term Memory (LSTM) networks, a class of recurrent neural networks well-suited for time series prediction, to model and forecast the volatility of climate benchmarks. We train the LSTM model on historical climate index data, capturing both short-term fluctuations and long-term dependencies. The model’s performance is evaluated against conventional models, such as GARCH and stochastic volatility models, demonstrating superior predictive accuracy and adaptability to regime shifts and extreme events. Our findings suggest that LSTM networks offer a powerful tool for investors and policymakers to assess climate-related financial risks, enhancing decision-making processes in sustainable finance. Numerical experiments performed on some climate benchmarks validate the proposed approach.
Keywords
- Artificial Intelligence
- Forecasting
Status: accepted
Back to the list of papers