2585. Optimizing high and low ESG portfolios with DCC-GARCH models and graph theory
Invited abstract in session TC-7: Portfolio Risk Management, stream Risk Management in Commodities and Financial Markets .
Tuesday, 12:30-14:00Room: Clarendon GR.01
Authors (first author is the speaker)
| 1. | Pilar Gargallo
|
| Facultad de Economía y Empresa, Universidad de Zaragoza | |
| 2. | Manuel Salvador
|
| Facultad de Económicas, Universidad de Zaragoza | |
| 3. | Jesús Miguel
|
| Universidad de Zaragoza | |
| 4. | Luis Lample
|
| Universidad de Zaragoza |
Abstract
This study proposes a network theory-based approach to portfolio optimization, combining advanced statistical techniques with graph analysis. To achieve this, we use VAR-DCC-GARCH models to capture the dynamic correlations between assets and construct an undirected graph where connections reflect significant relationships.
Additionally, we apply the Granger causality test to generate a directed graph that identifies predictive links between assets. The combination of both approaches results in a directed and weighted graph, on which we implement filtering and clustering techniques to select the assets that will compose the portfolio based on their ESG values.
The weights of the selected assets are assigned using the Minimum Variance Portfolio model, allowing for risk minimization without compromising expected returns. Furthermore, we compare the performance of portfolios with higher and lower exposure to ESG assets to assess whether investments with strong sustainability criteria yield better risk-return outcomes.
Finally, we analyze the temporal evolution of the networks and their impact on portfolio composition, evaluating performance through financial metrics.
Keywords
- Risk Analysis and Management
- Financial Modelling
- Graphs and Networks
Status: accepted
Back to the list of papers