Abstract
The urgency of climate change forces the real estate sector to address financial risks tied to decarbonization and regulations. This study examines climate transition risks affecting BPI Gestão de Ativos’ real estate portfolio, where trends favor sustainability. By extending MSCI’s building-level scenario analysis to the asset-level, it provides a framework to evaluate vulnerabilities and develop optimization strategies. The methodology integrates techniques to assess risks, translate them into profitability impacts, and identify actions between holding, selling, or acquiring assets. XGBoost predicted rental income for vacant properties, enabling a risk-adjusted return metric. A Gaussian Mixture Model simulated a synthetic portfolio representing market assets for acquisition, and the Simulated Annealing algorithm optimized profitability while minimizing climate risk. Findings confirm BPI GA’s concerns, revealing higher climate risks correlate with declining asset values, with the 1.5ºC scenario amplifying pressures. Vulnerabilities concentrated in specific regions and sectors, with high-risk assets facing reduced demand. Under stricter scenarios, optimization prioritized divesting, while moderate 2ºC scenarios enabled acquisitions. By accounting for demand for lower-risk assets, optimization achieved better performance, underscoring the importance of aligning strategies with market trends. This analysis equips BPI GA with a framework to navigate climate risks, ensuring maximized risk-adjusted profitability.| Date of Award | 11 Feb 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Miguel de Oliveiros Vieira de Albergaria e Castro Nogueira (Supervisor) |
UN SDGs
This student thesis contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 11 Sustainable Cities and Communities
-
SDG 13 Climate Action
Keywords
- Climate transition risks
- CVaR
- Climate risk management
- Scenario-based analysis
- Portfolio optimization
Designation
- Mestrado em Análise de Dados para Gestão
Cite this
- Standard