Introduction

Stranded assets—fossil fuel reserves that become uneconomical to extract due to climate policy or technology shifts—threaten long-term oil & gas company valuations. Machine learning models predict which companies face highest stranding risk by analyzing policy trends, renewable energy economics, and reserve characteristics.

Stranding Risk Factors

Features: global climate policy momentum (carbon prices, net-zero commitments), renewable cost trends (falling solar/wind costs), reserve characteristics (high-cost, remote reserves strand first), company diversification (pure-plays strand faster than diversified energy companies). Models trained on past stradings (tar sands economic collapse) predict future risk.

Application

Identify oil & gas companies with high stranding risk. Underweight or divest to avoid value destruction. Overweight low-stranding-risk companies with diversified portfolios and transition plans.

Conclusion

ML-based stranded asset prediction improves portfolio resilience by avoiding climate-vulnerable energy companies.