Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Reading the Winds: Climate Anomalies and Smarter Commodity Bets

Andersson, Elias LU and Lundgren, Emil LU (2025) NEKN01 20251
Department of Economics
Abstract (Swedish)
Can climate signals help investors make sharper moves in volatile commodity markets? This study explores how weather-related risk, specifically the El Niño-Southern Oscillation (ENSO), can be used to improve portfolio performance. We incorporate the Oceanic Niño Index (ONI) into a Random Forest forecasting model to capture the complex, non-linear effects of climate anomalies on commodity returns - signals often missed by traditional strategies. These machine learning-based portfolios are benchmarked against mean-variance optimization (MVO) and established factor models including market, momentum, value, and carry. The results show that ONI holds strong predictive power for commodities such as wheat, corn, and crude oil, while its influence... (More)
Can climate signals help investors make sharper moves in volatile commodity markets? This study explores how weather-related risk, specifically the El Niño-Southern Oscillation (ENSO), can be used to improve portfolio performance. We incorporate the Oceanic Niño Index (ONI) into a Random Forest forecasting model to capture the complex, non-linear effects of climate anomalies on commodity returns - signals often missed by traditional strategies. These machine learning-based portfolios are benchmarked against mean-variance optimization (MVO) and established factor models including market, momentum, value, and carry. The results show that ONI holds strong predictive power for commodities such as wheat, corn, and crude oil, while its influence is weaker for soybeans, cotton, coffee, gold, and silver. Across all configurations, the Random Forest portfolios deliver the highest return per unit of risk, outperforming traditional optimization approaches and passive market benchmarks. These findings provide a practical framework that investors can leverage to inform return expectations and manage risk in climate-exposed portfolios. (Less)
Please use this url to cite or link to this publication:
author
Andersson, Elias LU and Lundgren, Emil LU
supervisor
organization
course
NEKN01 20251
year
type
H1 - Master's Degree (One Year)
subject
keywords
ENSO, Weather Anomalies, Commodities, Random Forest, Machine Learning, Forecasting, Portfolio Optimization
language
English
id
9203240
date added to LUP
2025-09-12 09:58:17
date last changed
2025-09-12 09:58:17
@misc{9203240,
  abstract     = {{Can climate signals help investors make sharper moves in volatile commodity markets? This study explores how weather-related risk, specifically the El Niño-Southern Oscillation (ENSO), can be used to improve portfolio performance. We incorporate the Oceanic Niño Index (ONI) into a Random Forest forecasting model to capture the complex, non-linear effects of climate anomalies on commodity returns - signals often missed by traditional strategies. These machine learning-based portfolios are benchmarked against mean-variance optimization (MVO) and established factor models including market, momentum, value, and carry. The results show that ONI holds strong predictive power for commodities such as wheat, corn, and crude oil, while its influence is weaker for soybeans, cotton, coffee, gold, and silver. Across all configurations, the Random Forest portfolios deliver the highest return per unit of risk, outperforming traditional optimization approaches and passive market benchmarks. These findings provide a practical framework that investors can leverage to inform return expectations and manage risk in climate-exposed portfolios.}},
  author       = {{Andersson, Elias and Lundgren, Emil}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Reading the Winds: Climate Anomalies and Smarter Commodity Bets}},
  year         = {{2025}},
}