Time-frequency analysis and machine learning models for carbon market forecasting

Jules Sadefo Kamdem, Passy Miano Mukami and James Njong
Additional contact information
Passy Miano Mukami: AIMS - African Institute for Mathematical Sciences
James Njong: AIMS - African Institute for Mathematical Sciences

Post-Print from HAL

Abstract: As a follow-up to the recommendations of the Kyoto Protocol to reduces climate change, the European Union has created carbon financial markets (such as "European Union Emissions Trading Scheme") allowing the introduction of a cap and trade system where mandated companies are granted carbon dioxide emission permits. These markets allow companies to trade carbon permits. The Paris Agreement of December 2015 has also highlighted the importance of the allocation of capital from the carbon market to achieve a global reduction in CO2 emissions. Therefore, an in-depth analysis of carbon market mechanisms could provide some flexibility as to where and when greenhouse gas emissions are reduced, and thus could reduce the costs of climate change mitigation. climate change, but also enable anticipate the behavior of market stakeholders concerning risk and expected return. There are several variables or institutional characteristics that may explain the formation or volatility of carbon prices in these markets. Thus, carbon prices somtimes exhibit non-stationary characteristics which makes them particularly hard to predict. In this paper, we introduce an hybrid forecasting model, incorporating multiscale analysis techniques, and other machine learning models to improve prediction accuracy of carbon prices. The proposed model for carbon price prediction outperforms other comparative models. The mean squared error(MSE), goodness of fit (R2) and the Wilmott index of agreement of the proposed model are 0.0047, 0.9185 and 0.9801 respectively 5.06285, 0.97959 and 0.99480

Keywords: Carbon market; Climate change; Forecasting; Neural networks; Multiscale analysis (search for similar items in EconPapers)
Date: 2023-06-12
References: Add references at CitEc
Citations:

Published in Annals of Operations Research, 2023, ⟨10.1007/s10479-023-05443-x⟩

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-04134564

DOI: 10.1007/s10479-023-05443-x

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().