Title: On multi-factor state-space modelling and forecasting of EUA futures prices
Authors: Jun Seok Han - Macquarie University (Australia) [presenting]
Nino Kordzakhia - Macquarie University (Australia)
Pavel Shevchenko - Maquarie University (Australia)
Stefan Trueck - Macquarie University (Australia)
Abstract: The European Union Emissions Trading System (EU ETS) was introduced in 2005 to confront rising greenhouse gas emissions. The EU ETS covers all major CO2 emitting industries, and next to European Emission Allowance (EUA) spot contracts, there is also a wide range of futures contracts available for trading. A multi-factor state-space model for risk-neutral pricing of EUA futures is presented that can also be applied for the out-of-sample forecasting of futures prices. A comparative analysis of the performance of a state-space model in a general setup with correlated measurement errors versus reduced-form models is conducted. As we deal with unobservable factors, we use the Kalman filtering technique for the estimation of the state variables, subsequently estimating the model parameters by maximising a marginal likelihood function. We illustrate the developed model, using a cross-section of daily futures contracts for the sample period from January 2016 - April 2020 that corresponds to the Phase III period of the EU ETS.