A0261
Title: Factor-augmented functional regression with an application to electricity price curve forecasting
Authors: Sven Otto - University of Cologne (Germany) [presenting]
Luis Winter - University of Cologne (Germany)
Abstract: A function-on-function linear regression model is proposed for time-dependent curve data that is consistently estimated by imposing factor structures on the regressors. A novel integral operator based on cross-covariances identifies two components for each functional regressor: a predictive low-dimensional component, along with associated factors that are guaranteed to be correlated with the dependent variable, and an infinite-dimensional component that has no predictive power. In order to consistently estimate the correct number of factors for each regressor, we introduce a functional eigenvalue difference test. The model is applied to forecast electricity price curves in three different energy markets. Its prediction accuracy is found to be comparable to popular machine-learning approaches while providing interpretable insights into the conditional correlation structures of electricity prices.
