A fractional Brownian -- Hawkes model for the Italian electricity spot market: estimation and forecasting

TL;DR

This paper introduces a novel two-factor model combining Hawkes and fractional Brownian processes to improve the description and forecasting of Italian electricity spot prices, addressing seasonality and spike identification.

Contribution

It presents a new additive model for electricity prices that integrates Hawkes and fractional Brownian processes, with calibration, validation, and forecasting performance analysis.

Findings

Effective modeling of electricity price spikes.

Accurate forecasting demonstrated through evaluation metrics.

Incorporation of seasonality and Hurst coefficient estimation.

Abstract

We propose a model for the description and the forecast of the gross prices of electricity in the liberalized Italian energy market via an additive two-factor model driven by both a Hawkes and a fractional Brownian processes. We discuss the seasonality, the identification of spikes and the estimates of the Hurst coefficient. After the calibration and the validation of the model, we discuss its forecasting performance via a class of adequate evaluation metrics.