Modeling PJM Day-Ahead Forward Prices: A Mean Reversion, Jump Diffusion, Seasonality, and Stochasticity-Focused Approach

Abstract

The electricity market is a well-established and far-reaching financial market that facilitates the delivery of electricity produced from power plants to its end users. We begin by exploring current models for electricity prices and proceed to propose a new model for PJM electricity day-ahead forward prices. The model has several components and together these components are meant to accurately represent several phenomenon commonly observed in electricity prices - mean reversion, seasonality, jumps, and high volatility. The proposed model has several parameters and to a large extent, this thesis will focus on identifying the optimal model parameters. The model goes through several iterations before arriving at its final form. This study will use data from the PJM Interconnection, which operates in 13 states and is the most extensive data source available. Our goal is to understand the characteristics of PJM electricity prices and develop a model that simulates PJM day-ahead forward prices as accurately as possible.