Date of Original Version



Technical Report

Rights Management

All Rights Reserved

Abstract or Description

We introduce a discrete-time model for electricity prices, which accounts for both transitory spikes and temperature effects. The model allows for different rates of mean-reversion: One for weather events, one around price jumps, and another for the remainder of the process. We estimate the model using a Markov chain Monte Carlo approach with three years of data from Allegheny County, Pennsylvania. We show that our model outperforms existing stochastic jump-diffusion models for this data set. Results also demonstrate the importance of model parameters corresponding to both the temperature effect and the multi-level mean-reversion rate.