Valuation of electricity derivatives

Abstract

This paper aims to forecast the value of electricity derivatives in the MIBEL system. For that purpose, three different models based on the simulation of future price scenarios are developed. The project begins with an extensive analysis of prices in spot and forward markets, with the objetive of charactering market's behaviour. The obtained results are introduced in the simulation models, building their basis. The models aim to forecast all the possible future price scenarios and their probability. A Brownian motion and a mean reversion model have been used to simulate the spot market. A Monte Carlo algorithm forecasted the forward curve evolution. Finally, the models are applied to two practical cases. The first one consists on valuating a collar for a consumer. The second one, focuses in the valuation of a combined-cycle gas turbine asset. The calculated results are compared with the ones obtained with the Black-Scholes formula. As a conclusion, the Monte Carlo model obtains the best trade-off between simplicity and accuracy.