A Volatility Index for Bilateral Electricity Contracting Auctions
Keywords:
Electricity auctions, ARIMA, Metaheuristics, Time Series, MIBELAbstract
The use of electricity auctions in daily and intraday trading is one of the most common means of trading on electricity exchanges. Auctions are carried out aiming to reach a marginal price that will be charged for all agents involved. The identification of appropriate moments for placing bids interests consumers and generators, as negotiated demand is directly related to the marginal price established at the end of each auction. In the stock market several indexes assist investors in decision making on operations. However, energy auctions still lack this type of tool. In this context, this paper presents an index proposal based on aggregate supply and demand curves. Machine learning models and time series statistics are employed to predict bid value volatility trends. The experiments used real data from a 958 bid auction promoted by the Iberian Energy Market Operator (Operador del Mercado Ibérico de Energia - OMIE), the stock exchange responsible for the spot market in Spain and Portugal. Based on the results, it was possible to identify moments in which the parties involved in the auction chose to compromise on bid values with specification of the magnitude of these concessions.