Ainhoa Zarraga

Electricity consumption and economic growth in Spain

This article investigates linear and nonlinear causality between electricity consumption and economic growth in Spain for the period 1971 to 2005. We use the methodology of Toda and Yamamoto (1995) and Dolado and Lütkepohl (1996). We also apply the standard Granger causality test in a vector autoregression for the series in first differences. We find unidirectional linear causality running from real GDP to electricity consumption. By contrast, we find no evidence of nonlinear Granger causality between the series in either direction.

Common features between stock returns and trading volume

This article tests for the existence of features shared in common by daily stock returns and trading volume contributing to the empirical analysis of the relation between those series. Using Spanish data this study analyses this hypothesis looking at features such as seasonality, skewness, kurtosis, non normality and serial correlation. This study finds that monthly seasonalities and distributional features such as skewness are driven by a common factor in stock returns and volume. This study also finds a non-synchronized comovement between the cycles of both variables.

Economic growth and electricity consumption in 12 European countries: A causality analysis using panel data

We apply recent panel methodology to investigate the relationship between electricity consumption and real GDP for a set of 12 European Union countries using annual data for the period 1970–2004. Tests for panel unit roots, cointegration in heterogeneous panels and panel causality are employed. The results show a long-run relationship between the series. We estimate this relationship and test for causality. We find no shortrun causality in any direction. These results might help to design appropriate electricity consumption policies in the sample countries, as well as investment policies in interconnections to build a single European market for electricity.

GMM-based testing procedures of the mixture of distributions model

A direct test of the mixture-of-distributions model is conducted using daily Spanish stock return and trading volume for the period April 1990 to January 1996. Both the standard mixture-of-distributions model of Tauchen and Pitts (1983) and the modified version proposed by Andersen (1996) are estimated by GMM and tested using the overidentified restrictions. The results reject the models, that is, the variables are not related due to a common dependence on a factor, namely the flow of information, according to the specifications of the mixture models considered.

Volatility transmissions in the Spanish Intra-day electricity market

This paper analyzes the volatility transmissions between the different sessions of the Spanish Intraday electricity market for the period 2002-2013 using hourly prices. Based on the daily realized volatility and the organization of the market, volatility can only be transmitted from session 1 to 6 in chronological order, which allows to formulate six Autoregressive Distributed Lag models in which inference on volatility transmissions can be conducted. Within the Heterogenous Autoregressive Regression framework two different models are estimated. The first one (HAR-RV) explains the volatility in terms of its past, while the second (HAR-CV-JV) decomposes the volatility into continuous and jump components to capture extreme spikes. In-sample forecast criteria select HAR-CV-JV model for all Intraday sessions whereas out-of-sample criteria select HAR-RV.

Forecasting volatility in the EPEX market

This paper uses high-frequency intraday electricity prices from the EPEX market to estimate and forecast realised volatility. Variation is broken down into jump and continuous components using quadratic variation theory. Then several heterogeneous autoregressive models are estimated for the logarithmic and standard deviation transformations. GARCH structures are included in the error terms of the models when evidence of conditional heteroscedasticity is found. Model selection is based on various out-of-sample criteria. Under the logarithmic transformation the simplest model outperforms the rest. Under the standard deviation transformation, jump detection before model estimation is useful to improve forecasting.

Modelling Realized Volatility in Electricity Spot Prices: New Insights and Application to the Japanese Electricity Market

The paper analyzes volatility of the electricity prices in the Japanese day-ahead market using realized volatility. We use several jump tests to decompose total realized variation into jump and continuous components. Then, we estimate several HAR models that show the time-dependence structure of the volatility. Our results show that even though that market is narrow, it is relevant to identify jumps in volatility. Besides, modelling residuals improve estimation results. The time-dependent structure of the prices is present in volatility as well.