This paper analyzes hourly electricity prices in three day-ahead markets - the European Energy Exchange in Germany, the Paris Power Exchange in France and Operadora del Mercado Español de Electricidad in Spain - using a periodic panel model. The empirical results show that, for individual hourly price series, periodic autoregressive models fit the data better than standard autoregressive models. It is also found that the decrease in prices associated with public holidays effects is different between hours and across markets. There is wide variation among hours in each market, with no effect on prices in some hours but substantial reductions in other hours. When all hourly prices are modeled jointly as a panel, autoregressive periodic components models fit the data better than standard nonperiodic models. The results have implications both for price modeling and for the management of electricity price risk in these markets. Regarding modeling, our results suggest that periodic components are significant features of electricity prices. Regarding risk management, for example, the pricing of derivatives on hourly prices, it seems advisable to use the whole set of twenty-four hourly prices as the underlying process and not only the prices for delivery at a specific hour, which supports the use of rainbow options.