Sammendrag
Correct modelling and forecasting of price distributions is important in forming trading and bidding strategies as well as in risk management. This paper seeks to characterise the nonlinear effects of exogenous factors on peak hour wholesale electricity price formation as well as forecasting the price distribution. Using a dynamic quantile regression model adjusted for heteroskedasticity, we capture effects such as mean reversion, spikes, time varying volatility, and at the same time, estimate this price rather complex relationship to fundamentals. We demonstrate how lagged prices, prices of gas, coal and carbon, forecasts of demand and reserve margin, and weekend effects influence peak price distribution. For some variables, the price elasticity's can vary a lot according to whether prices are low, medium or high. We also show that a structural quantile regression model in generally outperform a skewed GARCH-t model and perform just as good as a CaViaR model regarding out-of-sample forecast of Value at Risk.