The Analysis of Spot Price Stochasticity in Deregulated Wholesale Electricity Markets
Tipping, James Patrick
Thesis DisciplineBusiness Administration
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Traditionally, time series of wholesale electricity market spot prices have been modelled either by mimicking market operation and equilibrating demand and supply, or by specifying an exogenous process for prices. More recently, a number of hybrid models have been developed, combining the merits of both methods. In this vein, we present an econometric model for daily spot prices in the New Zealand Electricity Market (NZEM) that utilises reservoir management theory to incorporate information on the hydro storage level, a recognised driver of NZEM spot price behaviour. In order to forecast future storage levels and prices, we also construct a model for daily reservoir releases that can be used in conjunction with time series of inflows. This analysis reveals that releases in New Zealand are driven primarily by hydrological factors, as opposed to market conditions. The combined price and storage forecasting model can be applied in a variety of contexts, and offers an alternative perspective to the traditional models of NZEM behaviour. Finally, we calibrate a Cournot model of market behaviour in the National Electricity Market of Australia during daily peak, shoulder and off-peak periods, adding credibility to the future application of such models. The resulting model parameters are, in general, consistent with conventional wisdom. Spot prices from this market are then modelled by combining the output of the analytical model with a stochastic price process.