Skip to Content

Banner for HDR Expo 2012

Rangga Handika

First Name: Rangga
Surname: Handika
Department Dept of Applied Finance and Actuarial Studies
Supervisor(s): A/Prof Stefan Trück ,

Email Student

View Rangga Handika's profile

Paper Title

Modelling Price Spikes in Electricity Markets' Impact of Load, Weather and Capacity


This paper proposes a model for price spikes in Australian electricity markets considering the impact of variables such as load, weather and capacity constraints.


This paper is the first to model the magnitude of price spikes in Australian electricity markets by considering load, weather and capacity variables and applying the Heckman two-stage regression model to consider sample selection bias.

Key literature/theoretical perspective

Because only certain prices are considered as spikes, we face a sample selection bias in modelling price spikes.
The Heckman two-stage regression contains a probit model for occurrence of spikes and a least square regression for modelling the size of spikes in electricity prices.


We use a probit model for the occurrence of price spikes and then a least square regression for the magnitude of price spike.


We find that most of the considered weather variables tend to be important factors in determining whether a price spike occurs or not.
We find evidence for a significant sample selection bias.
We find that for most of the considered regional markets, included variables such as load, relative air temperature and reserve margin are significant for explaining the magnitude of the observed price spikes

Research limitations/implications

Similar to most empirical studies that deal with heavy-tailed data, the explanatory power of the model for the magnitude of price spikes is rather low. However, the considered explanatory variables are still significant and explain some of the variation in the size of the spikes.

Practical and Social implications

A new approach for modelling the occurrence and magnitude of price spikes in electricity markets.


Electricity Markets, Price Spikes, Selection Bias, Inverse Mills Ratio