Macquarie University
01whole.pdf (6.81 MB)

Three essays on risk management in electricity markets

Download (6.81 MB)
posted on 2022-11-29, 02:24 authored by Lin HanLin Han

Electricity markets are significantly more volatile than other comparable commodity markets because of the non-storable nature of electricity. Thus, market participants are exposed to significant risks caused by extreme price outcomes, periods of heightened volatility, and their transmission between interconnected markets. In addition, uncertainties induced by the current transition from fossil fuel to renewable energy sources for power generation pose substantial challenges. For retailers or generators, these challenges include making decisions on effective investments and hedging, and for regulators, guaranteeing a reliable, secure supply of electricity. This thesis aims to shed light on risk management in the Australian National Electricity Market (NEM) in the above context. It follows the format of ‘thesis by publication’ and contains three research papers, each of which explores one or more related dimensions. 

The first study examines volatility spillover effects across interconnected regions in the NEM. This study applies the econometric framework originally proposed by Diebold and Yilmaz (2009, 2012) to assess market aggregate spillover effects and their directional decomposition between individual regions, both statically and dynamically. We find that spillover effects are significantly influenced by regional proximity and interconnectors, while exhibiting different patterns during peak and off-peak hours. We further relate the dynamic spillover patterns to specific short-term market events as well as to long-term changes in the renewable energy share and the generation capacity, and to the implementation of a carbon pricing mechanism. 

The second study extends the analysis of the first and focuses on extreme price outcomes. It examines the persistence of price spikes for individual markets and the dependence of extreme prices across different regions in the NEM. We apply the extremogram (Davis and Mikosch, 2009; Davis et al., 2011, 2012) to both 5-minute and half-hourly prices and find that extreme prices are more persistent in markets with a higher share of intermittent renewable energy, or more concentrated markets than in the other markets. We also find significant extremal dependence between physically interconnected adjacent regions, which shows asymmetric patterns. Further, based on the estimated extremograms, we show the effectiveness of the Australian Energy Market Commission’s 2016 rebidding rule change in reducing the strategic behaviour of market participants. 

The third study examines the profitability and investment risk of different generation technologies used in the NEM, by assessing dispatch-weighted prices (DWPs) and earnings-at-risk (EaR) of generators. We find that dispatchable generators, such as hydro, natural gas, battery storage, kerosene and diesel oil plants with controllable and flexible output, achieve significantly higher DWPs than less flexible generators, such as coal fired power plants. They also achieve higher DWPs than non-dispatchable renewable energy generation from solar and wind farms. In addition, although dispatchable and flexible generators that have stronger correlations with spot prices have significantly higher upside earnings potential compared to coal fired power plants and variable renewable energy generators, their EaRs are also much higher, which indicates greater uncertainty in profits. Further, DWPs and earnings are affected by seasonal patterns, and higher uncertainty and dispersion in summer are observed for both. 


Macquarie University

Australian Government Research Training Program (RTP) Scholarship

RoZetta Institute PHD Scholarship


Table of Contents

1 Introduction -- 2 Volatility spillovers in Australian electricity markets -- 3 Extremal dependence in Australian electricity markets -- 4 Average dispatch prices and earnings-at-risk for generators in Australia -- 5 Conclusion


A thesis presented for the degree of Doctor of Philosophy Thesis by publication

Awarding Institution

Macquarie University

Degree Type

Thesis PhD


Thesis (PhD), Macquarie University, Macquarie Business School, Department of Actuarial Studies and Business Analytics, 2020

Department, Centre or School

Department of Actuarial Studies and Business Analytics

Year of Award


Principal Supervisor

Stefan Trück

Additional Supervisor 1

Nino Kordzakhia


Copyright: The Author Copyright disclaimer:






259 pages