Testing for Market Integration in the Australian National Electricity Market.

• Author: Nepal, Rabindra

INTRODUCTION

The 1990s witnessed a remarkable global experiment with the introduction of market-based reforms and the restructuring of the electricity sector. Australia established the National Electricity Market (NEM) on 13 December 1998 as a consequence of the restructuring and liberalisation of its electricity supply industry (ESI). The NEM is a gross pool arrangement for wholesale electricity trade in Australia and operates a deregulated wholesale market in the electrically connected but separate regions

of New South Wales (NSW), Victoria (Vic), Queensland (QLD), the Australian Capital Territory (ACT), South Australia (SA) and Tasmania (TAS). One of the foremost objectives of wholesale electricity market reform in Australia, consistent with the electricity market reforms objectives in the European Union (EU), was to deepen competition in wholesale electricity trade and facilitate the fuller integration of the separate regional electricity markets into an efficient and integrated NEM (Worthington, Kay-Spratley and Higgs, 2005). In the EU, the aim of creating a common and integrated wholesale market for electricity remains an ongoing experiment that started since the second half of 1990s.

Economic theory on market integration, as developed by the famous economist Antoine Augustine Cournot (1838) suggests that markets become fully integrated when 'an entire territory of which the parts are so united by the relations of unrestricted commerce that prices take the same level throughout with ease and rapidity' (Stigler, 1969; Cournot, 1971). This resembles the law of one price where prices for a well-defined product traded in two separate geographic markets should constitute a single integrated geographic market. Prices across these two markets should be equal and show a common long-run equilibrium. In the case of electricity markets, the price differences should not exceed transmission charges, including transmission losses, assuming no physical transmission constraints (De Vany and Walls, 1999a; Productivity Commission, 2013). As such, the systematic testing of the wholesale electricity markets integration based on the 'law of one price' is only possible by undertaking an appropriate econometric approach. However, direct empirical evidence, based on econometric methods, as to whether the separate regional electricity markets in the NEM are fully integrated and well-functioning is missing. This is a major gap in the literature since electricity market integration studies based on econometric evidence already exist in the European context (for example, Zachmann (2008); Castagneto-Gissey; Chaves and Fellani (2014) and Menezes and Houllier (2014)) and in the North American context (for example, Woo, Lloyd-Zanetti and Horowitz (1997); De Vany and Walls (1999) and Dempster, Issacs and Smith (2008)).

Several studies such as Neuhoff and Newbery (2005) and Brunekreeft and Newbery (2006) have shown that direct physical interconnections among separate regional electricity markets can facilitate market integration and enhance competition in wholesale electricity trade, especially, in concentrated wholesale electricity markets with limited market participants. However, empirical evidence on the role of direct interconnections and regional price linkages in achieving fuller integration of regional electricity markets into a NEM is also missing in the Australian context. This is particularly relevant from a policymaking perspective as this kind of evidence can serve as additional input into decisions as to whether investment in new interconnectors should be undertaken. Concerns about under-investment in transmission networks and interconnectors, in contrast to broader concerns of excessive distribution network investments in Australia (Productivity Commission, 2013), must also be addressed. Hence, findings on the state and development of market integration can provide guidance in reforms of the existing wholesale market arrangements in the NEM.

The purpose of this paper is to empirically assess the existing degree of market integration and provide associated evidence on the role of interconnections in achieving an integrated wholesale market for electricity in Australia. We do so by obtaining a fuller understanding of the direct pricing relationships (price convergence /divergence) between the interconnected regional electricity markets in the NEM. (1) We analyse prices because they reveal and aggregate all available information in an efficient market (Grossman, 1976). Price convergence (divergence) shows how strongly prices are related (unrelated) across similar markets and cointegration analyses can detect if the prices share a common long-run equilibrium (Engle and Granger, 1987; Johansen, 1995). The lack of cointegrating relationships is indicative of significant underlying transfer constraints within the grid that connects these markets (De Vany and Walls, 1999). However, cointegration implicitly assumes a fixed structural relationship among prices over the considered time-period. This is not very appropriate for studying a relatively new market that has an emergent connective structure and poised for the large scale uptake of wholesale renewable power. So, we also apply the Kalman filter methodology based on a state-space model that allows us to estimate time-varying coefficients. This can account for possible dynamic structural changes that cannot be identified by cointegration analysis (Kalman, 1960). A time-varying filtered coefficient indicates how strong pricing relationships are among the markets considered and gives us a guide to the level and development of market integration.

The contribution of this paper is threefold. This is the first study to directly examine the state and development of market integration among the physically interconnected regional electricity markets in Australia using robust econometric approaches. By doing so, we aim to clarify to what extent the objective of creating an integrated NEM has been achieved since its establishment. Second, this paper adds to the limited number of studies analysing electricity market integration in island economies where the benefits of an integrated market are generally challenging because cross-border interconnections have, to date, been infeasible.

Third, the empirical evidence on market integration can provide a significant additional input into the on-going policy debate concerning transmission investment strategies and network regulation arrangements across Australia. The empirical findings can aid the existing regulatory tests for transmission expansion and network planning which is based on identifying investment option that maximises the net economic benefits. (2) Evidence of regional network constraint effects in the existing interconnectors, based on real-time price difference, are also important in the NEM given that technical constraints and grid limitations, if left unaddressed, could lead to significant curtailment of wind energy generated in Victoria (35%) and South Australia (15%) by 2020 (AEMO, 2013a).

Furthermore, the impact of growing large scale trade of wholesale renewable power on the market integration outcomes of 'energy-only' markets has largely remained unexplored (Tangeras, 2015). The NEM provides an appropriate case to study such outcomes as wholesale renewable energy trade (wind and solar) is expected to significantly increase. The NEM also remains a gross-pool and one of the most transparent 'energy-only' wholesale electricity markets globally. Thus, the findings of this study are relevant for markets like the EU in the quest towards creating a common and integrated 'energy-only' (predominantly) market for electricity since the start of the electricity sector liberalisation in the 1990s.

The remainder of the paper is structured as follows. Section I provides a brief overview on the operation of the NEM, including the currently operational interconnectors. The econometric methodology is described in Section II while Section III explains the data and summary statistics. Section IV discusses the results and relevant policy implications. Section V contains concluding remarks and discussion of future research possibilities.

1. OVERVIEW OF THE NEM

   The Australian ESI consisted of vertically integrated and state-owned entities prior to 1994. The Victorian power exchange started operating the first Australian power market in 1994 and transmission was unbundled from generation. New South Wales (NSW) launched a market in 1996. Queensland (QLD) mimicked the reforms that took place in NSW in 1998, while in South Australia (SA) accounting unbundling took place in 1997 following the early corporatization of the state-owned, vertically integrated utility in 1995. Thus, the liberalisation of the power sector separated the Australian ESI into competitive (generation and retail) and regulated monopoly (networks and supply) segments with increased private sector participation in the competitive segments. The NEM was created as a wholesale market for electricity in the Eastern jurisdictions with Tasmania joining the NEM in 2005. (3)

   Economic exchange between the electricity generators and consumers is facilitated through a gross-pool spot market where the output bids from all generators is aggregated and instantaneously scheduled to meet demand through a centrally-coordinated dispatch process. The scheduled generators submit offers every five minutes of the day and the system operator (the Australian Energy Market Operator (AEMO)) determines the generators required to produce electricity in the most cost-efficient way. AEMO then dispatches these generators into production by sending Automatic Generation Control (AGC) target signals to each generating unit.

   A dispatch (or spot) price is determined every five minutes and six dispatch prices are averaged every half-hour to determine the spot price for each trading interval for each of...