Supply function competition in a mixed electric power market

• Author: José Vicente‐Pérez,Carlos Gutiérrez‐Hita,Marc Escrihuela‐Villar
• Published date: 01 August 2020
• DOI: http://doi.org/10.1111/jpet.12432
• Date: 01 August 2020

J Public Econ Theory. 2020;22:1151–1175. wileyonlinelibrary.com/journal/jpet © 2020 Wiley Periodicals, Inc.

|

1151

Received: 11 October 2019

|

Accepted: 24 January 2020

DOI: 10.1111/jpet.12432

ORIGINAL ARTICLE

Supply function competition in a mixed

electric power market

Marc Escrihuela‐Villar

1

|Carlos Gutiérrez‐Hita

2,3

|

José Vicente‐Pérez

4

1

Departament d'Economia Aplicada,

Universitat de les Illes Balears, Palma de

Mallorca, Spain

2

Departamento de Estudios Económicos y

Financieros, Center of Operations Research

(CIO), Universitas Miguel Hernández,

Elche, Spain

3

ECEMIN Research Group, Universidad

Antonio de Nebrija, Madrid, Spain

4

Departamento de Matemáticas,

Universidad de Alicante, Alicante, Spain

Correspondence

Carlos Gutiérrez‐Hita, Departamento de

Estudios Económicos y Financieros, Center

of Operations Research (CIO), Universitas

Miguel Hernández, 03202 Elche, Spain.

Email: cgutierrez@umh.es

Funding information

Ministerio de Ciencia, Innovación y

Universidades, Grant/Award Number:

PGC2018‐097965‐B‐I00 (MCIU/AEI/

FEDER, UE); Ministerio de Economía y

Competitividad, Grant/Award Numbers:

ECO2015‐67901‐P, ECO2016‐77200‐P

Abstract

In this paper, we present a mixed oligopoly model

where electric power generators compete in supply

functions in a liberalized market. A former mono-

polist, the state‐owned generator, is assumed to be

(partially) privatized. First, we obtain that there is a

relationship between privatization and the number of

electric power generators concerning the level of

consumer surplus and total welfare. Indeed, a fully

state‐owned generator is socially optimal, lowering

private generators' profits and enhancing consumer

surplus; that is, if the degree of privatization de-

creases, consumer surplus increases compensating

the damage imposed on generators' profits. Second,

as the number of generators increases, full

privatization may provide similar levels of consumer

surplus and social welfare than those observed in a

mixed oligopoly. Moreover, it is also obtained that

price‐cost margins increase as marginal cost

increases. Overall, our results suggest that the

state‐owned generator should be privatized when

entry barriers are low enough, and competitiveness is

enhanced. Otherwise, a state‐owned generator may

protect consumers, enhancing consumer surplus.

1|INTRODUCTION

During the last decade of the 20th century, a number of countries began a liberalization process

of the electricity system. Traditionally, the electric power generation sector was a state‐owned

monopoly. As a result of huge demand increases, government budget constraints, and tech-

nology improvements, the sector was ready to introduce competition. At the same time, gov-

ernments decided a total (or partial) privatization of the state‐owned power generation

monopolist, while a reduced number of private electric power generators entered the market.

Thus, it is interesting to study the effects that two distinct policies (liberalization of the market

and partial or total privatization of the state‐owned power generator) have in the market.

1

Since

liberalization, wholesale electricity markets were mostly organized as bidding markets where

electric power generators compete by submitting price‐quantity bids in daily interaction.

2

Al-

though regulatory authorities expect both, competitive price‐cost margins and absence of

electric fall‐outs, wholesale electricity markets were found to have prices well above competitive

levels.

In this paper, we investigate the extent to which privatization, the cost structure, and the

number of competitors affect market outcomes in a context where electric power generators

strategically interact in supply schedules. In fact, such schedules can be seen as a continuum of

bids, that is, a supply function. Since the seminal paper by Klemperer and Meyer (1989), the

literature in electric power markets organized as bidding systems borrows this model to

characterize generators' bidding behaviour.

3

Moreover, from a technical point of view, the use

of supply function schedules to model electric power generation is also advised (see, for in-

stance, Day, Hobbs, & Pang, 2002). In this strand, a pioneering paper by Green and Newbery

(1992) analyzed the British wholesale electricity markets in a context of supply function

competition. It is shown that the Nash equilibrium in supply schedules implies a high markup

over marginal cost and substantial deadweight losses. Green (1996) introduces in the discussion

the effect of the government policies aimed to increase the amount of competition in the

electricity spot market in England and Wales. Under linear supply function competition with

asymmetric firms, it is found that partial divestiture should lead to a substantial reduction in

deadweight losses. In Green (1999), it is also analyzed the impact of long‐term contracts in spot

electricity prices for the England and Wales market. More recently, in Holmberg and Newbery

(2010) and Holmberg, Newbery, and Ralph (2013), various features of recent developments

concerning spot electricity markets, and technical features of supply function competition have

been studied. In addition, Newbery and Greve (2017) also emphasize that markup models may

explain competitive behaviour in electricity markets and thus, they can also be used as a

modeling approach. There are also in the literature some refinements of the supply function

model and its comparison with the traditional oligopoly models where the implications to

model real markets are presented. For instance, in Vives (2011) supply function competition

with random costs is introduced. A Bayesian supply function equilibrium is characterized

where supply functions are steeper with a higher correlation among the cost parameters.

1

Indeed, there can be two extreme cases: the policy‐maker may privatize a monopolistic state‐owned firm without opening‐up the market to competitors

(privatization without liberalization), or opening‐up the market retaining the full ownership of the public firm (liberalization without privatization).

2

Indeed, in most of the bidding systems, generators submit day‐ahead offers on an hourly basis according to the future demand forecast.

3

Other approaches use Cournot models to describe some features of the electricity generation market. See for instance Borenstein and Bushnell (1999) for

California's case.

1152

|

ESCRIHUELA‐VILLAR ET AL.