Investment Incentives and Electricity Spot Market Competition

• Date: 01 December 2013
• Author: Gregor Zoettl,Veronika Grimm
• Published date: 01 December 2013
• DOI: http://doi.org/10.1111/jems.12029

Investment Incentives and Electricity Spot Market

Competition

VERONIKA GRIMM

University of Erlangen,

Lange Gasse 20 Nuremberg, Germany

veronika.grimm@wiso.uni-erlangen.de

GREGOR ZOETTL

University of Erlangen,

Lange Gasse 20 Nuremberg, Germany

gregor.zoettl@wiso.uni-erlangen.de

In this paper we analyze investment decisions of strategic firms that anticipate competition on

many consecutive spot markets with fluctuating (and possibly uncertain) demand. We study how

the degree of spot market competition affects investment incentives and welfare and provide an

application of the model to electricity market data. We show that more competitive spot market

prices strictly decrease investment incentives of strategic firms. The effect can be severe enough

to even offset the beneficial impact of more competitive spot markets on social welfare. Our

results obtain with and without free entry. The analysis demonstrates that investment incentives

necessarily have to be taken into account for a serious assessment of electricity spot market design.

1. Introduction

Incentives to invest in generation capacity have been heavily debated in the recent lit-

erature on electricity market regulation. Many authors suspect that there is a trade-off

between low spot market prices and proper investment incentives if firms behave strate-

gically. As Paul Joskow (2008) puts it, “policymakers in many countries are concerned

that competitive wholesale markets for electricity do not provide adequate incentives

for investment in sufficient quantities of generating capacity.” In this paper we provide

a model which analyzes the interdependency of spot market competition and invest-

ment incentives. We illustrate how the degree of competition at the spot markets affects

investment and welfare. Wealso show how the degree of competition at the investment

stage determines the desirability of different spot market regimes. We finally provide an

application of the model to data of a specific electricity market.

In our model, investment takes place at a first stage prior to competition among

firms on spot markets where they face fluctuating production cost and fluctuating de-

mand. Due to limited storability of the good (e.g., electricity1), demand and supply have

We thank Rabah Amir, Claude d’Aspremont, Eric van Damme, Jean Gabszewicz, Yves Smeers, and Achim

Wambach,as well as seminar participants at Berlin, Brussles, and Cologne for helpful comments and sugges-

tions. Financial support by the Deutsche Forschungsgemeinschaft and the Instituto Valenciano de Investigaciones

Econ´

omicas (IVIE) is gratefully acknowledged.

1. Notice that the questions we analyze (with the main feature being limited storability of the good) are

relevant also for a series of other markets. Examples are oil and gas extraction, capacity choices of hotels and

hospitals (e.g., number of beds), or capacity choices of airlines (number of planes). Our main motivation for

this paper was, however, to get a deeper understanding of investment incentives in electricity markets.

C2013 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume22, Number 4, Winter 2013, 832–851

Investment and Electricity Spot Market Competition 833

to coincide at any point in time in those spot markets.2We analyze two scenarios of

spot market competition: either behavior at the spot markets is competitive or it is given

by the Cournot–Nash equilibrium. Note that the Cournot and the competitive solution

are typically the upper and lower bound of the set of supply function equilibria (Klem-

perer and Meyer, 1989), a concept which is commonly used in the literature to model

electricity spot market competition.3In our analysis, we adopt the view of Bushnell

et al. (2008) who also focus on those upper and lower bounds of possible spot market

outcomes, arguing that the exact “market rules and local regulatory differences” in the

end determine which of the many possible equilibria within those bounds arises.

Our analysis yields important insights on the desirability of spot market regulation

under different assumptions about firms’ behavior at the investment stage. We establish

existence and fully characterize all equilibria of the strategic investment game for both

regimes of spot market competition (Cournot and competitive prices). We then show

that a regulation of prices to the lower bound of the above-mentioned range of spot

market equilibria (the case of perfect competition, which is clearly more desirable from

a short run perspective) is potentially problematic in the long run: competitive prices

at the spot markets lead to strictly lower investment by strategic firms and might even

lead to a welfare reduction. In a model with entry where firms enter the market as long

as they expect to cover some positive fixed entry cost, less firms enter the market under

competitive (than under Cournot) spot market pricing. Because investment and welfare

are increasing in the number of active firms, the problems identified for the case of an

exogenously fixed number of firms thus also obtain in a model with entry.

As a benchmark we also establish the case where investment is chosen optimally

from a welfare perspective (which would obtain for perfectly competitive investment

decisions), an approach which is often used in the literature to approximate investment

decisions in liberalized electricity markets.4Opposed to our results for strategic in-

vestment decisions, imposing competitive spot market prices (as compared to Cournot

prices) is more desirable both, from a short- and from a long-run perspective. The rea-

son is that nonstrategic firms would not aim at increasing scarcity prices by withholding

capacity.

In the empirical part of the paper, we finally quantify the effects we identified

in the theoretical part using data of the German electricity market. In sum, the results

of our analysis demonstrate (i) that it is crucial to accurately account for investment

incentives in order to assess spot market design properly and (ii) that regulation, which

aims at implementing competitive spot market prices, may lead to reduced investment

2. A model that abstracted from this property by assuming constant demand would not only be less

realistic but most importantly would eliminate the central problem analyzed in this paper. As shown in

previouscontributions for the case of constant demand (compare e.g., Kreps and Scheinkman, 1983), the degree

of spot market competition is irrelevant for firms’ investment decisions, because firms can fully determine the

outcome at the spot market by choosing their capacities. This is not true under fluctuating demand, where

invested capacities are either binding or idle. Spot market outcomes are determined by investment decisions

in the former case and by the degree of spot market competition in the latter, which has an impact on firms’

investment incentives.

3. As shown in Klemperer and Meyer (1989), for an unbounded support of uncertainty a unique supply

function equilibrium obtains, when uncertainty regarding demand at each single spot market becomes very

small, the highest and the lowest supply function equilibrium approach the Cournot and the competitive

solution, respectively (see their section 3), in the absence of uncertainty they precisely coincide with those

solutions (see their section 2). Notice that in a more recent article on supply function competition, however,

Vive s (2011) extends the analysis to the case of private information with respect to production cost of each

firm. In this case, the resulting equilibrium prices might also be above those obtained in a regular Cournot

framework.

4. See, for example, Bushnell (2005), Cramton and Stoft (2005), or Joskow (2007).