Chapter 12

Energy Markets and Principles

of Energy Pricing

12.1 Introduction: Basic Competitive Market Model

Any standard economics textbook starts with the theoretical world of perfect
competition. In such a case, consumers maximise their utility subject to their
budget constraints and producers maximise their profits subject to the constraints
of production possibilities. There are numerous consumers and producers trying to
transact in the market place. In a competitive market condition, all agents are price
takers and there is no market power of any agent. In general, the demand for a
good reduces as prices rise (i.e. inverse relationship with price) and vice versa.
This gives rise to the familiar downward sloping demand curve. Similarly, pro-
ducers face an upward sloping supply curve. The higher the price, the more is the
supply, as at higher prices more producers become viable. The interaction of
supply and demand decides the market clearing price of the good and the quantity
of goods that will be sold (or purchased).
Consumers satisfy their utility (or preferences) by consuming a good. As utility
is not observable, an alternative parameter for measurement of their satisfaction is
the willingness to pay or accept to move from a situation to another. At any given
price, consumers spend an amount equal to the price times the quantity purchased.
No consumer is willing to pay for something that she does not want but some
consumers may be willing to pay more than the market price. Thus the total
willingness to pay at price P0 in Fig. 12.1 is given by the area ACq0O. But the
expenditure for the good at this price is given by the area P0Cq0O. The difference
between these two areas gives excess benefit consumers obtain, known as con-
sumer surplus. This is represented by the area left of the demand curve but above
the price actually being charged for the good.
The sellers on the other hand incur cost for producing the goods sold and as
long as the costs are recovered, they may be willing to sell for any given price.
However, even at that price, some sellers will receive more benefits due to low
cost production, while others will just break-even. Therefore, the benefits accrued

S. C. Bhattacharyya, Energy Economics, DOI: 10.1007/978-0-85729-268-1_12, 277

Springer-Verlag London Limited 2011
278 12 Energy Markets and Principles of Energy Pricing

Fig. 12.1 Willingness to pay

ABq1O = total willingness
A
to pay at price P1

ACq0O = total
B
P1 willingness to pay
at price P0
C
P0

O q1 q0 Q

to the producers are known as producer surplus. Total benefits to the producers
then include production costs and the surplus (see Fig. 12.2).
At the equilibrium, the willingness to sell equals the willingness to pay. At this
condition, the demand matches the supply. This is considered as an optimal
allocation in the sense that the equilibrium cannot be replaced by another one that
would produce higher welfare for some consumers without reducing welfare of
others. This is depicted in Fig. 12.3.

Fig. 12.2 Willingness to sell P

P1

P0 Total cost =
Area under the
curve

O q0 q1 Q
12.1 Introduction: Basic Competitive Market Model 279
Fig. 12.3 Competitive P
equilibrium Consumers surplus

E = equilibrium

Q
Producers surplus

Competition forces sellers to charge no more than their rivals. If one seller
charges more than the market clearing price, consumers will go to others offering
the same good at lower price. If someone charges less than the market price, the
demand will outweigh supply, forcing a return to the market price. Individual
buyers and sellers cannot affect the price. Buyers and sellers react to changes in the
market price. At lower prices, some sellers will leave the market while more
consumers enter it. Similarly, at higher prices more sellers are willing to offer their
goods while there will be fewer consumers. The participation in the market is
voluntarily and consumers or sellers are free to enter or leave the market in a
perfectly competitive case. Price is equal to the marginal cost of the last supplier.
In mathematical terms, the above can be presented as follows:
The aggregate consumer surplus from consumption of a good at the prevailing
price p* is
Z1
CS Q pdp 12:1
p

The producer surplus for supplying the good having a cost function
C = C(Q) is
p pQ p CQ p 12:2
The net economic welfare is the unweighted sum of aggregate consumer sur-
plus and producer surplus is given by
Z1
W p CS p Q pdp pQ p CQ p 12:3
p
12.2
280 Extension of the Basic Model 4
12 Energy Markets and Principles of Energy Pricing

The objective is to find the price at which the welfare is maximized. This is
obtained by setting the first order derivative of the welfare function with respect
price to zero.
dW d dp
CS 0; or
dp dp dp
12:4
dQ p dCQ p
Qp Qp p 0
dp dp

From where we obtain, p* = MC or the price is equal to the marginal cost.

Such a market has a number of properties:
(a) the participation is voluntaryboth consumers and producers enter and exit
the market freely, without any compulsion.
(b) Consumers who are willing to pay the market price enter the market (which
means that there could be some consumers who remain outside). Similarly,
only those producers will be called to supply whose marginal cost of supply is
lower or equal to the price. The marginal producer will recover only his
operating costs while other producers who are called to supply would earn
some additional profits (which might cover their fixed costs partly or fully
depending on their cost structure). This puts pressure on the suppliers to keep
their costs low to enter the market. Therefore, there is nothing wrong in a
market economy to find price excluding some consumers or producers. Sim-
ilarly, there is nothing wrong for some producers to earn large profits while
others are barely profitable.
(c) The relevant pricing principle is essentially a short run one, with an objective
of clearing the market.
However, certain basic conditions have to be satisfied to obtain such efficiency
outcomes: existence of freely competitive markets, perfect and costless flow of
information and knowledge, smooth transferability of resources and absence of
externalities. Clearly, most of these requirements are not satisfied by the todays
energy market. In addition, the energy sector is marked by certain specific char-
acteristics such as indivisibility of capital, tradability of some products and
depletion of some resources. Consequently, the basic model needs to be expanded
for any meaningful analysis. We consider these aspects below.

12.2 Extension of the Basic Model

Let us consider a number of characteristics of the energy sector and see how the
basic model outcome needs to be modified.
12.2
280 Extension of the Basic Model 5
12 Energy Markets and Principles of Energy Pricing

12.2.1 Indivisibility of Capital

Indivisibility of capital implies that capacity expansion takes place in discrete unit
sizes of plant units, and investments are lumpy in nature. In the energy sector, this
is a common feature. For example, oil fields or coal mines are developed for a
particular capacity. Refineries or power plants come in particular sizes and once
one unit is installed, increments are possible only in standard sizes, and not in
smooth, continuous increments as is assumed in the theory. The existence of
economies of scale often suggests that better cost advantages could be achieved by
installing bigger sizes. The indivisibility of capital changes the shape of the supply
curve, for instead of a continuous supply curve, we now have a supply curve for a
fixed capacity and the addition of new capacities brings abrupt changes (or kinks)
at the point where investments take place. This is shown in Figs. 12.4 and 12.5.1
In a fixed plant with a capacity of q0, the output cannot go beyond the installed
capacity. The marginal cost of supply is assumed to be constant at v for the entire
capacity and when the capacity constraint is reached, the vertical line shows the
supply schedule. Thus, at the capacity q0, there is a rupture in the supply curve.
Initially, when the demand is given by schedule D, the market clearing price is the
marginal cost (v), as at this point, there is excess capacity compared to demand. In
such a situation, the investor would recover his operating costs only. But as the
demand shifts to D0 (due to changes in income and other factors), the demand
exceeds supply if the price is maintained at the short run marginal cost (i.e. v). A
market clearing price would imply that the pricing mechanism would have to be
used to ration demand to bring it down to the available supply level, thereby
charging a price p0 [which lies between v and (a ? v)], thereby recovering a part of
the fixed cost (but not fully yet). When the demand grows sufficiently that the price
would equal (a ? v), then the producer would recover his full cost of supply. But
at this stage, entry would not be encouraged because of inadequate cost recovery in
the past. As the demand increases further and moves to D00 , the price would exceed
the long run marginal cost of supply and would provide high excessive profits to
producers. Sustained shortage of capacity, high prices and existence of excess
capacity would encourage new entry to the market.
With new capacity, the installed capacity increases to Q1, and brings excess
capacity to the system. The intersection of demand schedule D3 with the supply
curve brings the prices down to the short run marginal cost. Thus in the process the
price passes through a cycle of volatility, bringing boom and bust of the industry.
This sort of inherent price instability of the energy industry is a source of major
concern if the competitive market principle is applied strictly. Such instability
could affect long-term investment decisions of the consumers and would increase
economic uncertainties. Moreover, investors would not prefer such an environ-
ment for investment decisions. Some arrangements would be required to manage
such fluctuations.

1
This presentation follows Rees (1984). Also see Munasinghe (1985).
282 Extension of the Basic Model
12.2 12 Energy Markets and Principles of Energy Pricing
6

Fig. 12.4 Fixed plant case

D

p D

D a+v

q q0

D
D3
a+v

q0 q1

Fig. 12.5 Boom bust cycle

It is important to note here that in the literature long-run marginal cost prin-
ciples are suggested a solution in such cases. As indicated above, the relevant
pricing horizon is essentially a short-term one and the long-run marginal cost
principle encounters practical problems in determining the cost and price. Often
this requires a departure from the marginal cost principle in favour of average cost
basis of some sort.2

2
This is an area of continuous debate in the economic literature. A summary of the debate is
provided in Chap. 13.
282 Extension of the Basic Model
12.2 12 Energy Markets and Principles of Energy Pricing
7

12.2.2 Depletion of Exhaustible Resources3

As coal, oil and gas are non-renewable resources, consumption of one unit of these
resources implies foregoing its consumption at any future date. This brings in
another dimension of decision-making: whether to use the resource now or later.
The use decision is affected by choice of using it now or later. As discussed in
Chap. 9, the price should depart from the marginal cost and include an additional
item called the scarcity rent or user cost. This implies that finite resources have a
value over and above their cost of production, which is due to their scarcity. Our
time preference would require us to consume a bit more in period 1 than in period
2 but for this the price in period 1 has to be somewhat lower than that in period 2.
If the reserve is very large and if the prospect of export is negligible, the rent
component will be practically insignificant, though theoretically it will still exist.
Of the reserve is very limited, the estimation of the rent does not pose any problem
either. The difference between the extraction cost of the resource and the price of
the substitute fuel gives the rent cost. In all other intermediate cases, the rent can
be significant and its evaluation is more uncertain and complex.

12.2.3 Asset Specificity and Capital Intensiveness

The energy sector employs highly specific assets in the sense of transaction cost
economics. Assets are considered as highly specific if they have little alternative
use. For example, a power generating plant has little alternative use. Similarly,
investments made in an oil field could hardly be redeployed elsewhere in any other
use. The asset specificity can arise because of a number of reasonssite speci-
ficity, specific investments in human capital, dedicated investment (or idiosyn-
cratic investment) and physical (Williamson 1985). The level and nature of
transaction costs depend on the frequency of transaction, the extent of uncertainty
and the degree of asset specificity.
The theory of transaction costs also identifies a number of alternative
arrangements for performing transactions (Williamson 1985):
Classical contracting which includes the textbook exchanges in the market
place.
Bilateral contracting using long term contracts;
Trilateral relationship where a third party determines the damages/adaptation
following some specified procedures (such as arbitration);
Unified governance or vertical integration that internalises the transaction with
the firm.

3
Please refer to Chap. 9 for further details.
284 Extension of the Basic Model
12.2 12 Energy Markets and Principles of Energy Pricing
8

Depending on the transaction attributes it is possible to identify the governance

arrangements that would be most appropriate (see Table 12.1). In the energy
industry given the frequency of transactions and high asset specificity, the ten-
dency for vertically integrated arrangements prevailed. This was the case in all
energy industriesoil, gas, coal or electricity but there are some differences
according to the industry. In the gas industry, the trilateral contracts are more
common while in the electricity industry, unified governance prevailed.
In addition to specificity, energy sector assets tend to be capital intensive as
well. Often the capital cost accounts for a large part of the average cost and
consequently, per unit cost falls with higher sizes, showing economies of scale.
An implication of such capital intensiveness and economies of scale is that the
marginal costs tend to be low compared to the average costs and any pricing based
on marginal cost would then lead to financial losses (see Fig. 12.6). But once in
operation, as long as the firm is able to recover its variable costs, it would continue
operating expecting to make up for the capital cost recovery at a future date. Thus,
the firm would have a tendency to produce at its maximum capacity, considering
fixed costs as sunk costs. This would lead to excess supply and the energy industry
has an inherent tendency to be in excess supply situation. But continued over-
supply situation is not beneficial for the future of any industry as no new invest-
ment would be encouraged and continued financial loss could promote premature
abandonment of certain facilities.
It needs to be highlighted that a certain amount of excess capacity has to be
maintained in any energy industry to cater to the unforeseen circumstances (nat-
ural calamity, disruptions, etc.), normal demand/supply fluctuations, and to ensure
reliability of supply. Moreover, as storage is a problem for electricity, instanta-
neous supply and demand balancing is required, making the process technically
demanding as well.

Table 12.1 Governance structure for transaction characteristics

Frequency of transactions Specificity of assets
Non-specific Medium High
Rare Classical Trilateral Trilateral or unified
Frequent Classical Bilateral Unified
Source Williamson (1985)
Fig. 12.6 Relevant cost P MC
= F irm Supply Curve
curve
AC

Profit
Loss

Q
284 Extension of the Basic Model
12.2 12 Energy Markets and Principles of Energy Pricing
9

The energy industry used two approaches to manage the problems related to
indivisibility of capital and excess capacity: Oil industry used horizontal inte-
gration while the electricity and network industries used regulation. In regulation,
the tariff relates to the cost of providing the service by maintaining and operating a
certain mix of assets, including those required for ensuring reliability. However for
a non-regulated industry (like oil), horizontal integration can work. Horizontal
integration implies linking with firms at the same stage of the value chain either
through merger and acquisition or through the formation of a cartel. The oil
industry has seen significant merger and consolidation in the post oil-shock era,
where large international companies merged together to better manage their assets.
On the other hand, collusive behaviour has also been used in the oil industry to
manage the problems. The major oil companies formed an effective cartel in 1928
through the As-Is agreement and froze the respective market shares until this
policy became public and abandoned in the 1950s, as collusive behaviour is not
legally tenable in most jurisdictions. However, the Majors found another way of
influencing the marketjoint ventures in the Persian Gulf, which provided them
with a legal solution of perfect information exchange and thus control the market.
Later when the OPEC was created, the market was controlled through production
quotas and price targets in a collusive manner. But as sovereign countries are
involved in these decisions, such behaviour is not illegal.

12.3 Market Failures

The competitive market model discussed above assumes a set of strong assump-
tions. A market failure occurs when such assumptions cannot be satisfied. Some
elements of the energy sector have the technical or other characteristics that
amount to the violation of the basic assumptions of a competitive market model.
The common sources of market failure are discussed below.

12.3.1 Monopoly Problems

The capital intensiveness of the energy sector requires large investments and as
bigger installations provide economies of scale, few large suppliers tend to
dominate the market. A profit-maximising monopolist will set her price at the
intersection of marginal cost and marginal revenue. But as the monopolist faces a
down-ward sloping demand curve, the marginal revenue will be less than price.4

4
The total revenue is given by TR = P Q, where P = price and Q = output. Marginal revenue
is then dTR P Q dP , or MR = P(1 ? 1/e), where MR = marginal revenue and e =
price dQ dQ
elasticity of demand. As e is less than 1, MR is less than P.
286 12 Energy Markets and Principles of Energy Pricing

Fig. 12.7 Price $

determination in a monopoly
A
market

B
Pm

AC=MC
Pc C D

MR Demand

Qm Qc Q

As shown in Fig. 12.7, the profit maximising output is given by Qm, while the
price charged by the monopolist is Pm.
Mathematically, the problem to maximize the profit
Max p D p CD p 12:5
The first order condition is

p D0 p Dp C0 Dp D0 p 0 12:6

p C0 Dp =D0 p 12:7
But

e D0 p p=Dp 12:8
Hence,
p C0
1
12:9
p e
That is, to maximize its profit, the monopolist will charge consumers inversely
to their elasticity of demand. Inelastic the demand, higher the price will be.
If the monopoly results are compared with the competitive outcome, it is found
that the monopolist restricts the output to Qm compared to Qc obtained in the
competitive market. Similarly, the price paid by the consumers is Pm compared to
Pc in a competitive condition. Thus the consumers pay Pm-PC as monopoly rent.
The consumer surplus is reduced to APmB compared to APcD whereas the pro-
ducer surplus increases to PmPcCB which was non-existent in a competitive set
12.3 Market Failures 287

Fig. 12.8 Economies of $

scale
AC

Cost is
subadditive
for Q<Q

Q
Q

up. Monopolisation of the market leads to a deadweight loss equivalent to the

triangle BCD.
In addition, following Leibensteins observation (Leibenstein 1966), it could be
indicated that a monopolist might operate in the inefficient zone of the production
possibility frontier. This in other words means that a monopolist may choose the
factors of production in an inefficient manner, thereby operating at a point above
its theoretical cost curve. This is known as X-inefficiency. While X-inefficiency is
inconsistent with profit-maximising behaviour, the inefficiency is possible given
that managers may pursue their own objectives in place of owners objective of
profit maximisation.
A third source of monopoly-related inefficiency is the possibility of rent-
seeking. A monopolist by charging more than the competitive market price earns a
monopoly rent, which is equal to the producer surplus of PmPcCB in Fig. 12.8.
The existence of such rent will set in competition among firms to seek the rent by
lobbying and influencing the legislators or regulators, thus wasting resources and
causing welfare loss to the society.
A fourth source of complication is the possibility of product differentiation,
which allows a firm to increase its price without loosing all its sales to a com-
petitor. In such a case, the price exceeds marginal cost, which signals misallo-
cation of resources.

12.3.2 Natural Monopoly

This is a situation where production of a good or service by a single firm ensures

least cost supply. The typical example is a single product where the long-run
average cost declines for all outputs (see Fig. 12.9). As the average cost falls over
the entire range of output, the marginal cost also falls. This is a case of permanent
natural monopoly, because irrespective of market demand size, a single firm can
produce the good at least cost.