Developing Marginal Cost-Based Rates

1
Developing MarginalCost-Based Rates
Kelly Eakin Senior Vice President Christensen
Associates Energy Consulting APPA Business and
Financial Conference Austin, TX September 25,
2007
2
Objectives of Presentation

• Provide a framework to evaluate social gains from
  marginal cost pricing and address the challenge
  of fixed cost recovery
• Identify key determinants to marginal cost
  pricing gains
• Look for ways to incorporate pricing efficiency
  principles into cost of service analysis

3
Business Objectives

• Revenue Sufficiency
• Maximizing stakeholder value

4
Outline

• Economics Basics
• The Regulatory Dilemma
• Incorporating Marginal Cost Pricing Principles
  into Cost of Service Analysis
• A Simple Stylized Example
• Conclusions

5
Economics Basics
6
Demand

• Consumers are willing to pay for a good because
  it brings them some benefit or satisfaction
• As more of a good is consumed, the additional or
  marginal benefit decreases (law of diminishing
  returns)
• Consequently, the consumers willingness to pay
  for another unit of a good decreases as
  consumption increases
• Law of Demand Consumers will buy more of a good
  a lower prices, other things the same
• Price Elasticity of Demand
• A measure of customer price responsiveness
• eD change in quantity demanded change
  in price

7
The Demand Curve
Price
P1
P2
D marginal benefit
Q1
Q2
QuantityDemanded
8
Costs

• Costs reflect the supply side of a market
• For goods to come to market, a supplier needs to
  expect at least to recover (variable) costs

9
Cost Measures (1)

• Total Costs
• Variable Cost costs associated with the inputs
  that change as production levels change
• Fixed Cost costs that remain the same
  regardless of the production level, sometimes
  called sunk costs (e.g., capital costs)
• Total Cost Variable Cost Fixed Cost or
  TCVCFC

10
Cost Measures (2)

• Average Cost Concepts
• Average Variable Cost (AVC)
• Variable cost per unit of output AVCTVC/Q
• Average Fixed Cost (AFC)
• Fixed cost per unit of output AFC/Q
• AFC decreases as output increases (spreading out
  the overhead)
• Average Total Cost (ATC)
• Cost per unit of output
• ATC TC/Q
• ATC AVC AFC
• The utility industry term average embedded
  cost corresponds closely to the economic term
  average total cost

11
Cost Measures (3)

• Marginal Cost (MC)
• Marginal cost measures how cost changes as an
  additional unit of output is produced
• MC ?TC/?Q ?TVC/?Q
• Marginal cost is the supply schedule for a
  competitive profit-maximizing firm
• A supply schedule is more ambiguous if there is a
  lack of competition or the firm is not a profit
  maximizer

12
Cost Measures (4)

• Fixed Costs
• Common costoverhead costs that occurs regardless
  of product lines offered, production levels or
  customer classes served
• Class-specific fixed costcosts that do not vary
  with production levels but that are avoidable if
  a customer class is not served
• Product-specific fixed costcosts that do not
  vary with production levels but that are
  avoidable if a product line is not produced
• Fixed cost recovery can introduce price
  distortion and resulting social value loss
  (called economic inefficiency)

13
Cost Measures (5)

• Incremental Cost (ICi)
• Incremental cost indicates the additional cost of
  adding a product line or serving another customer
  class
• Subtle differences from marginal cost
• Discrete change instead of incremental change
• May involve some product/class specific fixed
  (but avoidable) costs
• ICi TC TC without Qi TC TC( Qi)
• Average Incremental Cost (AICi)
• AICi ICi/Qi
• Important concept in investigating cross-subsidies

14
Cost Concepts (6)

• Finally, pulling in some cost of service concepts
• Attributable costs (or directly assigned costs)
  are those costs that can be assigned as caused
  by serving a customer class
• Variable costs
• Class-specific fixed costs
• Product-specific fixed costs for products serving
  only one customer class
• Non-attributable coststhose costs that occur
  regardless of whether a particular customer class
  is served
• Common costs
• Product-specific fixed costs for products serving
  all customer classes

15
Cross-Subsidization

• Charging some more than attributable cost so that
  others pay less than their attributable cost
• Different criteria for cross-subsidization
• Price vs. Average Total Cost
• Price vs. Average Variable Cost
• Price vs. Marginal Cost
• Price vs. Average Incremental Cost
• Cross-subsidization involves
• Inefficiency
• Fairness issue

16
Cross-Subsidization

• Pi lt AICi
• Revenue from class less than its incremental cost
• Serving the class adds to the overhead
  contribution required from other classes
• Other classes would pay less if this class were
  not served
• The class is receiving a cross-subsidy from other
  classes
• Pi AICi
• Revenue just covers incremental cost but class
  makes no contribution to overhead
• No impact on other classes
• No cross-subsidy
• Pi gt AICi
• Revenue from class more than its incremental cost
• the class makes a contribution to overhead
• Other classes would pay more if this class were
  not served
• No cross-subsidy

17
Economic Efficiency

• Economic efficiency occurs when resources are
  used in a way that generates the greatest
  economic value
• Price Marginal Cost is the efficiency condition
• P gt MC too little produced additional value
  would exceed additional cost
• P lt MC too much produced additional cost of last
  unit more than offsets additional value
• PMC maximum net benefit no way to reallocate
  resources to increase economic value

18
The Efficiency of Competition
Firm
Industry
ATC
MC
S MC
P
P
D MB
q
Q nq
19
The Efficiency and Fairness of Competition

• Efficiency
• P MC ? economic output distributed to
  consumers in a manner that achieves the greatest
  economic value
• MC ATC ? production is allocated among
  producers so that total production cost is
  minimized
• Fairness
• P ATC ? producers just break even covering
  their variable costs and earning a fair rate of
  return on capital investment also called earning
  normal profit of zero economic profit
• This condition is the result of no barriers to
  entry or exit
• The result of individuals pursuing
  self-interest, but the outcome is as if an
  invisible hand of a benevolent planner
  allocated the resources

20
The Invisible Hand
21
The Regulatory Dilemma
22
Alas, competition and efficiency may not prevail

• Industry might not be competitive
• Barriers to entry
• Large economies of scale
• Firms might not be profit-maximizers
• Public power has stakeholders rather than
  shareholders
• Other non-profit organizations have objectives
  other than maximizing profit
• Revenue adequacy still a requirement, but
  efficiency may not be a result

23
Natural Monopoly

• Economies of scale exist if average cost
  decreases as a firms production increases
• A natural monopoly has economies of scale over a
  large range of production relative to market
  demand
• One firm can produce market output at a lower
  total cost than can two or more firms
• Often have large overhead costs resulting from
  heavy capital investment (i.e., capital intensive
  industries)
• Often involve basic needs such as water,
  electricity

24
Natural Monopoly
25
Rationale for Regulation

• Promoting competition is inefficient in a natural
  monopoly situation
• Instead rate regulation is the policy
  prescription
• Called public utility regulation
• Trying to achieve competitive-type (invisible
  hand) outcomes via regulation

26
The Not-So Invisible Hand of Regulation
Intervenor Discovery
gt
Staff Intervenors Prefile Testimony
Staff Intervenors Present Witnesses
Company Presents Witnesses
File Case
Company Files Rebuttal
Cross Examination on Rebuttal
Submission of Briefs
Commission Decision
Commission Order
gt
gt
27
The Natural Monopoly Dilemma

• A natural monopoly presents the regulator with a
  dilemma
• Left unregulated, the monopolist could charge
  high prices resulting in inefficiency and a
  transfer of wealth from customers to the
  monopolist
• Setting PMC results in efficiency but
  insufficient revenues
• Set PATC collects sufficient revenues but is
  inefficient
• Issue becomes more complex with multiple products
  and customer classes

28
The Natural Monopoly Dilemma PMC is efficient
but revenue inadequate

ATC
ATC
Losses
MC
P
D
Q
Q
29
The Natural Monopoly Dilemma PATC collects
enough revenue but is inefficient

Efficiency Loss
ATC
P ATC
MC
D
Q
Q
Q
30
Solutions to the Pricing Dilemma

• Simple markup pricing
• Absolute markup raise prices above marginal
  costs by the same amount to all classes
• Proportional markup raise prices above marginal
  costs by the same percent to all classes

31
Solutions to the Pricing Dilemma (2)

• Differential markup pricing
• Raise prices above marginal costs by different
  percentages to different classes
• Ramsey Pricing raises prices differentially to
  minimize inefficiency
• Price inverse to the elasticity of demand
• Same pattern as what a monopolist would do, only
  to lesser magnitude

32
Solutions to the Pricing Dilemma (3)

• Non-linear pricing
• Collect some fixed costs via a non-volumetric
  charge (i.e., not per kWh or per kW)
• If all fixed costs were collected
  non-volumetrically, then per unit charge could be
  set at marginal cost

33
Incorporating Marginal Cost Pricing Principles
into Cost of Service Analysis Cost of Service
Basics
34
Why investigate cost of service?

• Improve understanding of the business
• Help with rate design
• Requirement for rate case

35
Basic Steps for a Traditional Cost of Service
Study

• Determine the Overall Revenue Requirement
• Establish the customer classes
• Attribute the attributable costs
• Allocate the non-attributable costs
• Set rates to achieve the revenue requirements

36
A Modified Approach Marginal Cost-Based Cost
of Service

• Determine the Overall Revenue Requirement
• Establish the customer classes
• Attribute the attributable costs
• Set preliminary prices at marginal costs
• Conduct preliminary cross-subsidy analysis
• Mark up prices to subsidized classes to eliminate
  cross-subsidies
• Calculate revenue insufficiency
• Mark up prices to all classes to achieve revenue
  requirement
• Introduce revenue-neutral non-linear pricing to
  improve pricing efficiency

37
Challenges of the New Approach

• Estimating the marginal costs
• Costly to make numerous unbundled marginal cost
  estimates
• Estimating marginal costs of ancillary services,
  transmission and distribution services is not
  trivial
• Nevertheless, these marginal costs should be
  understood for business reasons beyond
  cost-of-service studies
• Incorporating demand response into cost of
  service analysis
• Essential for the pursuit of efficiency
• Not a comparative disadvantage vis à vis
  traditional approach
• Attributable costs may be a small fraction of
  total costs

38
Advantages of a Marginal Cost-BasedCost-of-Servic
e Approach

• Based on principles of economic efficiency
• Knowledge of marginal cost has many useful
  business purposes separate of a cost-of-service
  study
• Less arbitrary allocation of costs
• May decrease the extent of cross-subsidization

39
Simple Stylized Example
40
Three Customer Classes
Customers Additional Sites Average Customer Annual Usage (kWH)
Residential 1,000,000 0 10,000
Commercial 50,000 10,000 100,000
Industrial 1,000 0 1,000,000
41
Marginal Costs(Assuming constant marginal cost)
Customers (annual) Additional Sites (annual) Energy (/kWH)
Residential 100 - 0.10
Commercial 1,000 500 0.06
Industrial 5,000 - 0.04
42
Cost Structure(Millions)
Fixed Costs Customer Cost Extra Site Costs Energy Cost TOTAL
Common Cost 400 - - - 400
Residential 80 100 - 1,000 1,180
Commercial 15 50 5 300 370
Industrial 5 5 - 40 50
TOTAL 500 155 5 1,340 2,000
43
Approaches to Allocating Common Cost

• Method A Divide total fixed costs by total
  usage
• Ignores that some fixed costs are attributable
• Ignores that marginal costs differ across classes
• Allocation implicitly achieved by charging same
  price per kWh across all classes
• Method B Assign attributable fixed costs and
  allocate common cost on a per kWh basis
• Method C Assign attributable fixed costs and
  allocate common cost according to shares of
  attributable costs

44
Alternative Allocations of Common Cost
Method A Method B Method C
Residential 70 250 295
Commercial 255 125 93
Industrial 75 25 13
TOTAL 400 400 400
45
Cost Recovery Through Energy Pricing Only(/kWh)
Method A Method B Method C
Residential 0.125 0.143 0.148
Commercial 0.125 0.099 0.093
Industrial 0.125 0.075 0.063
46
Incorporating Demand Response

• Assume the following price elasticities of demand
  (eD)
• eD
• Residential -0.05
• Commercial -0.10
• Industrial -0.20
• Also assuming linear demand curve and using
  Method B prices and stipulated usage as the
  reference point

47
Recall the Efficiency Loss Triangle from Pricing
Away from Marginal Cost

Efficiency Loss
ATC
P ATC
MC
D
Q
Q
Q
48
Economic Efficiency Analysis

• The efficiency loss or deadweight loss (DWL) in
  each market can be approximated as
• DWL -½ eD (Q/P) (P-MC)2
• Pricing inefficiency requires both
• Price differing from marginal cost
• Existence of price responsiveness (eD?0)
• Determinants of the magnitude of value loss
• Amount of fixed cost to be recovered
• Size of the price distortion
• Marginal cost
• Price responsiveness

49
Efficiency Loss from Markup Pricing(Millions)
Method A Method B Method C
Residential 1.09 3.23 3.94
Commercial 10.67 3.84 2.67
Industrial 9.63 1.63 0.68
TOTAL 21.40 8.71 7.29
50
Efficiency Loss as a Percentage of Class Revenue
Method A Method B Method C
Residential 0.09 0.23 0.27
Commercial 1.71 0.78 0.58
Industrial 7.71 2.18 1.08
TOTAL 1.07 0.44 0.36
51
Non-linear Marginal Cost Prices
Monthly Customer Charge Monthly Additional Site Charge Energy Charge (/kWh)
Residential
Method A 20.83   0.10
Method B 35.83   0.10
Method C 39.58   0.10
Commercial
Method A 533.33 41.67 0.06
Method B 316.67 41.67 0.06
Method C 262.50 41.67 0.06
Industrial
Method A 7,083.33   0.04
Method B 2,916.67   0.04
Method C 1,875.00   0.04
52
Non-linear Marginal Cost Prices

• The price schedules listed in the preceding table
  are all economically efficient
• The fixed charges are rolled into the customer
  charge
• The customer charge may be too steep to for some
  customers within a class
• For those classes can reduce the customer charge
  and increase the energy charge
• Use the marginal cost pricing as starting point
  for adjustment
• Inefficiency introduced by raising the energy
  charge might not be great for very inelastic
  demand

53
Conclusions
54
Foundation of a Foundation

• Cost of Service analysis provides a necessary
  foundation for a rate case or rate setting
• Marginal cost pricing principles provide a
  foundation for the COS foundation
• Structurally sound revenue recovery without
  cross-subsidies
• Level achieve competitive-type efficiency
• Earthquake proof framework that allows easy and
  appropriate response to competitive threats and
  other market dynamics

55
Summary

• Marginal cost pricing principles and
  cost-of-service objectives are compatible
• Simultaneous pursuit of revenue sufficiency and
  maximum stakeholder value
• Recovery of fixed cost is the big challenge
• Recovery of fixed cost via fixed charges the most
  efficient
• High fixed charges may not be feasible for some
  customer classes

56
Summary

• Being diligent on figuring out all variable costs
  reduces the fixed cost recovery burden
• Getting prices aligned with marginal cost is more
  important for classes with greater price
  responsiveness, which also tend to be the
  competitive at-risk customers
• Foundation of a Foundation

57

• QUESTIONS AND COMMENTS
• kelly_at_CAEnergy.com