Transcript Post-lecture version, including scribbles

BSc Economics and related
programmes
Economics of Competition and
Regulation EC 3015
Week 9: Costs in Price Cap
Regulation
Cost components
Cost drivers
Efficiency analysis
Main question
How do we work out financing requirement
for a regulated natural monopoly?
=> we look at :
• cost components and basics
• cost drivers
• efficiency analysis
Costs in RPI-X regulation
2
BASICS
Costs in RPI-X regulation
3
Cost components and basics
Cost drivers
Efficiency analysis
“Building blocks” of cash flow model
• Operating expenditures
– Maintenance, repairs, operation, customer
transactions, central operations
• Capital expenditures
• Initial and closing regulatory asset value
• Future values discounted at cost of capital,
avoiding accounting rates of return
• These are forward-looking costs, so costs need
to be modelled
Costs in RPI-X regulation
4
Cost components and basics
Cost drivers
Efficiency analysis
Modelling costs
• Starting point is existing costs.
– Are they “correct”?
– Are they temporarily high or low for specific
reasons? => initial adjustments
• Future costs depend on future events
– Need for forecasting elements
– Especially “cost drivers”
Costs in RPI-X regulation
5
If the average costs of delivering letters
falls when more letters are delivered to
each address this reflects...
48%
1. An economy of
network scale
2. An economy of
scope
3. An economy of
density
4. None of the above
38%
10%
bo
ve
sit
y
ea
th
of
No
ne
ec
on
om
yo
yo
An
ec
on
om
fd
en
fs
co
pe
...
or
k
w
An
fn
et
yo
ec
on
om
An
Market Investigations
5%
6
The difference between cost of
service and price cap is...
1. Price cap adjusts for
inflation
2. Price cap has a longer
regulatory lag
3. Cost of service uses
accounting rate of
return, but price cap
uses DCF
4. All of the above
5. None of the above
63%
26%
0%
bo
ve
of
th
ea
ab
ov
No
ne
of
t
he
o.
..
Al
l
us
es
ac
c
vic
e
fs
er
Co
st
o
ca
p
ha
s
a
lo
ng
er
re
...
in
fl.
..
fo
r
ad
ju
st
s
p
Pr
ice
ca
Pr
ice
Market Investigations
5%
e
5%
7
THE FACTORS THAT DRIVE
COSTS
Costs in RPI-X regulation
8
Cost components
Cost drivers
Efficiency analysis
Cost drivers: generic
Efficiency
improvements
Demand
growth
Financial
environment
Improved quality
Environmental
regulation
Risk
Operating Expenditures
Capital Expenditures
Revenue requirement
Costs in RPI-X regulation
Cost of capital
Arrows to show main
effects
9
Cost components
Cost drivers
Efficiency analysis
So let’s look at:
• Demand growth
• Quality and environmental regulation
• Efficiency
Costs in RPI-X regulation
10
Cost components
Cost drivers
Efficiency analysis
Demand growth
• Assumptions based on:
– Econometric analysis esp. income elasticities
– Surveys
– Other forecasts
• Impact on:
– New Capital expenditure
– Fixed operating cost
– Variable operating cost
Costs in RPI-X regulation
11
Cost components
Cost drivers
Efficiency analysis
Specific cost drivers: electricity
distribution – impact of growth
Electrical system costs:
Optimal layout given demand patterns:
– number and distribution of customers
– maximum demand at various points
– provision for responding to faults, repairs, damage, etc.
– deviations of actual from optimal:
» growth, churn, etc.
Non system costs:
Billing, finance, regulation:
– some fixed elements, other related to customer numbers
Some of the above are related to number of customers,
some to demand or network complexity/length, some to
overheads
Costs in RPI-X regulation
12
Cost components
Cost drivers
Efficiency analysis
Quality
• The produce itself:
– Water purity,
– compliance of electricity with standards
– Gas calorific value (Wobbe Index), water content, etc.
• The quality of service:
– Response times
• Problems, billing, etc
– Interruptions to service
• Frequency, length, warnings, compensation
• Some related to environmental considerations:
Costs in RPI-X regulation
13
Cost components
Cost drivers
Efficiency analysis
Environmental regulation
• Examples:
– CO2 and SO2 from power stations
– Water discharged from sewage treatment
stations
– Pesticide and nitrates in drinking water
Costs in RPI-X regulation
14
Which of the following is not a
building block in RPI-X?
1. Operating cost
2. Capital expenditure
3. Allowed return on
RAV
4. Annual depreciation
5. All of the above
6. None of the above
52%
38%
10%
0%
0%
le
Costs in RPI-X regulation
Al
lo
ita
Ca
p
Op
e
ra
tin
gc
os
t
xp
e
w
nd
ed
itu
re
re
tu
rn
An
on
nu
RA
al
de
V
pr
ec
ia
Al
tio
lo
n
ft
h
e
No
ab
ne
ov
of
e
th
ea
bo
ve
0%
15
Which of the following is not a building
block in cost-of-service regulation?
38%
Operating cost
Capital expenditure
Allowed return on RAB
Annual depreciation
All of the above
None of the above
29%
24%
5%
5%
le
Costs in RPI-X regulation
Al
lo
Ca
p
ita
ra
tin
gc
os
t
xp
e
w
nd
ed
itu
re
re
tu
rn
An
on
nu
RA
al
de
B
pr
ec
ia
Al
tio
lo
n
ft
h
e
No
ab
ne
ov
of
e
th
ea
bo
ve
0%
Op
e
1.
2.
3.
4.
5.
6.
16
The Averch-Johnson effect
33% 33%
1. is designed to ensure
that firms earn a fair and
reasonable return
2. means that opex is
above the optimal level
3. does not apply to oneshot price caps
4. all of the above
5. none of the above
14%
bo
ve
ov
e
th
of
no
ne
of
th
e
ea
ab
sh
..
all
to
pl
y
sn
ot
ap
10%
on
e-
ab
o
is
ex
op
ha
t
ea
do
e
is
Market Investigations
m
de
s
ign
ns
t
ed
to
en
su
re
th
a
v..
.
t..
.
10%
17
EFFICIENCY OF COSTS
Costs in RPI-X regulation
18
Cost components
Cost drivers
Efficiency analysis
Efficiency
Requirement is usually to allow an efficient
firm to finance its activities
What if a firm is inefficient?
(and what do we mean by this anyway?)
Productive Efficiency =(Output/input)actual
(Output/input)potential
=> need concept of frontier
Costs in RPI-X regulation
19
Cost components
Cost drivers
Efficiency analysis
The frontier
• Minimum possible costs, given the cost
drivers
(The position of the frontier depends on the
level of prices. Technical change may move
the frontier down.)
Cost
Empirical frontier
Theoretical frontier
Cost driver
Costs in RPI-X regulation
20
Cost components
Cost drivers
Efficiency analysis
Investigating the frontier
• Engineering (bottom up) investigation of
practices:
– intrusive, subjective
• Comparative analysis
– limited by paucity of observations
• Transco, NGC: no real comparators
– but some use of inter-zonal comparisons in gas distribution.
• Electricity distribution companies: 14 observations per year
• International benchmarking? Difficulties
• (Make deductions about relative importance of
cost drivers from foreign studies?)
Costs in RPI-X regulation
21
Cost components
Cost drivers
Efficiency analysis
Types of comparative data analysis
• Simple cost ratios
• Regression analysis
• Data envelope (DEA) and other frontier techniques
such a stochastic frontier analysis
Combination of cross section and time series? (panel
data)
Some scope for international comparisons, limited by
data definition issues.
Costs in RPI-X regulation
22
Cost components
Cost drivers
Efficiency analysis
Example of econometric analysis of costs: postal delivery services
Dep var = log(delivery
costs)
Stochastic
Frontier
OLS
(Halfnormal)
Variable (in logs
usually)
Coeff
t-ratio
Coeff
t-ratio
Constant
-2.78
-6.71
-2.63
-10.45
Wage rate paid
1.09
10.69
1.03
10.67
Local Wage level
0.12
1.58
0.1
1.36
Volume/Delivery point
0.67
18.15
0.66
34.32
No delivery points (log)
1.02
51.52
1.01
135.76
Road length per
delivery point
0.08
5.66
0.08
5.89
DELZONE1
-0.1
-1.28
-0.1
-1.95
DELZONE2
-0.13
-1.99
-0.13
-2.58
DELZONE3
-0.1
-1.57
-0.09
-2.19
DELZONE4
-0.11
-1.88
-0.11
-2.87
Business delivery
points
0.11
7.34
0.1
7.43
REDIRECTIONS
0.03
1.59
0.04
12.41
Costs -0.002
in RPI-X regulation
-1.41
-0.001
FRAMES
-1.29 23
Cost components
Cost drivers
Efficiency analysis
Postal services findings
Delivery costs:
• savings of £220m or 10.9% of relevant
costs
Mail Centre costs
• savings of around £100m or 14% of
relevant costs
Costs in RPI-X regulation
24
Cost components
Cost drivers
Efficiency analysis
Example 2: Distribution costs
Costs in RPI-X regulation
25
Cost components
Cost drivers
Efficiency analysis
Composite output 1999
Component
Customers
kWh
Network length
Relative weight
 1.00
 0.25
 0.25
(Arrived at through a process of consultation, not econometrics:
too few observations)
Costs in RPI-X regulation
26
Cost components
Cost drivers
Efficiency analysis
Engineering analysis 1
Electricity example of “bottom up” approach:
In order to assess the potential savings available to each PES,
a number of techniques were applied as follows:
— a cost per network kilometre benchmark of £575 per km,
based on costs from four "top" PESs;
—"engineering cost" based on a profile of its network assets
using a best practice cost per asset;
—comparison of historic savings achieved -- four of the top
PESs achieved savings in engineering costs of up to 40 per cent
from 1994/95 to 1997/98: in addition, the extent of savings in
costs from 1990/91 to 1994/95 was also considered;
—a review of each PES’s engineering organisational
structures, field efficiency and operating practices;
Costs in RPI-X regulation
27
Cost components
Cost drivers
Efficiency analysis
Engineering analysis 2
1) methods of cost reduction in past
2) plans for future
Examples:
• new terms and conditions of employment
• increased condition monitoring of assets
• staff multi-skilling
=> range of estimated savings feeding into targets
Costs in RPI-X regulation
28
Cost components
Cost drivers
Efficiency analysis
Movement of the frontier
• Total factor productivity analysis
– compares movements of outputs and of
inputs
– long term trend
– Energy industry plus other “similar” industries
– Overseas industries
TFP = Δlog (Q-M) - Δαlog K - Δβlog L
Q-M = Value added
K =capital input, α = share of capital in output
L = labour input, β = share of labour in output
Costs in RPI-X regulation
29
Cost components
Cost drivers
Efficiency analysis
Key issues for operating costs in
RPI –X (1)
• How much productivity gains for the whole
sector?
• How much weight to put on “efficiency" findings?
–How much of efficiency gap to be made up?
• How quickly should companies approach frontier?
Costs in RPI-X regulation
30
Cost components
Cost drivers
Efficiency analysis
Key issues for operating costs (2)
• How long should companies be allowed to keep
efficiency gains?
– P0 versus X
– five year profile issues
p0
X
-1
0
1
2
3
4
Informed by analysis of past reviews
How well did companies forecast?
How far did they all surpass targets?
How well did efficiency scores predict efficiency gains?
Costs in RPI-X regulation
31
Cost components
Cost drivers
Efficiency analysis
Summary
• As with so many aspects of competition
and regulation, the key to cost modelling is
finding an objective basis for decisions
• Economics has many tools that are used
in order to help
Costs in RPI-X regulation
32