Client Logo

SingaporeJune 10 – 14 2008

STRATEGIES to ACHIEVE CONNECTIVITY& CONVERGENCE : Executive Course on Telecom Reform

The Use of Benchmarks in RegulationWhat Operators & Regulators Need to Think About

Helani Galpaya - galpaya@lirne.net

1

Indicators – useful. A lot available.

�Telco sector has LOTS of indicators

� Indicators Used all the time

– To analyze historical performance

– E.g. India’s famous graph

0.88 1.20 1.88 3.586.50

13.00

33.60

52.17

90.00

166.00

0

2

4

6

8

10

12

14

16

18

Mar-98 Mar-99 Mar-00 `Mar-01 `Mar-02 `Mar-03 `Mar-04 `Mar-05 `Mar-06 `Mar-07

Eff

ec

tiv

e t

ari

ff (

Rs

./m

in)

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

160.00

180.00

Mo

bil

e S

ub

sc

rib

er

ba

se

(m

Fixed Tariff Cellular Tariff M obile Subs

NTP-99

Telecom

Tariff Order

3rd & 4th

cellular

operator

WLL

Introduced

CPP

introduced

Lowering of

ADC from 30%

to 10% of

sector revenue

2

But COMPARATIVE INDICATORS (= BENCHMARKS) are most useful

�You may think you are doing well until you compare yourself against others

�BENCHMARK: a comparison

– The same indicator, compared across countries, organizations, regions, etc.

Total number of mobile subscribers in India

and Bangladesh

0

50

100

150

200

250

2002 2003 2004 2005 2006 2007

Year

In m

illi

on

s

India Bangladesh

Mobile subscribers per 100 in India and

Bangladesh

0

5

10

15

20

25

2002 2003 2004 2005 2006 2007

Year

Pe

r 1

00

India Bangladesh

3

But can we do more than “after-the-fact”benchmarking?

�2 possible examples

– Price

– Quality

4

1. Price regulation – in the light of imperfect market conditions

5

What the conditions for perfect/well functioning markets?

�Perfect Information

�No barriers to entry and exit

�No Market power

– Multiple buyers and sellers

�Substitutable products

�Rational market players

6

In such markets, prices set by interaction of supply and demand.Role of regulator could be minimal

� If enough suppliers and buyers

– Harder to collude

– Suppliers set rational prices – i.e. prices that reflect costs

– Price competition occurs

– Usually prices drop

�Regulators can often step back

�E.g. [though not perfectly competitive] India has highest level of competition

– Each region/circle (not just country) has seen increase competition

– Widely perceived to have lowest prices in the region

� Indian Regulator (TRAI) has forborne from tariff regulation (for the most part)

�Rewarded with highest TRE scores

7

e.g. India has high competition with many players.

8

Forbearance in Tariff Regulation seen positively by stakeholders

2.8

3.7

2.6 2.7

2.92.9

3.9

2.8 2.9 2.9

2.22.3

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Pakistan India Indonesia Sri Lanka Philippines Thailand

Fixed Mobile

Highest Fixed

& Mobile

TRE for Tariff Regulation

9

But telecom markets are NOT perfect markets.

�Significant Information Asymmetries

– Certain players have more information than others.

� Access to scarce resources a barriers to entry

– E.g. rights of way to cable, towers, spectrum

� Increasing returns

– Large chunks of investment

– Essential facilities

�Ability to extend market power from one market to another

– E.g. fixed to mobile

10

Therefore regulatory intervention in price setting through various tools/methods….e.g. Rate of Return Regulation

�Regulates (limits) the PROFITS

�Find out costs

– Prudently incurred expense.

– Actual.

– Not forward looking (past for accounting period)

�Add a reasonable/”fair” rate of return

– based on weighted avg. cost of capital

�Determine Revenue Requirement

– a function of operating expenses, depreciation, taxes,. Book value of capital assets, Rate of Return

�Set prices so that

– Sum of expected revenue from all services <= Revenue Requirement

11

…and Price Cap Regulation…and other variations on it

�Regulates (limits) PRICES

�Regulator determines formula

�Tells how much prices must change in each period (year)

�Typically, allowed revision = CPI – x

– CPI = Consumer Price Index/Inflation

– x=efficiency factor

�So allowed new price = previous price * (1+ CPI-x)

�Other variations on these

12

These techniques often have perverse incentives and very resource intensive to implement

�E.g. ROR Regulation:

�Determining cost (e.g. used and useful) difficult

– E.g. CEO’s holiday bungalow vs. cost of new switching equipment

– Who has better information about such cost? NOT the regulator

�Lack of uniform system of accounts

– E.g. rules on depreciation of assets: unclear

�No incentive to lower costs/become efficient

– Any increased profits would be taken away by regulator

�Requires rate-rebalancing regularly

– Very resource heavy

– Not useful for today’s fast changing telecom environment

13

�E.g. Price Cap Regulation

� Incentive to lower costs

– Get to keep any extra profit (since price capped, but profits are not)

�But calculating X difficult.

– Doesn’t work for high-inflation countries

– e.g. LK CPI ~27%. Some licenses define x = 2%.

– Means prices may actually increase 25%!

14

The resource problem might be solved by Asymmetric Regulation..but that too has problems

�Asymmetric Regulation regulates dominant operator (~ the one with SMP)

– They have to file tariffs and obtain approvals

�But de-regulates all other players

– They can do what they like

– Their tariff’s do not require approval/filing

�Solves the resource problem only partially

– Dominant operator still has to be regulators

– How? Same difficulties as before (resource heavy. perverse incentives).

�AND end up with a very unhappy Dominant Operator

�Needs a high level of competition to work

– Not useful in oligopoly if the few competitors are shadows of the dominant operator

15

Micro-states often unlikely to get conditions of perfect (or even workable) competition

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

Com

oro

sB

aha

mas

Bhut

an

Cape V

erd

eD

ijibout

i

Equa

torial G

uin

ea

Fre

nch P

oly

nesia

Kirib

ati

Mars

hall

Isla

nds

Mayo

tte

Mic

ronesi

a, Fed. Sts

.

New

Cale

donia

São

Tom

é a

nd P

rincip

e

Solo

mon Is

lands

Vanu

atu

Barb

ados

Bru

nei D

aru

ssa

lam

Dom

inic

aG

renad

aG

uam

G

uyana

Mald

ives

Malta

Sain

t K

itts a

nd N

evi

s

Sam

oa

St. L

ucia

St. V

ince

nt and

the G

rena

din

es

Tonga

Icela

nd

Seyc

helle

sM

aca

o

Neth

erlan

ds A

ntille

s

Siz

e o

f econom

y G

DP

in U

S $

millions

1 operator

2 operators

3 o

pe

rato

rs

4 o

pe

rato

rs

5 o

pe

rato

rs

16

Fewer competitors and regulatory agencies with limited resources. Requires a new approach?

�Unlikely to have conditions of workable competition

– Smaller number of competitors supported by market

– Even though no monopoly/duopoly

�Therefore even Asymmetric Regulation may not be feasible

�Smaller regulatory agencies

– Can’t do resource intensive regulation.

– Need to allocate scarce resources to priorities

�E.g. ensure good Interconnection conditions vs. approve/regulate every tariff?

17

LIRNEasia’s proposal: Banded Forbearance. Based on using BENCHMARKS for regulation

�General principle of Benchmark regulation

– make regulatory decisions based on comparison with others

�1. Pick the indicator

– Fixed/Mobile: A “modified” OECD price basket

– Internet: Monthly price of selected service plan

�2. Identify the peer group

– Regional: often culturally similar, belong to same regional organizations

– Economic: similar ability to pay, similar levels of development

– Demographic: similar population numbers (particularly relevant for microstates?)

– Geographic: island nations; land-locked countries

– Other

�3…contd.

18

�3. Define benchmark period

– Every quarter/every year etc

�4. Definite the Band (upper and lower limit)

– Peer average +/- x%

– Lower limit = lowest peer; Upper limit = highest peer

– Other variations

Benchmark country / region Country being regulated

Pri

ce

p

p - x%

p + x%Benchmark

region

19

After these things are defined:

�Anything goes within the band

– Simply inform the regulator

– But don’t require approval

�All new tariff plans MUST fall within

– Else not approved; plan cannot be introduced to market

� If plans go outside band:

– Investigated by regulators

– Lower than band: possible anti-competitive practices ?

– Higher than band: problems with cost structure?

20

Advantages of Banded Forbearance

�Less resource intensive

– If band is defined well, most players stay within band most of the time

– Fewer investigations, faster (automatic) approvals

�Operators have certainty

– Rules known before hand

– Able to check themselves if price within band

– Easier planning. Less unknowns

�Can be applied to ALL operators, including SMP

– Essentially de-regulates incumbent/dominant player

– But provides safeguards

21

2. Another possible regulatory tool using Benchmarking: Broadband QoS

22

The traditional way of Regulating Quality of Service (if at all)

� Issue guidelines (the regulators)

�Wait for operators to measure and report to regulators

– In PSTN: “average wait time for a connection”, “% of faults fixed within X hours”

– In Broadband, historically not much QoS regulation. But lately some “truth in advertising” or minimum upload/download speed specifications

�Reported to regulator annually, quarterly etc

�Any penalty for not meeting specified standards well after offense

– By this time problem may have been rectified

�No incentive for operator to improve performance immediately

�Little/no visibility on QoS indicators to consumers – even if regulator releases the data

23

Is there another way? Using benchmarks?

�An application downloaded to computers/servers of a LARGE number of users

�Sample measures on QoS automatically measured

– Several times a day

– For multiple measures (upload speed, download speed, jitter, round-trip-time, etc ….)

– For different types of access (local, international etc)

– Data adjusted for various factors

�Measures uploaded (automatically) to a popular website

�Various analysis done on it. Various types of benchmarks displayed

– E.g. all ADSL operators in a country, running 30/60/120 days upload speeds

– E.g. all BB operators (ADSL, HSDPA, other) in a country…..

– All of Brazil’s providers against ALL Honduras’s providers of BB, past 30 days upload

24

A different model

�For the user…

– Choices based on real performance vs. marketing

– Opportunity to diagnose problems (just my connection? Or is everyone suffering?)

�For Operators…

– “how am I doing in comparison to others”: benchmarking

– Use this data in advertising “the fastest HSDPA provider for the past 6 months”

– Offer differentiated products (different prices for different QoS levels)

– Long term planning: increase capacity in the right place (int’l vs. domestic capacity)

�For regulator…

– Real time data for regulation

– No opportunity for operators “doctor” the data

25

Benchmarking can be a useful tool…not just in the traditional way

�Not just analysis, but tool for decision making.

�Not just after the fact analysis, but forward looking decision making