IN THE COMPETITION TRIBUNAL OF SOUTH AFRICA
CT Case No: 92/CR/OCT11
An economic assessment of the
Competition Commission’s
allegations of predatory pricing by
Media24
Neil Dryden and Jorge Padilla
6 September 2013
Non-confidential version
COMPASSLEXECON.COM
Table of contents
Section 1
Introduction
1
Section 2
Executive summary
3
Section 3
Background
8
Section 4
The Commission’s allegations
19
Section 5
Legal framework
21
Section 6
Economics of predation
27
Section 7
European approach to predatory pricing assessments
31
Section 8
Economic framework
38
Section 9
Assessment of Forum’s incremental profitability
42
Section 10
Comparison of Forum’s revenues and its avoidable costs
50
Section 11
Anti-competitive effect of Media24’s pricing conduct
53
Section 12
Media24’s intent to exclude Gold-Net News
57
Section 13
Recoupment
62
COMPASSLEXECON.COM
Section 1
Introduction
1.1
This expert report has been prepared jointly by Mr Neil Dryden and Dr Jorge Padilla. Mr
Dryden is an Executive Vice President of Compass Lexecon Europe. Dr Padilla is a Senior
Managing Director and Head of Compass Lexecon Europe. Compass Lexecon is a global
economic consultancy, a wholly owned subsidiary of FTI Consulting, Inc.
1.2
The authors’ CVs are attached in Appendix A.
Background
1.3
The Competition Commission of South Africa (the “Commission”) referred a complaint
against Media24 Limited (“Media24”) to the Competition Tribunal (the “Tribunal”) on 31
October 2011.
1.4
The Commission’s complaint concerns the Goudveld Forum (“Forum”), a community
newspaper formerly published by Media24 in Welkom, South Africa. The Commission
alleges that during the period January 2004 to February 2009 (the “complaint period”),
Media24 operated Forum as a ‘fighting brand’, setting low prices and operating the
newspaper at a loss with the intention of excluding a competing community newspaper,
Gold-Net News, from the market. The Commission contends that this constituted predatory
pricing conduct in contravention of Section 8(d)(iv), or alternatively Section 8(c), of the South
African Competition Act 1998 (the “Act”).
Instructions
1.5
We have been instructed by the Commission to carry out a price-cost test of Forum during
the complaint period, to examine the anti-competitive effects of Media24’s pricing and to
examine any other economic evidence that may be relevant to assessment of the
Commission’s allegations.
1.6
Assessment of market definition and dominance is outside the scope of our instructions. Our
understanding of the Commission’s position is that the relevant market consists of the
market for community newspapers in the Goldfields area, in which market Media24 is
dominant under Section 8 of the Act. We have adopted the Commission’s position in carrying
out our analysis.
1
Assistance
1.7
We have prepared this expert report assisted by Compass Lexecon staff working under our
direction, supervision and review. We have discussed the issues relevant to the matter with
the Competition Commission staff. The opinions expressed in this expert report are,
however, our own.
Restrictions
1.8
The data and source materials provided by Media24 and presented in this report have not
been subject to independent audit or verification by Compass Lexecon or us.
1.9
This report must not be construed as expressing opinions on matters of law, which are
outside our expertise. We are instructed to act as expert economic witnesses and not
witnesses of fact.
Sources of information
1.10
We list the sources of information that we have reviewed in Annex B and the data that we
have relied on in Annex C.
1.11
We have relied on several datasets provided by Media24 which are split across multiple
discovery documents. For brevity, we have referred to these datasets as single sources in
the main body of the report (e.g. “Forum management accounts”). Annex C sets out the full
references for each data source referred to in the report.
2
Section 2
Executive summary
Background and instructions
2.1
The Commission alleges that during the period January 2004 to February 2009 (the
“complaint period”), Media24 operated Forum as a ‘fighting brand’, setting low prices and
operating the newspaper at a loss with the intention of excluding a competing community
newspaper, Gold-Net News, from the market.
2.2
We have been instructed by the Commission to carry out a price-cost test of Forum during
the complaint period, to examine the anti-competitive effects of Media24’s pricing and to
examine any other economic evidence that may be relevant to assessment of the
Commission’s allegations.
Price-cost tests
Test of Forum’s incremental profitability
2.3
From an economic point of view, the correct price-cost test for the purposes of assessing the
Commission’s allegations is a test based on Forum’s economic costs.
2.4
The economic costs of Forum include the relevant accounting costs of Forum as well as the
opportunity costs for Media24 of operating Forum and the avoided exit costs of not closing
Forum, i.e. they incorporate all benefits and disbenefits incurred by Media24 in choosing to
operate Forum compared to closing it. We refer to the overall profitability of Forum given by
its economic costs so defined as Forum’s incremental profitability.
2.5
For the purposes of assessing Forum’s incremental profitability, we consider that it is
appropriate to examine the following two scenarios in assessing the Commission’s
complaint:
The actual scenario, where Media24 operated both Forum and Vista throughout the
complaint period; and
A counterfactual scenario, where Media24 closed Forum at the start of the complaint
period and continued to operate Vista throughout the complaint period.
2.6
The difference between the profits of Media24 in the actual and counterfactual scenarios can
be separated into two factors:
The additional revenues received by Media24 in the actual scenario compared to the
counterfactual, i.e. the incremental revenues of Forum over the complaint period; and
3
The additional costs incurred by Media24 in the actual scenario compared to the
counterfactual, i.e. the incremental costs of Forum over the complaint period.
2.7
We assess Forum’s incremental profits during the complaint period, calculating Forum’s
incremental costs based on two definitions:
Definition 1: All costs that could be eliminated over a period of one year according to
Media24’s cost classifications; and
Definition 2: All costs of Forum as shown in its management accounts.
2.8
Table 1 shows Forum’s incremental profitability during the complaint period.
Table 1: Forum’s discounted incremental profits (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
2.9
As set out above, Forum was incrementally loss-making in NPV terms over the entire
complaint period, under both definitions of avoidable costs. The extent of losses amounts to
R '''''''''''''''' under the first definition of Forum incremental costs (all costs that could be
eliminated over a period of one year) and R '''''''''''''' under the second definition (total costs).
A predatory pricing assessment based on this test would conclude (subject to other
evidentiary requirements) that Media24 was engaged in predatory pricing.
Comparison of Forum’s revenues against its avoidable costs
2.10
We have also considered a test of predatory pricing that only takes into account Forum’s
relevant accounting costs and not its economic costs.
2.11
While this is not the correct approach from an economic point of view, it may be an
appropriate test for the purposes of minimising potential errors in competition enforcement,
since it focuses on more directly measurable costs (i.e. from the accounts in the complaint
period) and is, in the present case, a more difficult cost standard to fail.
2.12
As before, we consider that it is appropriate to examine the following two scenarios in
assessing the Commission’s complaint:
The actual scenario, where Media24 operated both Forum and Vista throughout the
complaint period; and
4
A counterfactual scenario, where Media24 closed Forum at the start of the complaint
period and continued to operate Vista throughout the complaint period.
2.13
In the absence of opportunity costs, the difference between the revenues earned by Media24
in the two scenarios equates to the accounting revenues of Forum during the complaint
period.
2.14
Similarly, in the absence of avoided exit costs, the difference between the costs of Media24
in the two scenarios is the portion of Forum’s accounting costs that was avoidable during the
complaint period.
2.15
The predatory pricing test that excludes opportunity costs and avoided exit costs therefore
involves a comparison of Forum’s total accounting revenues against its accounting costs in
the complaint period that were avoidable.
2.16
Table 2 shows the results of this comparison.
Table 2: Forum’s revenues and avoidable costs (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
2.17
As set out above, Forum’s revenues were lower than Forum’s 1 year avoidable costs in NPV
terms during the complaint period. Over the entire complaint period, the magnitude of the
difference amounted to R ''''''''''''''''''''' in NPV terms as of January 2004. A predatory pricing
assessment based on this test would conclude (subject to other evidentiary requirements)
that Media24 was engaged in predatory pricing.
Anti-competitive effects of Media24’s pricing conduct
2.18
We have examined the anti-competitive effects of Media24’s pricing conduct based on two
approaches:
Actual ex post evidence of anti-competitive effects observed during the post-complaint
period; and
The degree to which any exclusion found under the two tests described above is
indicative of anti-competitive effects ex ante, given the characteristics of the market.
5
2.19
On an ex post basis, we observe that Gold-Net News exited the market soon after the end of
the complaint period. This would have caused anti-competitive harm to consumers in the
market due to loss of consumer choice. We also find evidence that after the exit of Gold-Net
News, Media24 engaged in recoupment of the losses incurred by Forum, causing further
anti-competitive harm.
2.20
On an ex ante basis, we consider that the exclusion of Gold-Net News as a result of
Media24’s pricing conduct likely resulted in anti-competitive effects. This is because, given
the barriers to entry already present in the market (due to the two-sided nature of the market)
and arising from Media24’s conduct (via the reputation effect), the exclusion of Gold-Net
News likely resulted in increased market power for Media24.
Media24’s intent to exclude Gold-Net News
2.21
We have examined whether Media24 could have forecast Forum to be incrementally
profitable if it had based its forecasts on extrapolations on any one of a number of potential
‘representative years’. Forum’s incremental profitability based on these extrapolations shows
whether it would have been a rational business decision for Media24 to continue to operate
Forum if it had based its forecasts on Forum’s performance over the relevant time period.
2.22
We find that:
Under all forecasts based on actual performance of Forum, Forum would have been
loss-making on an incremental basis; and
Under forecasts based on Media24 budget, Forum would have been generally
profitable based on the first definition of Forum incremental costs (all costs that could
be eliminated over a period of one year), but generally loss-making based on the
second definition (total costs).
2.23
Taken together, these results suggest that the incremental losses of Forum could not have
been due to unexpected underperformance by Forum. The fact that Media24 continued to
operate Forum for an extended period despite these losses suggests that the decision to
operate Forum was a result of exclusionary intent rather than normal business rationale.
Recoupment
2.24
We have examined whether it would have been viable for Media24 to recoup Forum’s losses
via price increases after the exit of Gold-Net News. We find that:
On an ex ante basis, the price increases required after the complaint period in order for
Media24 to recoup Forum’s losses were modest. Full recoupment would have been
possible based on a small but significant and non-transitory increase in price (“SSNIP”)
of 5 to 10 per cent; and
On an ex post basis, Vista’s financial performance improved after the exit of Gold-Net
News, and the improvement was large enough to make full recoupment of Forum’s
losses viable.
6
2.25
In addition, by forcing Gold-Net News out of the market based on an extended period of
predatory pricing, Media24 may also have gained a reputation for being a tough competitor
against entrants. Benefits arising from this reputation, in the form of reduced threat of new
entry, are another potential source of recoupment for Media24.
7
Section 3
Background
Introduction
3.1
This section sets out background information on community newspapers in the Goldfields
area and some descriptive statistics on Forum’s business during the complaint period.
Community newspapers in the Goldfields area
3.2
Community newspapers are free newspapers published at regular intervals and delivered
directly to homes, businesses and certain outlets for distribution such as petrol stations.
3.3
During the relevant period, there were three community newspapers operating in the
1
Goldfields area: Goudveld Forum (“Forum”), Vista and Gold-Net News.
Forum
3.4
Forum started publication in September 1983. It was operated by Media24 (via Volksblad
Group, its community newspapers division) throughout its existence. During the complaint
period, it was published on Wednesday of each week.
3.5
Media24 ceased publishing Forum in January 2010 and the title exited the market as a
result.
Vista
3.6
Vista started publication in 1971. One of its founders was Mr Hans Steyl, who also managed
Gold-Net News at a later date. The ownership of Vista changed several times during the
period 1980 to 1999.
3.7
During the complaint period, Vista was operated by Media24 (via Volksblad Group, its
community newspapers division), which had purchased Vista from the Caxton Group in
1999.
3.8
At the time of its purchase by Media24, Vista had two weekly editions, one published on
Tuesdays and the other on Thursdays. In December 2001, the Tuesday edition was
discontinued. Thereafter, Vista published one edition per week on Thursdays.
1
Background information on the three newspapers presented below is taken from the Respondent’s
Answering Affidavit dated 19 January 2012, paras. 16-29 and the Factual Witness Statement of Mr
Steyl, paras. 8-24.
8
3.9
Media24 continues to publish Vista in the Goldfields area.
Gold-Net News
3.10
Gold-Net News has its origins in the “Net News (Goldfields)” newspaper, which started
publication in 1999, as part of the Net News group of community newspapers established by
Ms Leda Joubert. Mr Hans Steyl became the manager of Net News (Goldfields) in October
1999.
3.11
Net News (Goldfields) closed in November 2000 due to financial difficulties. Immediately
following its closure, Mr Hans Steyl and other staff members reopened Net News
(Goldfields) as Gold-Net News.
3.12
Gold-Net News exited the market in April 2009.
Forum’s business
3.13
We set out below some relevant descriptive statistics on Forum’s business in and around the
2
complaint period.
Revenues
3.14
As a free newspaper, Forum’s sole source of revenue was from sales of advertising space.
The advertisements carried by Forum can be split broadly into national advertisements, local
advertisements and classified advertisements.
3.15
Media24 sells advertising space across all of its community newspapers to national
3
advertisers via Ads24, its national advertising sales arm.
3.16
Local advertisements were marketed to local businesses in the Goldfields area by Forum’s
4
sales team located in Welkom.
3.17
Classified advertisements are small advertisements sold to local businesses and individuals.
3.18
Table 3 shows a breakdown of Forum’s total net revenues during the complaint period.
2
All quantitative information presented is based on Forum’s management accounts and advertising
volume data.
3
Factual Witness Statement of Ms Van Eck, para. 25.
4
Factual Witness Statement of Ms Van Eck, para. 25.
5
The data provided by Media24 does not break down discounts and rebates by types of advertisements.
For the purposes of presenting net revenues or net prices by advertisement type, we allocate Forum’s
discounts and rebates between national and local advertisements in proportion to gross revenues, on
the assumption that classified advertisers would not receive discounts or rebates. The same approach
is taken for net revenue or net price figures shown in the remainder of the report.
5
9
Table 3: Total Forum net revenues, January 2004-February 2009
Category
Rands
%
National advertisements*
'''''''''''''''''''''''
''''%
Local advertisements
'''''''''''''''''''''''
''''%
Classified advertisements
'''''''''''''''''''''''
''''%
Total revenues
'''''''''''''''''''''''
100%
* Revenue lines marked '''''''''''''''''''''''''' ''''''''''''''''''''''' ''''''''''''''''''''''''''''' ''''''''''' '''''''''''''''' '''''''''''''''''''' ''''''''''''''''''''''''''''' ''''''''''' '''''''''''' ''''''''''''''''''
'''''''''' ''''''''''''''''''''''''' '''''''''''''' ''''''''''''''''''' '''''''''''''''''''''' in management accounts.
Note: Revenues shown at March 2010 prices after adjusting for inflation based on PPI. Forum’s discounts and
rebates allocated proportionally between national and local advertisements.
Source: Forum management accounts
3.19
Figure 1 shows Forum’s monthly national, local and classified net revenues over time at
6
March 2010 prices.
6
Since the majority of Forum’s sales were to other businesses, we have used PPI to deflate Forum’s
revenues. ''''''''''' ''''''''''' '''''' '''''''''''''' '''''''''''' '''''''''''''''' ''''''' '''''''' '''''''''''''' '''''' ''''''''''''' ''''''''''''''''''''' ''''''''''''''''''''''''''''''''' ''''''''''''''''''''
'''''''''''' '''''''''''''
10
Figure 1: Forum net revenues by type (Rands)
Note: Revenues shown at March 2010 prices after adjusting for inflation based on PPI. Forum’s discounts and
rebates allocated proportionally between national and local advertisements. Dotted lines show start and end of
complaint period.
Source: Forum management accounts
3.20
As shown above, local advertisements accounted for the majority of Forum’s revenues
throughout the period for which data are available. National advertisements declined in
importance during the complaint period compared to the years before the complaint period.
Costs
3.21
The most significant items in Forum’s costs were allocated central costs excluding printing
costs, printing costs, staff salaries and paper distribution costs. Table 4 shows a breakdown
of Forum’s total costs during the complaint period.
11
Table 4: Total Forum costs, January 2004-February 2009
Category
Rands
%
'''''''''''''''''''''''
''''%
'''''''''''''''''''''''
''''%
'''''''''''''''''''''''
''''%
'''''''''''''''''''''''
''''%
Other costs
'''''''''''''''''''''''
''''%
Total costs
'''''''''''''''''''''''
100%
Allocated central costs excluding printing costs
Printing costs
2
3
Staff salaries
Paper distribution
4
1
1. ''''''''''''''''''''''''''''' '''''''''''''''' ''''''''''''''''''''''''' ''''''''''''''' '''''''' ''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' ''''''''''' ''''''''''''' ''''''''''''''''' '''''
''''''''''''''''''''''''''''''''' ''''''''''''''''''''''''' ''''''''''''''' ''''''''''''''' ''''''''''''''''''''' ''''' ''''''''''''''''' ''''''''''''' ''''''''''''''' ''''''''''''''''' ''''''''''''''''' '''''''''''''''''''''''''''''''' ''''''''''''''''''''''''''
'''''''''''''''''''''''''''''''' '''''''''''''' '''''''''' ''''''''''''''''''' ''''''''''''''''''' '''''''' '''''''''''''''''''' ''''''''''''''''''''''''''''''' ''''''''''''''' '''''''''''''''''''''' '''''''' '''''''''''''''' '''''''''''''''''''''''''''''''''
''''''''''''''' 2. ''''''''''''''''''''''''' ''''''''''''''''''''''' ''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''' ''''''''''''''''''''''' ''''''''' '''''''''''''''' '''''''''
'''''''''''''''' ''''' '''''''''''''''''''''''''''''' ''''''''''''''''' 3. '' '''''''''''''''''''''''' '' '''''''''''''''''''''' '''''''''''''''''''' ''''''''' '''''''''' 4.'''''''''''''''''''''''''''''''' ''''''''''''''''''''''''' ''''''''''''''' '''''''''
'''''''''''''''''''''''''''' '' ''''''''''' ''''''''''''''''''''''''''' ''''''''''''' '''''''''' '''''''''''''
Note: Costs shown in March 2010 prices after adjusting for inflation based on PPI.
Source: Forum management accounts
3.22
Media24 has not provided the split of allocated central costs into printing and non-printing
costs prior to the complaint period. Figure 2 shows Forum’s allocated central costs
(excluding printing costs) and printing costs from January 2004 onwards in March 2010
prices.
12
Figure 2: Forum costs by type – allocated central costs (excluding printing costs) and
printing costs (Rands)
Note: Costs shown at March 2010 prices after adjusting for inflation based on PPI. January 2004 is the start of the
complaint period. Dotted line shows end of complaint period.
Source: Forum management accounts
3.23
''''''''''''''''''' ''''''''''''''''''''''''''''''' '''''''''''''''''''' ''''''''''''' '''''''' ''''''''''''''''''''''' ''''' '''''''''''''''''''''' ''''''''' '''''''''''''''''''''' '''''''''''''''''
''''''''''''' '''''''''''''' ''''''''''' '''''''''' '''''''''''''' '''''''''''''' '''''''''''' ''''''''''''''' '''''''''''''''' '''''' '''''''''''''''''''''''' ''''' '''''''''''''''''''''''' '''''
'''''''''''''''''''''' '''''''''''''''''''''''''''''''' '''''''''''''
3.24
Figure 3 shows the other major cost categories of Forum over a wider time horizon in March
2010 prices.
13
Figure 3: Forum costs by type – staff salaries, distribution costs and other costs
(Rands)
Note: Costs shown at March 2010 prices after adjusting for inflation based on PPI. Dotted lines show start and end
of complaint period.
Source: Forum management accounts
Advertising volumes
3.25
Forum’s total advertising volumes measured in single column centimetres (“sccm”) can be
classified into national, local and classified advertisements as set out above, plus editorials
and ‘unpaid advertisements’ (“Onbetaalde”).
3.26
‘Unpaid advertisements’ appear to have included free advertisements as well as various
newspaper spaces which could not be allocated to other categories, such as fillers (“Vullers”)
and white spaces and headlines (“Witruimte en Koppe”).
3.27
Table 5 shows Forum’s total advertising volumes over the complaint period, split into the
categories described above.
14
Table 5: Total Forum advertising volumes, January 2004-February 2009
Category
sccm
%
National advertisements*
'''''''''''''''''''''''
''''%
Local advertisements
'''''''''''''''''''''''
''''%
Classified advertisements
'''''''''''''''''''''''
''''%
Unpaid advertisements**
'''''''''''''''''''''''
''''%
Editorials
'''''''''''''''''''''''
''''%
Total revenues
'''''''''''''''''''''''
100%
* '''''''''''''''' ''''''''''' ''''''''''''''''''' '''''''''''''''''''''''''' ''''''''''''''''''''''' '''''''''''''''''''''''''''''''' ''''''''''''' ''''''''''''''' '''''''''''''''''''''''' ''''''''''''''''''''''''''''' ''''''''''''' '''''''''''''
''''''''''''''''''''' '''''''''' '''''''''''''''''''''''' ''''''''''''''' ''''''''''''''''' '''''''''''''''''''' ''''' '''''''''''''''''''''''''' '''''''''''''''' ''''''''''''
** '''''''''''''''''''' ''''''''''' '''''''''''''''' ''''''''''''''''''''''''''''''''''''''''''''' ''''''''''''''' ''''''''''''''''''''''''''''''' '''''''''''''''''''''' '''''''''''''''''''' ''''''''''''''' ''''''''''''''''''''''''''''' '''''''''''
'''''''''''''''''''''''''''''''''''' ''''''''' '''''''''''''''''''''''''' '''''' '''''''''''''''' ''''''''''''''''''''''''''''' ''''''''' ''''''''''''''''''''''''' ''''' '''''''''''''''''''''''' '''''''''''''''''' '''''''''''''
Source: Forum advertising volume data
3.28
Figure 4 shows Forum’s monthly total advertising volumes over time by type.
Figure 4: Forum advertising volumes by type (sccm)
Note: Dotted lines show start and end of complaint period.
Source: Forum advertising volume data
15
Prices
3.29
For the majority of advertisements, the prices charged by Forum to local advertisers were in
the form of a tariff charged on a per single column centimetre (sccm) basis. The exceptions
were advertising inserts (“invoegsels”), as they do not occupy space on the pages of the
newspaper.
3.30
Media24 maintained a rate card for Forum which listed Forum’s prices for advertising
7
space. The list prices varied based on whether the advertisement was local or national and
the number of colours used. The rate card was updated annually.
3.31
''''' '''''''''''''''''' '''''''' ''''''''''''''''' ''''' ''''''''''''''''''' '''''''''''''''''''''''' '''''''''''''''' ''''''''' ''''''''' '''' '''' ''''''''''''''''''''' ''''' '''''''' '''''''
'''''''''''''''' ''''''''' '''''''''''' ''''' '''''''''''''''''''''' ''''''''''''' '''''''''''''''' ''''' '''''''''''''''''''''''''' '''''''''' '''''''''''''''''''''''''' '''''' ''''''''''''
'''''''''''''''''''''''' '''''''' '''''''''''''''' '''''''''''''''' ''''''''''''''''''''''''''''' ''''''''' ''''''''''' '''''''''''
3.32
Figure 5 shows Forum’s monthly average net prices over time, in nominal terms and real
8
terms shown at March 2010 prices.
Figure 5: Forum average net prices over time (R/sccm)
Note: Prices shown at March 2010 prices after adjusting for inflation based on PPI. Dotted lines show start and end
of complaint period.
Source: Forum advertising volume data
7
For examples see discovery documents RD134-140.
8
The average prices shown exclude revenues from inserts, since inserts do not take up space in the
newspaper pages and would therefore inflate the average prices if included. The same approach is
taken for price figures shown in the remainder of the report.
16
3.33
Average prices differed between national, local and classified advertisements, ''''''''' '''''''''''''''''
'''''''''''''''''''''''''''''''''''' ''''''''''''''''' '''''''' '''''''''''''''' ''''''''''''''' Table 6 shows Forum’s average national, local
and classified advertisement prices over the complaint period.
Table 6: Average Forum net prices, January 2004-February 2009
Category
Rands/sccm
National advertisements*
'''''''''''''''
Local advertisements
'''''''''''''''
Classified advertisements
'''''''''''''''
All advertisements
'''''''''''''''
* ''''''''''''''''''''''' '''''''''''' '''''''''''''''' ''''''''''''''''''''''''' '''''''''''''''''''''' ''''''''''''''''''''''''''''''' '''''''''''' ''''''''''''''''' ''''''''''''''''''''' ''''''''''''''''''''''''''''' ''''''''''' '''''''''''''
''''''''''''''''''''' ''''''''' ''''''''''''''''''''''''''''' ''''''''''''' ''''''''''''''''' '''''''''''''''''' ''''' ''''''''''''''''''''''''''''''' ''''''''''''''''''''
Note: '''''''''''' ''''''''''''''''''''''''''''' ''''''''''''''''' ''''''''''''''''''''''' ''''''''''''' '''''''''' ''''''''''''''' ''''''''''' ''''''''''''''''' ''''''''''''' '''''''''''''''' ''''' ''''''''''''''''''''''' ''''''''''''''''''
''''''''''''''''' '''''''''''''''''''''''''' '''''''''''''''''''''' ''''' ''''''''''' ''''''''''''''''''''''''' '''''''''' ''''''''''''' '''''''''''''''''''''''''' '''''''' ''''''''''''''''''''' ''''''''''''''' '''''' ''' ''''''''''''''' Prices
shown at March 2010 prices after adjusting for inflation based on PPI. Forum’s discounts and rebates allocated
proportionally between national and local advertisements.
Source: Forum advertising volume data
3.34
Figure 6 shows Forum’s monthly average net prices for national, local and classified
advertisements over time at March 2010 prices.
17
Figure 6: Forum average net prices over time by advertising type (R/sccm)
Note: ''''''''''''''''''''''''''' ''''''''''' '''''''''' ''''''''''''''' '''''''''''' '''''''''''''''' '''''''''''''''' ''''' '''''''''''''''''''''''''' ''''''''''''''''''''' '''''''''''''''''' '''''''''''''''''''''''' '''''''''''''''''''''' '''''
'''''''''''' '''''''''''''''''''''''''' '''''''' ''''''''''''' ''''''''''''''''''''''''''' ''''''''' ''''''''''''''''' ''''''''''''''' '''''' '''' ''''''''''''''' Revenues shown at March 2010 prices
after adjusting for inflation based on PPI. Forum’s discounts and rebates allocated proportionally between national
and local advertisements. Dotted lines show start and end of complaint period.
Source: Forum advertising volume data
18
Section 4
The Commission’s allegations
Introduction
4.1
This section summarises the Commission’s allegations regarding Media24’s predatory
pricing conduct via Forum.
Complaint Referral
4.2
The Commission’s Complaint Referral contains the following allegations.
Over the complaint period of January 2004 to February 2009, Media24 engaged in
9
predatory pricing via Forum.
During the complaint period, Media24 set Forum’s price to below average variable
costs, defined as either (i) costs variable in relation to Media24’s decision to operate
10
Forum as a “fighting brand”; or (ii) costs that are variable over 12 months.
Media24’s predatory pricing conduct caused Gold-Net News’ exit from the market in
11
April 2009. In addition, potential new entrants were deterred from entering the market.
Media24’s conduct constituted a contravention of Section 8(d)(iv), alternatively Section
12
8(c), of the Act.
Supplementary founding affidavit
4.3
In a Supplementary Affidavit dated 25 February 2013, the Commission set out the following
clarification with regards to its allegations:
“As indicated above and in the complaint referral, the Commission contends
that Media24’s decision to continue to operate Forum during the complaint
period was motivated by its intention to exclude Gold-Net News from the
market, and that the continued operation of Forum did not make economic
sense other than as a fighting brand to achieve that objective.
9
Commission’s Complaint Referral, para. 31.
10
Commission’s Complaint Referral, paras. 32-47.
11
Commission’s Complaint Referral, paras. 51-60.
12
Commission’s Complaint Referral, para. 65.
19
In those circumstances, the Commission submits that it is relevant from an
economic perspective to compare the incremental costs of operating Forum
(being costs that are “variable” or “marginal” to the decision to operate Forum)
and the associated incremental revenues obtained by Media24 from operating
Forum for purposes of:
– the Commission’s complaint against Media24 under section 8(d)(iv) of the
Act; and
– the below-cost pricing aspect of the Commission’s complaint against
13
Media24 under section 8(c) of the Act.”
13
Commission’s Supplementary Affidavit dated 25 February 2013, paras. 24-25.
20
Section 5
Legal framework
Introduction
5.1
This section sets out the relevant provisions of the South African Competition Act and
summarises certain South African precedents which, we understand, are relevant to the
present matter.
South African Competition Act
5.2
Section 8(d)(iv) of the Act prohibits predatory pricing. It specifies the relevant test of
predatory pricing as a comparison of prices against “marginal or average variable cost”:
“8. It is prohibited for a dominant firm to – […]
(d) engage in any of the following exclusionary acts, unless the firm
concerned can show technological, efficiency or other pro-competitive, gains
which outweigh the anti-competitive effect of its act: […]
(iv) selling goods or services below their marginal or average variable cost”
5.3
14
In addition to Section 8(d)(iv), Section 8(c) of the Act states that conduct not specified in
Section 8(d) may nevertheless constitute an abuse if the anti-competitive effect of the
conduct outweighs its pro-competitive gains:
“8. It is prohibited for a dominant firm to – […]
(c) engage in an exclusionary act, other than an act listed in paragraph (d), if
the anti-competitive effect of that act outweighs its technological, efficiency or
15
other pro-competitive, gain”
South African precedents
5.4
We have been instructed to have regard to the Tribunal’s decision regarding predatory
pricing in Nationwide, and to the Tribunal’s decisions regarding other types of abuses in
South African Airways and Mandla-Matla.
14
Competition Act, Chapter 2, Section 8(d)(iv).
15
Competition Act, Chapter 2, Section 8(c).
21
16
Nationwide
5.5
Nationwide is a decision of the Tribunal relating to an application for interim relief brought by
Nationwide Airlines against South African Airways. Nationwide alleged that South African
Airways had engaged in predatory pricing, in contravention of Section 8(d)(iv), during the
17
period August 2000 to October 2000, as well as several other types of abuse.
5.6
In its decision, the Tribunal clarified the level of evidence required for a finding of predatory
pricing under Section 8(d)(iv) and Section 8(c).
5.7
The Tribunal considered that, given the pricing conduct described in Section 8(d)(iv), there
must be a rebuttable presumption of abuse:
“Predatory pricing or ‘predation’ is a term of art in competition law and
economics and its meaning is precisely captured in Section 8(d)(iv) of the Act
which prohibits dominant firms from engaging in the practice of ‘selling goods
or services below their marginal or average variable cost’. This price-cost
relation must be established for a predatory pricing charge to be sustained in
terms of this section. Once this relationship is established an anti-competitive
act is presumed and the respondent then has the burden of showing that
there are ‘technological, efficiency or other pro-competitive, gains which
outweigh the anti-competitive effect of its act.’ […]
In this respect section 8(d)(iv) assists a complainant with the aid of the
presumption. The logic of the presumption is that once a firm is pricing below
these costs its behaviour is prima facie either unlawful or irrational and the
18
onus should shift to the firm to explain which it is.”
5.8
Given such presumption, the Tribunal considered that Section 8(d)(iv) should be narrowly
applied to cases where the below-cost pricing is beyond doubt, and that other predation
pricing allegations should be assessed under Section 8(c):
“Our approach is to limit the scope of [Section 8(d)(iv)] by critically construing
any evidence when considering a complaint of predation under this section.
Unless the record shows unequivocally that a respondent is pricing below the
prescribed cost levels the Tribunal should not make a finding under section
19
8(d)(iv) but consider the complaint in terms of section 8(c).”
16
South African Airways (Proprietary) Limited, South African Express Airways (Proprietary) Limited and
South African Airlink (Proprietary) Limited v Nationwide Airlines (Proprietary) Limited, Nationwide Air
Charter (Proprietary) Limited, Nationwide Aircraft Maintenance (Proprietary) Limited and Nationwide
Aircraft Support (Proprietary) Limited, Competition Tribunal, Case no. 92/IR/Oct00 (“Nationwide”).
17
Other abuses alleged by Nationwide concerned South African Airways’ sales agreements with travel
agents and South African Airways’ practice of soliciting and recruiting Nationwide’s pilots. These
abuses are not relevant to the present case and are not discussed further.
18
Nationwide, p. 9.
19
Nationwide, p. 10.
22
5.9
The Tribunal considered that in comparison to Section 8(d)(iv), the assessment of abuse
under Section 8(c) places a higher evidentiary burden on the complainant in that (i) the onus
is on the complainant to show that the anti-competitive effect of the conduct outweighs the
pro-competitive gains; and (ii) requires evidence of predation beyond evidence of costs:
“Section 8(c) is the residual category or “catch all” of abuse practices. Unlike
section 8(d) where a closed list of abuses is catalogued this section is nonspecific and flexible. The crucial difference is that under 8(c) the onus is on
the complainant to establish that the anti-competitive nature of the act
‘outweighs the technological, efficiency or other pro-competitive gain.’
The burden on the complainant in a complaint of predatory behaviour is
higher under this section therefore than under section 8(d)(iv). On the other
hand the complainant is not bound to follow the prescribed cost formula
suggested in 8(d)(iv). In other words if a complainant, relying on section 8(c),
can show that a respondents [sic] costs are below some other appropriate
measure of costs not mentioned in the section it may prevail provided it
20
adduces additional evidence of predation beyond mere evidence of costs.”
5.10
The Tribunal therefore specified a two-part test for predatory pricing under Section 8(c):
“In our view, the correct approach to a predatory pricing claim under section
8(c) is a two fold test.
First the complainant must establish that the respondent is pricing below cost
for a sustained period. This price-cost relationship need not be the one
referred to in 8(d)(iv) but should have some support in the literature as an
appropriate measure of costs. Secondly there must be some additional
evidence of predation. We do not wish to be prescriptive as to what this
21
should be, but evidence of recoupment would meet this second test.”
5.11
The Tribunal found that neither of these two tests had been satisfied by Nationwide, and
dismissed the application for interim relief.
20
Nationwide, p. 10 (emphasis added).
21
Nationwide, p. 13 (emphasis added).
23
South African Airways
22
and Mandla-Matla
23
5.12
South African Airways concerned certain incentive schemes operated by South African
Airways for travel agents. The Commission alleged that these schemes discouraged travel
agents from offering airline tickets from South African Airways’ competitors and were abusive
24
under Section 8(d)(i) of the Act.
5.13
Mandla-Matla concerned a complaint of exclusionary practices brought by Mandla-Matla
against Independent Newspapers. The complaint was made at about the time that the
agreement between Mandla-Matla and Independent Newspapers, in terms of which
Independent Newspapers was responsible for the printing and distribution of the magazine
title Ilanga (owned by Mandla-Matla) was coming to an end. Mandla-Matla’s allegations were
twofold:
Under the ‘exclusivity complaint’, Mandla-Matla alleged that certain agreements
between Independent Newspapers and local distributors, which prohibited the
distribution of rival magazines to Independent Newspapers’ own titles, were abusive
25
under Section 5(1) or Section 8(d)(i) of the Act.
Under the ‘information complaint’, Mandla-Matla alleged that Independent Newspapers’
refusal to provide Mandla-Matla with certain information regarding local distributors was
abusive under Section 8(c) of the Act.
5.14
While South African Airways and Mandla-Matla did not involve allegations of predatory
pricing, the Tribunal’s decisions in both cases examined the issue of evidentiary
requirements for a finding of abuse under Section 8 of the Act.
5.15
In South African Airways, the Tribunal stated that the conduct referred to in Section 8(d) of
the Act is presumed to be exclusionary:
“Paragraph (d) makes specific reference to five exclusionary acts, whereas
paragraph (c) refers generally to an ‘exclusionary act’. It would follow that any
exclusionary act not captured in the list in paragraph (d) would fall to be
considered in terms of paragraph (c).
22
The Competition Commission v South African Airways (Pty) Ltd, Competition Tribunal, Case no.
18/CR/Mar01 (“South African Airways”).
23
Mandla-Matla Publishing (Pty) Ltd v Independent Newspapers (Pty) Ltd, Competition Tribunal, Case
no. 48/CR/Jun04 (“Mandla-Matla”).
24
Section 8(d)(i) of the Act prohibits “requiring or inducing a supplier or customer to not deal with a
competitor.”
25
Section 5(1) of the Act states that “An agreement between parties in a vertical relationship is prohibited
if it has the effect of substantially preventing or lessening competition in a market, unless a party to the
agreement can prove that any technological, efficiency or other pro-competitive, gain resulting from
that agreement outweighs that effect.”
24
It would seem from the manner in which the section is drafted, that conduct in
(d) is presumed to be exclusionary, whereas conduct not in the list would still
have to be established as exclusionary. It is established under (c) only if it
meets the definition of an exclusionary act.
The reason for these differences in treatment is that the exclusionary acts in
8(d) are listed, presumably evidencing the legislature’s view that these are the
more egregious of the exclusionary acts and so firms who are dominant are
on notice that they must behave with due caution in relation to this conduct,
whereas in 8(c) no acts are listed and hence the complainant would have to
26
prove that the conduct sought to be impugned is indeed exclusionary.”
5.16
However, the Tribunal found that there was a distinction between an exclusionary act and an
anti-competitive act, with the latter requiring to be separately established:
“The question is then whether conduct deemed (8(d)) or found to be (8(c)) an
exclusionary act, is, for that reason, impugnable (subject of course to the
conclusion on the efficiency justification). […]
One approach is to say that if the act is exclusionary, it is deemed to have an
anti-competitive effect. On this approach the only issue that remains to be
decided is the balancing of the efficiency justification against this deemed
anti-competitive effect.
The problem with this approach is that it can lead to the outlawing of conduct
that has no anti-competitive effect. The definition of an exclusionary act is
very broad indeed. […]
The term [exclusionary act] is not a useful filter for determining whether
conduct is competitive or anti-competitive; both can sensibly be included in
the definition. If, however, we do not regard the notion of anti-competitive
effect, referred to in both paragraphs (c) and (d), as inferentially linked to an
exclusionary act, this danger can sensibly be averted. It means that that [sic]
the notion of what [is] anti-competitive is something different to an
exclusionary act.
At a purely textual level they appear to be notionally different. If they were
congruent notions, then the legislature need not have troubled itself with
introducing the language anti-competitive effect into the paragraphs, but
would instead have referred to exclusionary effects. We suggest that there is
a difference between an exclusionary act as defined and the inference that it
has an anti-competitive effect. […]
For this reason the Act requires a finding both in terms of paragraphs (c) and
(d) that the complainant not only establishes that there has been an
27
exclusionary act, but also that it has an anti-competitive effect.”
26
South African Airways, paras. 100-102.
27
South African Airways, paras. 106-111 (emphasis added).
25
5.17
In Mandla-Matla, the Tribunal cited South African Airways and confirmed the requirement for
evidence of anti-competitive effects under Sections 8(c) and 8(d):
“However section 8(c) and 8(d) require that both the elements of the
exclusionary act and its alleged anti-competitive effect are proved in order to
undertake the balancing required by the pro-competitive defence that is
permitted in respect of these exclusionary acts but which is not provided for in
28
respect of the conduct described in Sections 8(a) and 8(b).”
Instructions
5.18
In the light of the above, we have been instructed:
To consider what tests of exclusion based on Media24’s revenues and costs would be
appropriate, as a matter of economics, for the purposes of assessing the Commission’s
allegations under Section 8(d)(iv) and Section 8(c);
To carry out these tests;
To carry out an assessment of the anti-competitive effects of Media24’s pricing; and
To present other economic evidence that may be relevant to assessment of the
Commission’s allegations.
28
Mandla-Matla, para. 78 (emphasis original).
26
Section 6
Economics of predation
Introduction
6.1
This section briefly sets out the economic literature on predatory pricing and the relevant
economic concepts for assessing predatory pricing.
Economic models of predation
6.2
The simplest model of predation is the ‘deep pockets’ or ‘long purse’ model of predation,
where a dominant firm drives a smaller firm out of the market by charging loss-making
prices. While this requires losses on the part of the dominant firm as well the smaller firm, it
is the smaller firm that exits the market because the dominant firm has the financial strength
to endure the losses for longer.
6.3
However, this simple model of predation is not capable of explaining several aspects of
predatory pricing. For instance, it does not explain why the smaller firm would not be able to
use access to credit markets to endure the predatory pricing, since it should be clear that the
loss-making strategy is not sustainable for the dominant firm. Given this possibility, it may
29
not be in the dominant firm’s interest to attempt predatory pricing.
6.4
One way of addressing this criticism is to introduce imperfect information into the model via
‘reputation’. Milgrom and Roberts (1982) and Kreps and Wilson (1982) construct ‘reputation’
models of predatory pricing based on imperfect information regarding the degree to which
the incumbent will fight or accommodate entry. According to the models set out by these
authors, if an incumbent active in multiple markets faces a new entrant in one of those
markets, the incumbent has incentives to use predatory pricing to force the new entrant out
of the market. This is because an instance of predation on a new entrant gives the
incumbent a reputation as a ‘tough competitor’. Future potential entrants will expect that the
incumbent firm would also respond aggressively to their entry. As a result, potential entrants
are deterred from entering the market to compete against the incumbent, who can therefore
exercise market power after the predation conduct.
29
This criticism and several others were raised in McGee (1958).
27
6.5
The criticism of the ‘deep pockets’ model of predation can also be addressed by
incorporating imperfect information in financial markets into the model. Bolton and
Scharfstein (1990) construct a model incorporating imperfect information on the part of
30
investors about whether managers of a firm are acting in the investors’ best interests. In
such a case, it may be rational for investors to commit to an incentive mechanism of
withdrawing funding if the firm performs poorly. However, the existence of this incentive
mechanism introduces a financial constraint on the firm that makes predatory pricing by its
competitors possible.
Tests of predation
6.6
Predation is ultimately defined by legislation and case law. We set out in Section 5 the
relevant South African legislation and case law which inform the appropriate tests for
predatory pricing. We also set out in Section 7 the relevant European legislation, guidelines
and case law.
6.7
From an economic perspective, the appropriate test for the purposes of competition
enforcement must balance the risks of type I and type II errors:
Type I errors are errors of over-enforcement. These refer to cases where conduct which
does not have anticompetitive effects is found to be abusive.
Type II errors are errors of under-enforcement. These refer to cases where conduct
which does have anticompetitive effects is not found to be abusive.
6.8
In order to find an efficient rule that minimises the risks of type I and type II errors, a possible
31
approach is to adopt a two-tier test of predatory pricing:
A test of predation based on a demanding cost standard, which results in a
presumption of abuse if satisfied; and
An alternative test based on a more inclusive cost standard, which, in conjunction with
further evidence of predatory pricing, results in a finding of abuse.
Cost standards
6.9
Tests of predatory pricing in the economic literature, legislation, guidelines and case law are
often based on economic measures of the costs incurred by a firm. We explain below some
of the commonly used cost standards in predatory pricing assessments.
30
This is an example of a ‘principal-agent’ problem which has been widely studied by economists in other
contexts.
31
According to Motta (2004), the first of such two-tier tests of predatory pricing was suggested by Joskow
and Klevoric (1979). See Motta (2004), p. 442. As set out in Section 7, this approach has been adopted
in European case law in AKZO and Tetra Pak.
28
Variable costs
6.10
Variable costs are costs that vary with changes in the level of output. Costs that are not
variable are defined as fixed costs.
6.11
Variable costs may be defined with respect to an increment or a decrement in output. As
such, they can be equivalent to incremental or avoidable costs.
Marginal costs
6.12
Marginal costs are the additional costs of producing one additional unit of output.
6.13
While an important concept in theoretical economics, marginal costs may be of limited
usefulness in applying tests of predatory pricing. This is because a single additional unit of
output is usually not a relevant quantity of output for a firm.
6.14
Instead, economists usually consider the additional costs of producing a larger increment of
32
output, known as the incremental cost. When defined over the relevant unit of economic
33
decision, marginal costs are equivalent to incremental costs.
Avoidable costs
6.15
Avoidable costs are costs that can be avoided or recovered by reducing a certain level of
output.
6.16
The degree to which costs can be avoided depends on the relevant timescale. Over a short
timescale, many costs may not be avoidable based on scaling back production. Over a
sufficiently long timescale, however, all incremental costs of a given increment in output are
likely to be avoidable.
Incremental costs
6.17
Incremental costs are the additional costs of producing a certain incremental level of output.
6.18
Incremental costs are equal to avoidable costs if the decisions to increase and decrease the
level of output are symmetrical, and all additional costs incurred in producing the incremental
level of output can be avoided by reversing the output decision. However, incremental costs
would be higher than avoidable costs if some costs incurred in increasing the level of output
cannot be avoided by decreasing the level of output to the original level. Such costs are
referred to as sunk costs.
32
See O’Donoghue and Padilla (2006), p.237.
33
See O’Donoghue and Padilla (2006), p. 237. “The difference between MC and incremental cost is
partly definitional and many economists therefore treat them as equivalent.”
29
6.19
Incremental costs may vary depending on the timescale involved. For instance, a decision by
a firm to increase output over a long timescale may involve significant investments in
additional assets, which would be part of the incremental costs defined over a longer
timescale. Once the firm has acquired the assets, however, then the short-term decision to
increase or decrease output may not have a significant incremental impact on the firm’s
costs because the assets have already been acquired and cannot be sold quickly.
Economists refer to the incremental costs defined over a sufficiently long timescale for all of
the factors of production to vary as the long-run incremental cost (“LRIC”) or, when defined
on a per-unit basis, as the long-run average incremental cost (“LRAIC”).
Application of cost standards
6.20
As set out above, marginal costs, incremental costs and avoidable costs are related cost
standards that all refer to costs specific to a subset of a firm’s output. These cost standards
should be defined by reference to the relevant quantity of output, the timescale over which
costs are to be assessed and whether an increment or a decrement in output is being
considered.
30
Section 7
European approach to predatory
pricing assessments
Introduction
7.1
Unlike the South African Competition Act, Article 102 of the Treaty on the Functioning of the
European Union does not contain a specific clause on predatory pricing. However, European
competition authorities and courts have established some principles of predatory pricing
assessment via guidelines and precedents.
7.2
In this section we summarise the relevant parts of:
AKZO and Tetra Pak, two influential predatory pricing cases examined by the European
Commission (“EC”);
The EC Guidelines on Article 102, which set out further principles on the assessment of
predatory pricing; and
Several additional predatory pricing precedents and one margin squeeze precedent
examined by competition authorities and courts in the European and UK jurisdictions,
which may usefully inform the Tribunal’s approach.
AKZO34 and Tetra Pak35
7.3
In AKZO, the European Court of Justice set out two alternative tests for determining whether
36
a dominant firm has engaged in predatory pricing:
Pricing below average variable costs; or
Pricing below average total costs, in combination with evidence that the prices were
37
“determined as part of a plan for eliminating a competitor.”
34
Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359 (“AKZO”).
35
Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755 (“Tetra Pak”); Case C333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951 (“Tetra Pak II”).
36
These two tests are an implementation of the two-tier assessment of predatory pricing explained in
paras. 6.7-6.8 above.
37
AKZO, paras. 71-72.
31
7.4
The Court of First Instance confirmed these tests in Tetra Pak.
38
EC Article 102 Guidance
7.5
The EC Guidelines on Article 102 give guidance on the EC’s approach to assessing pricebased exclusionary conduct in general, as well as predatory pricing in particular.
Price-based exclusionary conduct
7.6
The EC states that it is likely to use average avoidable cost (“AAC”) or long-run average
39
incremental cost (“LRAIC”) cost standards in assessing price-based exclusionary conduct.
7.7
The EC also notes that in applying the AAC and LRAIC cost benchmarks, it may be
necessary to examine incremental revenues of the relevant product as opposed to the total
40
revenues.
Predatory pricing
7.8
The EC characterises predatory pricing as conduct that results in deliberate sacrifice of
profits in the short term. The concept of sacrifice includes charging lower prices or
41
undertaking an expansion of output which results in losses that could have been avoided.
7.9
AAC will be used as a starting point for the assessment of profit sacrifice, but the concept of
profit sacrifice is wider than this and encompasses all conduct that leads to lower profits than
42
a reasonable alternative. In some cases average variable costs (“AVC”) may be a suitable
proxy for AAC, although this will not hold in cases where predation requires investment in
43
capacity or other fixed costs. If sufficient data are available, the EC will apply the LRAIC
44
cost standard.
7.10
Direct evidence of predatory intent in the form of internal documents may be relevant to the
45
EC’s analysis. Conversely, in cases where the firm can provide conclusive proof that the ex
ante decision to engage in the conduct was based on a reasonable expectation that the
conduct would be profitable, then the firm would not be penalised by the EC for incurring ex
46
post losses.
38
Tetra Pak, paras. 148-149 and Tetra Pak II, para. 41.
39
EC Article 102 Guidelines, para. 26.
40
EC Article 102 Guidelines, para. 26, footnote 3.
41
EC Article 102 Guidelines, paras. 63-64.
42
EC Article 102 Guidelines, paras. 64-65.
43
EC Article 102 Guidelines, para. 64, footnote 3.
44
EC Article 102 Guidelines, para. 67. See also para. 26.
45
EC Article 102 Guidelines, para. 66.
46
EC Article 102 Guidelines, para. 65, footnote 3.
32
7.11
The EC defines anti-competitive foreclosure as a situation where the conduct of the
dominant firm hampers or eliminates actual or potential competitors, such that it would likely
47
be in a position to profitably increase prices, causing consumer harm.
7.12
The EC’s assessment of anti-competitive effects will take into account evidence of actual
foreclosure and evidence of exclusionary intent, as well as a consideration of how the
predatory pricing conduct reduces the likelihood that competitors will compete with the
48
dominant firm. This may include consideration of entry barriers and the possibility of re49
entry. However, the EC does not consider it necessary to show the actual exit of
50
competitors from the market.
7.13
The EC will consider claims that predatory pricing conduct results in efficiencies, although it
51
notes that the likelihood of efficiencies arising from predatory pricing conduct is low.
Other precedents
Aberdeen Journals
52
7.14
In Aberdeen Journals, the UK Office of Fair Trading (“OFT”) alleged that Aberdeen Journals
had predated Aberdeen & District Independent (“Independent”) by deliberately incurring
losses on the Aberdeen Herald & Post (“Herald & Post”), one of three newspapers that it
operated in the Aberdeen area. Aberdeen Journals’ other newspapers were the Press &
53
Journal and the Evening Express.
47
EC Article 102 Guidelines, para. 19. See also para. 70.
48
EC Article 102 Guidelines, para. 68 (including reference to para. 20).
49
EC Article 102 Guidelines, para. 71.
50
EC Article 102 Guidelines, para. 69.
51
EC Article 102 Guidelines, para. 74.
52
OFT ‘Predation by Aberdeen Journals Ltd’ CA98/5/2001, Decision dated 16 July 2001; Case
1005/1/1/01 ‘Aberdeen Journals Limited v. Director General of Fair Trading’, CAT Judgment dated 19
March 2002; OFT ‘Predation by Aberdeen Journals Limited (Remitted case)’ CA98/14/2002, Decision
dated 16 September 2002; Case 1009/1/1/02 ‘Aberdeen Journals Limited v. The Office of Fair
Trading’, CAT Judgment dated 23 June 2003 (“Aberdeen Journals”).
53
Aberdeen Journals, paras. 4-7
33
7.15
It was common ground between the parties that the Herald & Post had incurred substantial
54
losses between 1996 and mid-2000. However, in its analysis the OFT considered the
period beginning on 1 March 2000, being the date that the relevant UK legislation, the UK
55
Competition Act 1998, came into force. The OFT’s complaint ultimately concerned a period
of one month between 1 March 2000 and 29 March 2000, although the UK Competition
Appeal Tribunal (“CAT”) found that this short complaint period should be considered as part
56
of a continuous period of abuse beginning in March 1996.
7.16
In assessing the OFT’s allegations, the CAT accepted the two alternative tests set out in
AKZO and Tetra Pak, namely: (i) prices below average variable costs; and (ii) prices below
57
average total costs along with evidence of intent.
7.17
It also set out additional clarifications for the two cost standards used in the AKZO tests as
follows.
7.18
With regard to variable costs, the CAT made clear that variable costs should be determined
according to whether or not each type of cost is variable over the length of the alleged abuse
period. The CAT noted that this definition of variable costs approaches total costs as the
time period grows longer. It accepted that this is a reasonable approach for the purposes of
58
enforcement.
7.19
Applying this principle to the case, the CAT stated that the correct definition of variable costs
for Aberdeen Journals was the costs that were variable over the period mid-1996 to January
59
2001, being the alleged period of below-cost pricing. Given that the OFT had in fact
examined costs that were variable over a period of one month, the CAT considered that the
60
OFT’s approach had been conservative and favourable to Aberdeen Journals.
7.20
With regard to total costs, the CAT stated that any measure of total costs must include
61
contributions towards overheads that are shared with other titles.
7.21
In summary, Aberdeen Journals confirmed the two-tier test of predatory pricing test set out in
AKZO of (i) prices lower than average variable costs, and (ii) prices lower than average total
costs with evidence of intent. It also clarified that variable costs should be defined over the
complaint period, that variable costs approach total costs as the complaint period grows
longer and that total costs should be defined to include contributions towards shared
overheads.
54
Aberdeen Journals, para. 8.
55
Aberdeen Journals, para. 10.
56
Aberdeen Journals, paras. 3, 416-422 and 447-449.
57
Aberdeen Journals, paras. 351-352.
58
Aberdeen Journals, paras. 353-355.
59
Aberdeen Journals, para. 384.
60
Aberdeen Journals, paras. 393-396.
61
Aberdeen Journals, paras. 365-366, 370.
34
Cardiff Bus
62
7.22
In Cardiff Bus, the OFT alleged that Cardiff City Transport Services (“Cardiff Bus”) had
engaged in predatory conduct against 2 Travel, a new entrant which operated ‘no-frills’ bus
services in Cardiff. The alleged abuse was the launch of Cardiff Bus’ own ‘no-frills’ services,
63
referred to as the ‘white services’, on the routes served by 2 Travel.
7.23
Cardiff Bus did not deny that its white services had been loss-making. However, it claimed
that the introduction of the white services had been intended to test the market demand for
‘no-frills’ services, and that it withdrew the services when it found that there was insufficient
64
demand for ‘no-frills’ services.
7.24
The OFT based its finding of abuse on an examination of “all the available evidence ‘in the
65
round’”, including:
Documentary evidence of Cardiff Bus’ intent to exclude 2 Travel;
66
An analysis of the revenues and costs of Cardiff Bus’ white services which showed the
67
white services to be loss-making; and
Lack of evidence supporting, and evidence contradicting, Cardiff Bus’ claims that the
68
white services had been introduced as a market test.
7.25
In considering the legal framework for assessing the abuse, the OFT noted the test of prices
against average variable costs as set out in AKZO, which would give a rebuttable
69
presumption of abuse.
7.26
The OFT considered that the most appropriate cost benchmark in that case was the average
avoidable costs of the white services. These included the variable costs of running the white
70
services, plus the fixed or sunk investments required to run the services. The OFT
71
considered that this was a logical extension of the test set out in AKZO.
62
OFT ‘Abuse of a dominant position by Cardiff Bus’ CA98/01/2008, Decision dated 18 November 2008
(“Cardiff Bus”).
63
Cardiff Bus, para. 1.4.
64
Cardiff Bus, paras. 7.36-7.38.
65
Cardiff Bus, para. 7.18. See also Cardiff Bus, para. 7.224.
66
Cardiff Bus, paras. 7.86-7.149.
67
Cardiff Bus, paras. 7.150-7.223.
68
Cardiff Bus, paras. 7.51-7.85.
69
Cardiff Bus, paras. 7.10-7.11.
70
Cardiff Bus, paras. 7.156-7.158.
71
The OFT did not specify whether it considered the test to be a logical extension of the AKZO test
based on the average variable cost or the average total cost. Cardiff Bus, paras. 7.160-7.161.
35
7.27
The OFT considered that avoidable costs should include all costs that are avoidable over the
time period of the alleged predatory conduct. Accordingly, the OFT assessed the costs of the
white services that were avoidable over a period of 10 months, being the period of time that
72
the white services were operational.
7.28
In summary, the OFT in Cardiff Bus adopted a test of predatory pricing based on average
avoidable costs over the length of the complaint period. The OFT reached a finding of abuse
based on the results of this test and additional evidence of intent.
Wanadoo
73
7.29
In Wanadoo, the EC found that Wanadoo Interactive had engaged in predatory pricing in the
market for high-speed Internet access services in France between January 2001 and
October 2002. The European Court of First Instance and the European Court of Justice
upheld the Commission’s findings.
7.30
In assessing its allegations, the EC applied the two alternative tests set out in AKZO and
Tetra Pak: (i) prices below average variable costs; and (ii) prices below average total costs
74
along with evidence of intent.
7.31
The costs included by the EC in its definition of variable costs included the network costs of
providing the broadband services and customer acquisition costs, including some sunk costs
75
such as advertising expenditure.
7.32
The EC noted that during the alleged abuse period, there were some specific developments
in the market for high-speed Internet access in France which could not have been predicted
by Wanadoo. Those developments had the effect of improving Wanadoo’s financial
performance. In its main analysis, the EC did not account for these unforeseen
developments and instead based its analysis on the actual performance of Wanadoo.
However, it noted that Wanadoo’s ex ante forecast of its profitability could also “throw further
light on the matter”, and examined Wanadoo’s forecast level of profitability as a
76
supplemental analysis. In that case, the forecast performance of Wanadoo was worse than
its actual performance. The EC did not therefore need to address the question of how
evidence of ex ante forecast profitability would affect a finding of predatory pricing or the
objective justification.
72
Cardiff Bus, paras. 7.172-7.175.
73
Case COMP/38.233, Wanadoo Interactive, EC Decision of 16 July 2003 (“Wanadoo”); Case T-340/03,
France Télécom SA v Commission of the European Communities, Judgment of the Court of First
Instance, 30 January 2007; Case C-202/07 P, France Télécom SA v Commission of the European
Communities, Judgment of the European Court of Justice, 2 April 2009
74
Wanadoo, para. 256.
75
Wanadoo, paras. 42-68.
76
Wanadoo, paras. 97-106 and 108. See also paras. 70-72.
36
Telefónica
77
7.33
Telefónica was a margin squeeze case rather than a predatory pricing case, but it set out a
relevant principle regarding ex ante and ex post losses that are relevant to assessment of
predatory pricing.
7.34
In assessing allegations of margin squeeze by Telefónica, the EC contended that the pricing
test for margin squeeze must be based on actual historical data rather than business
forecasts. However, the EC considered that business forecasts may be relevant to the
assessment of objective justification. The General Court of the European Union upheld the
78
EC’s findings.
Conclusions
7.35
AKZO and Tetra Pak set out a two-tier test of predatory pricing which has been widely
adopted by European competition authorities and courts. This is a relatively simple test
based on two cost standards (variable costs and total costs) and evidence of intent to
exclude competitors.
7.36
The EC Article 102 Guidelines, however, set out a wider definition of predatory pricing based
on the concept of profit sacrifice, and appear to give more weight to predatory pricing
assessments based on avoidable or incremental cost standards.
7.37
The predatory pricing cases since AKZO and Tetra Pak have examined predatory pricing in
more complex contexts. Competition authorities and courts in those cases have applied a
test of predatory pricing based on avoidable costs as a natural extension of the AKZO test
(Cardiff Bus) and confirmed that variable costs would approach total costs over the long run
(Aberdeen Journals).
77
Case COMP/38.784, Wanadoo España vs. Telefónica, EC Decision of 4 July 2007 (Telefónica); Case
T-336/07, Telefónica SA and Telefónica de España SA v. European Commission
78
Telefónica, paras. 378-382.
37
Section 8
Economic framework
Introduction
8.1
This section sets out the economic framework that we will apply for our assessment of the
79
Commission’s complaint referral.
Price-cost tests
Application of the marginal and variable cost standards
8.2
The Act specifies that the assessment of predatory pricing under Section 8(d)(iv) should be
based on a comparison of the price of a good or service against its “marginal or average
80
variable cost”.
8.3
As set out above, marginal and variable cost standards should be defined precisely as a
matter of economics having regard to the relevant quantity of output, the relevant timescale
81
and whether an increment or a decrement in output is being considered.
8.4
In the present case, the Commission’s complaint concerns Media24’s decision to continue to
operate Forum during the complaint period. The relevant costs in this case from an
economic perspective are therefore costs that are marginal or variable to the decision to
operate Forum during the complaint period, which are given by the avoidable costs
associated with hypothetically not operating Forum during the complaint period.
8.5
We have considered two tests that apply this definition of marginal and variable costs to the
present case.
Test of Forum’s incremental profitability
8.6
From an economic point of view, the correct approach to assessing the marginal or variable
costs of Forum is based on its economic costs rather than its accounting costs.
79
See para. 5.18 above.
80
See para. 5.2 above.
81
See paras. 6.9-6.20 above.
38
8.7
The economic costs of Forum include the relevant accounting costs of Forum as well as the
opportunity costs for Media24 of operating Forum and the avoided exit costs of not closing
Forum, i.e. they incorporate all benefits and disbenefits incurred by Media24 in choosing to
operate Forum compared to closing it. We refer to the overall profitability of Forum given by
its economic costs so defined as Forum’s incremental profitability.
8.8
For the purposes of assessing Forum’s incremental profitability, we consider that it is
appropriate to examine the following two scenarios in assessing the Commission’s
complaint:
The actual scenario, where Media24 operated both Forum and Vista throughout the
complaint period; and
A counterfactual scenario, where Media24 closed Forum at the start of the complaint
period and continued to operate Vista throughout the complaint period.
8.9
The difference between the profits of Media24 in the actual and counterfactual scenarios can
be separated into two factors:
The additional revenues received by Media24 in the actual scenario compared to the
counterfactual, i.e. the incremental revenues of Forum over the complaint period; and
The additional costs incurred by Media24 in the actual scenario compared to the
82
counterfactual, i.e. the incremental costs of Forum over the complaint period.
8.10
The incremental revenues and costs of Forum may differ from the total revenues and costs
of Forum shown in its management accounts. Incremental revenues would be expected to
be lower than the total revenues of Forum since Forum and Vista competed in the same
market, and would have been cannibalising each other’s revenues to a certain degree.
8.11
Incremental costs of Forum may be lower than the total costs of Forum due to costs of
Forum that cannot be avoided by closing Forum. Incremental costs of Forum may also be
lower because closing Forum in the counterfactual scenario involves additional costs (i.e.
exit costs) that were not incurred in the actual scenario.
8.12
A comparison of the incremental revenues and incremental costs of Forum gives the
incremental profits of Forum, which indicates whether Media24’s decision to operate Forum
resulted in an economic loss.
8.13
We are instructed that it is a matter of law whether the above approach satisfies the
provisions of Section 8(d)(iv). For the Tribunal to accept this test under Section 8(d)(iv):
Marginal or variable costs must be interpreted as costs that are marginal or variable
with respect to Media24’s decision to operate Forum; and
82
Strictly speaking, since Forum was already operational and this case concerns Media24’s decision to
continue to operate Forum, this should be understood as the avoidable costs of Forum. However, as
set out above (see paras. 6.15-6.19), the two concepts are equivalent in the absence of sunk costs,
and we will refer to these as Forum’s incremental costs throughout this report for ease of exposition.
39
The price-cost test specified in Section 8(d)(iv) must be understood as incorporating
considerations of revenue diversion and exit costs.
8.14
In the alternative, if the Tribunal were to examine Media24’s conduct under Section 8(c),
Forum’s incremental profitability would also form the price-cost test required for a finding of
predatory pricing under that section.
Comparison of Forum’s revenues against its avoidable costs
8.15
We have also considered a test of predatory pricing that only takes into account Forum’s
relevant accounting costs rather than its economic costs. While this is not the correct
approach from an economic point of view, such an approach could be seen in economic
83
terms as an approach that minimises the risk of enforcement errors. This is because it
focuses on more directly measurable costs (i.e. from the accounts in the complaint period)
and is, in the present case, a more difficult cost standard to fail.
8.16
As before, we consider that it is appropriate to examine the following two scenarios in
assessing the Commission’s complaint:
The actual scenario, where Media24 operated both Forum and Vista throughout the
complaint period; and
A counterfactual scenario, where Media24 closed Forum at the start of the complaint
period and continued to operate Vista throughout the complaint period.
8.17
In the absence of opportunity costs, the difference between the revenues earned by Media24
in the two scenarios equates to the accounting revenues of Forum during the complaint
period.
8.18
Similarly, in the absence of avoided exit costs, the difference between the costs of Media24
in the two scenarios is the portion of Forum’s accounting costs that was avoidable during the
complaint period.
8.19
The predatory pricing test that excludes opportunity costs and avoided exit costs therefore
involves a comparison of Forum’s total accounting revenues against its accounting costs in
the complaint period that were avoidable. This comparison indicates whether Media24’s
decision to continue to operate Forum resulted in a loss in accounting terms, without taking
into account the full likely impact of the decision via revenue cannibalisation and exit costs.
8.20
While this test does not take into account some factors which form part of the correct
economic test, this test may be viewed as an application of the “marginal or average variable
cost” standard specified in the Act and therefore may be considered by the Tribunal as a
basis for a finding of predatory pricing under Section 8(d)(iv).
83
See paras. 6.7-6.8 above.
40
Anti-competitive effects
8.21
We have been instructed to examine the anti-competitive effects of Media24’s pricing based
on two approaches:
Actual ex post evidence of anti-competitive effects observed during the post-complaint
period; and
The degree to which any exclusion found under the two tests described above is
suggestive of anti-competitive effects ex ante, given the characteristics of the market.
Other evidence
8.22
We have also been instructed to examine any other economic evidence that may be relevant
to assessment of the Commission’s allegations. We have therefore examined:
Whether Media24’s intended to exclude Gold-Net News from the market; and
Whether Media24 recouped its predation losses in the post-complaint period.
Structure
8.23
The remainder of this report is structured as follows.
Section 9 sets out our assessment of Forum’s incremental profitability, which is our
preferred test of Media24’s pricing conduct;
Section 10 sets out our comparison of Forum’s revenues against its avoidable costs,
which is our intermediate test of Media24’s pricing conduct;
Section 11 examines the anti-competitive effect of the alleged predation;
Section 12 examines whether Media24 intended to exclude Gold-Net News from the
market; and
Section 13 examines whether Media24 recouped its predation losses in the postcomplaint period.
41
Section 9
Assessment of Forum’s incremental
profitability
Introduction
9.1
This section sets out our assessment of Forum’s incremental profitability. We consider this to
be the correct test of predation as a matter of economics because it takes into account all
relevant factors in the business decision to operate Forum, and indicates whether that
decision involved a profit sacrifice against a reasonable and clear alternative of closing down
Forum.
Methodology
9.2
As set out above, our framework for assessing Forum’s incremental profitability involves a
comparison of Media24’s profits under the following two scenarios:
The actual scenario, where Media24 operated both Forum and Vista throughout the
complaint period; and
A counterfactual scenario, where Media24 closed Forum at the start of the complaint
period and continued to operate Vista throughout the complaint period.
9.3
The differences in Media24’s profits between these two scenarios can be broken down into
the following revenues and costs:
Revenues that were earned by Forum during the complaint period which would not
have diverted to Vista in the counterfactual;
Costs that were actually incurred by Forum during the complaint period but would have
been avoided in the counterfactual;
Increased costs of Vista in the counterfactual compared to its actual costs; and
Any one-off exit costs involved in closing Forum.
42
Incremental revenues of Forum net of diversion to Vista
9.4
Since Media24 operated Vista as well as Forum during the complaint period, the actual
revenues of Forum cannot simply be taken to be the incremental revenues of Forum. This is
because, if Media24 had closed Forum, some of the advertisers previously served by Forum
would have diverted to Vista. An estimate of Forum’s incremental revenues therefore
requires an estimate of the proportion of Forum’s advertising sales that would have been
recaptured by Vista (the “diversion ratio” from Forum to Vista).
9.5
Based on an estimate set out in an internal Media24 document ''''''''''''' ''''''''''' ''''''''''''' we have
assumed a diversion ratio of '''''' per cent from Forum to Vista throughout the complaint
84
period.
9.6
Another internal Media24 document ''''''''''''''' '''''''''''''''''''''' ''''''''''''' estimates a higher diversion
85
ratio of '''''' per cent from Forum to Vista. Since Gold-Net News had exited the market by
that date, this diversion ratio estimate is not directly comparable with the previous estimate
during the complaint period.
9.7
However, a diversion ratio of '''''' per cent after the complaint period is broadly consistent with
a diversion ratio of '''''' per cent during the complaint period given the relative market shares
of Vista and Gold-Net News during the complaint period. Media24’s internal estimates of
advertising volume market shares suggest market shares of ''''''''' per cent for Vista, ''''''''''' per
cent for Forum and ''''''''''' per cent for Gold-Net News on average during the complaint
86
period. If '''''' per cent of Forum customers would not switch to other community
newspapers after closure of Forum (as indicated by '''''''' '''''''''''''''''''''''' ''''''''''''' '''''''''''''''''''''') and
diversion to other community newspapers is proportional to market shares, this would imply
diversion from Forum to Vista during the complaint period of ''''''''''' per cent.
9.8
For completeness, we set out calculations of Forum’s incremental profitability based on this
alternative estimate of diversion ratio of ''''''''''' per cent in Annex F.
Avoided Forum costs
9.9
We have modelled the Forum costs that could have been avoided by closing Forum based
on two definitions:
84
'''''''''''''''''' ''''''' '''''''''''' ''''''''''''''' '''''''''''''''''''''''''''''''' '''''''''' '''' ''''''''' ''''''''''''''''' '''''''''''''''''''' '''''''''''''''''''' ''''''''''''''''''' '''' '''''''''''''''''''''''' '''''''
''''''''''''''''' ''''' '''''''''''' ''''' '''''''''''''' ''''''''''''''' Discovery document RD103. As we understand it, this document
indicates that in the event of Forum closure, Vista’s revenues would increase by the equivalent of ''''''
per cent of Forum’s revenues. A possible alternative interpretation would be that ''''''' per cent of
Forum’s advertising volume would transfer to Vista and pay Vista’s higher tariffs, meaning that the
increase in Vista revenues would be larger than '''''' per cent of Forum’s revenues. On our
interpretation, the diversion of revenue from Forum to Vista after closure of Forum is lower, and the
incremental revenue of Forum is higher. Our interpretation is therefore conservative and favourable to
Media24.
85
Discovery document RD261.
86
Average of market shares shown in discovery documents RD604-630.
43
Definition 1: All costs that could be eliminated over a period of one year according to
87
Media24’s cost classifications; and
Definition 2: All costs of Forum as shown in its management accounts.
9.10
We consider that Definition 1 understates the true magnitude of costs that could have been
avoided during the complaint period by closing Forum for two reasons.
Firstly, the complaint period lasted 5 years and 2 months, during which there would
have been greater scope for Media24 to make changes to its business to eliminate
additional costs compared to the cost savings that could be achieved over a period of 1
year.
Secondly, based on the explanations provided by Media24 as to how costs can be
varied or eliminated, we understand that Media24’s definition of costs that can be
eliminated over 1 year does not include costs that could be reallocated to alternative
88
productive uses within Media24 after Forum’s closure. From an economic viewpoint,
avoidable costs should include costs that could be eliminated or reallocated to
alternative productive uses.
9.11
Definition 2 would overstate the true magnitude of costs that could have been avoided by
closing Forum if any of Forum’s costs shown in its management accounts could not be
avoided over the complaint period. However, given the extended length of the complaint
period of 5 years and 2 months, we consider Definition 2 to be an appropriate scenario for
89
our analysis.
Increased Vista costs
9.12
As set out above, we have examined the diversion of revenues from Forum to Vista that
would result from the closure of Forum. Accordingly, we also take into account the increase
in Vista’s costs that would accompany the increase in Vista’s revenues.
9.13
In order to examine the impact of the revenue diversion on Vista’s cost base, we have
estimated the percentage increase in Vista’s total advertising volumes that would result from
the revenue diversion, assuming that the diversion of ''''''' per cent would apply to the
national, local and classified revenues of Forum, and that each type of advertisement would
90
be sold at Vista’s average prices for that advertising type.
87
See Annex H for details of cost classifications.
88
Discovery document RD22 explains various ways in which costs can be reduced or new contracts
negotiated, but does not mention any reallocation of resources within Media24.
89
The possibility that an extended period of alleged abuse would result in almost all costs being treated
as variable was noted in Aberdeen Journals. The UK Competition Appeal Tribunal considered this to
be reasonable. See para. 7.18.
90
This is in line with Media24’s internal estimates, which concluded that '''''''''''''''''''''''''''''''''''''' '''''''''''' '''' '''''''''
'''''''''''''''''' ''''''''''''''''''' ''''''''''''''''''' ''''''''''''''' ''''' ''''''''''''''''''''''''' '''''''' '''''''''''''''' '''''' ''''''''''''' ''''' ''''''''''''''''' ''''''''''''''' Discovery
document RD103 (emphasis added). See footnote 84 for an explanation of how we have interpreted
this source.
44
9.14
We applied this percentage increase to Vista’s total variable costs in order to estimate the
91
total increase in Vista’s costs in the counterfactual.
Exit costs
9.15
We have also taken into account potential one-off exit costs that would be incurred in closing
down Forum.
9.16
We have estimated these exit costs based on the observed costs at the time of Forum’s
actual closure in January 2010. Table 7 shows the revenues and costs of Forum shown in its
management accounts around the time of its exit.
Table 7: Forum revenues and costs, December 2009-March 2010
January
February
March
2010
2010
2010
Total revenues
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Allocated costs excluding printing costs
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Printing costs
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Staff salaries
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Paper distribution
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
'''''''''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Other costs
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Total costs
''''''''''''''''''''''
''''''''''''''''''''''
''''''''''''''''''''''
Category
Source: Forum management accounts
9.17
As shown above, Forum’s management accounts included two significant one-off costs
during the last month in which Forum’s management accounts were maintained:
'''' '''''''''''''''''' ''''' ''''''''''''''''''''''''''''''''''''' ''''''''''' ''''''''''''' '''''''''''''''''''''''''''''' ''''''' '''''''''''''''''''''''''''''' and
''''' ''''''''''''''''''' ''''' '''''''''''''''''''''''''' '''''''''''''''''''''' ''''''''''''''''''''''''' ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
9.18
Under our counterfactual scenario, Media24 would close Forum and incur these exit costs in
January 2004. Accordingly, we incorporated these costs in our counterfactual calculations by
deflating the total value to January 2004 prices based on PPI.
91
Variable costs of Vista were defined according to Media24’s classifications shown in Discovery
documents RD20-RD22.
45
Adjustments to Forum’s printing costs prior to April 2008
9.19
The printing costs of Forum shown in its management accounts were not market
transactions, but transfer prices to other Media24 entities. ''''''''' ''''''''''''''''''' ''''''''''''''''' ''''''''''' ''''''
'''''''''''''''' '''''''''''''''' ''''''''''''''''''''' '''' ''''''' ''''''''''''''''''''''''' '''''''''''''''' '''''''' '''''''''''''''''''''' '''''''''''''''' ''''''''''''''' '''''''''''''''''''''''
'''''''''''''''''' '''''''''' '''' '''''''''''' ''''''''''''' '''''''''''' ''''''''''' ''''''''''' ''''''''''''''''''''' '''''''' '''''''''''''' ''''''''''' ''''''''''''''''' '''''' '''''''''''''
''''''''''''''' ''''''''''''''''' '''''''''''''''''''''''' ''''' ''''''''''''''''''''''
9.20
'''''''' ''''''''''''''''''''''''' '''''''''' ''''''' '''''''''''''''' ''''''''''''''''' ''''''''''''''''''''''''' ''''' '''''' ''''''''''''''' ''''''''' '''''''''''''''' '''''''' '''''''''''''' ''''''''''''
'''''''' '''''''''''''''' ''''''''''''''''' ''''''''''''' ''''''''''''''''' '''''' '''''''' '''''''''''''''''''''' ''''''''''''''' '''''''' '''''''''''''''' ''''''''''''''''''' ''''' ''''''''''''''
''''''''''''' ''''''''''''''''''''' ''''' ''''''''''' ''''''''''''' ''''''' '''''''''''''''''' '''''''''''''''' ''''''''''''' ''''''''''''''''''''''''' ''''' '''''''''''' '''''''''''''''' ''''''''
92
''''''''''''''''''' ''''''''''''''''''''' ''''' '''''''''''''' ''''''''''''''''''''' ''' ''''''' '''''''' '''''''''' '''''''''''''''''' ''''''' ''''''''''''''''
9.21
''''''''''' ''''' ''''''' '''''''' '''''''''' ''''''''''''''''''' ''''''''''''''''''' ''''''''''' ''''''''''''''''''' ''''' '''''''''''''''' '''' ''''''''' '''''''''''' ''''' '''''''''' '''''''''''''
'''''''' '''''''''''''''''' ''''''''''''' '''''''''''''''' ''''' ''''''''''''''''''' '''''''''''''''''''''''''''''''' ''''''''''''''''''' ''''''''''' '''' ''''''''''' ''''''''''' '''''' ''''''''
'''''''''''''''' '''''''' ''''''''''''''''' ''''''''''''''' ''''' '''''''''''''''''' ''''''''''''''''''''''' '''''''''' ''''' '''''''''''''''''''''' ''''' '''''' ''''''''''''''''''' '''''''''''''''''''''
''''' ''''''''''''''''''''' '''''''''''''''''''''' '''''''''''''''' '''''''''''''' '''''''''''''''' ''''''''''
''''' '''' '''''''''''''''''''''' ''''' ''''''''''' '''''''''' ''''''''''' '''''''' '''''''''''' '''''''''''''''''' '''''''''''' ''''''''' '''''''''''' '''''''''''''''''''
'''''''''''''''' ''''''''''''''''''''' ''' '''''''''''''''''' '''''''''''''''''''''' '''''' '''''' '''''''''' '''''''' ''''''''''' ''''''''''''' ''''''''''''''' '''''''''''''
'''''''''''''''''''' ''''''''''' ''''''''''''''''' '''' '''' '''''''''''''''''''' ''''''''''''' '''''''''''''''''' ''''''''' '''' ''''''''''''' ''''''''''''''''' '''''''''''''
''''''''''' '''' ''''''''''' '''''''''''''' '''''' '''' ''''''''''''''''''''' '''''''''''''' ''''''''' ''''''''''''''''''''''' '''''''' ''''''''''''''''' '''''''''' ''''''
93
''''''''''' '''''''''' '''''''''''''''''''''''''' ''''' ''''''''''''''''' '''''''''''''''''' ''''''''''''''''
9.22
Transfer prices should be adjusted to market prices for the purposes of competition
assessment. Accordingly, we have made an upward adjustment to the printing costs shown
in Forum’s accounts prior to April 2008. '''''''''''''' '''''' '''''''' '''''''''''''''''''''''''''' ''''''''' '''''''' '''''''''''''''''' '''''''''''
''''''''''''''''''''' ''''' ''''''''''''''' '''''' ''''''''''''' ''''''''''''''''' '''''''''''''''''''''' '''''''' ''''''''''''''''' ''''''''''' ''''' ''''''''''''''''' ''''''''''''''''''' '''''' '''''''''''
'''''''''''''''''''' ''''''''''''''''''' '''''''''''''''''' ''''''''''' ''''''''''' ''''' '''''''''' ''''''''''''' ''''''''''''''''''' '''''' ''''''' '''''''' '''''''''''
Discounting
9.23
We have compared the relative profits under the actual and counterfactual scenarios on net
present value (NPV) bases. We have used a discount rate of 18 per cent per annum, based
on the pre-tax nominal discount rates shown in the FY2005 financial accounts of Naspers for
94
its print media subsidiaries.
Base case results
9.24
Table 8 sets out our results based on the methodology set out above.
92
''''''''''''''''''''''' ''''''' '''''''''''''''''' ''''''''''''''''' ''''''''' ''''''''''''''''''' ''' '''''''''' ''''''''''''''''''' ''''''' '''''''''''' ''''''''' ''''''''' '''''''' ''''''''''''''' '''''' ''''''''''''''''''''''
''''''''''' ''''' ''''''' ''''''''''''''''''' ''''' '''''''''''''' '''''''''''''''''' '''''''' ''''''''' '''''''''''' ''''''''''''''''' ''''''''''''''''''''' ''' '''''''''' ''''''''''''''''''''' ''''' '''''''''''' ''''''''''''
''''''''' ''''''''''''''''' ''''' ''''''''''''''''''''''''' '''''''''''''''''''''''''''''''' Factual Witness Statement of Mr Botha, paras. 7-8.
93
Discovery Document RD20.
94
See Annex I.
46
Table 8: Forum’s discounted incremental profits (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
9.25
As set out above, Forum was incrementally loss-making in NPV terms over the entire
complaint period, under both definitions of avoidable costs. The extent of losses amounts to
R '''''''''''''' under the first definition of Forum incremental costs (all costs that could be
eliminated over a period of one year) and R '''''''''''''' under the second definition (total costs).
9.26
Forum was also loss-making during each year of the complaint period under both definitions,
apart from 2004 under the first definition, when the significant exit costs avoided by keeping
Forum open meant that its incremental profits were positive by R ''''''''''''''''. In later years,
when Media24 would not have incurred exit costs in the counterfactual scenario, the
incremental profits of Forum were consistently negative.
Sensitivity based on adjustments to allocated costs
9.27
''''''''''''''''''' ''''''''''''''''''''' '''''''''''''' ''''''''''''''''''''''''''''''''' '''''''''''''''''' ''''''''''' '''''''''''' ''''''''''''''''''''' ''''' '''''''' '''''''''''' ''''''''''''''' '''''
'''''''''''''''''''' ''''' '''''''''''''''''''''''' '''''''''''''' ''''''''''' ''''''''''''''''''''''' '''' ''''''''''''''''''''' '''''''''''''' '''''''''''''''''''''' '''''''''''''''
9.28
The Factual Witness Statement of Mr Bonthuyzen alleges that the allocated costs shown in
Forum’s management accounts did not reflect the true costs related to operating Forum and
were significantly understated:
'''''''''''''''''''''''''' ''''''''''''''' ''''''''''''''''''''' '''''''''''''''''' ''''''''''''' '''''''''''''''''''''' '''''' '''''' ''''''''''''''''''''
'''''''''''''''''''''''' '''''''''''' '''''' '''''''''''''''''''''''' ''''''''''''''''''''''''' ''''''''''''''' ''''''' '''''''''''''''''''''' '''''''''''''''''''''''''''''''''
'''''''''''''''''''' '''''''''''''''''''''''' ''''''''''''''' ''''''''''''''''''''''''''' ''''''''''''''''''''''''''' ''''''''''' '''''''''' '''''''''''''''''''''''''''''''''
'''''''''''''' '''''''''''' ''''''''''''''''' ''''''''''''''''''''' '''''''' ''''''''''''''''''''''''''''' ''''' '''''''' '''''''''''''''''''''' ''''''''''''''''
47
'''''''''' ''''''''''''''''' '''''''''''' '''''''''''''''''''''' ''''' '''''''''''''''' ''''''''' ''''''''''''''''''' ''''' '''''''''''''''''''''''''' ''''''''''''''''''''''
'''''' '''''''' '''''''''''''''''''''''' '''''''' '''''''' '''' '''''''''''''''' '''''''''''''''''''''' ''''' ''''''' '''''''''''''''' ''''''''''''' ''''' ''''''''''''''''
''''''''''''''''' ''''''''''''''''''''' ''''''''''''''''''''''' '''''''''''''''''''''' ''''''''''''' '''''''''''''''''''' '''''''''''''''''''''' ''''''''''''''''' '''''
''''''''''''''''' '''''''''' ''''''''''''''''' ''''''''''''' ''''' '''''''''''''''' '''''''''''' ''''''''''''' ''''''''''' '''''''''''' '''''''''''''''''''' ''' ''''''''
''''''''''''''''''''' ''''''''' ''''''''''' '''''''''''' '''''''''''''' '''''' '''''' ''''''''''''''''''''''' '''''''''''''''''''''''' '''' '''''''''''''''' ''''''''''''''
''''''''' ''''''''''''''''''''' ''''''''''''''''''''' '''''''''''''''''''' ''''''''''''''''' '''''' ''''''''''' ''''''''''''''''' '''''''''''''''''''''''''' '''''''''
'''''''''''''' '''''''''''''''''''''''''''' ''''''''' ''''''''''''' ''''''''''''''''''''''' '''''''''''''' '''''' ''''''''''''' ''''''''''''''''''''''''''' ''''''''''
''''''''''''''''''' '''''''''''''''''' '''''''''''''' '''''''''''''' ''''''''''''' '''''' '''''''''''''''''''''''''' ''''' '''''' ''''' ''''''''''''''''' ''''''''''''' '''''
''''''' ''''''''''' '''' '''''''''''''''''''''' '''''''''''''''''''''''' '''''''''''''''''''''''''''' ''''''''''''' ''''''''''''''''''''''''' ''''''''''''''''''''''''''' '''''''
''''''''''''''''' ''''''''' ''''''''' '''' '''''''''''''' ''''''''''''' ''''' ''''''''''''''''''' '''''''''''''''''''''''''' ''''''''''' '''''' ''''''''''''''''''
''''''''' ''''''''''''''''''' ''''''''' ''''''' ''''''''''''' ''''' ''''''''''''''''''''''''''' ''''''''''''''''''''' '''''''''''''''''''''' '''''''''''''''' '''''''''''
'''''''' '''''''''''''''''' '''''''''''''''' ''''''''''''''' '''''''''''''' '''''''''''''''''''''''' '''''''''''' ''' '''''''''''''''''' ''''''''''' '''''''''' '''''''''''''
'''''''''' ''''''''''''''''''''''''' '''''' '' ''''''''''''''' ''''''' ''''''''''''''''''''''''' '''''''''''''''''''' ''''''' '''''''''''''''''''' ''''''''''''''''''
'''''''''''''' '''''''''''''''''''' '''' '''''''''''''''''''''''''' ''''''''''''''''''''''''' '''''''''''''''' '''''''''' '''''''' ''''''''''''''' '''''''''''''''''''
''''''''''' ''''''' ''''''''''''''' ''''''''''''''' '''''''''''''''' ''''''''''''' ''''''''''' ''''''''''''''''''' ''''''''''' ''''' '''''''' '''''''''''''' '''''
95
''''''''''''''''
9.29
We have not independently verified these allegations. However, in order to examine their
potential impact on our results, we have modelled the impact of increasing the allocated
costs shown in Forum’s management accounts by ''''''' per cent. Table 9 sets out our updated
results based on this sensitivity.
Table 9: Forum’s discounted incremental profits after allocated cost adjustments (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: Forum’s allocated costs (“toegedeelde koste”) increased by ''''''''''. All figures shown on discounted basis as of
January 2004.
95
Factual Witness Statement of Mr Bonthuyzen, paras. 29-31 (emphasis added).
48
9.30
Increasing the allocated costs of Forum has the impact of increasing the incremental losses
of Forum, especially under the second definition of Forum incremental costs (total costs)
which takes into account allocated costs to a greater extent. Based on the adjustment,
Forum was incrementally loss-making in NPV terms over the complaint period, under both
definitions of avoidable costs. The extent of losses amounts to R '''''''''''''' under the first
definition of Forum incremental costs (all costs that could be eliminated over a period of one
year) and R '''''''''''''''' under the second definition (total costs).
9.31
Based on annual results, Forum showed positive incremental profits of R '''''''''''''''' in 2004
under the first definition of avoidable costs, due to the significant exit costs avoided by
keeping Forum open. In later years, when Media24 would not have incurred exit costs in the
counterfactual scenario, the incremental profits of Forum were consistently negative.
Other sensitivities
9.32
In addition to the corrections to transfer prices set out above, we have also carried out
sensitivity analyses on several other inputs in order to assess the sensitivity of our
conclusions to changes in inputs. We find that our conclusions are not affected by
reasonable changes to the diversion ratio from Forum to Vista and the discount rate. The
results of these analyses are set out in Annex F.
Conclusion
9.33
Our test of Forum’s incremental profitability shows that Forum was loss-making on an
incremental basis during the complaint period. A predatory pricing assessment based on this
test would therefore conclude (subject to other evidentiary requirements) that Media24 was
engaged in predatory pricing.
49
Section 10
Comparison of Forum’s revenues and
its avoidable costs
Introduction
10.1
This section sets out our comparison of Forum’s revenues and its avoidable costs, our
intermediate price-cost test.
Methodology
10.2
As set out above, our framework involves a comparison of Forum’s total revenues against
Forum’s costs that could have been avoided over the complaint period.
Forum’s revenues
10.3
Forum’s revenues for the complaint period are set out in its management accounts.
Forum’s avoidable costs
10.4
Given that we do not have information on Forum’s avoidable costs over a time period
equivalent to the complaint period, we have implemented the comparison based on Forum’s
1 year avoidable costs. For our base case, we have based this definition on the cost
96
classifications indicated by Media24.
10.5
We understand that Media24’s classifications reflect the cost factors that can be eliminated
over a period of 1 year. This is likely to be an underestimate of the total 1 year avoidable
costs of Forum because from an economic viewpoint, costs can be avoided by either
eliminating them entirely or reallocating them to an alternative productive use.
10.6
The magnitudes of Forum’s 1 year avoidable costs were calculated based on Forum’s
management accounts.
Adjustments to Forum’s printing costs prior to April 2008
97
10.7
For the reasons explained the previous section, we have adjusted Forum and Vista’s
printing costs prior to April 2008 upwards by ''''''' per cent in order to reflect the market prices
of printing services.
96
See Annex H for details of cost classifications.
97
See paras. 9.19-9.22.
50
Discounting
10.8
We have compared Forum’s revenues and 1 year avoidable costs based on net present
value (NPV) bases. We have used a discount rate of 18 per cent per annum, based on the
pre-tax nominal discount rates shown in the FY2005 financial accounts of Naspers for its
98
print media subsidiaries.
Base case results
10.9
Table 10 sets out a comparison of Forum’s revenues and 1 year avoidable costs based on
the methodology set out above.
Table 10: Forum’s revenues and avoidable costs (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
10.10 As set out above, Forum’s revenues were lower than Forum’s 1 year avoidable costs in NPV
terms during the complaint period. Over the entire complaint period, the magnitude of the
difference amounted to R '''''''''''''''''' in NPV terms as of January 2004.
10.11 Based on year-by-year comparisons, Forum’s revenues were lower than Forum’s 1 year
avoidable costs in 2004, 2005, 2007 and 2008, and higher than Forum’s 1 year avoidable
costs in 2006 and the first two months of 2009.
Sensitivities based on alternative definition of 1 year avoidable costs
10.12 The Commission has previously set out its own views on which of Forum’s costs could be
99
avoided over a 1 year time period, which differ from the submissions of Media24.
10.13 We are not able to comment on whether Media24’s or the Commission’s view on the 1 year
avoidable costs of Forum is more appropriate. However, we have repeated our assessment
of Forum’s pricing based on the Commission’s definition of Forum’s 1 year avoidable costs.
Table 11 sets out our updated results based on the Commission’s definitions.
98
See Annex I.
99
See Commission’s Complaint Referral, para. 44 and TF6.
51
Table 11: Forum’s revenues and avoidable costs based on Commission’s 1 year
avoidable cost definitions (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
10.14 As set out above, based on the Commission’s definition of 1 year avoidable costs, Forum’s
revenues are shown to have been significantly and consistently lower than the 1 year
avoidable costs of Forum throughout the complaint period. Over the entire complaint period,
the magnitude of the difference amounted to R '''''''''''''' in NPV terms as of January 2004.
Other sensitivities
10.15 In addition to the adjustments to transfer prices set out above, we have also carried out
sensitivity analyses on the discount rate used in our analysis in order to assess the
sensitivity of our conclusions. We find that our conclusions are not affected by reasonable
changes to the discount rate. The results of these analyses are set out in Annex G.
Conclusion
10.16 Based on the results set out above, we conclude that Forum’s revenues during the complaint
period did not cover the 1 year avoidable costs of Forum based on either Media24’s
definition or the Commission’s definition. A predatory pricing assessment based on this test
would conclude (subject to other evidentiary requirements) that Media24 was engaged in
predatory pricing.
52
Section 11
Anti-competitive effect of Media24’s
pricing conduct
Introduction
11.1
In this section we examine the anti-competitive effects of Media24’s pricing conduct based
on two approaches:
Actual ex post evidence of anti-competitive effects observed during the post-complaint
period; and
The degree to which any exclusion found under the two tests described above is
100
indicative of anti-competitive effects ex ante, given the characteristics of the market.
Actual evidence of anti-competitive effects
11.2
Gold-Net News exited the market in April 2009, soon after the end of the complaint period.
Based on the input of factual witnesses, we understand that there has been no material entry
into the market since the exit of Gold-Net News and therefore that Media24 has become the
101
monopoly supplier in the relevant market.
11.3
The Commission contends that the exit of Gold-Net News was the direct result of Media24’s
pricing conduct. The results of our price-cost tests set out above are consistent with this
insofar as they show that Media24’s pricing of Forum involved a profit sacrifice and would
have been capable of excluding an as-efficient competitor.
11.4
The exit of Gold-Net News would have caused anti-competitive harm via two channels:
Reduction in consumer welfare caused by recoupment on the part of Media24; and
The loss of consumer choice due to the exit of Gold-Net News.
11.5
We analyse both of these channels below.
100
The ex ante effects are the anti-competitive effects of Media24’s pricing conduct that would have been
expected by Media24 before and during the complaint period. The ex post effects refer to evidence of
actual anti-competitive effects of Media24’s pricing conduct. We also make this distinction between ex
ante and ex post when assessing evidence of recoupment in Section 13.
101
See Factual Witness Statement of Mr Steyl, paras. 25-26.
53
Recoupment by Media24 after the complaint period
11.6
We find evidence that after the exit of Gold-Net News, Media24 engaged in recoupment of
the losses incurred by Forum during the complaint period. This would have had the effect of
reducing consumer welfare. We examine this issue fully in Section 13.
Loss of consumer choice due to the exit of Gold-Net News
Previous studies of consumer welfare impacts of new product introductions
11.7
In a differentiated market, consumers benefit from having a wider variety of products to
choose from. This is referred to as the ‘variety effect’. Several studies have estimated the
magnitude of the variety effect resulting from the introduction of new products, and found
102
them to be significant.
11.8
While the magnitude of the welfare impacts found in these studies to not apply directly to the
present case, they illustrate that additional choice in the market can significantly benefit
consumers. Conversely, a loss of choice in the market can potentially cause significant
consumer harm.
Consumer welfare impacts resulting from the exit of Gold-Net News
11.9
The exit of Gold-Net News in April 2009 reduced the number of community newspapers
available to newspaper readers in the Goldfields area and would therefore have caused
consumer harm via the variety effect.
11.10 We do not have a basis for directly estimating the magnitude of the consumer harm arising
from the loss of choice. However, it is likely that the consumer harm in the present case is
significant given the following considerations:
Newspapers are highly differentiated products due to differences in editorial content
and the range of advertisements carried. As a result, consumers are likely to value the
existence of additional choices highly, and the consumer harm arising from the removal
of a competitor is likely to be greater.
The number of alternative choices available to consumers in the relevant market is
small. After the exit of Gold-Net News, there were only two community newspapers
remaining in the market (Forum and Vista), both of which were owned by Media24 and
one of which also closed shortly afterwards.
102
These include studies of new product introductions in breakfast cereals (Hausman 1997), tissues
(Hausman and Leonard 2002), motor vehicles (Petrin 2002), satellite TV services (Goolsbee and Petrin
2004) and grocery retail (Hausman and Leibtag 2007).
54
Ex ante likelihood of anti-competitive effects
11.11 On an ex ante basis, we consider that the exit of Gold-Net News from the market as a result
of Media24’s pricing conduct would likely have resulted in anti-competitive effects. Given the
barriers to entry already present in the market, and those arising from Media24’s conduct,
the exit of Gold-Net News would likely result in increased market power for Media24.
11.12 Potential barriers to entry in this case include the reputation effect, and the network effects
arising from the two-sided nature of the market.
Reputation effect
11.13 By forcing Gold-Net News out of the market based on an extended period of predatory
pricing, Media24 would likely have gained a reputation for being a tough competitor. This
reputation acts as a barrier to entry for potential entrants, which gives Media24 additional
market power following its predatory pricing conduct. This issue is relevant to the
assessment of Media24’s recoupment and is discussed fully in Section 13.
Two-sided market
11.14 Two-sided markets are markets where firms serve two distinct groups of customers and
each group benefits from the firm’s ability to attract the other group of customers. In the case
of community newspapers, firms serve both readers and advertisers, and both readers and
103
advertisers benefit from the ability of newspapers to attract the other group of customers.
This feature of two-sided markets creates a network effect. For both readers and advertisers,
the value of a community newspaper product increases as the number of the other group of
customers of that newspaper increases.
11.15 This creates barriers to entry for a potential entrant, since it faces the “chicken and egg”
problem that it requires a wide readership in order to attract advertisers but may not be able
to establish a wider readership in the absence of advertisers. In this regard, entrants face the
particular difficult of having to attract readers and advertisers away from an incumbent which
104
already has an established base of readers and advertisers.
103
Advertisers want to reach a wide readership and benefit from additional readers. Readers may read the
newspaper primarily for editorial content but may also benefit from informative advertisements.
104
Network effects in two-sided markets were recognised as a barrier to entry to markets for operating
systems in Microsoft (United States v. Microsoft Corporation, Civil Action No. 98-1232 (CKK)). In this
market, operating system manufacturers face a similar problem of attracting both end users and
application developers simultaneously. The Court found in Microsoft that the established base of
applications available for Microsoft Windows acted as a barrier to entry for potential entrants into the
operating system market. This was referred to as the “applications barrier to entry”. See the Court’s
Findings of Fact in United States v. Microsoft Corporation, 5 November 1999.
55
Conclusion
11.16 The structure of the relevant market and the nature of Media24’s pricing conduct suggest
that Media24’s pricing conduct would likely have resulted in anti-competitive effects ex ante.
11.17 The available evidence on the post-complaint period shows that there has been no new
entry into the market and that Media24 has engaged in recoupment of the losses incurred by
Forum during the complaint period, showing anti-competitive effects ex post.
56
Section 12
Media24’s intent to exclude Gold-Net
News
Introduction
12.1
In this section, we have examined the evidence of Media24’s intent to force Gold-Net News
out of the market.
12.2
We are aware that the Commission has put forward certain evidence of Media24’s intent to
exclude Gold-Net News from the market, including:
105
Several internal Media24 documents;
and
The Factual Witness Statement of Mr Bonthuyzen, a former employee of Media24.
12.3
In this section, we examine the available economic evidence of Media24’s intent.
Economic evidence of intent
12.4
We set out in the previous section evidence showing that Forum’s incremental revenues did
not cover its incremental costs. However, such evidence may not be indicative of predatory
intent if it was due to unexpected underperformance by Forum or overoptimistic business
forecasts by Media24. We have therefore examined whether Media24 could plausibly have
expected Forum to become profitable on an incremental basis.
12.5
As discussed above, ex ante forecasts cannot be a basis for the price-cost test of predation
106
but may be a relevant factor in the assessment of objective justification. Accordingly, we
consider the analyses set out below to be relevant to the assessment of Media24’s intent but
not a potential alternative to our test of predation set out in Section 9.
Overview
12.6
We examine below whether Media24 could have forecast Forum to be incrementally
profitable if it had based its forecasts on any one of a number of potential ‘representative
years’.
105
For instance, '''''' ''''''''''''''''' ''''''''''''''''''' '''''''''''''''''''''' '''''''''''''' '''''' ''''''''' ''''''''''' ''''''''''''''' ''''''''' '''''''''''''' ''''''''''''''''''''' '''''' '''
'''''''''''''''''''' '''' ''''''''' ''''''''''''''''' ''''''' ''''''''''' '''''''''''''''''''''' ''''''''' '''''''''''''''''''''''' ''''''''''''''''''' ''''''''''''' ''''''''''' '''''''''''''''''''' ''''' '''''''''''''''''''.
See discovery document RD103.
106
See paras. 7.33-7.34.
57
12.7
In each case, we take Forum’s performance over a selected 12 month period and examine
whether Forum would be profitable on an incremental NPV basis if Forum’s performance
over that 12 month period were to be repeated (i) over a period of 5 years and 2 months, the
length of the complaint period; and (ii) over perpetuity.
12.8
The NPV of Forum’s incremental profit under these scenarios shows whether it would have
been a rational business decision for Media24 to continue to operate Forum over a period of
5 years and 2 months or over perpetuity, as the case may be, if it had based its forecasts for
Forum on Forum’s performance over a certain time period.
2004 actuals extrapolated
12.9
We have considered whether Forum would have been profitable on an incremental NPV
basis if Forum’s actual revenues and costs during the calendar year 2004 had been
repeated either over the complaint period or to perpetuity. This addresses the possibility that
Media24’s management could forecast Forum’s performance over 2004 accurately but were
uncertain regarding Forum’s performance after this period, and therefore made the decision
to continue to operate Forum on the basis of extrapolating Forum’s 2004 performance over
the complaint period or for an indefinite period of time.
12.10 Table 12 shows the results of our analysis based on this extrapolation.
107
Table 12: Forum’s discounted incremental profits based on extrapolation of 2004
actual revenues and costs (R millions)
Incremental cost
definition 1
Incremental cost
definition 2
Incremental revenues
''''''''''''''
''''''''''''''
Incremental costs
''''''''''''''
''''''''''''''
Incremental profit excluding exit costs
''''''''''''''
''''''''''''''
Avoided exit costs
''''''''''''''
''''''''''''''
NPV incremental profit
(Extrapolation to complaint period)
''''''''''''''
''''''''''''''
NPV incremental profit
(Extrapolation to perpetuity)
''''''''''''''
''''''''''''''
2004 actuals:
Note: NPV based on Media24’s real discount rate at the start of the complaint period.
12.11 As shown above, based on the extrapolation described above, Forum would have been lossmaking on an incremental NPV basis under both definitions of incremental costs.
107
The NPV calculations in this extrapolation and all following extrapolations are based on Media24’s
nominal discount rate for FY2005 (see para. 9.23) and PPI inflation rate for CY2004. They are
therefore based on Media24’s real discount rate at the start of the complaint period.
58
2005, 2006, 2007 or 2008 actuals extrapolated
12.12 We have also considered whether Forum would have been profitable on an incremental NPV
basis if the extrapolation exercise described above had been applied to the performance of
Forum in 2005, 2006, 2007 and 2008. These analyses examine whether Media24’s
management could have forecast Forum to have been incrementally profitable on a longterm basis at any point during the complaint period if they had based their decisions on the
short-term forecasts of Forum’s performance.
12.13 Table 13 shows the results of our analysis based on extrapolation of 2005, 2006, 2007 and
2008 Forum performance.
Table 13: Forum’s discounted incremental profits based on extrapolation of 2005,
2006, 2007 and 2008 actual revenues and costs (R millions)
Basis for
extrapolation
Extrapolation period
Incremental cost
definition 1
Incremental cost
definition 2
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
2005 actuals
2006 actuals
2007 actuals
2008 actuals
Note: NPV based on Media24’s real discount rate at the start of the complaint period.
12.14 As shown above, based on the extrapolation described above, Forum would have been lossmaking on an incremental NPV basis for all extrapolations.
FY2005 budget extrapolated
12.15 We have considered whether Forum would have been profitable on an incremental NPV
basis if the extrapolation described above had been applied to Forum’s FY2005 budget,
which covers the period April 2004 to March 2005. This examines whether operating Forum
for the complaint period would have made business sense based on a budget which would
have been prepared around the time of the start of the complaint period.
12.16 Table 14 shows the results of our analysis based on this extrapolation.
59
Table 14: Forum’s discounted incremental profits based on extrapolation of FY2005
budget (R millions)
Incremental cost
definition 1
Incremental cost
definition 2
Incremental revenues
'''''''''''''
''''''''''''''
Incremental costs
'''''''''''''
''''''''''''''
Incremental profit excluding exit costs
'''''''''''''
''''''''''''''
Avoided exit costs
'''''''''''''
''''''''''''''
NPV incremental profit
(Extrapolation to complaint period)
'''''''''''''
''''''''''''''
NPV incremental profit
(Extrapolation to perpetuity)
'''''''''''''
''''''''''''''
FY2005 budget:
Note: NPV based on Media24’s real discount rate at the start of the complaint period.
12.17 As shown above, based on the extrapolation of the FY2005 budget, Forum would have been
profitable on an NPV basis based on extrapolation to the complaint period under the first
definition of Forum incremental costs (all costs that could be eliminated over a period of one
year). However, under all other sensitivities, Forum would have been loss-making on an
incremental NPV basis.
FY2006, FY2007, FY2008 and FY2009 budgets extrapolated
12.18 We have repeated the extrapolation exercise based on Media24’s budgets over financial
years 2006, 2007, 2008 and 2009, which together cover the period April 2005 to March
2009. As before, this examines whether Media24’s management could have forecast Forum
to have been incrementally profitable on a long-term basis at any point during the complaint
period if they had based their decisions on Forum’s annual budgets.
12.19 Table 13 shows the results of our analysis based on extrapolation of FY2006, FY2007,
FY2008 and FY2009 Forum performance.
60
Table 15: Forum’s discounted incremental profits based on extrapolation of FY2006,
FY2007, FY2008 and FY2009 budgets (R millions)
Basis for
extrapolation
Extrapolation period
Incremental cost
definition 1
Incremental cost
definition 2
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
5 years 2 months
'''''''''''''''
'''''''''''''''
Perpetuity
'''''''''''''''
'''''''''''''''
FY2006 budget
FY2007 budget
FY2008 budget
FY2009 budget
Note: NPV based on Media24’s real discount rate at the start of the complaint period.
12.20 As shown above, based on extrapolations of the FY2006, FY2007, FY2008 and FY2009
budgets, Forum would have been profitable on an NPV basis under the first definition of
Forum incremental costs (all costs that could be eliminated over a period of one year) but
loss-making under the second definition of Forum incremental costs (total costs).
Conclusion
12.21 We have analysed the incremental profitability of Forum based on actual and budgeted
performance of Forum on the basis that the management of Media24 may have used these
indicators to forecast Forum’s performance. We find that:
Under all forecasts based on the actual performance of Forum, Forum would have been
loss-making on an incremental basis; and
Under forecasts based on Media24’s budget, Forum would have been generally
profitable based on the first definition of Forum incremental costs (all costs that could
be eliminated over a period of one year), but generally loss-making based on the
second definition (total costs).
12.22 Taken together, these results suggest that the incremental losses of Forum could not have
been due to any unexpected underperformance by Forum. While Forum’s budgets may have
predicted marginal incremental profits for Forum based on costs that could be avoided over
a period of several years, a reasonable forecast based on the actual performance of Forum
would have suggested that Forum would not be profitable on an incremental basis and that
Media24 would be better off closing Forum.
12.23 The fact that Media24 continued to operate Forum for an extended period despite these
losses suggests that the decision to operate Forum was a result of exclusionary intent rather
than normal business rationale.
61
Section 13
Recoupment
Introduction
13.1
In accordance with our instructions, we have examined the evidence for recoupment of
Forum’s losses during the complaint period by Media24.
13.2
Evidence of recoupment may take the form of price increases after the competitor has been
forced to exit the market. We have examined whether this would have been a viable form of
recoupment ex ante by calculating the minimum levels of additional profits after the
complaint period that would be required to make Forum’s incremental losses during the
complaint period worthwhile. We then compare those levels with the ex post actual
108
performance of Media24 after the complaint period.
13.3
In the present case, the nature of the abuse and the structure of Media24 allow potential
additional recoupment via ‘reputation effects’. In other words, predatory pricing by Media24
may give it a reputation for being a tough competitor against local entrants, which benefits
Media24 in all areas where it operates a community newspaper.
Recoupment in the Goldfields area
Additional post-complaint period profits required to recoup Forum’s losses
13.4
We set out our estimates of Forum’s incremental losses during the complaint period in Table
8. We found that the incremental losses of Forum amounted to R '''''''''''''' under the first
definition of Forum incremental costs (all costs that could be eliminated over a period of one
year) and R ''''''''''''''' under the second definition (total costs) in NPV terms as of January
2004.
13.5
If these losses are to be recouped in NPV terms, they need to be exceeded by the NPV of
additional profits after the complaint period. Based on the discount rate of 18 per cent used
in our analyses, the required recoupment in NPV terms as of March 2009 would be R ''''''''''''''''
under the first definition of Forum incremental costs (all costs that could be eliminated over a
period of one year) and R '''''''''''''''' under the second definition (total costs).
108
The ex ante recoupment refers to the likely level of recoupment expected by Media24 before and
during the complaint period. The ex post recoupment refers to evidence of actual recoupment by
Media24 after the complaint period. We also make this distinction between ex ante and ex post in
assessing the anti-competitive effects of Media24’s pricing conduct in Section 11.
62
13.6
Table 16 sets out the levels of additional monthly profits that would be required in order to
recoup these losses.
Table 16: Additional monthly profits required to recoup Forum’s incremental losses
Forum incremental cost
definition 1
Forum incremental cost
definition 2
Required recoupment in
January 2004 NPV
R '''''''''''''
R '''''''''''''
Required recoupment in
March 2009 NPV
R '''''''''''''
R '''''''''''''
1 year
R '''''''''''''
R '''''''''''''
2 years
R '''''''''''''
R '''''''''''''
3 years
R '''''''''''''
R '''''''''''''
4 years
R '''''''''''''
R '''''''''''''
5 years
R '''''''''''''
R '''''''''''''
Perpetuity
R '''''''''''''
R '''''''''''''
Required level of monthly
profits if recoupment carried
out over:
Note: All monthly profit figures shown in March 2009 prices.
13.7
The level of additional monthly profits required for recoupment is lower under the first
definition of Forum incremental costs (because the required level of losses is lower) and
when recoupment is carried out over longer periods (because the recoupment can be spread
out over a greater number of months).
13.8
We understand that there has been no new entry into the community newspapers market in
the Goldfields area since the exit of Gold-Net News more than 4 years ago. With a
recoupment period of 4 years, Media24 could have recouped the entire losses incurred by
Forum during the complaint period based on additional profits of between '''''''''''''''''' and
''''''''''''''''''''''' per month. This equates to between '''''''' per cent and ''''''''''' per cent of Vista’s
average monthly revenues during the period March 2009 to December 2010.
13.9
Given that no new entry has yet occurred in the Goldfields area, the potential recoupment
period is longer than 4 years. If recoupment is possible over perpetuity, Media24 could
recoup the entire losses incurred by Forum during the complaint period based on additional
profits of between ''''''''''''''''''''' and '''''''''''''''''''''' per month. This equates to between '''''''' per cent
and '''''''' per cent of Vista’s average monthly revenues during the period March 2009 to
December 2010.
63
13.10 As an illustrative example, if Media24 had been able to implement a ‘small but significant
109
and non-transitory increase in price’ (“SSNIP”) of 5 to 10 per cent after the exit of Gold-Net
News, this would have generated additional revenues of between ''''''''''''''''''''' and ''''''''''''''''''''''
110
per month for Vista. Based on this price increase, Media24 could recoup the entirety of
Forum’s losses over a period of less than two years (based on the first definition of Forum’s
incremental costs) or over perpetuity (based on the second definition of Forum’s incremental
costs).
13.11 Based on this evidence, we therefore consider that on an ex ante basis, there was a high
likelihood that Media24 would be able to recoup the losses incurred by Forum during the
complaint period after the exit of Gold-Net News.
Vista’s actual performance after the complaint period
13.12 Calculating the actual level of recoupment requires a comparison of Media24’s actual profits
in the post-complaint period against counterfactual profits that Media24 would have earned
during the same period if it had not engaged in predatory pricing against Gold-Net News. In
this case, Forum would have exited the market at the start of the complaint period and,
insofar as that the exit of Gold-Net News was a direct result of Media24’s pricing conduct,
Gold-Net News would have continued to operate in the post-complaint period.
13.13 Given the extended length of the alleged abuse, we would consider that it is highly
speculative to directly estimate the state of the market in the post-complaint period if
Media24 had not engaged in the allegedly abusive conduct. As such, we have not tried to
directly estimate the actual level of recoupment achieved by Media24.
13.14 In order to inform the assessment of recoupment, we have instead examined the actual
performance of Vista after the complaint period and compared that with Vista’s performance
during the complaint period. While this is not a strictly accurate measure of recoupment, it
indicates the degree to which the exit of Gold-Net News is likely to have increased Vista’s
profits. This can be compared to the levels of required recoupment calculated in the previous
section as a further check on whether full recoupment of Forum’s losses is plausible.
13.15 Table 17 compares the monthly average profits of Vista during and after the complaint
period.
109
SSNIP is a minimum threshold typically used in competition cases to determine whether a firm has
market power. Given Media24’s dominance in the relevant market, it would be reasonable to assume
that it is capable of implementing this price change, especially in response to the exit of the only
competitor in the market.
110
Calculated based on Vista’s revenues during the period March 2009 to December 2010.
64
Table 17: Vista’s average monthly profits
Time period
Complaint period
(Jan 2004-Feb 2009)
Post-complaint period
Average monthly profits (R)
Increase compared to
complaint period average
'''''''''''''''
'''''''''''''''
'''''''''''''''
Mar 2009-Feb 2010
'''''''''''''''
'''''''''''''''
Mar 2010-Dec 2010
'''''''''''''''
'''''''''''''''
(Mar 2009-Dec 2010)
Notes: All revenue and price figures deflated by PPI and shown in March 2010 prices.
Source: Forum management accounts
13.16 The data provided by Media24 sets out Vista’s monthly profits to December 2010. A
comparison of Vista’s profits during the complaint period and the post-complaint period
shows that Vista’s average monthly profits increased by ''''''''''''''''''' in real terms after the end
of the complaint period.
13.17 This comparison, however, hides a significant variation in Vista’s profits during the postcomplaint period. As shown above, during the first 12 months after the end of the complaint
period in February 2009, Vista’s average monthly profits decreased by '''''''''''''''''''''' compared
to the previous 12 months. After the first 12 months of the post-complaint period, however,
Vista’s profits recovered to levels much higher than the complaint period. During the period
March 2010 to December 2010, Vista’s average monthly profits were ''''''''''''''''''''' higher than
the complaint period average.
13.18 For the purposes of assessing recoupment by Media24, we consider that it is more
appropriate to examine the period March 2010 to Dec 2010 than the period March 2009 to
February 2010. This is for two reasons:
Vista’s profitability for the period March 2009 to February 2010 may have been
disproportionately affected by the onset of the financial crisis, and may not be a good
basis for assessing Vista’s long-term profitability; and
Media24 continued to operate Forum until January 2010. As a result, the period March
2009 to February 2010 may not be representative of the long-term business model of
Media24, which involved operating Vista alone in the Goldfields area.
13.19 Compared to the required levels of recoupment shown in Table 16, an improvement in
Vista’s profitability of '''''''''''''''''''' per month is:
Large enough to recoup all incremental losses incurred by Forum under the first
definition of Forum’s incremental costs over a period of less than 2 years; and
Large enough to recoup all incremental losses incurred by Forum under the second
definition of Forum’s incremental costs over perpetuity.
65
13.20 Based on this evidence, we consider that Media24’s increased market power after the exit of
Gold-Net News would likely allow it to recoup Forum’s losses during the complaint period.
Reputation effect
13.21 By forcing Gold-Net News out of the market based on an extended period of predatory
pricing, Media24 may also have gained a reputation for being a tough competitor against
entrants. Benefits arising from this reputation, in the form of reduced threat of new entry,
care another potential source of recoupment for Media24.
13.22 The possible impact of Media24’s conduct on its reputation to potential entrants was noted
by Mr Steyl:
“Since [Gold-Net News] was forced out of the market in 2009, I have not reentered the market with another community newspaper title, and am unaware
of any other material entry. I have no reason to believe that Media24 would
behave any differently in forcing other competitors out of the market as it did
with [Gold-Net News]. Media24 has shown itself to be a firm that is willing and
able to engage in highly aggressive pricing in order to protect its market
position, and that has disincentivized me (and presumably also others) from
111
attempting to compete in the market with a new title.”
13.23 This is consistent with the academic literature on ‘reputation’ models of predatory pricing,
where a single instance of predatory pricing establishes an expectation among future
potential entrants that the incumbent firm would also respond aggressively to their entry.
This discourages potential entrants from entering the market in the future and allows the
112
incumbent to exercise market power.
13.24 The benefits of the reputation effect arise in a wider area than the Goldfields. Media24 may
see lower likelihood of new entry in all areas where it operates a community newspaper, as
potential entrants consider the possibility of an aggressive response from Media24 based on
the example in the Goldfields area and decide on that basis against launching new
community newspapers.
111
Factual Witness Statement of Mr Steyl, para. 25.
112
See para. 6.4.
66
Annex A
Authors’ CVs
Neil Dryden
Neil Dryden is an Executive Vice President in Compass Lexecon’s European competition
policy practice, based in the firm’s London office. He has worked as a professional
economist for over 15 years and during that time he has advised on numerous mergers,
agreement cases, dominance cases and market investigations.
Neil’s significant cases since 2010 include acting as an expert in the pay TV and tobacco
cases (both at the Competition Appeal Tribunal), and advising in outdoor advertising (OFT),
Sports Direct/JJB (Competition Commission and Competition Appeal Tribunal), Asda/Netto
(OFT) and Level 3/Global Crossing (OFT).
In addition to numerous cases in the European Union, Neil has advised on cases in India
and South Africa. Neil also has extensive experience in regulatory economics, including a
series of projects for the UK postal regulator. Neil has prepared submissions in the context of
a number of UK government inquires including the Barker review of land use planning.
Neil has analysed cases in sectors including advertising, banking and financial services,
chemicals, energy, FMCG, grocery retailing, healthcare, manufacturing, media and
broadcasting, mining, petroleum, pharmaceuticals, postal services, publishing, scientific
instruments, sports, technology, telecommunications, tobacco, transport, and water.
Neil was educated at Oxford University where he obtained a B.A. in Philosophy, Politics and
Economics (first class) and an M.Phil. in Economics, and held a college lectureship for two
years. At King’s College, London, he obtained a postgraduate diploma in EC competition law
(with distinction). Neil co-authored “What makes firms perform well?” published in the
European Economic Review.
Prior to joining Compass Lexecon, Neil worked as a Director at LECG and prior to that as an
Associate Director in NERA’s European competition policy practice for seven years. Neil
spent the first six years of his career in Arthur Andersen’s economic and financial consulting
practice, where he was a Senior Manager.
Professional experience
April 2013 – Present, Executive Vice President, Compass Lexecon
March 2011 – March 2013, Senior Vice President, Compass Lexecon
2010 – 2011, Director, LECG
2003 – 2010, Associate Director, NERA Economic Consulting
67
2002 – 2003, Assistant Director, Deloitte & Touche LLP
1996 – 2002, Senior Manager, Arthur Andersen
Education
King’s College
Postgraduate Diploma in EC Competition Law (distinction), 2004
University of Oxford,
MPhil in Economics, 1996
University of Oxford
BA Philosophy, Politics and Economics (first class), 1994
Full CV is available from Compass Lexecon website (www.compasslexecon.com).
68
Dr. Jorge Padilla
Dr. Jorge Padilla is a Senior Managing Director and Head of Compass Lexecon Europe.
Dr. Padilla earned M. Phil and D. Phil degrees in Economics from the University of Oxford.
He is Research Fellow at the Centro de Estudios Monetarios y Financieros (CEMFI, Madrid)
and teaches competition economics at the Barcelona Graduate School of Economics
(BGSE).
He has advised on various cases and given expert testimony before competition authorities
and courts of several EU member states, as well as in cases before the European
Commission. Dr. Padilla has submitted written testimony to the European General Court and
the UK Competition Appeals Tribunal in cartel, merger control and abuse of dominance
cases. He has also given expert testimony in various civil litigation (damages) and
international arbitration cases.
Dr. Padilla has written numerous papers on competition policy and industrial organization in
the Antitrust Bulletin, the Antitrust Law Journal, the Economic Journal, the European
Competition Journal, the European Competition Law Review, the European Economic
Review, the Fordham International Law Journal, Industrial and Corporate Change, the
International Journal of Industrial Organization, the Journal of Competition Law and
Economics, the Journal of Economics and Management Strategy, the Journal of Economic
Theory, the RAND Journal of Economics, the Review of Financial Studies, the University of
Chicago Law Review, and World Competition. He is also co-author of The Law and
Economics of Article 82 EC, Hart Publishing, 2006.
Professional experience
2011 – Present, Senior Managing Director, COMPASS LEXECON
2007 – Present, Associate Professor of Competition Economics, Barcelona Graduate
School of Economics (BGSE)
2003 – Present, Research Fellow, Centro de Estudios Monetarios y Financieros
(CEMFI)
2004 – Present, Member of the Scientific Committee of FEDEA
2009 – 2011, European Chief Executive Officer, LECG
2004 – 2011, Senior Managing Director, LECG
1998 – 2004, Director and Managing Director, NERA
1993 – 2011, Research Fellow (Industrial Organization), Centre for Economic Policy
Research (CEPR),
1991 – 2003, Professor of Economics, Centro de Estudios Monetarios y Financieros
(CEMFI)
69
Education
University of Oxford, Nuffield College
D Phil in Economics, 1992
Thesis title: “Essays on the Economics of Consumer Switching Costs” Advisor: Dr. Paul
D. Klemperer
University of Oxford, St. Antony’s College
M Phil in Economics. (George Webb Medlay Medal for Best Performance in
Written Exams, Prize for Best M. Phil Thesis), 1990
University of Alicante, Spain
BA in Economics (with Honours), 1988
Full CV is available from Compass Lexecon website (www.compasslexecon.com).
70
Annex B
Sources of information
B.1
We have reviewed the following sources of information in preparing this report.
Competition law and guidelines
South African Competition Act 1998
Article 102 (formerly Article 82) of the Treaty on the Functioning of the European Union,
Official Journal of the European Union, C115.89, 9.5.2008
Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC
Treaty to abusive exclusionary conduct by dominant undertakings, Official Journal of
the European Union, 2009/C45/02
Case law
South African Airways (Proprietary) Limited, South African Express Airways
(Proprietary) Limited and South African Airlink (Proprietary) Limited v Nationwide
Airlines (Proprietary) Limited, Nationwide Air Charter (Proprietary) Limited, Nationwide
Aircraft Maintenance (Proprietary) Limited and Nationwide Aircraft Support (Proprietary)
Limited, Competition Tribunal, Case no. 92/IR/Oct00 (“Nationwide”)
The Competition Commission v South African Airways (Pty) Ltd, Competition Tribunal,
Case no. 18/CR/Mar01 (“South African Airways”)
Mandla-Matla Publishing (Pty) Ltd v Independent Newspapers (Pty) Ltd, Competition
Tribunal, Case no. 48/CR/Jun04 (“Mandla-Matla”)
Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359 (“AKZO”)
Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755 (“Tetra
Pak”); Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951
(“Tetra Pak II”)
UK Office of Fair Trading ‘Predation by Aberdeen Journals Ltd’ CA98/5/2001, Decision
dated 16 July 2001; Case 1005/1/1/01 ‘Aberdeen Journals Limited v. Director General
of Fair Trading’, CAT Judgment dated 19 March 2002; UK Office of Fair Trading
‘Predation by Aberdeen Journals Limited (Remitted case)’ CA98/14/2002, Decision
dated 16 September 2002; Case 1009/1/1/02 ‘Aberdeen Journals Limited v. The Office
of Fair Trading’, CAT Judgment dated 23 June 2003 (“Aberdeen Journals”).
71
UK Office of Fair Trading ‘Abuse of a dominant position by Cardiff Bus’ CA98/01/2008,
Decision dated 18 November 2008 (“Cardiff Bus”)
Case COMP/38.233, Wanadoo Interactive, EC Decision of 16 July 2003 (“Wanadoo”);
Case T-340/03, France Télécom SA v Commission of the European Communities,
Judgment of the Court of First Instance, 30 January 2007; Case C-202/07 P, France
Télécom SA v Commission of the European Communities, Judgment of the European
Court of Justice, 2 April 2009
Case COMP/38.784, Wanadoo España vs. Telefónica, EC Decision of 4 July 2007;
Case T-336/07, Telefónica SA and Telefónica de España SA v. European Commission
(“Telefónica”)
United States v. Microsoft Corporation, Civil Action No. 98-1232 (CKK); Court’s
Findings of Fact in United States v. Microsoft Corporation, 5 November 1999.
Academic articles and books
Bolton, Patrick and David Scharfstein (1990), “A Theory of Predation Based on Agency
Problems in Financial Contracting” American Economic Review 80(1), pp. 93-106.
Goolsbee, Austan and Amil Petrin (2004), “The Consumer Gains from Direct Broadcast
Satellites and the Competition with Cable TV” Econometrica 72(2), pp. 351-381.
Hausman, Jerry (1997), “Valuation of New Goods under Perfect and Imperfect
Competition” in Bresnahan, Timothy and Robert Gordon (eds), The Economics of New
Goods, University of Chicago Press 1997.
Hausman, Jerry and Ephraim Leibtag (2007), “Consumer Benefits from Increased
Competition in Shopping Outlets: Measuring the Effect of Wal-Mart” Journal of Applied
Econometrics 22, pp. 1157-1177.
Hausman, Jerry and Gregory Leonard (2002), “The Competitive Effects of a New
Product Introduction: A Case Study” Journal of Industrial Economics 50(3), pp. 237263.
Joskow, Paul and Alvin Klevorick (1979), “A Framework for Analysing Predatory Pricing
Policy” Yale Law Journal 89, pp. 213-270.
Kreps, David and Robert Wilson (1982), “Reputation and Imperfect Information” Journal
of Economic Theory 27(2), pp. 253-279.
McGee, John (1958), “Predatory Price Cutting: The Standard Oil (NJ) Case” Journal of
Law and Economics 1, pp. 137-169.
Milgrom, Paul and John Roberts (1982), “Predation, Reputation and Entry Deterrence”
Journal of Economic Theory 27(2), pp. 280-312.
Motta, Massimo (2004), Competition Policy: Theory and Practice, Cambridge University
Press 2004.
72
O’Donoghue, Robert and A. Jorge Padilla (2006), The Law and Economics of Article 82
EC, Hart Publishing 2006.
Petrin, Amil (2002), “Quantifying the Benefits of New Products: The Case of the
Minivan” Journal of Political Economy 110(4), pp. 705-729.
Documents related to the Commission’s case against Media24
Commission’s Complaint Referral, 31 October 2011
Respondent’s Answering Affidavit, 19 January 2012
Commission’s Supplementary Affidavit, 25 February 2013
RD20: ''''''''''''''''''''' ''''''''''''''''''''''''''''' '''''' '''''''''' ''''''''''''''''''''''''''''''' ''''''''''''''''' '''''' '''''' '''''''''''''
RD22: '''''''''''''''''''' ''''''''''''''''''''''''''''' ''''''' ''''''''' '''''''''''''''''''''''' '''''''''''''''''' '''''''''''''''' '''''' '''''' '''''''''''''''
RD103: ''''''''''''''''''' '''''''''''''''''' ''''''''''''''''''''' ''''''''''' ''''''''''''''''''''' '''''' '''''''''''''''''' '''''''''''' '''''' ''''''''' '''''''''''
RD134-140: '''''''''''''''' ''''''''' '''''''''''''''
RD261: ''''''''''''' '''''''''' ''''''''''''''''''''''''' ''''''''''' ''''''''' ''''''''''''' ''''' '''''''''''''''' '''''''''' '''''''' '''''''''''''' '''' '''''''''''''''''''''''''
''''''''''''
Factual Witness Statement of Ms Betts
Factual Witness Statement of Mr Bonthuyzen
Factual Witness Statement of Mr Steyl
Factual Witness Statement of Mr Botha
Factual Witness Statement of Ms Van Eck
73
Annex C
Data sources
C.1
We have relied on the following data sources for our analysis.
Data discovered by Media24
Management accounts
Forum management accounts (RD16, 120-129, 243-244 and 795)
Vista management accounts (RD7-16, 243-244, 796)
Advertising volumes
Forum advertising volumes data (RD141-150)
Vista advertising volumes data (RD30-39 and 794)
Market shares data
Media24’s market share monitoring data (RD604-630)
Cost classifications
Media24’s cost classifications (RD20-22)
Other data
Media24’s discount rate
Naspers Annual Report 2005
PPI
PPI for domestic output, all groups (PI000002), Statistics South Africa
74
Annex D
Timeline of events
D.1
Table 18 sets out a summary of relevant events in the community newspapers market in the
Goldfields area from the entry of Vista in 1971 to the present.
Table 18: Timeline of events
Date
Event
1971
Vista established as a community newspaper by Mr Steyl
1980
Vista becomes free newspaper
June 1980
September 1983
Vista sold to Perskor
Forum starts publication
1994
Mr Steyl leaves Vista
1996
Vista sold to Caxton Group
January 1999
1999
Media24 buys Vista
Net News (Goldfields) established by Ms Joubert
November 2000
Ms Joubert closes Net News and newspaper reopens as Goldnet News
under Mr Steyl
December 2001
Vista Tuesday edition (Vista Dinsdag) ceases circulation
''''''''''''''''' '''''''''''
'''''''''' '''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''
January 2004
Start of complaint period
February 2009
End of complaint period
April 2009
January 2010
Gold Net News exit
Forum exit
75
Annex E
European legislation, guidelines and
precedents on predatory pricing
European legislation
E.1
Article 102 of the EC Treaty does not contain a specific clause on predatory pricing.
However, the article prohibits any conduct by a dominant firm that has the effect of “limiting
113
production, markets or technical development to the prejudice of consumers”.
European guidance
E.2
The European Commission (“EC”) Guidelines on Article 102 give guidance on the EC’s
approach to assessing price-based exclusionary conduct in general, as well as predatory
pricing in particular.
Price-based exclusionary conduct
E.3
The EC states that it is likely to use average avoidable cost (“AAC”) or long-run average
incremental cost (“LRAIC”) cost standards in assessing price-based exclusionary conducts:
“The cost benchmarks that the Commission is likely to use are average
avoidable cost (AAC) and long-run average incremental cost (LRAIC). Failure
to cover AAC indicates that the dominant undertaking is sacrificing profits in
the short term and that an equally efficient competitor cannot serve the
targeted customers without incurring a loss. LRAIC is usually above AAC
because, in contrast to AAC (which only includes fixed costs if incurred during
the period under examination), LRAIC includes product specific fixed costs
made before the period in which allegedly abusive conduct took place. Failure
to cover LRAIC indicates that the dominant undertaking is not recovering all
the (attributable) fixed costs of producing the good or service in question and
114
that an equally efficient competitor could be foreclosed from the market.”
E.4
The EC also notes that in applying the AAC and LRAIC cost benchmarks, it may be
necessary to examine incremental revenues of the relevant product as opposed to the total
revenues.
113
Official Journal of the European Union, C115/89, 9.5.2008.
114
EC Article 102 Guidelines, para. 26.
76
“In order to apply these cost benchmarks it may also be necessary to look at
revenues and costs of the dominant company and its competitors in a wider
context. It may not be sufficient to only assess whether the price or revenue
covers the costs for the product in question, but it may be necessary to look at
incremental revenues in case the dominant company's conduct in question
negatively affects its revenues in other markets or of other products. Similarly,
in the case of two sided markets it may be necessary to look at revenues and
115
costs of both sides at the same time.”
Predatory pricing
E.5
The EC characterises predatory pricing as conduct that results in deliberate sacrifice of
profits in the short term.
“In line with its enforcement priorities, the Commission will generally intervene
where there is evidence showing that a dominant undertaking engages in
predatory conduct by deliberately incurring losses or foregoing profits in the
short term (referred to hereafter as ‘sacrifice’), so as to foreclose or be likely
to foreclose one or more of its actual or potential competitors with a view to
strengthening or maintaining its market power, thereby causing consumer
harm.
Conduct will be viewed by the Commission as entailing a sacrifice if, by
charging a lower price for all or a particular part of its output over the relevant
time period, or by expanding its output over the relevant time period, the
dominant undertaking incurred or is incurring losses that could have been
116
avoided.”
E.6
The EC states that AAC will be used as a starting point for assessment of profit sacrifice, but
that the concept of profit sacrifice is wider than this and encompasses all conduct that leads
to lower profits than a reasonable alternative:
“The Commission will take AAC as the appropriate starting point for assessing
whether the dominant undertaking incurred or is incurring avoidable losses.
[…]
However, the concept of sacrifice does not only include pricing below AAC. In
order to show a predatory strategy, the Commission may also investigate
whether the allegedly predatory conduct led in the short term to net revenues
lower than could have been expected from a reasonable alternative conduct,
that is to say, whether the dominant undertaking incurred a loss that it could
117
have avoided.”
E.7
The EC also discusses the use of average variable costs (“AVC”) and the long-run average
115
EC Article 102 Guidelines, para. 26, footnote 3.
116
EC Article 102 Guidelines, paras. 63-64.
117
EC Article 102 Guidelines, paras. 64-65.
77
incremental cost (“LRAIC”) standards:
“In most cases the average variable cost (AVC) and AAC will be the same, as
118
often only variable costs can be avoided.”
“If sufficient reliable data are available, the Commission will apply the equally
efficient competitor analysis, […], to determine whether the conduct is capable
of harming consumers. Normally only pricing below LRAIC is capable of
119
foreclosing as efficient competitors from the market.”
E.8
The EC notes that direct evidence of predatory intent in the form of internal documents may
be relevant to its analysis:
“In some cases it will be possible to rely upon direct evidence consisting of
documents from the dominant undertaking which clearly show a predatory
strategy, such as a detailed plan to sacrifice in order to exclude a competitor,
to prevent entry or to pre-empt the emergence of a market, or evidence of
120
concrete threats of predatory action.”
E.9
Conversely, the EC notes that in cases where the firm can provide conclusive proof that the
ex ante decision to engage in the conduct was based on reasonable expectation that the
conduct would be profitable, then the firm would not be penalised for incurring ex post
losses:
“However, undertakings should not be penalised for incurring ex post losses
where the ex ante decision to engage in the conduct was taken in good faith,
that is to say, if they can provide conclusive evidence that they could
121
reasonably expect that the activity would be profitable.”
E.10
The EC states that assessment of anti-competitive effects will take into account evidence of
actual foreclosure and evidence of exclusionary intent, as well as a consideration of how the
predatory pricing conduct reduces the likelihood that competitors will compete with the
dominant firm. This may include consideration of entry barriers and possibility of re-entry:
“The Commission will normally intervene under Article 82 where, on the basis
of cogent and convincing evidence, the allegedly abusive conduct is likely to
lead to anti-competitive foreclosure. The Commission considers the following
factors to be generally relevant to such an assessment: […]
118
EC Article 102 Guidelines, para. 64, footnote 3.
119
EC Article 102 Guidelines, para. 67. See also para. 26.
120
EC Article 102 Guidelines, para. 66.
121
EC Article 102 Guidelines, para. 65, footnote 3.
78
- possible evidence of actual foreclosure: if the conduct has been in place for
a sufficient period of time, the market performance of the dominant
undertaking and its competitors may provide direct evidence of anticompetitive foreclosure. For reasons attributable to the allegedly abusive
conduct, the market share of the dominant undertaking may have risen or a
decline in market share may have been slowed. For similar reasons, actual
competitors may have been marginalised or may have exited, or potential
competitors may have tried to enter and failed,
direct evidence of any exclusionary strategy: this includes internal documents
which contain direct evidence of a strategy to exclude competitors, such as a
detailed plan to engage in certain conduct in order to exclude a competitor, to
prevent entry or to pre-empt the emergence of a market, or evidence of
concrete threats of exclusionary action. Such direct evidence may be helpful
122
in interpreting the dominant undertaking's conduct.”
“[T]he Commission will generally investigate whether and how the suspected
conduct reduces the likelihood that competitors will compete. For instance, if
the dominant undertaking is better informed about cost or other market
conditions, or can distort market signals about profitability, it may engage in
predatory conduct so as to influence the expectations of potential entrants
and thereby deter entry. If the conduct and its likely effects are felt on multiple
markets and/or in successive periods of possible entry, the dominant
undertaking may be shown to be seeking a reputation for predatory conduct. If
the targeted competitor is dependent on external financing, substantial price
decreases or other predatory conduct by the dominant undertaking could
adversely affect the competitor's performance so that its access to further
123
financing may be seriously undermined.”
“Likely consumer harm may be demonstrated by assessing the likely
foreclosure effect of the conduct, combined with consideration of other
factors, such as entry barriers. In this context, the Commission will also
124
consider possibilities of re-entry.”
E.11
However, the EC does not consider it necessary to show the actual exit of competitors from
the market.
122
EC Article 102 Guidelines, para. 20.
123
EC Article 102 Guidelines, para. 68.
124
EC Article 102 Guidelines, para. 71.
79
“The Commission does not consider that it is necessary to show that
competitors have exited the market in order to show that there has been anticompetitive foreclosure. The possibility cannot be excluded that the dominant
undertaking may prefer to prevent the competitor from competing vigorously
and have it follow the dominant undertaking's pricing, rather than eliminate it
from the market altogether. Such disciplining avoids the risk inherent in
eliminating competitors, in particular the risk that the assets of the competitor
are sold at a low price and stay in the market, creating a new low cost
125
entrant.”
E.12
The EC states that it will consider claims that predatory pricing conduct results in
efficiencies, although it notes that the likelihood of efficiencies arising from predatory pricing
conduct is low.
“In general it is considered unlikely that predatory conduct will create
efficiencies. However, […] the Commission will consider claims by a dominant
undertaking that the low pricing enables it to achieve economies of scale or
126
efficiencies related to expanding the market.”
European precedents
AKZO
E.13
127
The European Court of Justice outlined in AKZO v Commission the two alternative tests by
which a dominant firm may be found to have engaged in predatory pricing:
“Prices below average variable costs (that is to say, those which vary
depending on the quantities produced) by means of which a dominant
undertaking seeks to eliminate a competitor must be regarded as abusive. A
dominant undertaking has no interest in applying such prices except that of
eliminating competitors so as to enable it subsequently to raise its prices by
taking advantage of its monopolistic position, since each sale generates a
loss, namely the total amount of the fixed costs (that is to say, those which
remain constant regardless of the quantities produced) and, at least, part of
the variable costs relating to the unit produced.
125
EC Article 102 Guidelines, para. 69.
126
EC Article 102 Guidelines, para. 74.
127
Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359 (“AKZO”).
80
Moreover, prices below average total costs, that is to say, fixed costs plus
variable costs, but above average variable costs, must be regarded as
abusive if they are determined as part of a plan for eliminating a competitor.
Such prices can drive from the market undertakings which are perhaps as
efficient as the dominant undertaking but which, because of their smaller
financial resources, are incapable of withstanding the competition waged
128
against them.”
Tetra Pak
E.14
129
In Tetra Pak, the Court of First Instance accepted the two tests set out in AKZO:
“[T]he existence of gross or semi-gross margins – obtained by subtracting
from the sale price the variable direct costs or the average variable costs,
being the costs relating to the unit produced – which are negative suggests
that a pricing practice is eliminatory. As the Court of Justice held in AKZO v
Commission, […] an undertaking in a dominant position has no interest in
applying prices below average variable costs (that is to say, those which vary
depending on the quantities produced) except that of eliminating competitors
so as to enable it subsequently to raise its prices by taking advantage of its
monopolistic position, since each sale generates a loss equal to the total
amount of the fixed costs (that is to say, those which remain constant
regardless of the quantities produced) and at least part of the variable costs
relating to the unit produced.
The Court of Justice also held in AKZO v Commission that if the net margin is
negative and the gross margin positive, that is to say, if the prices are below
average total costs (fixed costs plus variable costs), but above average
variable costs, those prices must be regarded as abusive if they are
130
determined as part of a plan for eliminating a competitor.”
E.15
This was also confirmed by the European Court of Justice in Tetra Pak II:
“In AKZO this Court did indeed sanction the existence of two different
methods of analysis for determining whether an undertaking has practised
predatory pricing. First, prices below average variable costs must always be
considered abusive. In such a case, there is no conceivable economic
purpose other than the elimination of a competitor, since each item produced
and sold entails a loss for the undertaking. Secondly, prices below average
total costs but above average variable costs are only to be considered
131
abusive if an intention to eliminate can be shown.”
128
AKZO, para. 71-72.
129
Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755 (“Tetra Pak”); Case C333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951 (“Tetra Pak II”).
130
Tetra Pak, paras. 148-149.
131
Tetra Pak II, para. 41.
81
Aberdeen Journals
E.16
132
In Aberdeen Journals, the UK Competition Appeal Tribunal accepted the two predatory
pricing tests used in AKZO and Tetra Pak:
“AKZO […] and Tetra Pak II […] show that pricing below average variable
costs by a dominant firm is normally to be regarded as an abuse. “Variable
costs” are those which vary with the unit of output produced as distinct from
“fixed costs” which do not vary with the output produced. An example of a
“fixed cost” might be the monthly rental of a company’s premises. Examples
of “variable costs” in the present case include, but are not necessarily limited
to, newsprint (the paper on which the newspapers are printed), distribution
(the cost of distributing the copies produced), and other costs such as ink,
plate and film charges, electricity, fuel for transport, overtime and “pre-press
production” costs […]. Thus, for example, to sell the Herald & Post below
average variable cost, as so defined, is to sell each copy of the newspaper for
less than the average cost of producing that copy. […] When undertaken by a
dominant firm, such conduct will normally constitute “recourse to methods
different from those which condition normal competition” within the meaning of
133
Hoffman-La Roche.”
“Similarly, AKZO […] and Tetra Pak II […] show that it may be an abuse by a
dominant undertaking to price between average variable cost and average
total cost, if the intention is to eliminate a competitor. Total costs are variable
costs plus fixed costs including, in our view, where appropriate, a share of
general overheads. In our view, pricing between average variable cost and
average total cost is likely to be abuse when undertaken in anticipation of
134
competitive entry or in order to undercut a new entrant.”
E.17
The CAT examined the distinction between fixed and variable costs in more detail, and
suggested that variable costs should be defined according to whether each type of cost is
variable over the course of the alleged abuse period:
132
OFT ‘Predation by Aberdeen Journals Ltd’ CA98/5/2001, Decision dated 16 July 2001; Case
1005/1/1/01 ‘Aberdeen Journals Limited v. Director General of Fair Trading’, CAT Judgment dated 19
March 2002; OFT ‘Predation by Aberdeen Journals Limited (Remitted case)’ CA98/14/2002, Decision
dated 16 September 2002; Case 1009/1/1/02 ‘Aberdeen Journals Limited v. The Office of Fair
Trading’, CAT Judgment dated 23 June 2003 (“Aberdeen Journals”).
133
Aberdeen Journals, para. 351.
134
Aberdeen Journals, para. 352.
82
“One particular aspect of the distinction between average variable costs and
average total costs, not yet explored in the case law, but relevant in the
present case, is the period of time over which costs are to be assessed as
“fixed” rather than “variable”. The longer the period that is taken, the more
likely it is that cost will be classified as variable since, for example, over a
longer timescale, employees can be dismissed or plant closed in response to
changes in output. Indeed, in the long run, almost all costs are “variable”.
[…] What the director should do, in the first instance, is to identify provisionally
the period over which pricing below cost is suspected. He should then take
that period and examine whether costs are variable over that period. We do
not exclude the possibility of the Director taking other periods, for example a
year or a period of months, or even less, as a cross check, if to do so would
be reasonable from a business perspective. Whether the period taken is a
reasonable period will be a matter of fact and degree, to be judged in the
circumstances of each particular case.
The consequence is that the longer a dominant undertaking prices at some
level below total costs, the more likely it is that costs which might be treated
as “fixed” in the short run should be treated as “variable” for the purpose of
applying the AKZO test. In applying the Chapter II prohibition, that would not
seem to us an unreasonable approach. In order to survive in the market a
competitor needs to cover total costs, including overheads. The longer a
dominant undertaking prices below total costs, the more likely it is that an
equally efficient competitor will be forced to exit the market. That risk is not
averted simply because the dominant firm may be covering its average
variable costs as measured on a short-run basis.
Similarly, it seems to us, the longer the prices of a dominant undertaking
remain below total costs the easier it is likely to be to infer an intent to
eliminate competition, in accordance with the AKZO test, absent special
circumstances such as recessionary conditions. Such an intention may be
inferred, of course, from other circumstances, such as selective price
135
cutting.”
E.18
From this viewpoint, the CAT considered that the OFT had been conservative in defining
variable costs as costs that can be varied within a month. The judgment also suggested that
any costs that vary according to the total volume of advertising carried by the newspaper can
be considered to be variable costs:
135
Aberdeen Journals, paras. 353-356.
83
“[The Director] has sought to take as ‘variable’, costs which are variable over
as short a period as one month. Had he taken ‘the time period over which the
alleged predatory pricing prevailed or could reasonably have been expected
to prevail’ […], he would have been entitled to ask himself which costs were
‘variable’ over the period from mid-1996 to January 2001, i.e. the whole period
when the Herald & Post was making a negative contribution. On such an
approach, a very high proportion, if not all, of the costs of the Herald & Post
136
could probably have been categorised as ‘variable’.”
“This part of the Director’s analysis seems to us to give rise to a difficulty,
namely that the figures supplied by Aberdeen Journals appear to relate to
costs that are variable within a month, rather than what the Director appeared
to be seeking, namely an estimate of costs that are variable from month to
month. Thus, on the figures supplied, editorial costs such as freelance
photographers are treated as variable, while all other editorial costs are
treated as fixed. Similarly, for advertising, the only elements that are treated
as variable costs are staff bonuses, stationary, office consumables etc, rather
than the salaries of the staff.
However, if one examines the underlying figures it does appear that a
different approach could have been taken. For example, on the Director’s
approach, a large proportion of advertising costs (mainly salaries of sales
staff) are classified as “fixed costs”. But if the volume of advertising carried by
the Herald & Post is reduced from month to month, it is likely that a
corresponding reduction will be made in the numbers employed in the
advertising department, or that employees working on the Herald & Post will
be switched to other titles. Depending on such matters as the terms of the
relevant employment contracts, it seems to us that staff salaries in the
advertising department could be regarded as “variable” over a relatively short
timescale. That conclusion could be borne out by the figures supplied by
Aberdeen Journals to the Director on 23 August 2000, relating to the period
March to July 2000 […]. Those figures show that “fixed” advertising costs
progressively fell, from £14,080 in March 2000 to £8,003 in July 2000, a
decline of 43 per cent. In the same period, average weekly advertising
volumes fell from 10,440 sccm to 5,620 sccm, a decline of 46 per cent. The
apparently close correlation between these changes in volume and the
changes in costs, suggests that almost all the advertising costs classified as
“fixed” costs could be reclassified as a “variable” cost […]. Similar
considerations apply to editorial and circulation costs.
If, as a matter of fact, costs such as advertising and editorial costs do vary
with output, the whole amount of such costs are to be treated as “variable” for
137
the purposes of the AKZO test.”
136
Aberdeen Journals, para. 384.
137
Aberdeen Journals, paras. 393-395 (emphases original).
84
E.19
The CAT also indicated in its judgment that any measure of total costs must include
contributions towards shared costs with other titles which are not alleged to have been lossmaking:
“Paragraphs 171 to 174 of the decision give details of what are described as
“the losses”, by which is meant the negative contribution, of the Herald & Post
in the period from 1996 to July 2000. The use of the word “losses” in this
context is not, in our view, technically correct. In paragraph 171 the figures in
question relate only to the Herald & Post’s negative contribution to Aberdeen
Journals on the basis of the management accounts, and while those in
paragraph 174 take account also of certain other costs attributed to the Herald
& Post, neither set of figures make any allowance for other operations and
overhead costs which are incurred by Aberdeen Journals but not charged to
the Herald & Post in the management accounts. “Operations” not included in
the management accounts include pre-press and printing staff costs.
Overheads could include, for example, the cost of central management and
administration, premises, maintenance and depreciation of machinery,
financial controls and audit, information technology, finance charges and so
on.
The information before the Tribunal does not enable us to determine whether,
or at what point, the Herald & Post ever reached a position in which it could
be said that the revenues of the Herald & Post exceeded its average total
costs, including an allocated share of operations costs and general
overheads. More generally, as a matter of terminology, it seems to us
incorrect for Aberdeen Journals to contend that the Herald & Post returned to
“profitability” in about February 2001. We would not ourselves regard a
product as “in profit” in a normal commercial sense if it is barely breaking even
on its directly attributable costs, let alone making an adequate contribution to
138
other operations costs and general overheads.”
“[I]n order to survive in the market, a competitor to a dominant firm must
normally cover its total costs (including overheads) and earn a return on its
investment. Moreover, in our view, in normal commercial business, each
product line is expected not merely to cover its variable costs, but to make an
appropriate contribution to general overheads. If a dominant firm prices below
average total costs, including a proportionate share of general overheads, for
a prolonged period, sooner or later an equally efficient competitor will be
139
forced out of the market.”
E.20
Based on this viewpoint, the CAT indicated that the OFT had been conservative in excluding
from its predation test the shared fixed costs of printing presses. Aberdeen Journals had
argued that this approach was justified, since the fixed costs of operating printing presses
would have been incurred whether the allegedly loss-making title had been published or not.
138
Aberdeen Journals, paras. 365-366.
139
Aberdeen Journals, para. 370.
85
While the CAT expressed doubts about this argument, it did not reach a final decision on the
issue:
“[T]he argument presented by Aberdeen Journals is, in effect, that a dominant
firm, using its spare capacity, is entitled to price against a competitor on a
marginal basis and that, so long as its marginal revenues exceed its marginal
costs (for which average variable costs are a proxy) it is acting lawfully, and
does not need to allocate any costs, other than marginal costs, to the
particular activity in question.
This argument has not, so far as we know, yet been considered by the Court
of Justice, or the Court of First Instance. The Tribunal has not heard
submissions on this issue, and it is not necessary for the Tribunal to rule on it
for the purpose of deciding the present case. Nonetheless, it does seem to us
that if a dominant undertaking is able to exclude from its computations
significant elements of cost which have to be borne one way or another, and
which any equally efficient competitor would have to bear, there is a risk that
the dominant firm will always be able, sooner or later, to undercut an equally
efficient competitor and drive it from the market. That, in our view, is a
particular risk where the marginal cost of a particular strategy, such as use of
a “fighting title” (as in this case) or a “fighting ship” (as in Compagnie Maritime
140
Belge, cited above) may well be very low.”
Cardiff Bus
E.21
141
In considering the legal framework for assessing the abuse in Cardiff Bus, the OFT noted the
tests set out in AKZO:
“A dominant undertaking may be found to have set prices in a way that allows
for predation to be presumed. In AKZO, for example, the European Court of
Justice stated that, as a general rule, pricing below average variable costs
(AVC), by means of which a dominant undertaking seeks to eliminate a
competitor, must be regarded as abusive, since it would not normally be
commercially rational for an undertaking to price at levels that did not even
cover AVC.
As the CAT held in Aberdeen Journals, however, a dominant undertaking
may, exceptionally, be able to rebut this presumption. Thus, depending on the
facts of the case, pricing below AVC might not be found to constitute an
142
abuse if, for example, there is an objective justification for that pricing.”
E.22
The OFT considered that the most appropriate cost benchmark in this case was the average
avoidable costs:
140
Aberdeen Journals, paras. 378-379.
141
OFT ‘Abuse of a dominant position by Cardiff Bus’ CA98/01/2008, Decision dated 18 November 2008
(“Cardiff Bus”).
142
Cardiff Bus, paras. 7.10-7.11.
86
“In the present case, the OFT has considered two bases for assessing costs:
average variable costs (AVC) and average avoidable costs (AAC). This
approach has been used in several OFT cases, such as in Aberdeen
143
Journals.
In past cases these two bases have been shown to be very
similar, since any cost that is variable over the period is also avoidable.
Often the AAC benchmark will be the same as the AVC benchmark, because
in many cases only variable costs can be avoided. However, where the
dominant company makes specific investments, such as expanding capacity
in order to predate, then the fixed or sunk investments made for this extra
capacity will be included within AAC, causing AAC to exceed AVC.
Given the circumstances of this case the AAC benchmark is the most
appropriate. The assessment includes the cost of restoring the buses used on
the white service routes. This preparatory work would not be considered a
variable cost. However it is the OFT’s view that these costs were avoidable.
The cost could have been avoided if Cardiff Bus had not restored the buses.
Alternatively, once restored, Cardiff Bus could have recovered at least part of
the cost of restoration had the white services not run or ceased to run, either
by transferring these, now serviceable, buses onto other routes or leasing the
144
buses.”
E.23
The OFT considered that this was a logical extension of the test set out in AKZO:
“In the medium term, pricing below AAC is not in the economic interest of an
undertaking, since by not providing the relevant output it would save more in
costs than it would forego in revenue. In addition, the longer a dominant
undertaking prices below AAC, the more likely it is that an equally efficient
competitor would be forced to exit the market. This follows the same logic as
the AKZO test as it has been applied in recent cases.
The OFT therefore concludes that AAC is the most appropriate measure in
this case, as it is the logical extension of the AKZO test, and it is relevant to a
consideration of whether Cardiff Bus’ conduct resulted in a loss. It also
identifies the extent to which an equally efficient competitor would be able to
145
operate profitably.”
E.24
Consistently with the CAT Judgment in Aberdeen Journals, the OFT considered that the
definition of variable costs should be based on the length of the alleged abuse period:
143
In Aberdeen Journals, average avoidable costs had been raised in the OFT’s initial complaint but not
used in the final decision. See Aberdeen Journals, para. 385.
144
Cardiff Bus, paras. 7.156-7.158.
145
Cardiff Bus, paras. 7.160-7.161.
87
“Generally, the longer the period considered in an investigation of predation,
the more costs become avoidable as commitments expire and/or alternative
uses of the existing resources become available. It is therefore important to
determine the appropriate time period to be used.
The relevant time period is usually that over which the alleged predatory
conduct prevailed or could reasonably be expected to prevail. Whether the
period taken is reasonable will be a matter of fact and degree to be assessed
in the circumstances of each particular case.
Applying those principles in this case, the OFT has considered the period over
which the white services were run. […] [I]n total, the activity on the white
services spanned 10 months.
The OFT therefore considers that a 10-month period is appropriate for
assessing avoidable costs in this case. Thus, avoidable costs include any cost
that would have been avoided within the 10-month period, 19 April 2004 to 18
146
February 2005, had the white services not run.”
Wanadoo
E.25
147
In Wanadoo, the European Commission confirmed the two tests set out in AKZO and Tetra
Pak:
“Community case-law has given rise to two separate methods of analysis
when it comes to examining whether an undertaking has practised predatory
pricing, one based on variable costs, and the other on full costs. Thus, the
existence of predatory prices is established in the following situations:
– the non-recovery of average variable costs per unit, the establishment of
which in itself suffices to justify a finding of abuse;
– the non-recovery of average full costs where this is accompanied by a plan
148
indicative of an intention to eliminate competitors.”
E.26
In implementing this test, the Commission considered three different approaches based on
the actual revenues and costs of Wanadoo, and one approach based on ex ante forecasts.
The Commission considered that its main approach should be based on one of the
approaches based on actual revenues and costs, although it also considered the approach
based on ex ante forecasts as a supplemental analysis:
146
Cardiff Bus, paras. 7.172-7.175.
147
Case COMP/38.233, Wanadoo Interactive, EC Decision of 16 July 2003 (“Wanadoo”); Case T-340/03,
France Télécom SA v Commission of the European Communities, Judgment of the Court of First
Instance, 30 January 2007; Case C-202/07 P, France Télécom SA v Commission of the European
Communities, Judgment of the European Court of Justice, 2 April 2009
148
Wanadoo, para. 256.
88
“In this section the Commission explains and comments on various
approaches that a test of predation might adopt […]:
– recovery of the full instantaneous costs as recorded in the undertaking’s
accounts;
– recovery of the full costs and of the variable unit costs as recorded in the
accounts after they have been adjusted by spreading certain costs over
several years;
– recovery of the variable costs for the different generations of subscribers
separately;
– recovery of the costs which the undertaking would have foreseen ex ante,
excluding any incidental or fortuitous components that may have affected
initial forecasts.
[…] At the end of the analysis the Commission finds that only the adjusted
costs approach allows any valid conclusions to be drawn; the others can help
149
to throw further light on the matter, but no more.”
“The Commission has applied a method designed to assess whether
Wanadoo Interactive’s ADSL services had a production margin sufficient to
cover the costs of acquiring new subscribers if the acquisition costs are
depreciated over 48 months. To supplement this analysis, given that fortuitous
circumstances arose at the end of 2001, the Commission has also evaluated
the extent to which Wanadoo Interactive could have made a positive forecast
of the profitability of its ADSL services in the second half of 2001 and at the
beginning of 2002 on the basis of the working hypotheses available in the
150
company at that time.”
151
Telefónica
E.27
In Telefónica, the General Court of the European Union upheld the European Commission’s
argument that pricing tests of abuse of dominance must be based on actual historical data
rather than business forecasts:
“There are two possible approaches to assessing the profitability of
Telefónica’s downstream activity with the DCF method; a backward-looking
and a forward-looking analysis. The backward-looking approach would consist
in calculating the NPV on the basis of Telefónica’s historical costs while a
forward-looking analysis would be based on the forecasts made by Telefónica
in its initial business plan.
149
Wanadoo, paras. 70 and 72.
150
Wanadoo, para. 108.
151
Case COMP/38.784, Wanadoo España vs. Telefónica, EC Decision of 4 July 2007 (Telefónica); Case
T-336/07, Telefónica SA and Telefónica de España SA v. European Commission
89
In the SO, the Commission used Telefónica’s historical revenues and costs
when available and the forecasts of the company when there was no historical
data available.
Telefónica argued in its Reply that this analysis is flawed because it implicitly
assumes that Telefónica had perfect foresight in 2001 when choosing the
level at which it set its prices. Telefónica considers that the Commission
should use the cash flows projected in its business plan of October 2001, i.e.
those based on the information available to Telefónica at the time it entered
the retail market in 2001 and set its prices.
There is no case law on unfair prices under Article 82 in which the existence
of an abuse would have been analysed on the basis of the forecasts of the
very company whose conduct is being investigated. In accordance with case
law and the practice of the Commission, also in the present case has the
Commission based its DCF calculation on Telefónica’s historical costs and
152
revenues in the period 2001-2006.”
E.28
However, the European Commission accepted that reasonable business forecasts would be
relevant to an assessment of objective justification:
“The question could nonetheless arise as to whether Telefónica’s pricing
policy could be objectively justified if its initial business plan indicated a
positive NPV with reasonable forecasts that are compatible with a competitive
environment. Therefore, and exclusively for the purpose of assessing whether
they indicate that Telefónica’s pricing policy could be objectively justified, the
153
Commission has considered Telefónica’s forward NPV calculations.”
152
Telefónica, paras. 378-381.
153
Telefónica, para. 382.
90
Annex F
Additional sensitivities on test of
Forum’s incremental profitability
Sensitivities on diversion ratio to Vista
F.1
As set out in paragraphs 9.4-9.5, we based our analysis on an internal Media24 estimate of
the diversion ratio from Forum to Vista of '''''' per cent.
F.2
We are aware of another internal Media24 document ''''''''''''' ''''''''''''''''''''''''' ''''''''''''' which
estimates a higher diversion ratio of ''''''' per cent. However, this cannot be used directly in
our analysis because this estimate is from a time when Gold-Net News was not present in
the market.
F.3
We have therefore adjusted the alternative diversion ratio of '''''' per cent based on the
relative market shares of Vista and Gold-Net News during the complaint period. Media24’s
154
internal estimates of advertising volume market shares suggest market shares of '''''''''' per
cent for Vista, '''''''''' per cent for Forum and ''''''''''' per cent for Gold-Net News on average
during the complaint period. If '''''' per cent of Forum customers tend to switch to outside
goods after closure of Forum (as evidenced by '''''''' ''''''''''''''''''''''''''' ''''''''''''' ''''''''''''''''''''') and
diversion to other community newspapers is proportional to market shares, this would imply
diversion from Forum to Vista during the complaint period of '''''''''''' per cent.
F.4
Table 19 shows our updated results based on a diversion ratio of '''''''''' per cent from Forum
to Vista.
154
Discovery documents RD604-630.
91
Table 19: Forum’s discounted incremental profits based on diversion ratio of ''''''''% (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
Sensitivities on discounting rates
F.5
As set out in paragraph 10.8, we used a discount rate of 18 per cent per annum in our
analysis, based on information set out in Naspers’ Annual Report.
F.6
In order to assess the sensitivity of our results to changes in discount rates, we have
modelled the sensitivity of increasing the discount rate by 50 per cent to 27 per cent and
decreasing the discount rate by 50 per cent to 9 per cent. Table 20 and Table 21 show our
updated results based on discount rates of 27 per cent and 9 per cent respectively.
Table 20: Forum’s discounted incremental profits based on discount rate of 27% (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
92
Table 21: Forum’s discounted incremental profits based on discount rate of 9% (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
Sensitivities on printing costs
F.7
As explained in paragraphs 9.19-9.22, we adjusted the printing costs shown in Forum and
Vista’s management accounts prior to April 2008 upwards by '''''' per cent. For completeness,
Table 22 shows our results without this adjustment to printing costs.
Table 22: Forum’s discounted incremental profits with no adjustment to printing costs
(R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
Results shown in Commission’s Supplementary Affidavit
F.8
The Commission’s Supplementary Affidavit dated 12 July 2013 set out an analysis of
Forum’s incremental revenues and incremental costs. This analysis followed the structure of
the test of Forum’s incremental profitability set out in this report, with two differences in
approach:
93
The analysis set out in the Supplementary Affidavit did not make an upward adjustment
of ''''''' per cent to Media24’s printing costs prior to April 2008; and
The analysis set out in the Supplementary Affidavit assumed that the diverted demand
from Forum to Vista would be served at Vista’s overall average net price, calculated as
Vista’s total net revenues (including revenues from inserts) divided by total paid
advertising volume. This is in contrast to our base case, which calculated separate
prices for national, local and classified advertisements based on Vista’s average prices
excluding inserts.
F.9
Table 23 sets out the results shown in the Commission’s Supplementary Affidavit.
Table 23: Forum’s discounted incremental profits as set out in the Commission’s
Supplementary Affidavit (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Incremental revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental costs (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoided exit costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (1)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Incremental profit (2)
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
94
Annex G
Additional sensitivities on comparison
of Forum’s revenues and its avoidable
costs
Sensitivities on discounting rates
G.1
As set out in paragraph 10.8, we used a discount rate of 18 per cent per annum in our
analysis, based on information set out in Naspers’ Annual Report.
G.2
In order to assess the sensitivity of our results to changes in discount rates, we have
modelled the sensitivity of increasing the discount rate by 50 per cent to 27 per cent and
decreasing the discount rate by 50 per cent to 9 per cent. Table 24 and Table 25 show our
updated results based on discount rates of 27 per cent and 9 per cent respectively.
Table 24: Forum’s revenues and avoidable costs based on discount rate of 27% (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
Table 25: Forum’s revenues and avoidable costs based on discount rate of 9% (R
millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
95
Sensitivities on printing costs
G.3
As explained in paragraphs 9.19-9.22, we adjusted the printing costs shown in Forum’s
management accounts prior to April 2008 upwards by ''''''' per cent. For completeness, Table
26 shows our results without this adjustment to printing costs.
Table 26: Forum’s revenues and avoidable costs (R millions)
2004
2005
2006
2007
2008
JanFeb
2009
NPV at
Jan
2004
Revenues
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Avoidable costs
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Difference
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
'''''''''''
Note: All figures shown on discounted basis as of January 2004.
96
Annex H
Cost classifications
Introduction
H.1
Table 27 sets out the cost classifications used in our analyses:
1 year avoidable costs according to Media24;
1 year avoidable costs according to the Commission; and
Variable costs according to Media24.
H.2
Our analyses apply the 1 year avoidable cost definition to Forum’s costs. Table 27 therefore
shows the 1 year avoidable cost definitions for the costs incurred by Forum during the
complaint period only. For costs not incurred by Forum during the period, the 1 year
avoidable cost classifications are shown as “N/A”.
Sources
Commission’s cost classifications
H.3
The Commission’s cost classifications were taken from the Commission’s Referral of
Complaint, para. 44 and TF6.
Media24’s cost classifications
H.4
Media24’s cost classifications were taken from Discovery Document RD20, with the following
exceptions.
Allocated central costs
H.5
Discovery Document RD20 did not set out the individual components of allocated central
costs (“Toegedeelde koste” and “Bydraes/Verhalings”), a breakdown of which is shown in
Media24’s management accounts and in Discovery Documents RD243-244. Discovery
Document RD20 stated that ''''''''''''''''''''''' '''''''''''''''' '''''''''''''' '''''' '''' ''''''''''''' ''''''''''''' ''''''''''' '''''' ''''''''''''''''''''''''
'''''''''' ''' ''''''''''''''' '''''''''''''''' '''''''''' '''' ''''''''''''
H.6
'''''''''''''''''''''''' '''''''''''''''''''''''''' '''''''''''''' '''''''''' ''''''''''''''' ''''''''''''''''''''''''' '''''''''''''''''' '''''''''''''''' '''''''' ''''''''''''''' ''''''''''''''' '''
''''''''''''''' ''''' ''''''''''''''''''''' '''''''''''''' ''''''''''''''' ''''''''''''' '''''' '''''''''''''' ''''' '''''''''''''''''''''''' '''''''''''''' ''' ''''''''''' '''''''''''''''' ''''''
'''''''''''''''''''''''' ''''''''''''' '''''''' ''''''''''''''' '''' ''''''''''''' ''''''''''''' '''''' '''''''''''' '''''''''''''''''''''''' '''' ''''''' '''''''' ''''''' ''''''''' ''''''''
''''''''''''''''''''''''' '''''''''''''''''' ''''''' ''''''''''' ''''''''''''''''''''''''''''''' ''''' ''''''''''''''''''' '''''''''''''''' ''''''''''''''''
H.7
'''''''' ''''''''''''''''''''''''''' ''''''''' '''''''' ''''''''' '''''''' ''''''''''''''''' ''''''''''''' ''''' '''''''' ''''''''''''''''' ''''''''''''''''' '''''''''''' '''''' '''''''''''''''''''
''''''''''''''' '''' ''''''''''' '''''''''''''''''''''''''' ''''''''' '''''''' ''''''''''''''''''''''' '''''''''''''''''' '''''''''''''' ''''' '''''''''''''''' '''''''''''' ''''''''''' ''''''''''
97
''''''''''''''''' ''''' ''''''''''''''' '''''''''''''' ''''''''' '''''''''''''' '''''''''' '''''' ''''''''''''''''' '''''''''''''' '''' ''''''''''' Therefore, based on this
evidence, we have included allocated printing costs of Forum in Media24’s variable cost and
1 year avoidable cost definitions.
H.8
In the absence of evidence for other components of allocated central costs, we have taken
all other allocated central costs of Forum to be fixed and non-avoidable over 1 year.
Other costs
H.9
Discovery Document RD20 did not set out the cost classifications for the following costs
incurred by both Forum and Vista:
''''''''''''''' '''''''''''''' '''''''''''''''''''''''''' '' '''''''''''''''''''''''''''''' '''''''' ''''''' '''''''''''''''''''' '''' '''''''' '''''''''''''''''''' ''''' '''''''' ''''''''' '''''''
'''''''''''''''''''''''''''''' '''''''''''''''' ''''' '''''''''''''''''''''' '''''''''''''''''''''''' ''''''''''''' '''''' ''''''''''''''''' ''''''''''''' '''''''''''''' '''''''''' '''''''''
''''''''''''''''''''''''' '' ''''''''''''''''''''' '''''''' ''''''''''''' '''''''''''''' ''''''''''''' ''''''''''''' ''''' ''''''' '''''''''''' ''''''''' ''' '''''''''' ''''''''''''''''''''''''
''''''''''''''''''''''''''''''''''''' ''''''''''''' '''''''''''' '''''''''''''''''''''''''''' '' '''''''' ''''''''''''''''''''''''''''''' '''''''' '''''''' ''''''''''''''''''''' ''''' ''''''''
'''''''''''''''''''' ''''''' ''''''''''''' '''''''''''' '''''''''''' ''''''''''''' ''''' ''''''' '''''''''''' ''''''''' '''''''''''''''''''''''''''''' ''''''''' '''' '''''''''''''
H.10
Discovery Document RD20 did not set out the cost classifications for the following costs
incurred by Vista only:
'''''''''''''''''''' '''''''''''''''''''' '''''''''''''''' '' ''''''''''''''''''''''''
''''''''''''''''' '''''''''''''''''''' ''''''''''''''' '' ''''''''''''''' ''' ''''''''''''''''''''''' '''''''''''''''' ''''''''''
'''''''''''''''' '''''''''''''''' '' '''''''''''''''''
''''''''''''''''''' '''''''''''''''' ''''''''''''''' '' ''''''''''''''''''''' '''''''''
''''''''''''' ''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''
H.11
For the purposes of our analysis, we have taken these costs to be fixed.
98
Table 27: Cost classifications under Definitions 1 and 2
Account
number
Cost line
Description
1 year
avoidable
(Media24)
1 year
avoidable
(CCSA)
Variable
costs
(Media24)
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
99
Account
number
Cost line
Description
1 year
avoidable
(Media24)
1 year
avoidable
(CCSA)
Variable
costs
(Media24)
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
100
Account
number
'''''''''''''''''''
'''''''''''''''''''
1 year
avoidable
(Media24)
1 year
avoidable
(CCSA)
Variable
costs
(Media24)
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
Cost line
'''''''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''''
Description
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
101
Account
number
'''''''''''''''''''
1 year
avoidable
(Media24)
1 year
avoidable
(CCSA)
Variable
costs
(Media24)
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
Cost line
'''''''''''''''''''''''
'''''''''''''''''''''''
Description
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
102
Account
number
Cost line
Description
1 year
avoidable
(Media24)
1 year
avoidable
(CCSA)
Variable
costs
(Media24)
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
'''''''''''''''''''
'''''''''''''''''''''''
'''''''''''''''''''''
''''''''''
'''''''''''
'''''''''''
103
Annex I
Discount rate for assessment of
Forum
I.1
We have used a discount rate of 18 per cent per annum in calculating the NPV of Forum’s
incremental profits.
I.2
This was based on the discounts rates applied by Naspers to its print media subsidiaries in
its 2005 Annual Report, as shown below.
Table 28: Naspers discount rates applied to cash flows, newspaper subsidiaries
Cash-generating unit
Discount rate applied
to cash flows
Boland Newspapers (Proprietary) Limited
18.63%
Paarl Media Holdings (Proprietary) Limited
17.53%
Natal Witness Printing and Publishing Company (Proprietary) Limited
18.63%
Simple average
18.26%
Source: Naspers Annual Report 2005, p. 94.
I.3
As shown above, Naspers allocated an average discount rate of 18 per cent to its three print
media subsidiaries. Naspers’ accounting policies set out in the Annual Report state that this
is a pre-tax discount rate that takes into account the time value of money and the assetspecific risks:
“The estimated future cash flows are discounted to their present value using a
pre-tax discount rate that reflects current market assessments of the time
155
value of money and the risks specific to the asset.”
I.4
Based on this explanation we consider that the discount rate of 18 per cent is appropriate for
the evaluation of printed media businesses of Naspers, including Forum.
155
Naspers Annual Report 2005, p. 79.
104