COMPETITION TRIBUNAL OF SOUTH AFRICA
CT Case No. 15/CR/Feb09
In the matter between
COMPETITION COMMISSION Applicant
and
DPI PLASTICS (PTY) LTD First Respondent
PETZETAKIS AFRICA (PTY) LTD Second Respondent
MARLEY PIPE SYSTEMS (PTY) LTD Third Respondent
SWAN PLASTICS (PTY) LTD Fourth Respondent
AMITECH SOUTH AFRICA (PTY) LTD Fifth Respondent
FLO-TEK PIPES AND IRRIGATION (PTY) LTD Sixth Respondent
MACNEIL AGENCIES (PTY) LTD Seventh Respondent
ANDRAG (PTY) LTD Eighth Respondent
GAZELLE PLASTICS (PTY) LTD Ninth Respondent
GAZELLE ENGINEERING (PTY) LTD Tenth Respondent
__________________________________________________________________________________
Panel : Norman Manoim (Presiding Member),
Yasmin Carrim (Tribunal Member)
Takalani Madima (Tribunal Member)
2
Heard on : 13-17 and 20-23 September 2010
17-18 January 2011
1 February 2011
18 and 20 April 2011.
Judgment issued on : 4 July 2012
Judgment
[1] This case concerns the alleged involvement of ten manufacturers in a cartel to fix prices,
divide markets and collude over tenders in respect of various types of plastic pipes,
more specifically, Polyvinylchloride (‘PVC’) and High Density Polyethylene (‘HDPE’)
pipes1. The Competition Commission (‘Commission’) initiated the complaint on 18
March 2008 and referred it to the Tribunal on 2 February 2009.
[2] Since the n the fates of the respondent firms have varied. The first respondent DPI
Plastics (Pty) Ltd (‘DPI’), applied for and was granted conditional leniency by the
Commission on 27 February 2008 in terms of the Commission’s Corporate Leniency
Policy (‘CLP’). No relief is thus being sought by the Commission against it . Three other
firms have since the referral, entered into consent orders with the Commission and
hence the cases against them are concluded and are not dealt with in this decision.
These are Marley Pipe Systems (Pty) Ltd (‘Marley’), Swan Plastics CC (‘Swan’) and Flo-
Tek Pipes and Irrigation (Pty) (Ltd) (‘Flo-Tek’).
[3] Marley agreed to pay an administrative penalty of R31 078 213.02. This amount
represents 6% of Marley’s 2007 financial year, less the turnover attributable to
speciality products2. Swan paid an administrative penalty of R7 649 414.40. The amount
also represented 6% of Swan’s total turnover in the 2007 financial year 3, whilst Flo-Tek
1 See CT1(1) Form, Pleadings Bundle, page 1.2 Competition Commission v Marley Pipes Systems (Pty) Ltd in re: Competition Commission v DPI (Pty) Ltd
and Others [2010] 1 CPLR 81 (CT) at paragraph 6.3 Ibid, paragraph 6.1.
2
3
paid an amount of R5 049 433.26, representing 6% of its total turnover in the 2007
financial year4.
[4] Five of the respondents, did not reach agreement with the Commission and they are
therefore the subject of this hearing. Two of them have accepted liability for cartel
involvement, but differ with the Commission over the extent of their involvement which
has a bearing on their liability. These firms are the second respondent, Petzetakis Africa
(Pty) Ltd (“Petzetakis”) and the fifth respondent, Amitech South Africa (Pty) Ltd
(“Amitech”).
[5] The four remaining respondents contested their liability. They are the seventh
respondent McNeil Mouldings (Pty) Ltd (“McNeil”), the eighth respondent, Andrag (Pty)
Ltd (“Andrag”), and the ninth and tenth respondents Gazelle Plastics (Pty) Ltd and
Gazelle Engineering (Pty) Ltd (“collectively referred as Gazelle”). We refer to the ninth
and tenth respondents jointly as ‘Gazelle’, as by agreement with the Commission, they
have accepted that they are both controlled by the same shareholders and they would
be jointly and severally liable if found to have contravened the Competition Act, no. 89
of 1998 (‘the Act’).
[6] There is general acceptance by all that a cartel existed in this industry and that it has
existed for a considerable time, preceding the commencement of the Act. One witness
testified that when he joined the industry in 1969 the cartel was known to be in
existence.5 In its most recent form, which appears to date back to at least the 1990’s
the cartel comprised three firms at its core, DPI, Marley and Petzetakis or their
predecessors.6
[7] The firms presently before us, with the exception of Petzetakis, were minor players in
the industry. Even the Commission accepts that they played bit parts in the cartel. The
big players were national and the small players, regional in their participation. This has
big players were national and the small players, regional in their participation. This has
meant that there is a complicated inter-relationship in this case between national
agreements reached by the bigger players, which are common cause, and smaller
4 Supra, footnote 2, paragraph 6.2.5 Robert Haynes of DPI. See Transcript page 224; see also Haynes’ witness statement on pages 431-438 of the
Commission’s witness statement.6 The present DPI and Petzetakis firms have been the subject of prior merger activity.
3
4
alleged agreements at regional level, whose existence or extent is contested. Thus,
whilst Andrag, MacNeil and Gazelle all concede to contacts with the other respondents
and in particular the undisputed cartel members, they do not concede that the legal
and factual characterisations of these interfaces constitute agreements for the purpose
of the Act.
[8] According to the Commission the cartel consisted of several meetings at different
centres where the essential agreements were struck. The three major players were
typically represented at all these meetings, whilst the minor players attended only
some. For this reason, an analysis of the case for those firms requires an analysis of
what happened at these meetings and versions vary not only as to content, but, in
some cases, as to who was present. Since we have approached the case by considering
each respondent individually, this has necessarily led to some repetition as we analyse
common meetings from different vantage points, but where we can, we have tried to
avoid repetition.
[9] The products to which the cartel related were HDPE and PVC pipes. Another product
known as GRP was alleged to form part of the conspiracy and we deal with this when
we consider the case against Amitech , the only firm that manufactures this product.
Not all the firms manufactured both products and even those who did tended to
specialise in one or the other. This accounts for different incentives between the firms
at times. A further distinction was between customers for the products. The industry
recognised that the same product had different customer markets e.g. building,
agricultural, mining or local authorities. Some firms specialised in a particular class of
customer, others supplied all. Again the customer market had implications for how the
cartel constructed itself.
[10] The collusion comprised in the main agreements on price – typically agreed discounts
[10] The collusion comprised in the main agreements on price – typically agreed discounts
off a common price list, but also involved customer allocation and bid rigging. There
were also attempts to use the cartel to boycott customers or suppliers who undercut
the cartel, but this evidence was not conclusive.
4
5
Background to the complaint
[11] On 12 October 2007, DPI filed an intermediat e merger notification with the
Commission, the applicant in these proceedings. DPI had intended to merge with
Incledon Cape (Pty) Ltd 7. During the course of the merger investigation, the
Commission came across allegations that DPI and certain of its competitors were
involved in bid rigging and allocation of contracts in contravention of the Act.
[12] As already stated above , DPI applied for immunity from prosecution in terms of the
Commission’s Corporate Lenience Policy (“CLP”). In its application in terms of the CLP,
DPI furnished evidence to the Commission of the existence of collusion in the markets
for pipe products. The alleged collusion involved price fixing, bid rigging and allocation
of markets or customers. The Commission, on learning of the alleged activities,
prohibited the intended merger8.
Commission ’s case
[13] The Commission, in the notice of motion, in its complaint referral sought an
order in the following terms:
1. “Declaring that the second to ninth respondents commit (sic) prohibited practices
in contravention of section 4(1)(b)(i), (ii) and (iii) of the Competition Act, 89 of
1998 (‘the Act’);
2. Interdicting the second to ninth respondents from committing prohibited
practices as set out in paragraph 1 above;
3. Imposing an administrative penalty against each of the respondents in an
amount equal to 10% of the annual turnover of each of the respondents in the
Republic and their exports from the Republic during each of the respondents’
preceding financial year;
7 Commission Founding Affidavit, Pleadings Bundle, page 12, paragraph 18.8 The Commission prohibited the intermediate merger on 10 January 2008, a day after receiving the CLP
application from DPI.
5
6
4. Granting the applicant such further and/or alternative relief as the Competition
Tribunal considers appropriate9.”
MacNeil
[14] MacNeil, as already stated above, contested both the merits and administrative penalty
prayed for in the Commission’s notice of motion.
[15] MacNeil is a newcomer to PVC manufacture. The MacNeil Group of companies had
commenced business in 1993 as building material suppliers. In 1997 an associate
company, MacNeil Pipes, commenced production of PVC pipes which were not SABS
approved and were essentially for the domestic market. In 2004, the firm presently
before us as MacNeil, was formed to manufacture irrigation fittings for the agriculture
industry. In 2006, MacNeil purchased an extrusion line for the production of SABS
approved PVC pipes. Production commenced in late 2006 and another extrusion line
came on stream in 2007. In 2009 the pipe manufacturing business was taken over by
the group’s moulding company and renamed MacNeil Plastics.
[16] MacNeil’s main customers are Flo-tek and merchants who supply to the civil
engineering sector.
[17] The Commission’s case against MacNeil has not been consistent in certain respects. This
is because the Commission’s two key witnesses against MacNeil, Andre Auret and Rene
Le Riche, both of DPI, had difficulties in their recall of the dates of meetings, the
number of meetings attended by certain people and the content of discussions.
MacNeil have criticised this evidence for being sketchy and imprecise.
[18] It is not necessary for us to go into the detail of these versions or speculate on why they
have changed. The Commission sought, at the end of its case, to rely on a further
version advanced by Shaun Hart, the Cape Town branch manager of Flo-Tek. Hart did
not give oral evidence or provide a written witness statement, but his answering
affidavit forms part of the record which sets out his version. It must be recalled that
affidavit forms part of the record which sets out his version. It must be recalled that
Hart’s firm, the sixth respondent, eventually settled with the Commission. His affidavit
was therefore prepared on the basis that the Commission’s case was being opposed. It
9 Pleadings Bundle, page 4.
6
7
can therefore not be seen to be tailored to exculpate his firm for the purpose of
leniency by exaggerating the role of other respondents. MacNeil does not have any
quarrel with this version and therefore we can choose to rely on it.
[19] MacNeil’s evidence comes from the version offered in its answering affidavit deposed
to by Phillip Brink, its sales manager, with confirmatory affidavits from Neil Malherbe,
its managing director and Shawn Diab, its general manager. The same three testified on
its behalf at the hearing.
[20] It is common cause that MacNeil representatives attended three meetings with some of
the other respondents during the course of 2007. The first such meeting was in
February or March 2007. In attendance from MacNeil was Shaun Diab who was
MacNeil’s general manager at the time. Others in attendance were Trevor Lombard of
Petzetakis, Shaun Hart of Flo-Tek and Rene Le Riche of DPI.
[21] What is not common cause is how Diab came to be attending this meeting and what
was discussed at it. Malherbe’s evidence had been that where McNeil’s employees had
attended meetings with competitors, it was in the context of MacNeil’s toll agreement
with Flo-Tek.10 However, contrary to Malherbe’s version, in his examination- in- chief,
Diab explained that MacNeil was new in the market and this occasion had afforded a
great opportunity to meet competitors and to get a view of the DPI factory “... to see
how they do things” .11 Diab went further in cross-examination and confirmed that the
meeting he attended was not about the toll arrangement with Flo-Tek 12 and conceded
that it was about MacNeil’s pricing in the market.13
[22] The next issue was what was discussed at the meeting. Le Riche and Auret had testified
that the meeting had discussed whether suppliers should boycott a merchant called
Peakstar that supplied pipes at heavily discounted prices in the market undercutting the
Peakstar that supplied pipes at heavily discounted prices in the market undercutting the
prices of the suppliers. Diab denies this. Hart of Flo- Tek does not mention this. Given
that Le Riche and Auret’s recall of specifics was not convincing we will rely on the
version of Flo- Tek supported by the denial of Diab.
10 Commission’s Heads of Arguments, paragraph 55.11 Transcript, page 1710.12 Transcript, page 1715-1716.13 Transcript, page 1718.
7
8
[23] It is, however, common cause that pricing in the market was discussed at this meeting.
Diab stated originally that he was completely passive at this meeting. During cross-
examination he conceded that he contributed something, but could not recall what his
contribution was.14
[24] Diab’s version of what was discussed is contradictory and not credible. Denying that
the meeting had discussed Peakstar he stated; “ No, most definitely not, I would have
remembered that. To my knowledge it was just general day to day stuff.”
[25] Thus Diab is able to confidently recall that others had not discussed certain issues, but
cannot recall what he himself said. Further, a reference to the topic of conversation
being about “day to day stuff” is wholly unsatisfactory and evasive. It is common cause
that a meeting took place between representatives of three competing firms. Even
MacNeil admits that the other firms were its competitors. In its answering affidavit,
Brink, the deponent, admits that in a “... very broad sense it [MacNeil] can be said to be
competing with the other manufacturers, particularly those manufacturers who like the
first respondent [DPI] market their products to civil merchants.”15
[26] The context of all the other evidence in the case is that when competitors in the
respondent firms in this case met in this period they did so to discuss collusive
arrangements. The likelihood that discussions about “day to day stuff” concerned non –
collusive topics is so remote as to be fanciful and must be rejected.
[27] A further issue ar ose at this meeting: Le Riche told Diab to bring MacNeil in line. This
instruction was understood to mean that MacNeil must charge prices in line with prices
charged by other competitors. Diab’s version on this point is confusing – he could not
recall if Brink had conveyed this message to him or he had conveyed it to Brink. The
Commission argued that the probabilities were that this message was given to Diab who
Commission argued that the probabilities were that this message was given to Diab who
then conveyed it to Brink. We would agree that this is the more probable version.
[28] The meeting was there to discuss pricing. If it was the first meeting on this topic
discussed with a MacNeil representative and if MacNeil , on its own version, was pricing
more aggressively in the market than its competitors, it makes sense that this message
14 Transcript, page 1725.15 MacNeil Answering Affidavit, paragraph 28.
8
9
was given and passed on by Diab to his colleague. The reason that Diab struggled during
cross-examination was that he was attempting to hold on to a version that despite,
attending a meeting with competitors in which he was given a warning by a larger
competitor to price in line with an industry agreement, he did not do so, and on his
version, did not take it seriously. This is highly improbable. Such a meeting on Diab’s
version was the first one he attended. Second, MacNeil was dependent on a rival, Flo-
Tek for toll manufacturing work. Flo-tek was part of the meeting and party to the
collusive agreements on pricing. It seems improbable that Diab would not have passed
on this warning to others in the firm and indeed all that occurred at the meeting
including the agreements on pricing.
[29] MacNeil does admit that Brink attended two further meetings with competitors in
September and October 2007. Present were representatives of the same competitors,
namely, DPI, Petztakis and Flo-tek.
[30] Here again it is important to look at MacNeil ’s version of how its representative, Brink,
came to attend the meeting. Brink testified that he was invited by Hart of Flo- Tek, to
attend a meeting at Flo Tek, but did not know what the purpose of the meeting was
until he got there. He states that there was nothing unusual about being invited to a
meeting by Flo Tek given that MacNeil was toll manufacturing for it.
[31] Brink does not dispute that at these two meetings price lists were discussed. What he
says is that he did not agree to any pricing proposal although , he does concede that
there was an indication at the meeting that pricing should follow an agreed price list. 16
In other words, Brink admits there was an agreement concluded at the meeting but
denies that his firm acceded to it.
[32] None of the Commission ’s witnesses can recall specifically any contribution that Brink
may have made at these meetings. Nor does Hart of Flo-Tek mention anything in his
may have made at these meetings. Nor does Hart of Flo-Tek mention anything in his
affidavit. We therefore must accept Brink’s evidence on this point. In Diab’s case he
admits that he was not completely passive at the meeting, but states that he cannot
recall what he said.17
16 Transcript, page 1737.17 Transcript, page 1725.
9
10
[33] But neither Diab nor Brink indicated to the others at any of the three meetings that,
between them they attended, that their firm did not associate itself with the price
increases. Indeed, had they done so at any prior meeting, the probabilities are that
MacNeil would not have been invited again. Nor after any of these meetings did
MacNeil take any steps to communicate with the other firms that it would not follow
their proposals. Thus, whilst we accept that there is insufficient evidence to show
MacNeil implemented the price increases, the real question is whether, at any of the
three meetings its representatives attended or at any time subsequent thereto, it gave
any indication to the other firms, that it would not adhere to the price increases
discussed and agreed to by the others. The evidence leads to the conclusion that
MacNeil did not.
[34] As the Tribunal has held in Aveng and on the basis of the authorities relied on in that
decision, even passive attendance at a meeting by a firm with its competitors where
agreements are reached on pricing will lead to an inference that the firm formed part of
that understanding unless it has engaged in conduct to repudiate its involvement. 18
There is no evidence that MacNeil either orally or by some form of conduct did. (We
deal more fully with this aspect of the degree to which a firm must be held to assent
when we discuss the case against Andrag below).
[35] Instead MacNeil attended three meetings with its competitors at which the subject
matter under consideration was a collusive arrangement. By its actions, in particular its
repeated attendance at meetings of this nature, MacNeil held out to the other firms
present that their pricing proposals would not be opposed by it and gave them comfort
18 The Competition Commission vs Aveng (Africa) Ltd and Others, Case No: 84/CR/Dec09. In that case at
paragraph [81] we quoted with approval the following extract from a European Court of Justice decision in the
Aalborg Portland case: “According to settled case-law, it is sufficient for the Commission to show that the
undertaking concerned participated in meetings at which anti-competitive agreements were concluded, without
manifestly opposing them, to prove to the requisite standard that the undertaking participated in the cartel.
Where participation in such meetings has been established, it is for that undertaking to put forward evidence to
establish that its participation in those meetings was without any anti-competitive intention by demonstrating
that it had indicated to its competitors that it was participating in those meetings in a spirit that was different
from theirs (see Case C-199/92P Hüls v Commission [1999] ECR I-4287, paragraph 155, and Case C-49/92 P
Commission v Anic [1999] ECR I-4125, paragraph 96).[82] The reason underlying that principle of law is that,
having participated in the meeting without publicly distancing itself from what was discussed, the undertaking
has given the other participants to believe that it subscribed to what was decided there and would comply with
it.( Aalborg Portland A/S v Commission of the European Communities (Joined Cases C-204/00 P, C-205/00 P,
C-211/00 P, C-213/00 P, C-217/00 P and C-219/00 P).
10
11
that the price increases would be adhered to in the market. If MacNeil had
communicated otherwise the other firms may have had to re-consider their strategy to
increase prices. MacNeil has accordingly contravened section 4(1)(b)(i) of the Act.
MacNeil – Remedy
[36] The Commission submitted that MacNeil’s involvement in the cartel was of short
duration starting at the beginning of 2007 until about November 2007. According to the
evidence of Brink, MacNeil got involved with the full knowledge that this was
prohibited. It was during evidence that an entity known as MacNeil Plastics was
identified as the supplier of PVC and HDPE Pipes. The Commission submitted that the
total turnover of this entity should be utilised as a base for calculating the penalty. 19
There is no dispute as to the actual turnover.
[37] The Commission further submitted that there were no mitigating factors; instead there
were aggravating factors in that MacNeil denied any involvement in the cartel, against
evidence of Flo-Tek which allegedly invited MacNeil into the cartel. Flo-Tek has settled
with the Commission and admitted that it attended the same meetings at which
MacNeil was represented that section 4(1)(b) of the Act was contravened.
[38] The Commission submitted that any penalty for MacNeil should be calculated on total
turnover. MacNeil contends that there was no evidence that it participated in any price
fixing in relation to its nylon irrigation fittings and other non-PVC components of its
business.20
[39] MacNeil submitted that the infringing line of business is its PVC pipe turnover. It argued
that there is no evidence to show that any anti-competitive behaviour on its part in the
PVC product market conferred some leverage in another product market which it would
not otherwise have had.21
[40] MacNeil further submitted that its conduct went no further than making itself party to a
[40] MacNeil further submitted that its conduct went no further than making itself party to a
price fixing agreement over the period September to October 2007, at the stage when
19 Commission Heads of Argument, paragraph 127, page 79.20 MacNeil’s Heads of Argument, paragraph 59, page 29.21MacNeil’s Heads of Argument, paragraph 59, page 29.
11
12
the cartel was no longer operating. It was never a member of the cartel and did not
implement any price fixing agreement. MacNeil followed its own pricing structure.
[41] MacNeil submitted further that no damage or loss was suffered as a result of its action.
No profit was derived from any proscribed activity by MacNeil. MacNeil further
submitted that it cooperated fully with the Commission and provided full disclosure.
MacNeil has never previously been found to have contravened the Act.
[42] MacNeil’s involvement arose from the conduct of Brink, of which MacNeil was
unaware.22
Appropriate penalty for MacNeil
[43] In this decision we will follow the approach to calculating penalties that w as adopted by
the Tribunal in Aveng.23
[44] In that case the Tribunal identified the following six step approach to assessing an
appropriate penalty for the purpose of section 59(3). These steps are:
[44.1] Step One: determination of the affected turnover in the relevant year of
assessment;
[44.2] Step two: calculation of the ‘base amount’ being that proportion of the
relevant turnover relied upon expressed as a percentage of the affected
turnover obtained in step1;
[44.3] Step three : where the contravention exceeds one year, multiplying the
amount obtained in step 2, by the duration of the contravention;
[44.4] Step four: rounding off the figure obtained in step 3, if it exceeds the cap
provided for by section 59(2);
[44.5] Step five : considering factors that might mitigate or aggravate the
amount reached in step 4, by way of a discount or premium expressed as a
percentage of that amount, which is either subtracted from or added to it;
22MacNeil’s Heads of Argument, paragraph 67, page 31.23 See supra footnote 18 Aveng.
12
13
[44.6] Step six: rounding off this amount if it exceeds the 10% total turnover cap
provided for in section 59(2). If it does, it must be adjusted downwards so
that it does not exceed the cap.
[45] We start off by considering the affected turnover. In this case the Commission contends
that penalty should be based on total turnover. However, the total turnover includes
other products manufactured by MacNeil that have not been shown to have formed
part of the cartel arrangements or been indirectly influenced by the pricing of the PVC
products. For this reason we are in agreement with MacNeil that the affected turnover
should be limited to the turnover in respect of the PVC products. Given the evidence
that MacNeil’s involvement was limited to the year 2007, this turnover should be
assessed at financial year end 2007. This figure is twenty nine million two hundred
thousand rand (R 29, 2 million).24
[46] The next step is to calculate the base amount, i.e. that percentage of relevant turnover
to take into account to assess the gravity and extent of the contravention. The issue in
this case is whether to assess MacNeil from the point of view of a national cartel or a
regional cartel. If the cartel was national it would have included some players such as
Marley and Amitech; whom the evidence shows were part of the national agreements.
We know, however, that these firms did not attend any meetings involving MacNeil.
We can conclude from this that the cartel agreements to which MacNeil was party,
were regional and largely confined to the Western Cape. This reduces somewhat the
assessment of the harm it caused, given its limited scope, the numbers of firms involved
and its apparent lack of success. For this reason we would consider an appropriate
percentage for Step 1 of 15%, given this was a regional and not a national cartel. This
gives a figure of R 4.38 million.
[47] MacNeil’s involvement does not extend beyond one year and hence no multiplication
[47] MacNeil’s involvement does not extend beyond one year and hence no multiplication
need be applied in respect of duration. Whilst we reject the version that it was involved
from only September to October 2007 – its involvement was at least from February
2007 when Diab attended the meeting at DPI - it was nevertheless involved for less than
one year - approximately seven months. We would thus reduce the base figure by the
24 In 2007 total turnover was R 40 million and Malherbe says 73% of this was PVC, thus the R29, 2 million.
13
14
period of involvement by dividing the affected turnover figure, which was four million
three hundred and eighty thousand rand (R4,38 million), by 7/12 which gives a figure of
two million five hundred thousand rand (R2,5 million).
[48] There are some mitigating features about MacNeil’s behaviour. It appears that it was
pricing below the other cartel members and hence the warning issued to Diab to get
MacNeil’s pricing into line. This is an indication that the other competitors considered
that MacNeil, at least in February 2007, was pricing below cartel prices. The fact that its
turnover in PVC more than doubled between 2006 and 2007, albeit off a low base,
seems consistent with its version that it was taking some market share.25
[49] Further, the evidence that MacNeil did not implement cartel prices is not contradicted.
MacNeil argued that it was in the process at this time of expanding its market share and
that its margins had been cut severely because of this. Again, this evidence has not
been challenged. Nevertheless, access to the information provided by other cartel
members at meetings enabled MacNeil to appreciate where pricing was in the market
in order to gauge where to set its own prices. The fact that its prices were lower than
the other firms attending the cartel meetings does not mean that the prices were set at
a competitive price. Given the existence of the cartel we do not know what that pricing
would have been.
[50] Also taken into account has been the less than frank manner in which MacNeil has
approached its defence in the matter. Diab was a most unsatisfactory witness whose
attempts to minimise his firm’s role were exposed. Further, Malherbe’s assertion that
he was unaware of the meetings which his two subordinates had attended with rivals
on at least three occasions is difficult to accept; more particularly when threats were
made by the market leader for the firm to bring itself in line. It seems extraordinary that
made by the market leader for the firm to bring itself in line. It seems extraordinary that
Diab would not have passed this fact on to Malherbe. MacNeil was a small producer
and one of its significant customers was another self-confessed member of the cartel,
Flo-tek.26 A threat uttered by the largest firm in the presence of a major customer is
highly unlikely not to have been communicated by Diab to his managing director.
25 Total turnover for 2006 was R 24 million. Malherbe says 56% of this was on PVC sales thus 13.44 million. In
2007 total turnover increased to R 40 million and Malherbe says 73% of this was PVC thus 29,2 million.
Transcript, pages 1676-1679.26 See MacNeil’s Answering Affidavit, paragraph 25.1.
14
15
[51] As mitigating factors we take into account that MacNeil is a first time offender and that
market circumstances in which the firms operated were difficult; given that the pipes
were commodities which many firms could make and for which there were powerful
buyers. Level of profitability was low. Malherbe alleged that profit after tax had been
reduced in 2007 from 6% to 3 %.27
[52] For this reason we find that, on balance, the mitigating evidence outweighs the
aggravating evidence and we would reduce the amount of R 2, 5 million by 20%. Most
importantly, apart from this period, MacNeil seems to have been a firm that competed
in the market independently and priced aggressively.
[53] MacNeil is liable for a penalty of two million rand (R 2 000 000.00). This figure is less
than 10% of its total turnover in the 2009 financial year, the last year for which MacNeil
gave us a figure for total turnover.28
Andrag
[54] Andrag is also a Cape based firm. It is family owned and its core business is to supply
various products to the agricultural sector. Amongst the products and services it
provides are irrigation systems. This has led the firm to supply PVC pipes and small
quantities of non-SABS accredited HDPE pipes. Andrag claims that almost all of its
customers are in the agricultural sector and it sells to them directly through its branches
and not through dealers. 29 In this niche market Andrag competed with DPI and
Petzetakis. Andrag states that it started manufacturing PVC pipes in 1989 because of
the high cost of the product. It does not manufacture the full range of pipes; rather it
manufactures some classes and purchases the rest from other suppliers.30
[55] The case against Andrag is limited and concerns the attendance by one of its directors,
Walter Andrag, at a meeting in August 2006, which was also attended by
representatives of DPI and Petzetakis, and two subsequent telephone calls. At this
representatives of DPI and Petzetakis, and two subsequent telephone calls. At this
meeting pricing proposals were made by a DPI representative. It is not alleged that he
attended any further meetings, so the issue to be decided is whether his conduct at the
27 Transcript, page 1676 onwards.28 Total turnover for this year was R 72 million. The penalty would therefore be 2,8 % of its total turnover.29 Andrag’s Answering Affidavit, paragraphs 10-11.30 Andrag’s Answering Afidavit, paragraphs 13.
15
16
meeting and subsequently, was such as to make his firm a party to a collusive
arrangement by signifying an acceptance by his firm of the DPI’s pricing proposal.
[56] Most of the facts concerning this meeting are common cause , however, the inferences
that may be drawn from these facts, which are unusual, are not.
[57] Walter Andrag states that he was invited to this meeting by Peter Ke rr Fox, then of
Petzetakis. Andrag did toll manufacturing for Petzetakis as Petzetakis did not have the
necessary manufacturing capacity in the Western Cape. 31 Andrag was not told the
purpose of the meeting, but given the business relationship he had with Petzetakis, he
assumed the meeting related to the toll manufacturing agreement. When he arrived at
the meeting, which was at Petzetakis’ premises, he found Rene Le Riche and Keith Terry
of DPI present. He was surprised to see them as Kerr Fox had not told him they were
going to be present. Le Riche at the time was DPI’s sales director with responsibility for
the Western Cape.32
[58] At the meeting Le Riche proposed that the three firms (DPI, Petzetakis and Andrag) co-
ordinate over their discount structures. Specifically he proposed that they apply a
discount of 55% off the DPI list price. Walter Andrag states that he was taken aback by
the proposal, but that he agreed to take the proposals back to his sales force.
[59] During the course of the meeting , Le Riche confronted Andrag with a quotation that
Andrag had provided to a mutual customer, Bergrivier Irrigation, an irrigation
engineering firm that supplies farmers. Le Riche’s purpose in doing so, was to accuse
Andrag of not sticking to an apparent prior agreement on pricing. Andrag had
apparently quoted Bergrivier a price below that which Le Riche had understood Andrag
and Petzetakis to have agreed upon. Walter Andrag again deferred any comment on
this accusation other than undertaking that he would look into it.
this accusation other than undertaking that he would look into it.
[60] After the meeting, Andrag phoned his contact at Bergrivier to complain that his quote
had been given to a competitor. It appears that Andrag and DPI worked through
different contacts at Bergrivier. (What emerges from Le Riche’s evidence is that DPI has
31 Ibid, paragraph 12. In terms of this arrangement Petzetakis supplies capital equipment to Andrag who in turn
manufacture PVC pipes for Petzetakis on a cost plus basis. See also the witness statement of Peter Kerr Fox
paragraph 10 and see respondent’s witness statements, page 649.32 Transcript, page 327.
16
17
a long relationship with a sales representative at Bergrivier and that Andrag’s quotation
threatened to disrupt this cosy relationship).33
[61] That same afternoon Andrag says he received a telephone call from Andre Auret of DPI,
Le Riche’s superior. Auret told him to treat everything said at the meeting as
confidential. Walter Andrag understood this to refer to everything said at the meeting.
Auret’s version of the conversation, is that the confidentiality request was limited to a
discussion of Bergrivier, as Auret did not want Le Riche’s comments to get back to the
customer.34 Nevertheless what is important here is the response of Walter Andrag who
admits that he did not tell Auret that he was not part of the pricing arrangements.35
[62] Sometime later, Andrag was called by Le Riche on his private line and asked why Andrag
was not adhering to the prices agreed at the meeting. Walter Andrag informed him that
Andrag did not operate pricing centrally and that prices were set at a regional level by
branches. He testified that he tried to explain this as nicely as possible to Le Riche to
avoid confrontation.
[63] This phone call was the last contact that Andrag or his firm had with DPI and Petzetakis
over pricing. It is not clear how long after the meeting this phone call was made. An
attempt was made by Andrag’s counsel to use Le Riche’s phone records to date this
conversation, but this effort failed as no one has been clear on the exact date of the
meeting. The phone call, we can assume, came some time after the meeting – not that
long after, but presumably long enough for Le Riche to pick up information from the
market that Andrag was not adhering to the agreement. We can conclude that after this
call Le Riche, and hence DPI, no longer considered that Andrag was an adherent to the
agreement on discounting.
[64] Andrag denies that it was ever party to an agreement , either before this meeting or
[64] Andrag denies that it was ever party to an agreement , either before this meeting or
thereafter. It presented an unusual explanation for this, which entails the curious role
played in this whole saga by Peter Kerr Fox, then of Petzetakis, but who at the time of
the hearing had joined DPI.
33 Transcript, page 357.34 Transcript, page 302.35 Transcript, page 1463.
17
18
[65] Kerr Fox is the elephant in the room in respect of the August meeting, yet neither the
Commission nor Andrag called him as a witness.
[66] Instead, what we have from Kerr Fox is a witness statement that he provided to Andrag
prior to the hearing. By this time he was no longer working for Petzetakis, but DPI. In it
Kerr Fox confesses that he played a game of duplicity with both DPI and Andrag. Kerr
Fox explains that DPI was concerned about pricing from Andrag in the market and was
anxious to get it to agree to pricing arrangements. Le Riche had approached Petzetakis
to act as a go-between with Andrag, given their tolling relationship. Kerr Fox had
indicated to DPI that Andrag was willing to adhere to arrangements. In fact, he had
never discussed anything of the sort with Andrag. However, he was anxious to make DPI
believe that Andrag was part of the arrangement, because he feared that if it was not,
DPI- the bigger, more aggressive firm- might otherwise engage in a price war. If there
was a price war, Petzetakis would be the loser. For this reason, although there was no
direct communication between DPI and Andrag up until this meeting, his assurances
that Andrag were on board appeared to have pacified DPI for a time.
[67] As he put it in his statement:
“In short the impression I created with DPI was false. I created the impression with DPI
Plastics (i.e. that I had influence over Andrag) to ensure that Petzetakis was protected
from a price war by DPI Plastics.”36
[68] Kerr Fox never explains in his witness statement why he never discussed the pricing
with Andrag despite holding out to DPI that he had. However, the Bergrivier quotation
issue seemed to have exposed the fault line in this fragile strategy and so, at the
request of DPI, he arranged a meeting at which Andrag could be confronted by DPI. He
conceded further that he had misled Andrag as well. He had not told it of the
conceded further that he had misled Andrag as well. He had not told it of the
36 Kerr Fox’s witness statement, paragraph 11.2. See respondent’s witness statements, page 650.
18
19
arrangements with DPI nor did he tell Walter Andrag that DPI would attend the meeting
in August.
[69] The statement from Kerr Fox was put to Le Riche who confirmed much of it. He was not
in a position to comment on any communication between Andrag and Kerr Fox. Where
Le Riche differs with both is over the reason for the meeting and what was held out by
Andrag during the meeting. Le Riche says that the meeting was called at Andrag’s
request as it was concerned over the prices DPI was quoting Bergrivier and that it
wanted to confront DPI about this. 37 Later, during cross examination, Le Riche modified
this to state that DPI had called the meeting because it wanted to show Andrag “... that
it’s not us”.38
[70] Le Riche did not adhere to this position very confidently so we will accept Andrag and
Kerr Fox’s version, that the meeting had been called at DPI’s initiative.
[71] It is common cause that Walter Andrag had, after the meeting, complained to his
contact at Bergrivier about the quotation being given to DPI. The details of this phone
call had got back to DPI. Le Riche had phoned Kerr Fox and stated that this “ guy”
(meaning Walter Andrag ) “ was not in” .39 Again, we are not sure of the timing of this
call, but given that according to Le Riche it had been precipitated by Andrag’s call to
Bergrivier after the meeting, we can assume it was close in time to the meeting.
[72] Although the factual versions of what was said when and the sequence of events are
largely common cause there are two disputes between the Commission and Andrag
that we must resolve, one factual and one inferential.
[73] The factual dispute concerns whether Andrag was aware of the cartel prior to the
meeting. Despite in cross examination suggesting to Le Riche that Kerr Fox would
testify, Andrag, which had taken his witness statement, did not call him and instead,
engaging in a game of brinkmanship with the Commission offered him to them as a
engaging in a game of brinkmanship with the Commission offered him to them as a
witness – an offer the Commission declined.
37 Transcript, pages 507 – 508.38 Transcript, page 508.39 The suggestion was made in the Kerr Fox witness statement, but was put to Le Riche in cross examination
which he agreed with. Transcript, page 525.
19
20
[74] There was much debate about who in the circumstances had the duty to call Kerr Fox.
The Commission , which has the overall onus to prove its case, or Andrag which had
taken the witness statement and who relied on him as part of its defence. We need not
decide who had this duty and whether we should draw any adverse inference against
the party who did not call him.
[75] Two corroborative facts seem to suggest that Andrag’s version - that he was not party
to the pricing arrangement prior to the meeting - was genuine. First, the lower quote to
Bergrivier is consistent with his version of non-adherence to any prior agreement.
Whilst this might still be explicable as cheating, we have, secondly, the evidence of Le
Riche as to Andrag’s demeanour in the meeting which would suggest the latter was
indeed caught by surprise. According to Le Riche:
“He was a bit tentative in the beginning but then slowly started participating. We got
down to discussing a discount structure which would be list less 55 off the price list.
He agreed to take that information back to his sales team and discuss it with his sales
guys.”40
[76] Andrag’s version, that he was not party to any agreement prior to this meeting and that
he was taken by surprise by it, is reasonably possible and absent any evidence from the
Commission to refute it, we will accept it.
[77] More difficult is the issue as to whether Andrag was a party to the agreement on
discounting, proposed at the meeting by Le Riche, after the meeting. Since Kerr Fox did
not testify, we have to decide this point on the testimonies of Le Riche and Walter
Andrag.
[78] Le Riche concede d that Andrag did not expressly agree to participate in the cartel.
However, he says the fact that he agreed to take the discount proposal back to his sales
team amounted to an agreement in principle to be involved.41
40 Transcript, page 352.41 Transcript, pages 513-514.
20
21
[79] Le Riche concedes that Andrag was uncomfortable about the proposal, but nevertheless
thought he would look at it because it was in his business interests to do so, in the same
way as it was in DPI’s.42
[80] Counsel for Andrag and Le Riche sparred over whether this inference was one for him
to legitimately draw, but Le Riche remained adamant on this point.
[81] Walter Andrag maintained, with equal steadfastness, that he never became party to the
agreement. However, he had difficulty explaining, if this was the case, and if he had
been invited to the meeting under false pretences, why he had not said so expressly.43
[82] Andrag was asked in cross examination by the Commission’s counsel why he did not say
at the meeting that he was not part of the deal when there was no constraint upon him.
His answer was:
“No, the constraint was that one of the parties in the room was a substantial
customer of ours. The other constraint was that both of them were suppliers to us
and I didn’t want a big confrontation. That was the constraint.”44
[83] He was also cross-examined on why he had, at the meeting, undertaken to investigate
his firm’s pricing to Bergrivier in response to Le Riche’s accusations. His answer was “It
was the only sensible way of ending that, but you are right I agree. I said let me have a
look at that.”45
[84] This answer is also relevant to an argument made by Andrag that whatever impression
DPI was labouring under about Andrag being part of the agreement was a result of the
false impression created by Kerr Fox, for which Andrag cannot be blamed. Yet whatever
the merits of this argument are up until the time of the meeting, once Walter Andrag
was confronted with the Bergrivier quotation, and accused of not adhering to a pricing
agreement, he became responsible for how he responded. The Bergrivier arrangement
clearly refers to a prior understanding. Andrag, given his version, would have had no
clearly refers to a prior understanding. Andrag, given his version, would have had no
knowledge of this agreement until that moment. Instead of saying something to this
42 Transcript, pages 522-523.43 Transcript, pages 1461-1463.44 Transcript, page 1466.45 Transcript, page 1463.
21
22
effect at the very least, he remains passive and simply remarks that he will look into it.
At that moment he must accept responsibility for the reliance created in the mind of Le
Riche.
[85] This conduct continued; for when he was phoned later that day by Auret who criticised
him for phoning Bergrivier, he conceded in cross examination that he again did not
indicate that he was not part of any pricing arrangement.46
[86] Despite these concessions Walter Andrag was unwilling to concede that his behaviour
amounted to a representation to the others (i.e. DPI and Petzetakis) that he had agreed
with what had been discussed. He said it would be an assumption on their part if they
did, but it was definitely not the message he conveyed. 47 But the fact that Le Riche later
called him to ask why he was not adhering to the pricing arrangement, seems
inconsistent with Andrag’s view that the others should have known that he was not
party to the pricing agreement. Recall that even in this call, Andrag did not dispute that
his firm was a party to the agreement; instead he made the feeble excuse that non-
adherence by his firm was due to logistical problems. Le Riche’s call and conversation is
consistent with DPI’s understanding that Andrag had made itself a party to the
agreement after the meeting and he was phoning to try and understand why Andrag
was not adhering to the agreement. If his understanding had been any different, the
call makes no sense.
[87] This is the crucial point in the case against Andrag. It is not disputed that the meeting
had discussed a proposal on pricing for adoption by the three firms present. The only
question is whether the two other firms could have reasonably concluded that Andrag
had agreed to the proposal. It is also not in dispute that he neither stated affirmatively
that he was in agreement nor that he rejected the proposal. He stated on both issues
that he was in agreement nor that he rejected the proposal. He stated on both issues
that he would look into it. Certainly, this was at best for Andrag, an ambivalent
response.
46 See this extract in which the Commission’s counsel puts this question to him: “MAENETJE: Well you didn’t
say that, I don’t see why I should stick to the arrangements of the meeting and not speak to the customer, not
so? ANDRAG: Yes.” (Transcript, page 1464.)47 Transcript, page 1467.
22
23
[88] Was he, given the circumstances, under a legal duty to give an unambiguous response?
Expressed differently, even if he had been innocently placed in a situation not of his
own creation, once he was in it, did he have a positive duty to reject the proposal lest
the other would-be conspirators in the room assume his firm was in agreement with
them and place reliance on this lack of rejection to behave in the market accordingly- to
the detriment of consumers?
[89] Areeda and Hovenkamp , in their treatise, write about how a legally similar situation
would be regarded in United States law and conclude that even where a competitor
has been lured innocently into an occasion where an agreement is discussed with its
competitors that is unlawful that innocent competitor is obliged to depart:
[90] “Suppose, for example that one entrepreneur invites its major rivals to dinner. After
dinner, the host announces future price intentions and one or more of the guests does
the same. The only objection to finding an agreement to exchange this information is
the guests’ claim that they were taken by surprise. But we can reject this suggestion
because the guests had an option: although it might have been socially uncomfortable,
the guests could have departed as soon as the host began to transform the social
occasion into a conspiratorial one. The dinner guest can fairly be held to have entered
an information-exchange agreement when she chose to remain”.48
[91] What the authors are suggesting, and this is consistent with the approach that has been
taken by the European Courts in a number of cases which we cited in Aveng, is that
once a firm has been placed in such a situation by its rivals it has a positive duty to
repudiate the proposal, and that if it does not do so, it will not escape liability because
of some private, but unexpressed reservation.
[92] There is no reason for our law to adopt a less stringent approach to this issue than do
[92] There is no reason for our law to adopt a less stringent approach to this issue than do
the European and United States courts. Indeed , the far reaching provisions of sections
4(2) - 4(5) of the Act, which presume agreements between rival firms where one has a
substantial interest in another or common directors, illustrate the extent of the
legislature’s intent to discourage cartels, such that when competitors find themselves
48 See Areeda and Hovenkamp, Antitrust Law – An analysis of Antitrust Principles and their application, Second
Edition, paragraph 1406c.
23
24
in the same enterprise at board or shareholder level, they may be rebuttably presumed
to have agreed to engage in a restrictive horizontal practice. 49 As Areeda and
Hovenkamp observe in their treatise, the cases seldom resolve the question of what
degree of assent suffices for liability. They conclude the resolution of this problem does
not rely in reliance on the common law of contracts or vague words like “assurance”,
but in statutory policy. As they put it: “We must always return to the antitrust policy
underlying the statute’s focus on agreements.” They make the point that for antitrust
purposes:
“... there will be an agreement even though the challenged arrangement falls short of
forming a contract because, for example, the parties declare an intention not to be
legally bound, each party reserves the right to abandon the venture at will and
without notice, offer and acceptance are not full in accord, or the understanding is
too vague to allow a court to enforce it ( even though it were not illegal). ”50
[93] Thus, returning to our Act, if sub-sections 4(2)-(5) manifest the extent of competitor
interaction that the legislature seeks to condemn, based on the mere form of the
relationship, it is safe to assume that in South Africa antitrust policy imposes a heavy
burden on firms who find themselves in meetings where the subject of discussion
concerns matters set out in section 4(1)(b) of the Act. We derive from this an antitrust
policy, to follow Areeda and Hovenkamp, which has an overriding intention to condemn
and discourage such contacts. Note that in the definition of agreement in the Act it is
made clear that an agreement is not necessarily one that could be enforceable at law. 51
This suggests that the interpretation given to agreement for antitrust purposes
suggested by Areeda and Hovenkamp above is one equally applicable to our statute.
suggested by Areeda and Hovenkamp above is one equally applicable to our statute.
[94] The various respondents in this matter had relied on the decision of the Competition
Appeal Court in Netstar to suggest that for competition law purposes, the prerequisites
49 Section 4(2) provides that an agreement to engage in a restrictive horizontal practice is presumed to exist
between two or more firms if any one of those firms owns a significant interest in the other or if they have at
least one director or substantial shareholder in common. Section 4(3) states that the presumption may be rebutted
if a reasonable basis exist to conclude that the agreement was a normal commercial response to conditions
prevailing in the market. 4(4) provides for definitions, whilst 4(5) excludes the operation of the presumption
when firms form part of the same single economic entity or are wholly owned subsidiaries.50 See Areeda and Hovenkamp, op cit, paragraph 1404.51 Section 1(1)(ii) defines an agreement, when used in relation to a prohibited practice, as including “ ...a
contract, arrangement or understanding, whether or not legally enforceable.” (Our emphasis).
24
25
for agreements were close to those of their common law counterparts. Hence, if some
element of offer and acceptance to a specific agreement is missing than an agreement
for the purpose of section 4 of the Act was not established. 52
[95] In the passage relied upon , the Court, whilst acknowledging that the definition of an
agreement in the Act extends the concept beyond a contractual arrangement, still
required “...a form of arrangement that the parties regard as binding on both
themselves and the other parties to the agreement.” The Court had also held that “ Its
essence [meaning the agreement] is that the parties have reached some kind of
consensus....”
[96] As we interpret that decision , it is authority for the proposition that there must be an
element of consensus and secondly that there be some “ form” to the arrangement in
order for it to constitute an agreement. It is not authority, because the Court did not
have to decide that question, on what form that consensus needs to take nor what facts
would suffice to infer consensus. This latter issue the United States and European
courts, in the passages cited earlier, have considered. They have determined that an
agreement may be inferred to exist in certain circumstances for antitrust purposes,
even though they might be insufficient for a conclusion that an agreement existed at
common law. To hold otherwise, would not only defeat the statutory purpose to
prevent cartel activity, but also to deprive the definition of an agreement in the Act, of
an essential part of its meaning, when it suggests that an agreement may still exist
“...whether or not legally enforceable” . The reason for the distinction is vital to
appreciate. We are dealing in collusion cases with the behaviour of firms. If firms allow
a situation where others have an understanding that they can act in a collusive manner,
a situation where others have an understanding that they can act in a collusive manner,
because rivals will do so as well, a fortiori competition is harmed and consumers suffer.
It does not matter that privately, and unexpressed to the others, the respondent firm
did not regard itself as bound by the understanding. If it created a situation where
reasonable reliance that it was a party to the agreement might be inferred by the other
parties to the collusive design, it was under a positive legal obligation to rectify it.
52 See Netstar (Pty) Ltd and Others v The Competition Commission and Another CAC May 2010, paragraph 37
of that judgement.
25
26
[97] Contract law is concerned with the private consequences of agreements ; competition
law is concerned with the public consequences. Formal offer and acceptance, whilst
rigid concepts in the common law of contracts, become more supple in the competition
law form of the concept. Within the ambit of competition law, in a particular context,
where there is a duty to speak, silence or an ambivalent answer may suffice as well as
the witnessed signature, to signify assent.
[98] Thus when competitors meet, even on an unintended occasion for some, once the
conversation moves to proposals of an unlawful agreement, those attending must
repudiate the proposal by conduct in unambiguous terms, however awkward it may be
to do so, lest the other firms present reasonably infer that the accused firm had
assented.
[99] Walter Andrag found himself placed in a legally comprising situation. His failure to
affirmatively, by some form of co nduct, repudiate the proposal for an agreement made
at the meeting by his two competitors makes Andrag liable for entering into an
agreement in contravention of section 4(1)(b)(i) of the Act.
Appropriate r emedy for Andrag
[100] We have decided not to impose any penalty on Andrag, despite finding that it
contravened section 4(1)(b)(i) of the Act. Although a penalty for contravening this
section is normally imposed, it is not mandatory.
[101] In the circumstances of this case there are grounds for not imposing the penalty. Firstly,
the duration of the period in which competitors would have reasonably believed that
Andrag was party to the agreement , was relatively short, and whilst precise dates are
uncertain, this must be resolved in Andrag’s favour. It is not clear for how long the
other two firms relied on Andrag to adhere to the agreement, as DPI became suspicious
of the latter’s inadherence shortly after the meeting when, on a date unknown, Le
of the latter’s inadherence shortly after the meeting when, on a date unknown, Le
Riche had, in his call to Kerr Fox, stated ‘ this guy is not in’ . Finally, we can assume, that
at least after the date of the call from Le Riche, where Andrag stated, albeit somewhat
feebly, that his low prices were due to his branches doing their own thing, DPI would
have realised that Andrag was not going to adhere to the agreement. On a generous
26
27
construction for Andrag this would have been the first act of repudiation even though
Le Riche may have been sceptical about his adherence before this.
[102] We will also, for Andrag’s benefit, accept that he came to the meeting in August
innocently and that he had no prior involvement with the cartel. We accept further that
the meeting was awkward for him and whilst the failure to repudiate does not
exonerate his firm from legal liability, it is a factor that can be considered in mitigation.
[103] We accept further that Andrag was pricing more aggressively than the other firms and
may well, as it asserts, not have implemented the agreement. The fact that it had priced
aggressively before the meeting we know from the dispute over Bergrivier. Le Riche’s
later phone call is also consistent with Andrag’s own evidence that it did not price at the
discount price agreed. That was, after all, why Le Riche was complaining.
[104] Pricing evidence adduced by Andrag suggests that its discounts did not follow the
discount formula proposed by Le Riche at the meeting, which corroborates the other
evidence that Andrag did not implement the proposed discount. Le Riche also conceded
that Andrag might have had a lower cost structure than DPI.53
[105] What is not clear of course is whether it nevertheless raised prices beyond the level
they may have been , but for the information acquired at the meeting. As the
Commission had the pricing information from Andrag, but did not analyse it, we will
again give Andrag the benefit of the doubt on this issue.
[106] Andrag has not been found to have been previously in contravention of the Act. It was a
minor player in a niche segment of the industry. For all these reasons , but most
importantly, the issue of the short duration and the fact that Andrag came to the cartel
innocently on these facts, we find that it would not be appropriate to impose a fine on
innocently on these facts, we find that it would not be appropriate to impose a fine on
Andrag and a finding of a contravention suffices.
Gazelle
[107] Gazelle, like Andrag and MacNeil, is a relatively new entrant to the plastic piping
industry. In 1994 Gavin Warmback formed Gazelle Engineering CC as a producer of
53 Transcript, page 512.
27
28
HDPE tubing for conveyer idlers. Amongst its customers were several mining houses. In
2001 one of them suggested to Warmback that he enter the market to manufacture
HDPE pressure pipes for the mining industry. This customer considered that the sector
was in need of some price competition.
[108] Warmack did so, in the process converting the close corporation into a private company
of the same name. In 2004 he entered the PVC business using Gazelle Plastics as its PVC
arm.54 In practice the companies were run by the family as one and hence the issue of
joint and several liability in the event of liability in this case has not been contested.55
[109] Most of Gazelle’s pipes were produced for the mining industry.56 It states that 70% of its
pipes were HDPE, with the remaining 30%, being PVC. It has, however, also sold some
pipes to merchants. These include a firm called Peakstar, a merchant based in the
Western Cape, of whom more will be said later.
[110] By all accounts , not just its own, it was an aggressive entrant winning market share
from the incumbents. According to Warmback’s testimony they began charging prices
that were 40-50% cheaper than what the then incumbents were charging. 57 He testified
that Gazelle took away contracts from DPI and Petzetakis- which these firms had held
for fifteen and twenty years, respectively.58
[111] Gazelle’s competitors in the mining industry were Petzetakis, DPI and Marley.
Petzetakis had the most to lose by Gazelle’s entry. In order to justify to their customers
why they had been paying so much more to them, competitors, and in particular
Petzetakis, had to denigrate the Gazelle product and suggest that its lower prices were
a result of poor quality.
[112] The ‘bad mouthing’, as Warmback described the denigration, is important because it
led to the first key meeting that representatives of Gazelle had with a member of the
cartel. In July 2006, Gavin Warmback arranged a meeting with the Petzetakis managing
cartel. In July 2006, Gavin Warmback arranged a meeting with the Petzetakis managing
54 See Answering Affidavit for this history. In particular paragraphs 17-26.55 The Commission applied to the Tribunal for the joinder of Gazelle Engineering as tenth respondent as it is one
and the same firm with Gazelle Plastics, the ninth respondent. The application was not opposed and was granted
by the Tribunal.56 Gazelle states that 95% of its customers were in the mining industry. Answering affidavit, paragraph 21.57 Transcript page 1505. Warmback later suggests his prices were 30% less. Transcipt,1506.58 Transcript page 1506.
28
29
director, Michele Harding, at a coffee shop in Centurion. In attendance was his brother
Haydan. The meeting according to Warmback had been suggested by Roy Carver of
Marley. Carver and Warmback had a good relationship.59
[113] Versions of what happened at this meeting differ. The Commission alleged that the
Warmback brothers indicated that they were aware of industry discussions and
agreements regarding selling prices for HDPE and other pipe products, and wanted to
be party to the discussions, agreements and/or arrangements.60
[114] This version in the affidavit presumably came from Michelle Har ding. When she
testified as a Commission witness, Harding stated that Gavin Warmback enquired
whether there were prices set for HDPE in the industry. Harding denied that there were
any discussions in the industry around pricing. This of course was untrue. In her
testimony she explained that she gave a false account of what went on because she
could not trust Gazelle and did not want to expand the cartel. 61 Harding advised the
Warmback brothers to obtain Petzetakis’ standard price list and to follow it if they so
wished.62 Warmback, in his evidence, limits the conversation to a discussion around the
bad mouthing, but he admits to having a general discussion with Harding around recent
increases in raw material prices, although he denies that they discussed prices for
piping or discounts.
[115] The Commission does not allege that an agreement on prices was reached at this
meeting nor does it seek to rely on it to suggest that was the beginning of Gazelle’s
involvement with the cartel.
[116] What then, if no agreement was reached, does knowledge of this meeting contribute to
the Commission’s case against Gazelle? The Commission relies on it for two purposes.
Firstly, if Gazelle was interested in making pricing enquiries from a competitor it makes
Firstly, if Gazelle was interested in making pricing enquiries from a competitor it makes
it more probable that it would be willing to enter into agreements with rivals around
pricing as, on the Commission’s case, it did later. Secondly, the Commission seeks to
59 See answering affidavit paragraph 63.60 See Commission Supplementary Founding Affidavit, Pleadings Bundle paragraph 9.1, page 35 of the
pleadings.61 Transcript page 203.62 Commission Founding Affidavit, Pleadings Bundle, page 18, paragraph 35.
29
30
demonstrate that Warmback is an unreliable witness, whose version of events is
contradicted by other Commission witnesses.
[117] Sometime after this meeting, it is not clear when and in what sequence, Gazelle,
according to the Commission, is implicated in three further contacts with competitors at
meetings and has an ongoing relationship with Marley’s Roy Carver, as a go-between
over pricing.
[118] The first of these three meeting is not relied on to support the case for price collusion,
but serves again to buttress the Commission’s case on probabilities, as does the prior
meeting with Michelle Harding. It is common cause that sometime in 2006, Warmback
was involved in a commercial dispute with a company called Flexicon. Gazelle was
subcontracted to supply Flexicon with piping, the latter were in turn, supplying a mine
in Witbank. During the course of the contract, raw material prices increased. Gazelle
sought unsuccessfully to persuade Flexicon to pass on the increases to the customer.
When this stratagem failed, Gazelle attempted to win the contract for itself directly. To
do this it had to ensure that no rival firm competed for the contract. To this end he
arranged a meeting with Roy Carver and Mike Smart of Petzetakis. At the meeting
Warmback urged the other two to stand back and in return Gazelle would give them a
cut. On Carver’s version, Smart was not persuaded, as he did not trust Warmback to
honour the sub-contract. On Warmback’s version, Smart also does not go along with
the proposal, although the reason given for not doing so is different.
[119] At the hearing Carver produced an invoice which he said related to this sub-contract,
but there was nothing on the face of it to link it to the Flexicon sub-contract and Carver
concedes this.63
[120] Whilst this episode supports the Commission’s version on two aspects- firstly that
Warmback was not averse to attempting to collude with two rival firms when it suited
Warmback was not averse to attempting to collude with two rival firms when it suited
his commercial purpose to do so and secondly, that he was willing to accommodate
Petzetakis in such an arrangement despite the mutual animosity – it also supports that
of Gazelle. Gazelle’s, most convincing defence, as we examine later, was that the
63 Transcript, pages 599-600.
30
31
collusion could not be effected, because the other firms, particularly Petzetakis, did not
trust Warmback.64
[121] We know the sequence of the two remaining meetings , but we do not know if they
preceded or succeeded the meeting on Flexicon, discussed above. The one meeting
took place in Isando, Johannesburg and the second in Cape Town.
[122] The Isando meeting took place probably in late 2006 or early 2007. Carver does not
recall how the meeting was set up, but suggests that it had something to do with the
Cape based Andre Auret of DPI being in town.65
[123] Warmback’s version is also that Carver set up the meeting. But Warmback’s version
about why he did so was curiously inconsistent. His first version was that Carver had
proposed the meeting saying that Auret was going to be in Johannesburg and wanted
to meet with Gazelle to discuss a reciprocal tolling arrangement. Warmback was
enthusiastic about the proposal.66
[124] But during cross examination Warmback vehemently denied that Carver had discussed
with him before the meeting that it would be about toll arrangements:
“Not before the meeting, I mean how can Roy Carver speak on DPI’s behalf” 67
[125] There is also no agreement on who attended the meeting. Whilst all are agreed that
Warmback, Andre Auret of DPI and Carver attended the meeting, there is a dispute over
the presence of Peter Kerr-Fox of Petzetakis. Yet again Kerr Fox is a crucial figure at a
meeting who no side wished to call. The importance of Kerr Fox’s presence or absence
is that if he was there, it is consistent with the Commission’s case that the meeting was
to discuss an agreement on pricing for HDPE and hence representatives of four rival
companies (DPI, Marley, Petzetakis and Gazelle) were present; if he was not, his
absence is more consistent with the Gazelle version of the meeting – that it was limited
to a discussion of toll manufacturing between Gazelle and DPI and hence the presence
to a discussion of toll manufacturing between Gazelle and DPI and hence the presence
64 It’s important to distinguish between Petzetakis having an antipathy towards Gazelle and from them not
trusting it. Whilst the former did not inhibit the possibility of reaching an understanding, the latter did.65 Transcript, page 604.66 Transcript, page 1547.67 Transcript, page 1570.
31
32
of another person, from a firm unconnected to the purpose of the meeting, would be
improbable.
[126] Auret and Carver both place Kerr Fox at the meeting, whilst Warmback was adamant he
was not there. Harding testified that he may have attended. Given this testimony one
may have expected Gazelle to call Kerr Fox, who we know from the earlier discussion of
Andrag, was available – but it did not do so; despite the fact that it was put during
cross-examination to Carver that Kerr Fox would say that he was not at the meeting. 68
Kerr Fox in his witness statement given to Andrag does not mention this meeting in
Isando, so this instruction cannot have come from the statement on the record.
[127] On the Commission’s version discussion at the meeting concerned both the tolling
arrangement between DPI and Gazellle and a discussion of pricing for HDPE. This is
confirmed by Auret and Carver. Carver has the better recollection of the discussion and
states that the meeting discussed two different specifications of HDPE, the PE 80 and PE
100. When asked whether the meeting reached any agreement or understanding his
answer was:
“Nothing changed as far as the rand per kilogram pricing. That was still agreed that
that was the basis for pricing. I can’t still recall what the rand per kilogram was at
that stage but we agreed to continue on that basis.”69
[128] Auret stated that he recalled the meeting discussing pricing as this was the first time
that he h ad heard of per kilogram pricing. 70 Whilst he conceded that the pricing of
HDPE was not of great interest to DPI, it was the unusual manner of the pricing that he
remembered. Crucially, Auret is unable to help on whether an agreement was reached
at the meeting or the role of the other firms present. In his testimony his vagueness is
evident:
“...I’m not too sure what transpired with regard to the actual meeting or why we had
our competitors there to discuss it, but we did discuss reciprocal business with regard
to PVC.”71
to PVC.”71
68 Transcript, page 694.69 Transcript, pages 603-605.70 Transcript, page 317.71 Transcript, page 278.
32
33
[129] It is also not clear whether Warmback joined a meeting already in progress or left the
others who continued meeting. If this was the case it might be that the meeting to
discuss pricing for HDPE occurred in his absence. This seems unlikely. If Gazelle was not
to be trusted, why risk having Warmback see them meeting. Especially if, on his own
version, the meeting was limited to discussing the DPI/Gazelle tolling arrangements.
[130] Carver also adds that the meeting was intended to repair the relationship between
Gazelle and Petzetakis. As he put it “...to accept each other in the market place a bit
more.”
[131] Although Harding did not attend this meeting she was asked whether she knew if Ker r
Fox had attended it. She stated that she knew of the meeting and heard at a late stage
that Gazelle was invited to attend and for that reason she did not want Kerr Fox to go.
She did not, however, know definitively if he went or not.72
[132] Warmback limits the discussion at this meeting to a discussion of the tolling
arrangements. He says in his answering affidavit that he told Auret that the only basis
on which he was interested in a tolling arrangement was if there was reciprocity, i.e.
that Gazelle would manufacture for DPI in Johannesburg. Since no agreement was
reached on this aspect it led to a later discussion at a meeting in Cape Town.73
[133] For the Commission this meeting is a crucial plank in its case against Gazelle as it
establishes that an agreement was reached on pricing or that Gazelle acceded to an
already agreed formula on pricing of HDPE.
[134] The problem is that even if we accept the evidence of Auret and Carver as to who
attended and what was said , it remains unclear if any understanding was reached with
Gazelle. Auret recalls the discussion and not much else, since DPI was not concerned
about HDPE. Carver is unsure whether an agreement was reached or if an agreement
about HDPE. Carver is unsure whether an agreement was reached or if an agreement
had been entered into already, prior to the meeting, and was just being confirmed or
further discussed. Crucially, he could not recall how Warmback reacted to the pricing
discussion. Since Warmback did not attend one of these meetings again, we do not
know from his attendance at this one, if he signalled, expressly or tacitly, any
72 Transcript, page 217.73 See Gazelle’s answering affidavit, paragraphs 80-83.
33
34
agreement with the proposal or, if it was already an agreement, any intention to abide
by it. It is surprising given the level of suspicion against Gazelle that neither can recall
anything on this aspect.
[135] Thus, even if Warmback’s evidence about who attended the meeting and whether
discussions about HDPE pricing took place is unsatisfactory, the Commission’s
witnesses, on their own versions, do not, from their reconstruction of the meeting,
prove that Gazelle was party to an agreement on pricing of HDPE with its competitors.
[136] This does not dispose of the Commission’s case against Gazelle as an agreement
between Gazelle and the other firms might still be established inferentially by events
subsequent to this meeting and it is to these that we now turn.
[137] It is common cause that following the Isando meeting , Gazelle met with Auret of DPI in
Cape Town. This meeting appears to have taken place in February 2007.
[138] Warmback’s version of what happened at the Cape Town meeting is set out in his
answering affidavit. He says that Auret had wanted to have a discussion around a PVC
arrangement between the two firms. For this reason Warmback travelled down with his
factory manager Steve Ashley, who was the appropriate person to discuss the technical
aspects of the arrangement. At the meeting it turned out that Auret’s real purpose was
to discuss Peakstar, a merchant that supplied pipes to the building industry and which
had become a thorn in DPI’s side as it was selling pipes at retail prices lower than DPI’s
wholesale prices. Auret wanted to know how Peakstar could afford to sell at the low
prices it did and, in particular, whether Gazelle was supplying it. Warmback does not
tell us what he said at the meeting. However, he later met with Colin Brown of Peakstar
who informed Warmback that he had been supplied by Swan. Warmback later passed
this information on to DPI.
this information on to DPI.
[139] Auret’s version in his witness statement is consistent with this. He too does not
mention what the Gazelle representatives said at this meeting. Indeed he observes that:
“no agreement was reached with Gazelle to discontinue supply for these contracts
and no subsequent discussions with Gazelle involving me were held regarding PVC.”74
74 Auret’s witness statement, paragraphs 72-73.
34
35
[140] He did however indicate in his oral testimony what the purpose of the arrangement was
from the DPI perspective. A toll arrangement would be beneficial to both parties and by
doing that DPI would be able to maintain prices in the area, i.e. the toll arrangement
functioned as a form of market division to keep up prices. However, he goes on in his
testimony to concede that the discussions “...died a natural death after that”.75
[141] Warmback’s version of the discussion that took place at the Cape Town meeting is
consistent with either version of the Isando meeting - his or the Commission’s. If Isando
had been limited, as far as he was concerned, to tolling, then a further discussion on
this topic to flesh out the detail of the arrangement would not be surprising. It is
surprising that the discussion was limited to finding out the source of Peakstar’s pipes,
but Warmback says he was surprised by this turn. On the other hand, if the Isando
meeting had included setting prices for HDPE, Auret would have felt more confident to
have this type of conversation with Warmback whom he had only met on one prior
occasion. The Cape Town meeting thus takes the Commission’s case against Gazelle no
further, as it neither serves to establish an inference of a prior agreement at Isando nor
to refute one. Indeed the Commission did not rely on the Cape Town meeting in
argument as proof of its case.
[142] But attendance at these meetings is not the only evidence that the Commission led that
Gazelle was party to an agreement on pricing. Perhaps the most important was Carver’s
testimony that for a period he supplied a matrix of prices that had been agreed upon by
DPI, Marley and Petzetakis, at prior meetings between themselves, to Gazelle. He says
once agreements had been reached on pricing he would act as middleman and convey
this agreement to Gazelle. He would meet at Gazelle premises or in a pub. As he did this
this agreement to Gazelle. He would meet at Gazelle premises or in a pub. As he did this
on several occasions (he does not tell us how many) he assumed that Warmback, by
receiving the matrices, assented to them.
[143] Warmback vehemently denies this ever happened. Whilst admitting to meeting Carver
on many occasions and having a social relationship with him, he limits discussions to
supply arrangements when Marley supplied Gazelle on large contracts. Gazelle
75 Transcript, page 275.
35
36
subpoenaed Carver to produce an example of the matrix that he had given to Gazelle.
He did so, but there is nothing ex facie the document that would link it to Gazelle.
[144] We are thus faced with the dilemma as to whose evidence to accept. It is difficult to
reject either version on credibility alone. Both witnesses were unreliable , but in
different respects. Carver’s recollection of events is poor, whilst Warmback’s answers
were glib and in some instances, explanations for difficult aspects of his testimony, such
as how he came to attend the meeting in Isando, were unsatisfactory. We also find that
he did not tell the truth about his meeting with Harding.
[145] For this reason we need to examine the secondary evidence- to see whose version it is
more consistent with. The Commission did not call any witnesses who attended the
meetings at which the prices which Carver refers to were set. The closest we get to
corroboration is the evidence of Harding that Gazelle and Carver were involved in an
exchange of prices. Yet Harding’s version contradicts that of Carver. On Harding’s
version Carver and Gazelle were to agree prices and come back to the others with
them. This she said never happened. 76 On Carver’s version the other firms agreed the
prices in meetings that excluded Gazelle and he then took these prices to Gazelle, who
seemingly accepted them without demur. He cannot state whether Gazelle in fact
implemented these prices.
[146] But the versions are inconsistent. On Carver’s version the other firms reach an
agreement which is then taken by him to Gazelle who seemingly agrees by virtue of
their continuous interest in receiving the matrix. On Harding’s version there is no group
agreement, but an expectation that Gazelle and Carver will agree on something to take
back to the others. This version seems the least likely given that Gazelle was the
newcomer and the smallest of the firms. As counsel for Gazelle put it, the tail would be
wagging the dog.
[147] Whilst Carver’s version is the more plausible , the lack of corroboration is noteworthy.
No other witness can testify to this arrangement nor does he recall any specific reaction
by Gazelle to the pricing, nor whether anyone monitored its adherence. This is
surprising given the suspicion about Gazelle and the threat they had constituted.
76 Transcript, page 203.
36
37
[148] But what is most unsatisfactory about this version is the inability to place the matrix
exchange arrangements in time. By this we don’t mean an exact date or period , rather,
it was important to place this event against the sequence of other events. In a case
which depends on inference this is vital. What happens before explains what happens
after. When we don’t know whether an event happens before or after another of
importance and we are testing for evidence of the existence of an understanding a
legitimate inference cannot be drawn if we do not know whether event A preceded
event B.
[149] In this case we have to know if the matrix exchange had taken place before or after the
Isando meeting. The reason this is important is that it would explain the logic of what
took place at meetings. If Warmback was already receiving the matrices before the
Isando meeting, this would make him a cartel member prior to the meeting and hence
the meeting might have been there to firm up an existing agreement. Carver’s evidence
seems to suggest this. But if he was, why was it necessary to give him the tolling pretext
to attend the meeting.
[150] If he was not , then the tolling pretext makes sense. But then the meeting might have
been there to ascertain his firm’s willingness to join the cartel and then if he had
signalled some acceptance the matrix exchanges might have followed thereafter.
Neither Carver nor Auret testify to some act of acceptance by Warmback at the
meeting. Indeed neither are clear on what he said. If he had accepted at the meeting
and the matrices came later, why was this plan necessary? Why was Warmback, if he
had accepted to collude on pricing at Isando, not included in later meetings, or was he
not trusted to be party to other discussions?
[151] The descriptions of the Isando meeting from Auret and Carver are so vague that they
serve either possibility , but it had to be one or the other. In serving both they serve
neither.
neither.
[152] Other questions remain. Why is there no one from Petzetakis or DPI to testify that the
matrix arrangement constituted the understanding between Gazelle, themselves and
37
38
Marley on pricing? Both firms to a greater or lesser extent have admitted liability – why
do they fall short on this aspect?
[153] Given that the Commission must prove its case on a balance of probabilities we cannot
find that the onus has been discharged.
[154] We have evidence of Gazelle nibbling at collusion on several occasions, but never taking
the necessary bite that would establish liability. Gazelle clearly had no scruples about
entering into a collusive arrangement, but perhaps none of the other firms trusted it
sufficiently to let it join the club. The Commission’s burden is to prove the existence of
an agreement by way of an understanding between Gazelle and the other firms. We
have shown that there is insufficient evidence of an understanding. Whilst we have
difficulties in accepting aspects of Warmback’s testimony, it is not sufficient to draw
inferences from this alone that Gazelle is liable.
[155] It might have been open to us to draw an adverse inference about his testimony had
the Commission’s case, taken on its own, established liability. But it does not. The
Commission’s witnesses either do not give consistent versions of the same events or do
not provide testimony that proves an agreement between the other cartel members
and Gazelle. Thus, on the Commission’s case we cannot establish proof of an agreement
with the other firms by Gazelle to set prices for HDPE and hence we do not need to
consider whether the defence offered is credible. The Commission has not established a
prima facie case of contravention on a balance of probabilities against Gazelle.
[156] The case against Gazelle must be dismissed.
Amitech
[157] Amitech South Africa (Pty) Ltd is the parent company of two associates companies APS
South Africa (GP) Pty Ltd (“APS Gauteng”) and APS South Africa (EC) Pty Ltd (“APS
Eastern Cape or APS EC”).
[158] Amitech admits that its two associates participated in the cartel until May 2007.77
[158] Amitech admits that its two associates participated in the cartel until May 2007.77
77 Amitech’s Answering Affidavit, paragraph 5. See as well paragraph 41.1 where it admits that meetings which
it attended with other firms from 2004 to 2007 “... constitute breaches of the Act.”
38
39
[159] Amitech is the South African subsidiary of an international company called the Amiantit
Group which has subsidiaries in several countries internationally.
[160] Amitech commenced production in South Africa in 2001 , manufacturing PVC pipes. It
produced HDPE for a short period from 2002-2005, after which this equipment was sold
and it became a distributor, rather than a manufacturer, of HDPE.
[161] The third product it has produced , and around which there is some controversy in this
case, is GRP. GRP stands for Glass Reinforced Polyester pipes, a product that is
sometimes a substitute for PVC. Amitech maintains that is the only firm that locally
produces GRP.
[162] Amitech restructured its businesses in 2004 , creating three associate companies which
were majority black owned and which would become responsible for sales. (Note that
these became distribution companies, not manufacturing companies, a function which
remained with Amitech.) These are the two associates referred to earlier in Gauteng
and the Eastern Cape and another in Mpumulanga.78 The activities of the latter firm are
not of relevance to this case.
[163] Although Amitech describes these firms as associates it consolidates their financial
results into its group financial statements.79
[164] Unlike the case against some of the other respondents considered earlier, the case
against Amitech does not turn on consideration of what occurred at a single or a limited
number of meetings. Amitech and its associates all attended a number of meetings over
the period May or June 2004 to October 2007 that we have evidence of. There is no
evidence of them dissenting or being surprised by what took place at meetings. Rather
the material issues that divide Amitech and the Commission concern: 1) whether the
cartel concerned the GRP products, 2) whether agreements reached at meetings were
in fact implemented by Amitech and its associates, and 3) the strength of Amitech’s
in fact implemented by Amitech and its associates, and 3) the strength of Amitech’s
case on mitigation, in particular its emphasis on its current financial circumstances and
level of co-operation with the Commission.
78 Amitech owns a 49% share in the Gauteng business and a 35% share in the Eastern Cape one. Answering
affidavit, paragraph 23.79 Answering affidavit, paragraph 24.
39
40
[165] The first meeting that Amitech attended, that we have evidence of, took place in May
or June 2004 at Irene. It was attended by its erstwhile managing director Wally Van
Coller and Ashin Tasdhary, the managing director of APS Gauteng. Recall that the latter
had only been established that same year. At this meeting, attended by representatives
of Marley, DPI and Petzetakis, prices of HDPE and PVC were discussed and agreed.
Amitech states that at the time neither it nor APS Gauteng manufactured HDPE and
hence it was not a party to any agreement in respect of the latter. It does, however,
concede to being a party to an agreement concerning pricing for PVC products. The
essence of this agreement was that DPI’s price list for PVC was to be used as a
benchmark; PVC pricing would be increased by a certain percentage, discounts on the
price list would be reduced from 55% to 45% at the tender phase, with a maximum 5%
discount given to contractors at the award phase.80
[166] Despite this agreement , Amitech maintained that it never implemented the price
increases as suggested. The reason, according to its managing director Hein Momberg,
was that the parent was unable to get the distributors to agree to the agreed prices
with customers. Further the transfer price – the price at which the parent sold to its
distributor associates - was not during the time of Van Coller, Momberg’s predecessor,
rigid. It was only when Momberg took over that this tightened up. Momberg suggests
that despite the PVC agreement in 2004 other firms also cheated and set their own
prices and hence Amitech’s associates had to do likewise.81
[167] Van Coller is the person who brought Amitech into contact with the cartel , but he was
not called as a witness, so this earlier history is missing from the record. Momberg
admitted to attending a meeting in September 2006 at a restaurant attended by
admitted to attending a meeting in September 2006 at a restaurant attended by
representatives of the same firms, viz. DPI, Marley and Petzetakis. From Amitech’s
perspective the relevant agreement struck (or re-confirmed it is not clear which) related
to discounts on PVC at tender phase. To this end, and pursuant to the agreement,
Momberg sent an email to his staff and associates telling them that the selling price to
80 Firms in the industry would quote twice on a tender. If the contractor to whom they had first quoted won the
tender it would typically come back to the suppliers and demand a further discount. Collusion had to be agreed
to provide for both stages.81 Momberg mentions that there was one incident that Tasdhary has told him about where the latter was given a
specific instruction to keep to a specified discount structure. He does not know whether this discount instruction
was made pursuant to a cartel arrangement, but states that in any event, Tasdhary did not adhere to it.
(Answering affidavit, paragraph 36.8)
40
41
the market was list less 55% and that from 1 November that year it would be list less
45%. In an email he conveyed the pricing agreement reached at the September
meeting with the other manufacturers. As if to anticipate non-adherence by the
distributors Momberg wrote: “All orders will be checked and orders will not be accepted
with lower prices unless accepted by me.” 82 This email is important as we discuss later
in dealing with the degree of blameworthiness that attaches to Amitech.
[168] Momberg states that this instruction was ignored in Mpumulanga and that Gauteng and
Eastern Cape told him that competition in their respective areas prevented them from
implementing the agreed pricing structure.
[169] Another email from Momberg demonstrates a further attempt by the parent to get
pricing adherence from the associates. This email , which attaches what are termed the
new pricing guidelines (which came about as a result of a conversation between
Momberg and DPI), tells associates not to sell outside the guidelines without speaking
to Momberg.83
[170] Momberg admitted to attending a further meeting in May 2007 where the same firms
were in attendance. Despite discussions to change the pricing model, the parties agreed
to adhere to the status quo.84
[171] The final meeting that Momberg attend ed was on 11 October 2007. At this meeting,
which we consider in greater detail when we get to the case in respect of Petzetakis,
Harding announced that Petzetakis would no longer participate in meetings. Momberg’s
understanding was that each firm would then use its own price list going forward. He
attended no further such meetings.
[172] Momberg sought to minimise the extent of his firm’s involvement by suggesting that it
had practically abandoned the cartel price list in December 2006 and that it then
produced its own price list in September 2007. However, despite this, Amitech
produced its own price list in September 2007. However, despite this, Amitech
continued attending meetings until October 2007. Further, it is difficult to see how it
82 Annexure AM 9 to answering affidavit. The email is dated 11 October 2006.83 Annexure AM 10 to answering affidavit. The email is dated 13 December 2006.84 Answering affidavit, paragraph 38.2.
41
42
could have practically abandoned the cartel price list on 13 December 2006 when that
was the date on which Momberg was urging adherence to it.85
[173] Whatever its failure to implement cartel pricing, Amitech continued to hold out to other
firms in the cartel that it was adhering to agreements by its attendance at meetings
called until October 2007. It was Harding’s actions (which we get to discuss later) and
not conduct by Amitech that led to the cartels demise.
[174] Apart from the meetings referred to above attended by Amitech, APS Gauteng
representatives attended further meetings with the other firms , some of which also
included Swan. It is not necessary to go into detail on any of the meetings. The themes
are similar – the meetings got adherence to agreed prices, but APS Gauteng maintained
it never followed them and that all parties cheated. One variation on theme was a
meeting in September 2006 to allocate tenders. Amitech claims that because APS
Gauteng did not bring along its contract register it was excluded from the allocation.
[175] At no stage however d id APS Gauteng signal to any of the other firms at the meetings
its non-adherence to the agreements and it continued to attend meetings called with
one possible exception.86
[176] We now turn to APS EC. APS EC started attending meetings in November 2006 and the
last which it admits to attending was on 3 July 2007. APS EC admitted to attending nine
meetings during this period, which were variously attended by representatives of its
competitors. It admitted that of these nine, five dealt with price fixing and the principle
of how to allocate contracts. The other firms in attendance were DPI, Petzetakis, Marley
and Swan. Four meetings dealt with the actual allocation of contracts (as opposed to
just the principles of the collusive arrangement) and were attended by APS, Petzetakis
and DPI. Marley and Swan did not attend these meetings.
and DPI. Marley and Swan did not attend these meetings.
[177] APS EC accepts that its conduct in attending these meetings constitutes a breach of the
Act. It is therefore not necessary to go into each one in specific detail given this
admission.87
85 See email annexure AM 10 supra footnote 83.86 A meeting in October 2006, which Tashary and Gafoor claim to have known about but not attended. Yet not
much turns on the non-attendance as Amitech continues to attend meetings in 2007 as we have seen.87 See answering affidavit, paragraphs 57.1 and 57.8.
42
43
[178] As with its Gauteng counterpart , APS EC argues that these agreements were honoured
in the breach with frequent cheating taking place amongst all the firms. It does concede
that it attempted to implement the discount scheme, but that others cheated and
hence this was of no affect. It also contends that customer allocations were not
followed and that by August 2007 agreements were disregarded.88
[179] This version presented on the papers by Amitech and supported by its associates, has
assumed a rote form in respect of every meeting attended by the three entities during
the period. In effect Amitech wants the Tribunal to accept that all agreements reached
failed to produce their intended collusive outcome. Yet despite this the firms continued
to meet and conspire for a period of almost three- and-a- half years.
[180] Not surprisingly, when it came to oral evidence, the version proved less resilient.
Ragaval did not fare well when cross-examined in respect of APS EC and made a
number of concessions to the effect that the arrangements proved more beneficial than
the answering affidavit of Momberg would suggest. By way of example, when
challenged, Ragaval conceded that the cartel arrangements had improved prices in the
Eastern Cape and that the contract allocations were respected.89
[181] Indeed Ragaval admitted that it was beneficial to collude. 90 In respect of APS Gauteng a
similar admission, that it was beneficial to collude, was made by Tasdhary. 91 Gauteng
APS’s reasons for not submitting their register of contracts arose not out of a desire to
compete, but a failure to satisfy other cartel members that it knew what it was doing.92
[182] Amitech does reveal one tender arrangement – in the interests of transparency as it
puts it - although it denies this was unlawful. Amitech’s version, is that it did not
implement this arrangement.93
implement this arrangement.93
[183] Finally, before considering the remedy, we must deal with the issue of GRP. Although
the oral testimony shows that there was a probable link between GRP prices and prices
88 See answering affidavit, paragraph 57.2.89 Transcript, pages 1217-1218.90 Transcript, pages 1223 and 1230.91 Transcript, page 1276.92 Transcript, page 1278. 93 Amitech’s Answering Affidavit, paragraph 36.5.
43
44
of PVC and some types of HDPE pipe, since certain contracts for PVC described GRP as
an alternative, 94 the issue is whether we can make a finding against Amitech on the
basis that this allegation had not been made out fairly against it.
[184] It is common cause that the collusion in respect of GRP was not alleged in the
complaint referral, nor was it mentioned in the Commission’s witness statements, but
arose for the first time during oral testimony. Amitech objected to the late introduction
of this evidence against it. Whilst the cases relied on as authority for this objection are
no longer good law since the Constitutional Court decision in Senwes, that decision did
not rule out the Tribunal’s discretion around issues of fairness. 95 In our view this late
allegation, that was neither contained in the complaint referral or in the witness
statements or any other document prior to the hearing, did catch Amitech unawares
and it was entitled to object. For this reason only, we have not included GRP revenue in
our calculation of relevant turnover set out below.
Appropriate Remedy for Amitech
[185] Amitech was not a major player in the cartel and had a market share less than the three
larger firms. We don’t know how it came to join the cartel as Van Coller is no longer
employed by the firm. We can accept that it did not initiate the cartel. It is not clear if
Amitech implemented pricing agreements reached by the cartel- it certainly denies this-
but its witnesses conceded in testimony that even if it did not adhere to the agreed
pricing it benefitted from the arrangements. There is also evidence that it benefited
from customer allocation. Even if all firms in the market deviated from the terms of the
agreement they had the effect of raising prices from what they would have been absent
collusion. Thus, while the firm did not benefit excessively from the collusion it still
collusion. Thus, while the firm did not benefit excessively from the collusion it still
benefited and consumers suffered harm. As important is the fact that the other firms,
which included the three major firms in the industry, accepted that Amitech was part of
the cartel and could set prices based on this understanding. Amitech, by signalling its
acceptance by attendance at numerous meetings, both nationally and regionally,
helped the cartel implement its collusive design.
94 The Jaggersfontein project and certain of the Eastern Cape contracts.95 Competition Commission vs Senwes Limited Case CCT 61/11 [2012] ZACC6.
44
45
[186] Market circumstances were initially difficult for the firm and there was evidence of this-
particularly in the Eastern Cape. At the time that APS EC entered the market a price war
had broken out amongst producers, leading to the temporary closure of the Petzetakis
East London office. Since APS EC had been formed by three former DPI employees it
struggled in the beginning to get a foothold in the market.96
[187] Amitech’s main argument is one of hardship. It states that it has now exited the PVC
market that is has sold its interest in that part of the business. In the 2010 financial year
the group had accumulated losses of eighty eight million rand (R88 million) and its
liabilities exceeded its assets by thirty three million rand (R33 million.)
[188] Although represented by counsel and attorneys during the hearing of evidence the
attorneys withdrew on record prior to final oral argument and thus Amitech was not
represented at the Tribunal for final argument. The attorneys, in lieu of heads of
argument, prepared submissions on Amitech’s behalf. These serve perfectly adequately
as written heads of argument and thus Amitech has been disadvantaged only to the
extent that no-one appeared on its behalf to present oral argument at the hearing.
Although Mr Momberg was invited by the Tribunal to attend the hearing of final
argument he did not take up the invitation.
[189] We have received no subsequent communication from Amitech as to its fortunes. The
Commission argues that Amitech is owned by a large foreign group , the Saudi Arabian
Amiantit Company. The shareholder is capable, it argues, of financing the company
through its present difficulties. The last audited financial statements state that:
“The ability of the group to continue as a going concern is dependent on a number of
factors. The most significant of these is that the directors continue to procure
funding for ongoing operations of the group.” 97
funding for ongoing operations of the group.” 97
[190] But the same financial statements als o show that the parent company has extended a
loan to Amitech that is interest free. There is no reason if the parent has the ability to
fund the subsidiary that it will not continue to do so in the future, unless it chooses not
to for commercial reasons, extraneous to the levelling of the penalty. Expressed
96 Answering affidavit, paragraphs 48.3 -48.5.97 Amitech Annual Financial Statements for year ending 31 December 2010, page 51.
45
46
differently, if Amitech does exit the market it will not be consequent on the levelling of
a penalty, but because of other commercial reasons.
[191] The onus is on the firm to establish that the levelling of a penalty would result in its exit.
Beyond raising its present financial difficulties in 2010 it has not lead any further
evidence on this point. In United States v Nathan it was held that in evaluating the
inability-to-pay question, “[t]he sentencing court must take account of the corporate
defendant’s financial resources, putting the burden on the defendant to produce
relevant materials.”98
[192] We do not consider that this failing firm factor has been established properly and it
should not detract from the levelling of an appropriate penalty.
[193] We now consider the mitigating and aggravating factors as they affect Amitech. As an
aggravating factor we note that senior management was involved in the cartel
activities. First, Van Coller and then Momberg, who were respectively the managing
directors of Amitech. Momberg wrote two emails to distributors, including APS Gauteng
and APS EC, both stressing that there had to be compliance with the pricing
agreements.99 Thus, adherence to agreements was being urged upon the others from
the most senior person in the group. If adherence was honoured in the breach, as
Amitech now alleges, this was because others lower down in the company acted against
Momberg’s stipulations.
[194] Nevertheless we accept that Amitech was not a leader or initiator of the cartel, that its
managers and staff played minor roles in the cartel and that it has not been established
that it implemented the cartel agreements save in relation to certain customer
allocation agreements.
[195] We accept, as we did with MacNeil, that Amitech has not made excessive profits from
the collusion. Loss or damage caused by the collusion is not capable of quantification
the collusion. Loss or damage caused by the collusion is not capable of quantification
and therefore we do not consider this as an aggravating factor.
98 188 F. 3d 190, 213 (3d Cir. 1999).
99 Annexures AM9 and AM10 to the Amitech answering affidavit, pages 293-294 of Pleadings.
46
47
[196] Amitech also makes much of the fact that it co-operated with the Commission during
the investigation and admitted its involvement. This kind of co-operation is no less than
one would expect of a firm confronted with the facts of the case against them. This co-
operation is not such that it is worthy of consideration as a mitigating circumstance.
However, in Amitech’s favour is that it provided the Commission with evidence it
appears it did not have which would have assisted it in proving its case against other
firms.100 Where firms provide evidence, particularly documentary evidence, which
assists enforcement against other firms involved in the cartel this should be recognised
as an important mitigating factor. We have for this reason significantly discounted the
penalty which would have otherwise been imposed on Amitech.
[197] Finally we accept that Amitech has not previously been found to have been in
contravention of the Act.
[198] We now proceed with calculating the penalty amount.
[199] The affected turnover for Amitech should be the year ending 2007. If we exclude GRP
and confine this to PVC products produced in Gauteng and the Eastern Cape the
affected turnover would be in the order of eighty million three hundred thousand rand
(R 80,3 million), made up as follows:
[199.1] Affected turnover from PVC, APS Gauteng in 2007 was fifty two million
nine hundred thousand rand (R 52.9 million). 101
[199.2] Affected turnover from PVC, APS Eastern Cape in 2007 was ten million
two hundred thousand rand (R 10.2 million).102
[199.3] Affected turnover for PVC, Amitech in 2007 was seventeen million two
hundred thousand rand (R 17.2 million). 103
Total R 80, 3 million
100 Amitech provided inter alia minutes of meetings in the Eastern Cape.101 Amitech’s answering affidavit, paragraph 47.11.102 Ibid paragraph 57.5.103 Ibid paragraph 82.
47
48
[200] Amitech was involved in a national cartel and two regional cartels , but not in respect of
all products, its involvement we find was limited to PVC. As with MacNeil, an
appropriate percentage for step one would be 15%, as in some regions, such as
Mpumulanga, where it operated it did not participate in the cartel and it was not
involved in all products to which others in the cartel were.
[201] The evidence shows that Amitech was involved in the national cartel from June 2004 to
October 2007, a period of three years. In Gauteng it was involved for the same period.
In the Eastern Cape from late 2006 to August 2007, thus approximately one year. 104
Multiplying the national and Gauteng turnovers by three years and the Eastern Cape
taken by one year, and multiplying that by 15%, amounts to thirty three million and
seventy thousand rand (R 33.07 million). 105 Since this amount exceeds the 10% of 2010
turnover, of one hundred and eighty five million (R 185 million), the rounded off figure
is eighteen million five hundred thousand rand (R 18,5 million).106
[202] We would give Amitech a 40% discount for mitigation. This leaves the amount of
eleven million one hundred thousand rand (R 11.1 million). 107 This figure is 6% of total
turnover in 2010 and thus does not exceed the 10% limit set out in section 59(2) of the
Act.
[203] We accordingly impose a penalty of eleven million one hundred thousand rand (R 11.1
million.)
Petzetakis
[204] Petzetakis, like Amitech, does not contest liability, but only the amount of the penalty it
is required to pay.
[205] The dispute between Petzetakis and the Commission revolves around two issues : firstly,
which products should be included in calculation of its affected turnover and secondly,
104 Answering affidavit, paragraph 48.8.105 (52.9 + 17.2) (x3) + 10.2 = 220.5 x 15% = 33.07 106 Total sales in the 2010 financial statements rounded off amount to R 185 million. See page 9 of the 2010
annual financial statements.107 Note gross profit in 2010 was R 20, 4 million although net profit was R36, 4 million. See Amitech’s financial
statement for the year ended 31 December 2010, page 9.
48
49
whether the actions of its erstwhile chief executive should factor in as mitigation of any
penalty.
[206] The first dispute of fact is fairly trivial and not much turns on it for the purpose of the
penalty calculation. The second is not and much does turn on it.
[207] It is not necessary to go into any detail around Petzetakis role in the cartel. It is common
cause that Petzetakis was involved in the cartel for years , certainly even before the Act
came into effect. Secondly, that it was a national player and probably the second, and
at some times the most significant, player in this market.
[208] It manufactured both HDPE and PVC pipes. Petzetakis also manufactured another
product called Weholite. Whilst the Commission contends that this product formed part
of the cartel arrangement, Harding, who was called by the Commission as a witness (we
explain why later) denied this and limited the collusion to HDPE and PVC. Without other
contradictory evidence on this aspect we will accept her evidence that this was so.108
[209] Petzetakis is owned by a parent company that operates internationally and is based in
Greece. It entered the local market in 2000 when it acquired Main Industries, from
Murray and Roberts. It is clear that the cartel operated for many years, but due to
mergers in the industry the nature of the players varied. 109 We are concerned for the
purpose of this case with when Petzetakis became involved with the cartel
arrangements that form part of this case. Dating its entry is uncertain. Petzetakis admits
to being involved at least from the time that it entered the local market as a new
owner. This means that it would have been involved from 2001 until May 2007. 110 There
is some debate between Petzetakis and the Commission as to whether liability should
extend beyond the six years to the involvement of the firm under its previous
shareholder.
shareholder.
[210] Given how we approach the penalty for Petzetakis , this debate need not be resolved by
deciding the point of law. We will accept, given that we have little evidence of cartel
108 Transcript, page 119.109 Both Petzetakis and DPI have been the subject and objects of various mergers. See Auret’s witness statement
and Harding’s testimony, transcript page 42.110 In the heads of argument counsel for Petzetakis suggests a date of April 2001. In the answering affidavit for
Petzetakis this date is mentioned as 2001. See affidavit, paragraph 18.
49
50
activities involving its predecessor that Petzetakis’ period of liability commences in April
2001, as conceded by it.
[211] Petzetakis was clearly a key player in all regions of the country and one or more of its
employees attended most meetings of the cartel. Names mentioned in evidence were:
Michelle Harding and her predecessor Tony Dean, Peter Kerr Fox, Cindy Dixon in KZN
and Colin Murphy in the Eastern Cape.
[212] Along with DPI and Marley it constituted part of the glue that held this cartel together
despite its frequent ups and downs.
[213] For this reason one would have expected that an administrative penalty would be set at
a level that would be appropriate to reflect the seriousness of Petzetakis conduct.
[214] We have decided to impose a much lesser penalty than we ordinarily would have and
we explain why.
[215] Petzetakis can count themselves fortunate to have hired Michelle Harding and then
promoted her to managing director. They have since fired her , but that is an issue
between her and her erstwhile employer.111
[216] The dismissal explains a curious turn in this case alluded to earlier. When answering
affidavits were filed, Harding was still working for Petzetakis as its managing director
and she deposed to its answering affidavit in that capacity and prepared a witness
statement for them. 112 By the time of the hearing she had been dismissed, but she
nevertheless testified- not for Petzetakis, but for the Commission under subpoena.
[217] Harding testified to the most remarkable set of events, which demonstrated that
redemption is possible even in the most unlikely environments - an industry riddled
with collusive practices, betrayal and deception.
111 Transcript, page 84. When asked the reason for her dismissal Harding stated she is still unsure, see
Transcript, pages 83-84. Harding stated that she filed a case with the Labour Court against Petzetakis for unfair
dismissal, see Transcript, page 145. 112 See paragraph 2 and paragraph 3 of Harding’s witness statement.
50
51
[218] Harding had risen through the ranks at Petzetakis’ predecessor, whom she joined as a
bookkeeper in 1987.113 She became sales and marketing director in 2002 and eventually
managing director in 2006. Harding was not exposed to cartel activity in her earlier
years with the company as a bookkeeper, but she was aware of them. When she was
promoted to sales and marketing director in 2002 the then sales director Tony Dean
explained to her how a cartel between the three major firms, DPI, Petzetakis and
Marley was operating.
[219] Harding had participated over the years in many meetings of the cartel , both at sales
director level and later attended meetings of the managing directors. She testified that
she became increasingly uncomfortable with the illicit conduct of the company. 114 After
attending a conference on business ethics called “Unashamedly Ethical” where the
speaker challenged those in the audience, many of whom were from the building
industry, to clean up their ethics, in particular referring to practices such as price fixing,
she decided to take up the challenge. She did this first by telling senior people in her
company that they needed to change the way they operated and specifically to price
independently and not meet with competitors. She then contacted her group chief
executive in Greece, at that time a Mr Panadakis, who was willing to endorse her
approach provided, as he apparently put it, it didn’t compromise “the bottom line.”115
[220] She then arranged what turned out to be the most important of the meetings referred
to in this case. She invited Wally Van Coller of DPI, Dave King of Marley and Hein
Momberg of Amitech to a breakfast meeting on 18 May 2007. She told them that from
then on Petzetakis and its subsidiaries would no longer participate in any collusive
activities. All contact with her firm and staff from the other firms had to cease.
activities. All contact with her firm and staff from the other firms had to cease.
[221] Within her own firm Harding did the same. She called a meeting a few days later of the
national sales staff and gave them the same message. Sales staff were warned not to
participate in any discussions or telephone calls with “the opposition” and were told
that disciplinary steps would be taken against them if they did. Indeed Harding testifies
113 The predecessor was a company called Mega Pipe, which was then acquired by Main Industries, who were
then acquired by Petzetakis.114 Petzetakis' answering affidavit, paragraph 24.115 Transcript, page 85.
51
52
that two staff members who did not follow this instruction were disciplined and given
warnings.
[222] Despite taking all these steps there was one step Harding did not take. She did not
approach the Commission for leniency as part of its corporate leniency program.
Unfortunately for Petzetakis the firm that got in first , and thus effectively secured for
itself leniency, was DPI.
[223] Harding testified that she was not aware of the existence of the CLP and thus had not
explored this option. After the complaint was initiated she met with the Commission
during the investigation and said she co-operated in giving information that its
investigators were not aware of. Sometime after the Commission referred this case she
was dismissed.
[224] What Harding did required courage and integrity. In an industry dominated by illicit
practices for decades , she had the moral fibre to face up to all of them and say this
must stop. That most in this male dominated industry were very much her senior, both
in years and industry experience, makes it all the more remarkable.
[225] But what we have to decide is whether Harding’s remarkable conduct should count as
mitigation for Petzetakis and if so, to what extent.
[226] Two factors complicate matter s for Petzetakis. In the first place the evidence was that
despite Harding’s intervention, some members of her staff continued attending
meetings, albeit behind her back. The culture of an industry riddled with bad practices
does not change easily or dramatically.
[227] Secondly, it is clear from her testimony that her own company did not take ownership
of her actions. The parent company , it appears, was mean spirited about what she had
done and no doubt felt they were doubly done down by her: first having brought an end
to the collusion and thus damaging the ‘ bottom line’ and worse still, not getting
immunity from prosecution via the CLP, despite her ending the cartel, whilst rival DPI,
immunity from prosecution via the CLP, despite her ending the cartel, whilst rival DPI,
the cartel leader, got in first and has now escaped liability.
52
53
[228] Whilst the latter two aspects do not reflect well on Petzetakis this should not detract
from the significance of what Harding has done. What Harding did, and the evidence in
this case is common cause on this point, was to effectively destabilise the cartel. Whilst
it continued for some months after her dramatic meeting with the other managing
directors it eventually ground to a halt because she had made it too difficult for them to
continue operating in the way they did. Without a key player like Petzetakis as a party
to the conspiracy the cartel failed.
[229] Bringing cartel activity to an end should be recognised as the most important of
mitigation factors. 116 Because cartels only work if there is an understanding between
players who account for an overwhelming share of the market, the exit from the
consensus of a major player destabilises the cartel. However, what is important is how
Harding did it. She did not stealthily withdraw or cheat, so that the remaining firms
were still under the impression that Petzetakis was part of the club. Rather, she
announced to the most senior people at the three other major players in the market, at
a meeting which they jointly attended, that her firm was no longer part of their
collusive arrangement. This unambiguous step meant that they knew they could no
longer rely on Petzetakis to co-operate in the market and worse still that covert
meetings might now become known to others.
[230] Enforcement against cartels requires more Hardings who are willing to take a moral
stand and, as she put it in her testimony, ‘stop the cancer’. The leniency policy is
premised on the logic that cartel activity is best exposed and hopefully terminated if
firms are incentivised by the benefit of immunity if they come forward to the
Commission and confess. The approach to penalties must also ensure that those same
Commission and confess. The approach to penalties must also ensure that those same
incentives are given to firms, which, whilst not meeting the requirements for leniency,
behave in a manner with effects similar to those of the leniency applicant. Harding’s
actions in declaring her firm’s non-adherence with the collusive arrangements then in
existence, did more on the facts of this case to end the collusion than did the later
actions of the leniency applicant.
116 In terms of section 59(3)(c), one of the factors we consider is the behaviour of the respondent.
53
54
[231] It is important to signal to the business community that the public disavowal of cartel
arrangements will not go unrecognised when it comes to the imposition of a sanction.
Whilst such conduct does not lead to the firm escaping liability, it will , if honestly and
effectively executed, qualify for recognition as a significant mitigating factor.
[232] For this reason the administrative penalty that might otherwise have been levied on
Petzetakis is reduced by 80%. It would not be appropriate, however, not to levy a
penalty on Petzetakis. It benefited from the cartel for many years and was integral to its
success nationally.
[233] We set out this calculation below as follows:
[233.1] Affected turnover in year 2007 for PVC and HDPE was six hundred and
thirty eight million, one hundred and thirty eight thousand, three hundred
and seven rand (R 638 138 307.00).
[233.2] 15% of this figure for the purpose of step 1 is in the amount of ninety
five million, seven hundred thousand rand ( R95,7 million).
[233.3] Multiplied by the six years of participation we get: 95,7 x 6 = 574.2
[233.4] Rounded off because this figure exceeds 10% of the total turnover for
2010, which was four hundred and ninety six million, seven hundred and
twenty two thousand and eighty eight rand (R 496 722 088.00), we get
forty nine million six hundred thousand rand (R49,6 million).117
[233.5] Forty nine million six hundred thousand rand ( R 49 600 000.00) reduced
by 80 % for net mitigation, leads to a figure of nine million, nine hundred
and ninety two thousand rand (R 9.92 million).
[234] We accordingly impose a penalty of nine million, nine hundred and ninety two
thousand rand (R 9.92 million).
Conclusion
117 In 2010 the affected turnover was R262 999 322.
54
55
[235] The Commission’s case against Gazelle falls to be dismissed for the reasons discussed.
The Commission’s case against MacNeil succeeds and it is liable for contravening
section 4(1)(b)(i) of the Act and a penalty of R 2 million is imposed. The case against
Andrag succeeds, and it is liable for contravening section 4(1)(b)(i), but no penalty is
imposed on it. The case against Amitech, in which liability for contravening sections 4(1)
(b)(i), 4(1)(b)(ii) and 4(1)(b)(iii) of the Act was admitted, has led to the imposition of an
administrative penalty of R 11.1 million. The case against Petzetakis in which liability
for contravening sections 4(1)(b)(i), 4(1)(b)(ii) and 4(1)(b)(iii) of the Act was admitted,
has led to the imposition of an administrative penalty of R 9.92 million.
[236] Because, unfortunately, there has been a delay in completing these reasons, through no
fault of any of the parties, we have decided to give an extended period of time to those
firms liable for penalties to make payment.
ORDER
1. The case against Gazelle (the ninth and tenth respondents) is dismissed;
2. MacNeil is found to have contravened section 4(1)(b)(i) of the Act for a period of one
year from to February 2007 to October 2007;
3. MacNeil is accordingly ordered to pay an administrative penalty of two million rand
(R 2 000 000.00). Of this amount, one million rand (R 1 000 000.00) is to be paid
within 90 business days of the date of this order, and the balance within one year of
the date of the first payment, but no later than one year and 90 days after the date
of this order.
4. Andrag is found to have contravened section 4(1)(b)(i) of the Act, but no penalty is
imposed on Andrag.
5. Amitech is found to have contravened sections 4(1)(b)(i), 4(1)(b)(ii) and 4(1)(b)(iii) of
the Act, for a period of two years from May 2004 to October 2007;
6. Amitech is accordingly ordered to pay an administrative penalty of eleven million,
6. Amitech is accordingly ordered to pay an administrative penalty of eleven million,
one hundred thousand rand (R 11 100 000.00). Of this amount, five million, five
55
56
hundred thousand rand (R 5 500 000.00) is to be paid within 90 business days of the
date of this order, and the balance within one year of the date of the first payment,
but no later than one year and 90 days after the date of this order;
7. Petzetakis is found to have contravened sections 4(1)(b)(i) and 4(1)(b)(ii) of the Act
for a period spanning from 2001 to October 2007;
8. Petzetakis is accordingly ordered to pay an administrative penalty of nine million,
nine hundred and ninety two thousand rand (R 9 920 000.00). Of this amount, four
million, four hundred and forty six thousand (R 4 460 000.00) is to be paid within 90
business days of the date of this order, and the balance within one year of the date
of the first payment, but no later than one year and 90 days after the date of this
order;
9. There is no order as to costs.
_________________________________ 4 July 2012
Takalani Madima and Norman Manoim
Yasmin Carrim concurring
Tribunal Researcher: Londiwe Senona
For the Applicant: Adv H Maenetje instructed by Gildenhuys
Lessing Malatji
For the First Respondent : Adv J Wilson instructed by Nortons Inc.
For the Second Respondent : Adv MJ Engelbrecht instructed by Bowman
Gilfillan
For the Third Respondent : Edward Nathan Sonnenbergs
For the Fourth Respondent : Garlicke & Bousfield
For the Fifth Respondent : Adv M Van Der Nest SC instructed by
Werskmans
For the Sixth Respondent : Fullard Mayer Morrison Inc.
56
57
For the Seventh Respondent : Adv S.E Rosenberg SC instructed by Smith
Tabata Buchanan Boyes
For the Eighth Respondent : Adv J. C Butler SC instructed Werksmans
For the Ninth and Tenth
Respondents : Adv P McNally SC instructed by Deneys Reitz
57