Competition Commission of South Africa v Uniplate Group (Pty) Ltd (CR188Nov15) [2019] ZACT 61 (27 June 2019)

SOUTH

AFRICA

COMPETITION

TRIBUNAL OF SOUTH AFRICA

Case no: CR188Nov15

In the matter between:

THE

COMPETITION COMMISSION OF SOUTH AFRICA Complainant

UNIPLATE GROUP (PTY) LTD Respondent

Panel: Yasmin Carrim (Presiding Member)

: Enver Daniels (Tribunal Member)

: Mondo Mazwai (Tribunal Member)

Heard on: 08,09,10,13 and 14 November 2017; 04 to 06

December 2017 and 15 May 2018.

Order and Reasons issued on: 27 June 2019

REASONS

FOR DECISION AND ORDER

Introduction

1. In this case the Competition Tribunal ('Tribunal") was asked to determine whether Uniplate Group (Pty) Ltd ("Uniplate"),

allegedly a dominant firm in the manufacture and supply of embossing machines and number plate blanks, abused its dominance by

contractually obliging its customers, when purchasing an embossing machine to also purchase all of their number plate blanks from

Uniplate.

2. This contractual obligation was contained in exclusive supply agreements between Uniplate and its customers.

3. The Commission, after investigating the complaint, referred it to the Tribunal on 27 November 2015 in terms of section 50(1).

Relief Sought

4. The relief sought by the Commission in relation to the exclusive provisions was the following:

4.1 A declaratory order that Uniplate has contravened sections 8(d)(i), in the alternative 8(d)(iii), or in the further alternative 8(c) for the period 2010 up to and including 2014;

4.2 An order that the impugned provision (exclusivity) is void in terms of section 58(1)(a)(vi) of the Act;

4.3 An interdict in terms of section 58(1)(a)(i) to restrain the respondent from engaging in a contravention of section 8(d)(i), or in the alternative 8(d)(iii), or in the alternative 8(c) of the Act; and

4.4 An order for the payment of an administrative penalty in the sum of 10% of Uniplate's turnover in the Republic and its exports from the Republic during the preceding financial year.

5. However, during final argument, the Commission appeared to have abandoned the prayers for an interdict and the voiding of the exclusionary

provision.[1] The relief sought by the Commission is therefore confined to that of a declaratory order and the imposition of an administrative penalty.

Industry background: The value chain

6. The value chain in this industry is made up of (i) manufacturers of embossing machines; (ii) manufacturers of blanks; (iii) embossers

(firms that make the final number plate out of blanks), and (iv) end users (motorists and car dealerships).

Manufacturers of embossing machines

7. There are three types of embossing machines i.e. Acrylic, Type A, and Type B machines. Embossing machines are designed to only be compatible with the equivalent number plate blank. For example, an Acrylic embossing machine can only be used to make an Acrylic number plate. Similarly, a Type A embossing machine is only compatible with a Type A number plate blank.

8. Uniplate and New Number Plates Requisites CC ("NNPR") are the only two manufacturers of embossing machines and are the only players in the market who manufacture both embossing machines and number plate blanks.

9. Acrylic and Type A machines were in the market prior to 2007. In 2007 Uniplate innovated its machines to emboss Type B plate types in response to anticipated changes in regulations in Gauteng which were set to come into effect in 2010. Uniplate started selling these Type 8 embossing machines in 2008/2009.

10. While evidence suggests that a Type 8 (newer or recent) machine could, theoretically emboss Type A and Type B number plates, switching

between Type A and Type B embossing required significant adjustments, to the setting board and changes in dies[2] and involved additional costs.[3] In general substitution between Type A and Type B machines did not occur and machines were sold as such i.e. those which could emboss Type A or Type B blanks respectively.

11. Both Uniplate and NNPR manufacture all three types of embossing machines - Acrylic, Type A, and Type B. Until 2010, Arga manufactured Acrylic and Type A embossing machines but was no longer in the market as an embossing machine manufacturer at the time of the hearing.[4]

12. Uniplate supplies its embossing machines and number plate blanks directly to its customers throughout the country and is the largest manufacturer of blanks. In addition, Uniplate makes use of two third-party distributors, namely Teqplate (Pty) Ltd ("Teqplate") and Baleka Plates (Pty) Ltd ("Baleka") to exclusively on-sell its embossing machines and number plate blanks. NNPR distributes directly to its customers.

Manufacturers of number plate blanks

13. A number plate blank can be described as a number plate absent the numbers and letters that are used to identify the vehicle. Number plate blanks (or blanks for short) can be manufactured from two types of material i.e. Acrylic and aluminium.[5]

14. Type A and Type B number plate blanks are made of aluminium, and Acrylic blanks are made from plastic - all three types are sold in South Africa. Historically there had only been one type of aluminium blank produced in South Africa namely the Type A. However, with the change in regulations in Gauteng mentioned above, a new aluminium blank namely the Type B, was developed and introduced into the market by Uniplate.[6]

15. There are four manufacturers of blanks also referred to as "blankers" i.e. Uniplate, NNPR, Arga Plates & Signs (Pty) Ltd ("Arga") and Naicker Toolmaker and Metal Pressing ("Naicker Tools").[7] Arga manufactures only Type A and Acrylic blanks, while Naicker Tools only manufactures Type A blanks. Uniplate and NNPR manufacture all three types of blanks.

16. In order to manufacture blanks, every blanker is required to conform to certain regulations prescribed by the National Road Traffic Act 93 of 1996 ("the Traffic Act"). The Traffic Act provides that number plates must conform to SANS 1116 published by the South African Bureau of Standards ("SABS").[8] These standards make it an offence for vehicle owners to drive with number plates that do not conform to SANS 1116. In terms of these standards every blanker is issued with a permit to manufacture number plate blanks.[9]

17. Embossing machines and blanks are then sold by the manufacturers to embossers.

Embossers

18. Embossing refers to the process of imprinting or cutting out unique numbers and letters onto a number plate blank through the use of an embossing machine. Firms that emboss number plates are referred to as embossers.

19. The majority of embossers in South Africa are small businesses which typically do not have access to large capital resources. One example of such a firm is JJ Plates and Signs CC, one of the complainants in this matter ("JJ Plates").

20. It is estimated that there are approximately 1000 registered embossers in South Africa which compete with each other in various

regional and local markets.

End users

21. End users are members of the public who purchase the embossed number plate from embossers. This includes vehicle owners and car dealerships.

22. End users are free to choose which type of number plate to affix to their vehicles. However, in 2010, regulations required that Gauteng motorists use only aluminium number plates.

Procedural Background

23. On 26 June 2012, NNPR lodged a complaint with the Commission alleging that Uniplate had required its customers, being embossers, to exclusively purchase all types of blanks and all their embossing requirements from Uniplate when purchasing a Uniplate embossing

machine, irrespective of the type of machine bought.

24. On 07 February 2013, the Commission received a similar complaint, this time from JJ Plates a customer of Uniplate. JJ Plates complained that Uniplate was contractually obliging it to purchase all of its number plate blanks from Uniplate and thereby precluding it from potentially sourcing cheaper number plate blanks from other suppliers.

25. Based on the above, the Commission decided to consolidate these complaints under a common investigation and a referral was filed with the Tribunal on 27 November 2015.

26. The hearing commenced on 08 November 2017 and was heard over a period of 9 days.

27. The Commission led two factual witnesses, Mr Johannes Marthinus Steenekamp ("Steenekamp"), the current managing director of NNPR, and Mr Jan Johannes de Lange (ude Lange") the owner of JJ Plates. Uniplate led one factual witness, the current managing

director of Uniplate, Mr Devandran Naicker ("Naicker").

28. Both sides also called economic experts. Dr Liberty Mncube ("Mncube"), the Commission's Chief Economist was the expert

for the Commission and Mr Richard Murgatroyd ("Murgatroyd") of RBB, a private sector economics consultancy, testified on behalf of Uniplate.

Commission's complaint

29. Uniplate was established in 1957 under the name, United Reflective Company, as a manufacturer and distributor of number plate blanks and embossing machines. In 2009, Uniplate became a subsidiary of Tonnjes CARD International GmbH which gave it access to global experience and world class technology which it used to diversify its operations in Africa.

30. lt supplies its embossing machines on either cash sale (being either cash upfront or on instalment), loan or rental terms to its customers comprising mainly of embossers.

31. The basis of the Commission's case is linked to the above agreements and the exclusivity provision contained therein.[10] It reads as follows:

"For as long as this agreement subsists, you shall purchase solely and exclusively from us. all your requirements of retro reflective blanks made from steel, aluminium, acrylic blanks (reflective sheeting and transfers for acrylic), ink, solvent,

SABS stickers and the New Non Paint Supreme Aluminium Blanks, intended to be used for the manufacture of motor vehicle registration plates in accordance with SANS Specification 1116 parts II and IV, and any future amendments to the motor vehicle licence plates as legislated by the Government of RSA. Should we for any reason whatsoever not be in a position to supply your requirements, you shall be entitled to acquire your blanks from an alternate source, provided we have afforded you permission in writing, which permission shall apply only for such period we are not in a position to supply." (own emphasis)

32. This provision confirms that Uniplate required customers to exclusively purchase all their number plate blank requirements,

irrespective of the type of machine being bought, from Uniplate when acquiring a Uniplate embossing machine.

33. It is common cause that this provision formed the foundation of Uniplate's business model. Uniplate referred to it as the "bait

and hook" system. The embossing machines are the 'bait' and the supply contracts are used as a 'hook' to secure the sale of number plate blanks.

34. The Commission alleged that Uniplate's exclusive supply agreements had resulted: (i) in theforeclosure of the market to firms competing for the manufacture of number plate blanks. These rivals were unable to sell number plate blanks to embossers who signed Uniplate's exclusive supply agreements; (ii) further, embossers were harmed as these exclusive supply agreements deprived them of a choice to purchase their number plate blank requirements from competing manufacturers of blanks often at a cheaper price.

35. Given that Uniplate is the largest manufacturer of number plate blanks in the country, the Commission was of the view that Uniplate

had abused its dominance in the market through the use of these exclusive supply agreements.

36. The Commission therefore concluded that Uniplate's conduct had contravened section 8(d)(i), alternatively 8(d)(iii), or 8(c) of the Act.

Market definition: two markets vs one systems market

37. An area of debate which remained a contentious issue throughout the hearing concerned the relevant product market definition. While the economists were able to reach agreement regarding the relevant geographic market i.e. that this is a national market, they were less inclined to agree with each other regarding the definition of the relevant product market.

38. However, both economists agreed that irrespective of which way the market was defined, Uniplate was dominant for purposes of the Act and therefore the Tribunal did not have to make a finding on the issue of dominance.

39. The Commission defined the market as comprising two separate markets i.e. one for the manufacture and sale of embossing machines; and another for the manufacture and sale of number plate blanks. Uniplate defined the market as an 'overall embossing systems' market where the embossing machine and number plate blanks are sold as an integrated system and therefore constitute one market.

40. For its part the Commission relied on the following factors to define two separate markets:

40.1 The evidence of all three witnesses who testified before us that 'mixing' by embossers was possible. Mixing refers to a practice where the embossing machine of supplier X is used to emboss a number plate using a number plate blank of supplier Y. de Lange testified that there was no technical difficulty with mixing; as a matter of fact, he had used Uniplate's embossing machine to make number plates from number plate blanks supplied by NNPR previously and the SASS had never revoked his certification as a result of this. Steenekamp testified that NNPR (up until 2010) sold embossing machines without restricting embossers to use only NNPR number plate blanks on NNPR embossing machines to make number plates i.e. the embossers were free to mix NNPR embossing machines with competing suppliers' blanks. Naicker confirmed that there was no technical impediment to mixing and that it was common in the industry.[11] He acknowledged that absent exclusivity, embossers would be free to mix.[12]

40.2 Uniplate itself sells blanks to walk-in customers who have no Uniplate embossing machines.

40.3 Uniplate's exclusive contracts with embossers provides for the embosser to use alternative blanks on Uniplate's embossing machines provided the embosser has been granted permission by Uniplate.

40.4 The fact that some suppliers in the market only operate in the number plate blanks market indicates separate markets. In this regard the Commission pointed out that Arga and Naicker Tools were only involved in the number plate blanks market.

40.5 A systems market is appropriate where a customer considers 'whole life costing' when purchasing a product whose pricing is dependent on the pricing of another related product. The Commission relied on the OFT Guidelines on market definition which provide that:

"A system market may be appropriate where customers engage in whole life costing or where reputation effects mean that setting a supra competitive price for the secondary product would significantly harm a supplier's profit on future sales of its primary product”.

41. Since de Lange's testimony was that when buying an embossing machine (the primary product) he does not consider the price of number plate blanks (the secondary product), it was inappropriate to refer to the supply of number plate blanks and embossing machines as a systems market.

42. Uniplate, for its part supported an overall embossing systems market on the basis of the interdependence of the embossing machines and number plate blanks. This is because:

42.1 While there may be stand-alone customer demand for number plate blanks, there is unlikely to be stand-alone demand for embossing machines because embossing machines are once-off purchases with no regular demand and are thus unprofitable. Naicker testified that it was not sustainable in South Africa to be solely in the business of selling embossing machines. This explains Uniplate's "bait and hook" model described above, where the embossing machine is sold as 'bait' at a subsidised price and the supply contract is the 'hook' to secure the sale of number plate blanks. Steenekamp also acknowledged that the market for embossing machines is a difficult market because sales are occasional or periodic.

42.2 Insofar asmixing is concerned, Murgatroyd the expert economist for Uniplate submitted that the ability to technically mix Uniplate's embossing machine with NNPR's number plate blanks is not the correct threshold since it is one­ sided. He submitted that while technical incompatibility would clearly indicate a systems market, it does not follow that technical compatibility means separate markets.

42.3 Murgatroyd submitted that whole life costing was not the correct threshold either to determine if the market was an embossing systems market or two separate markets. He submitted that just because de Lange testified that when buying an embossing machine, he does not take into account the number of blanks required and their prices over the lifespan of the embossing machine (whole life costing) it does not mean the market is not a systems market.

Our Findings

43. lt bears mention that apart from this delineation of the relevant product market as discussed, the Commission and Uniplate differed on whether within the broad embossing machine market, there were further distinct markets based on the type of embossing machine (Acrylic, Type A and Type B). The same applies to blanks as to whether the different types of blanks each constituted a distinct market.

44. On the embossing machine side, the Commission argued for a broad market encompassing all three types of embossing machines since it considered supply­ side substitution to be possible, whereas Uniplate argued for distinct markets on the basis that a new entrant would have to incur significant R&D costs to match Uniplate and NNPR's Type A and B offerings, and therefore supply-side substitution was not possible in the short-term.

45. It is not necessary for purposes of these reasons to conclusively decide whether there are distinct markets for each type of embossing machine or a broad market for machines in general. The core concern for an assessment of competition effects is whether there are two separate markets for embossing machines and number plate blanks or one market for embossing systems as contended for by Uniplate.

46. Turning then to this discussion, we have found that there is a primary market for the manufacture and supply of embossing machines

and a secondary market for the manufacture and supply of number plate blanks. This is because:

46.1 The factual witnesses all testified that there was no technical incompatibility between a suppliers' embossing machines and number plate blanks. The evidence shows that de Lange had used Uniplate's embossing machine to make number plates using NNPR's number plate blanks without difficulty.[13] The SABS confirmed in a letter to the Commission dated 6 March 2014 that its process for the accreditation of embossers is based on a standard (SANS1116) which does not prevent mixing.[14]

46.2 Uniplate itself allows its customers to use alternative suppliers' blanks on its machines in certain circumstances. ln addition, the exclusivity provision requires the embosser to purchase all its number plate blanks from Uniplate regardless of the type of machine purchased (despite the fact that a Type A machine cannot produce a Type B or Acrylic number plate). Had the market been a systems market Uniplate would not be selling all types of blanks to an embosser who only purchased a Type A machine.

46.3 As we discuss below, some of the contracts contained buy-back clauses giving Uniplate the first right of refusal to purchase its machine back if the customer requires it. This suggests that there was a market for machines because the embossing machine could be sold to a third party which is not consistent with the notion of a systems market.

46.4 de Lange's first entry in the number plate blanks market was by purchasing an embossing machine online without exclusivity and he was able to purchase blanks from various suppliers. His second purchase was from Teqplate (Uniplate's distributor) who supplied the embossing machine with the said requirement for exclusivity.

46.5 Steenekamp's evidence was that NNPR entered the number plate blanks market in 1995 supplying only Type A number plate blanks. Arga and Naicker Tools also operated only in the number plate blanks market when the Gauteng regulation changed in 2010.[15] According to Steenekamp NNPR only started producing embossing machines for the first time a few months later in 1995 because it

could no longer sign up customers because they were locked into Uniplate's exclusive contracts.[16]

46.6 In his witness statement, Naicker also referred to instances where Uniplate supplied embossing machines without the requirements to exclusively purchase all number plate blanks requirements (and other consumables) from it. The sale agreements were concluded in 2009 with Dampier GM and 3DX-ACT respectively.[17] Naicker said the above two instances were not in line with Uniplate's strategy of 'bait' and 'hook' but were borne out of special

relationships that the Managing Directors of Uniplate, at the time, had with these customers.

46.7 Murgatroyd accepted that absent exclusivity, embossers would purchase number plate blanks from a different supplier to their embossing machines if they could find a cheaper price.[18]

47. These facts indicate that there exist two markets. It may well be that these markets are interdependent, but their interdependence is no basis to deviate from the normal approach to market definition. As the European Guidelines on Exclusionary Abuses indicate, the interdependence of markets is relevant in assessing dominance, not market definition.[19] The fact that Uniplate has chosen a particular business model for its own commercial interests does not make the business model a market for competition law purposes. In fact, the very rationale of the exclusivity clause (the "bait and hook" model) suggests that there are two markets, one for machines (low volume of sales) and the other for blanks (large volumes). Were it otherwise there would be no need for Uniplate to require customers to use only Uniplate blanks.

Legal Framework

48. As mentioned, it has been accepted that Uniplate is dominant for purposes of section 8. The Commission's case is that Uniplate has contravened sections 8(d)(i); or in the alternative 8(d)(iii) or in the further alternative 8(c) all of which require dominance.

49. Although the Commission has sought an order declaring that Uniplate has contravened these sections of the Act, based on the evidence before us we have assessed this matter under section 8(d)(i) of the Act. Section 8(d)(i) provides as follows:

"It is prohibited for a dominant firm to-

(d) engage any of the following exclusionary acts, unless the firm concerned can show technological, efficiency or other pro-competitive, gains which outweigh the anti-competitive effect of its act;

(i)requiring or inducing a supplier or customer to not deal with a competitor; ..."

50. Murgatroyd considered that from an economics perspective, section B(d)(i) and 8(d)(iii) were two sides of the same coin in that while section 8(d)(i) is concerned with exclusive dealing, 8(d)(iii) was concerned with vertical tying which has the same economic effect as exclusive dealing under B(d)(i).

51. Whether we decide this case under section 8(d)(i) or 8(d)(iii) makes no material difference in this sense. It is trite under both sections that once the Commission has proved anti-competitive effects, the onus shifts to a respondent to justify its conduct by showing pro-competitive gains that outweigh the anti-competitive effects shown by the Commission. If we find that the conduct is a contravention of section 8(d)(i) there would be no need for us to assess whether the same conduct contravenes section 8(d)(iii) of the Act.

52. The Commission's theory of harm was foreclosure. It argued that Uniplate has sought to leverage its dominance in the supply of embossing

machines by requiring embossing m chine customers to also purchase number plate blanks exclusively from it, in contravention of section 8(d)(i).[20]

53. The parties agreed that section 8(d)(i) was effects-based in that it required a showing of competitive harm. The main contention between them was whether the Commission had met the standard required to prove this harm.

54. The Commission largely relied on the seminal case of Competition Commission/South African Airways ("SAA")[21] where the Tribunal provided guidance on the standard of anti-competitive harm required: "this question will be answered in the affirmative if there is (i) evidence of actual harm to consumer welfare or (ii) if the exclusionary act is substantial or significant in terms of its effect in foreclosing the market to rivals." (own emphasis)

55. The Competition Appeal Court later confirmed this test in South African Airways vs Comair Limited and Another.[22] It held:

"if the exclusionary act is substantially significant, in terms of its effect in foreclosing the market to rivals, the section applies. This approach can be established either by way of evidence of actual competitive harm or by evidence that the exclusionary practice is substantially significant. that is the practice has the potential to foreclose the market to competition, in which case an anti-competitive effect can be inferred."

56. The Commission focused on the second method of showing an anti-competitive effect i.e. that the harm was substantially significant. This was because it was difficult to construct a counterfactual without the exclusive agreements. The Commission therefore relied on indirect proxies, such as evidence on the overall incidence and nature of the contracts, and the extent of their foreclosure of competitors.

57. Despite the clear case law cited above, Uniplate persisted that the cases were not authority for the proposition that the Commission does not need to prove actual as opposed to likely foreclosure but offered no other authority contradicting the above. Uniplate agreed however that the foreclosure must be significant and substantial.

58. Later case law following SAA makes it clear that proof of actual foreclosure is not required.

59. In Nationwide Airlines (Pty) Ltd v South African Airways (Pty) Ltd ("Nationwide'? the Tribunal stated that in order to establish likely or actual anticompetitive effects, it is not necessary to show that the conduct

"completely foreclosed rivals from entering or accessing a market". Rather it was sufficient to show that the conduct "prevents or impedes a firm from expanding in the market".[23]

60. The Tribunal further held in Telkom that "In order to show harm for purposes of section B(d)(i) it is not necessary to show that competitors must first exit a market or even that they lost market share before harm. All that is required to be shown is that Telkom's conduct was likely to result in preventing or lessening competition which would include the impeding of competition".[24]

61. In light of this case law, we now turn to consider the evidence the Commission relies on for its submissions that there has been substantial foreclosure and Uniplate's rebuttal of this.

The contracts

62. As previously indicated, Uniplate enters into three types of contracts with embossers namely, cash sale agreements (which include cash upfront or an instalment sale);[25] rental agreements;[26] and loan agreements. A total of 514 contracts were submitted by Uniplate during the Commission's investigation which covered the period 1994 to 2014.

63. The contentious terms which featured in all of Uniplate's agreements were the exclusivity requirement and the duration of the agreements. We turn to consider these aspects.

Commission's view. Uniplate's responses and Tribunal Assessment Exclusivity

64. The Commission found that 99% of contracts entered into by Uniplate in the period 1994-2014 provided for exclusivity. In terms of this exclusivity clause, embossers were required to purchase all of their number plate blank requirements from Uniplate, regardless of the type of embossing machine purchased. This meant that even if for example a Type A embossing machine was purchased, an embosser was still required to purchase their Type B and Acrylic number plate blank requirements from Uniplate.

65. Recall the discussion earlier that a Type A embossing machine can only make a Type A number plate i.e. there is no substitutability

between machine type and number plate blank type. Despite this lack of substitutability, Uniplate required customers buying say a Type B machine to buy all their blanks e.g. Type A blanks even though these cannot be used on the Type B machine.

66. The Commission found that even when the ownership of the embossing machine had transferred i.e. in a cash sale agreement (where the recoupment of costs justification by Uniplate does not apply), Uniplate still required embossers to purchase their number plate

blank requirements from Uniplate.

67. Uniplate did not deny that it required exclusivity in its agreements but denied that its exclusive supply agreements have harmed

competition. It put up three justifications for its exclusivity clause.

68. The first was that exclusivity allowed Uniplate to offer its embossing machines at a subsidised price. By doing so embossers were able to afford the machine and Uniplate was able to recoup its investment costs incurred to produce the machine.

69. The second defence put up was that exclusivity prevented free riding and ensured against reputational damage i.e. it ensures that the number plates are of the requisite quality and adhere to regulations.

70. Finally, exclusivity enabled Uniplate to offer free servicing and maintenance of embossing machines to its embosser customers.

71. We will deal with these defences in more detail under the Efficiencies/Defences section of these reasons. We turn first to the

duration of the exclusive purchase obligation.

Duration

72. The majority of Uniplate's contracts reviewed by the Commission are rental agreements at 55%, followed by cash and instalment sales at 30% and loan agreements at 15%.[27]

73. 91% of the contracts i.e. across all three types of agreements, in the period 1994 to 2014, endured for 120 months being 10 years.[28] This was reinforced by the fact that other than the loan agreements; the contracts did not provide for termination. According to the Commission, the absence of a cancellation or termination clause in cash (upfront or on instalment) and rental agreements further

contributed to foreclosing the market to competitors since customers were locked into long term contracts with no option to terminate.

74. Moreover, 93% of the cash and rental agreements provided for the automatic renewal of the contract for a further 120 months after the expiry of the initial 120 months. As we discuss later Uniplate denies that 91% of the contracts endured for 10 years. lt claimed that the contracts could be terminated before the expiry of the 10-year period.

75. Uniplate conceded that thebasis for the ten-year period was not scientific in nature. It justified the duration as being a reasonable period to enable it to recoup the cost of its investment as well as to provide value added services to the customer.[29] However, as we discuss below, this claim was not substantiated.

76. As to the absence of a termination clause, Naicker testified that the majority of Uniplate's contracts were loan agreements (and contained termination clauses) which are entered into in order to enable embossers to try out Uniplate's system without having to fully commit to purchasing the embossing machine outright.[30] However, as mentioned earlier the Commission's contract analysis shows that loan agreements only constituted 15% and not the majority

of the contracts. In his witness statement[31] Naicker stated that loan agreements constituted the majority of the agreements in respect of Type A and Acrylic machines but were

insignificant for Type B machines.

77. He provided three responses regarding the absence of termination clauses in cash and rental agreements. Firstly that, although the contracts contained no provision for cancellation, Uniplate would not as a matter of philosophy insist on continuing with an agreement when a customer wanted to cancel. Secondly, although there may be no express cancellation, Naicker submitted that the embosser could still exit the contract via the buy-back clause. Thirdly, that agreements without termination were old agreements prior to 2011; since then Uniplate has not had contracts without a termination clause.

78. However, the alleged philosophy of cancelling when a customer requests it, is not borne out by the evidence. On the contrary, the evidence supports the conclusion that Uniplate and its distributors resisted their customers' attempts to cancel.

79. de Lange testified about his numerous attempts to cancel his rental agreement with TeqPlate a Uniplate distributor. He was told that he had to wait the ten-year period.[32]

CHAIRPERSON: We have a few questions. Mr de Lange, I want to go back to your ... you said that you had tried to cancel your agreement with TechP/ate many times?

MR DE LANGE: Yes.

CHAIRPERSON: And when you did try to do that, what was their response?

MR DE LANGE: Their response was that I cannot cancel the agreement, I was bound to the conditions of the agreement and that was their answer, yes. I have to ... the term is 10 years and I have to wait until the term has expired.

80. It was only during the hearing that Uniplate made an offer to de Lange to cancel. Naicker explained his belated offer on the basis that he was unaware that JJ Plates had attempted to cancel the Teqplate agreement. In this explanation Naicker seemingly tried to distance Uniplate from the business of Teqplate. […] Naicker's attempt to distance Uniplate from de Lange's failed efforts to cancel the contract cannot be sustained in light of the relationship between Uniplate and TeqPlate.[33]

81. Turning then to Uniplate's claim that despite no cancellation provision embossers could exit the agreements through the buy-back

clauses. The relevant provisions state that:

Cash sale agreement:

"In the unlikely event that you wish to sell the equipment, Uniplate reserves the right to be given fist option to re-purchase the equipment on the following scale:

0-1 year 75% of purchase price excluding VAT

1-2 years 50% of purchase price excluding VAT

2-3 years 25% of purchase price excluding VAT

3-8 years 5% of purchase price excluding VAT

8-12 years 2.5% of purchase price excluding VAT

12 years and on 0.5% of purchase price excluding VAT".

Instalment sale agreement:

"In the unlikely event that you wish to sell the equipment, the supplier reserves the right to be given the first option to re-purchase at a price negotiated relevant to the time period already lapsed. The above also assumes the equipment to be in good working order Jess normal wear and tear. This arrangement will not apply should you sell your number plate business and any new owner enters into .a dealership Agreement with the said supplier".

82. In oral evidence, Uniplate provided figures to show the total number of Uniplate agreements allowing for early termination or buy-back, over the period 2010 to 2014 across all types of agreements.[34] The table (reproduced below) shows that in the relevant period on average, 45% of contracts allowed for termination, meaning that

customers would have been able to exit Uniplate's agreements.[35]

83. However, we have found the above table unhelpful since it excludes contracts concluded prior to 2010 which would still probably be in force in the relevant period of the complaint (2010-2014). In other words, the number of contracts concluded e.g. in 2009 which would be in force until at least 2019 are not included in this table.

84. Significantly, the table only reflects contracts in which termination or buy back clauses were included. It doesn't serve as evidence of the actual number of machines sold and the number of terminations and buy-backs that actually occurred during the relevant period. Uniplate conceded as such that they could not put up any evidence of the number of terminations or returned machines as demonstrated during the testimony of Naicker below:

CHAIRPERSON: The question I have is, and if the answer to that is confidential, then you can tell me. You say that while those contracts did not have cancellation clauses, the policy has always been that when the customer ... if the relationship fails and you can't persuade them, that you would certainly buy back the machine and cancel the agreement. Would you be able to give the Tribunal a sense of in how many instances you've done that and particularly during the 2010 to 2014 period and for what reason? Because I saw Mr Chagan saying somewhere and I can't locate the document, that that would be only when the embosser exits the market, not when the embosser wants to switch. So, do you have any slats like those in your business since you do have a contract management unit?

MR NAICKER: I don't have the spreadsheet, this wonderful spreadsheet. How much of dependence I have on it now is questionable. It's a human thing. So, I stand under correction. I know I did something for the economist I (sic) terms of understanding all the contract periods of equipment returned and customers that were allowed to cancel and move over. So, I don't have the exact figures here, but I'm sure that can be noted by the team and they can present that information to you, ma'am.

ADV WESLEY: Chair, we can certainly try and do that analysis."

85. This spreadsheet was never provided.

86. As mentioned, Naicker also said there were no longer any contracts without a cancellation clause since 2011. He provided examples of contracts post-2011 which expressly allow for termination.[36]

87. While it may be that the insertion of a cancellation clause in the newer agreements is an improvement, it does not nullify the impact of the agreements as they existed before this change. What we know of the agreements so far is they contained exclusivity for ten years and up until 2011 did not expressly provide for cancellation. Additionally, they provided for the automatic renewal of the contracts.

88. Naicker stated that automatic renewal was inserted to safeguard customers should they wish to continue the contract after the initial ten years. He mentioned that Uniplate had an internal system which alerts it to the upcoming termination dates, following which a Uniplate representative will visit the customer to determine whether the customer wishes to terminate or renew the contract.[37] Should the customer wish to renew, only then will the contract be renewed. As such the renewal was not automatic as the contract

suggests.

89. However, we have no factual evidence of cancellation by any customer either in terms of an express cancellation or buy back clause during the term of the contract. Nor at the end of the term of the contract (when automatic renewal kicks-in).

90. In fact, the evidence supports the contrary situation. When asked whether he had had any experience of a customer cancelling the contract at the end of the initial 10-yearterm Naicker responded:"/ haven't seen a customer cancel the agreement at the end of term.[38]"

91. This means in reality the majority of the contracts endured for longer than 10 years. The evidence of Mr Nizoo Chagan ("Chagan"),

Naicker's predecessor as CEO of Uniplate, confirms that cancellation by a customer was a rare occurrence. In an e-mail to the Commission

dated 2 August 2013 he stated that: " ... might I add in my 40 years' experience with Uniplate. I can safely say I have not had to cancel more than 20 agreements. mainly because of the customer not paying his account or has gone into liquidation or has used blanks that were not supplied by Uniplate."[39]

92. An analysis of the contracts and the evidence traversed so far shows that the majority of the contracts were at a minimum 10 years long and invariably longer since customers once signed up either could not cancel because there was no provision for such, or as de Lange testified, were refused cancellation when they tried to cancel. Thus customers were locked into a period of 10 years or longer in an arrangement which required that they purchase all their blanks and embossing requirements from Uniplate such as solvents, dies and even stickers, irrespective of the type of machine they had bought or the price of blanks and other materials.

93. A further point linked to customers' inability to switch out of the Uniplate purchase obligation is demonstrated in cash sales where the customer is obliged not only to purchase all of their number plate requirements but also the dies.

94. As explained dies are consumables used to produce alphanumeric number plates. They typically wear out faster than the embossing machine, becoming blunt due to usage, and have a shorter lifespan to the machines. Hence a customer was likely to require new dies prior to the expiry of the 10-year exclusivity period.

95. In such a situation Uniplate requires the cash customer to extend this exclusivity for a further ten years by signing a new supply agreement for dies.[40] This exchange addresses this:

ADV QUILLIAM: Well, let's go to your witness statement on page 131 of the bound bundle. It's paragraph 61. Mr Naicker, this is your witness statement. It says ·the purchase agreement includes a requirement for exclusivity, purchase blanks for a period of 120 months. Although notionally a customer will then be free at the end of the period to buy blanks from any supplier, in fact. the lifetime of the machine and particularly the dies for the aluminium blank plates is less than 10 years and so the customer would have to acquire a new machine or at least the material components of that machine. being the dies before the expiry of the 10-year period and would then be required to agree to a new period of exclusivity?" So. when I purchase replacement dies, I will be required to sign a new contract of exclusivity.

MR NAICKER: That's correct.

ADV QUILLIAM: And this period of exclusivity is for a further 120 months. MR NAICKER: That's correct. for this system. (own emphasis)

96. Uniplate provided no economic rationale for extending the contract for a further 10 years for a cash customer who already owns a machine and simply wants to purchase dies in the normal course as dies are outlasted by the machine.

97. Based on the contracts and the evidence discussed above, it is clear that the duration of the contracts is excessively long. A customer was bound for a period of 10 years which could be extended automatically for a further 10 years (in some instances) or with the purchase of additional dies in other instances. Even where the contracts provided for early termination or buy-back before the expiry of the 10

-year period, the evidence was that in fact there were very few, if any at all terminations or buy-backs. Where customers did try to terminate they were met with fierce resistance as discussed later.

Contestable Demand

98. We now turn to the Commission's claim that because of the nature of the agreements as discussed above, the size of the contestable demand for which NNPR could compete was limited.

99. This was compounded by the staggered nature of the contracts, meaning that the contracts would come to an end at different times. Because of this, when the contracts would end was unknown and it did not follow therefore that there would be sufficient contestable demand for competitors when contracts end.

100. According to the Commission, the staggered nature of the contracts played a significant role in raising barriers to entry.

101. Murgatroyd accepted the Commission's claim that the demand that NNPR could compete for was reduced as a consequence of Uniplate's existing exclusivity arrangements.[41] However, he submitted that NNPR could turn to new embossers; or those who switch away from Uniplate during the term of the contract; or those who cancel at the end of the contract.

102. He submitted that there could not be substantial foreclosure in light of the market share data provided below. To the contrary, he submitted that there was effective rivalry between Uniplate and NNPR as the market shares allegedly show.

Market shares

Table 2: Market Shares by company. of blanks supplied over the period 2010 to 2016 (%)

[See PDF for table]

103. Murgatroyd submitted that an analysis of the market shares above shows that NNPR maintained its market shares and in respect of Type 8 blanks, even having entered the market only in 2012, grew market share which shows effective rivalry between the two firms.

104. However, as Mncube correctly submitted the market share data[42] is meaningless without the counterfactual, which is - what would the competitive landscape look like absent Uniplate's exclusivity. Since the market shares are incapable of showing whether more firms could have come into the market or not, or whether any firms did come in and exit or what the true competitive dynamics would be absent exclusivity, they are unreliable.

105. But even if we were to rely on the market shares, they do not support Uniplate's claim that there remains a sufficient contestable demand despite exclusivity. On the contrary the market shares demonstrate Uniplate's enduring dominant position in all blanks increasing over 5 years from 72% to 76%, and NNPR's share staying stable within the low range of 17-21%.

106. A closer look shows Uniplate's market share increasing by 5% between 2011 and 2012, while NNPR's increases by 0% in the period. In the subsequent period (2012-2013), NNPR's share increases by 1% while Uniplate drops by 2%, (and they both retain their respective market shares in 2013-2014). This can be explained as Steenekamp[43] and Mncube did,[44] by the increase in demand for Type B blanks in Gauteng since NNPR started supplying Type B machines in 2012, rather than effective rivalry.

107. In light of Uniplate's dominance in all blanks and in Type B (in Gauteng which is the largest regional market), the size of the contestable demand would be miniscule if not absent.

Have Uniplate's contracts foreclosed the market?

108. Against the backdrop of the contracts we turn to consider the evidence of the witnesses regarding how the contracts have resulted

in foreclosure.

109. Recall, the Commission's case has been that Uniplate's exclusive agreements have had the effect of substantially foreclosing the market to manufacturers of number plate blanks. Put differently, competitors of Uniplate in the manufacture of number plate blanks have been foreclosed or unable to gain access to a significant number of embossers.

110. As mentioned, Steenekamp, Uniplate's competitor and de Lange, Uniplate's customer testified before us.

Steenekamp

111. Steenekamp's evidence was that Uniplate's exclusive contracts, which were present in the market when NNPR entered in 1995, have hindered its ability to compete in the market for number plate blanks. This is because NNPR has been unable to access embossers to achieve a sufficient scale of economies.

112. This was exacerbated in 2008/2009 when Uniplate began to enforce its exclusive agreements more aggressively in the market. This aggressive stance was driven by a change in regulations in 2010 requiring vehicles in Gauteng to be fitted with Type B number plates.

113. At this stage Uniplate had already developed a Type B machine tomeet the change in regulations in Gauteng. It was the only manufacturer of Type B blanks at the time and it started selling its Type B equipment to embossers in the market on an exclusive basis. Many of these customers had been loyal customers of NNPR at the time.

114. It appears to us from Steenekamp's evidence that Uniplate's contracts previously contained exclusivity, but this was limited to the supply of blanks compatible with the type of embossing machine purchased i.e. when an embosser purchased a Type A embossing machine they were required to exclusively purchase their Type A blanks from Uniplate. This exclusivity was not aggressively enforced.

115. However, the exclusivity requirements changed when Type B embossing machines were introduced. Uniplate now required the embosser to purchase fil!_of their number plate requirements from Uniplate regardless of the type of embossing machine purchased. This foreclosed these embossers from continuing to purchase blanks from NNPR, including Type A or Acrylic blanks.

116. Since NNPR had not yet at that time entered the market for the manufacture of Type B embossing machines, NNPR was foreclosed from continuing to supply these customers their requirements for Type A or Acrylic blanks. Uniplate continued to sign up customers for its Type B equipment and soon had signed up some of NNPR's top customers to NNPR's detriment.[45]

117. In order to retain its customer base, NNPR introduced an exclusivity provision into its rental agreements in 2010 as a defensive strategy. This exclusivity provision was similar to that of Uniplate's; however, it only endured for a period of 5 years and was contained only in NNPR's rental agreements and not for example in the cash sale agreements.

118. NNPR's justification for the inclusion of this provision was simple - it needed to create its own demand for its blanks and secure the loyalty of customers. However, Steenekamp's evidence was that exclusivity cannot endure for an unreasonable length of time such as in the case of Uniplate's.[46] He was of the view that a 5-year period was sufficiently reasonable to recoup any investment incurred for the manufacture of the embossing machine.

119. According to Steenekamp this (recoupment period) was calculated by taking the cost of the machine and doubling it so that this

investment was recouped within the 5-year period.

120. Further he suggested that if he were selling blanks to the embosser, this investment would be recovered in a shorter period of say 3 years. We will return to Steenekamp's evidence when we consider Uniplate's recoupment justification.

121. Naicker countered Steenekamp's allegation that he had lost customers to Uniplate as a result of Uniplate's exclusive contracts and claimed instead that Uniplate had lost customers to NNPR. The customers claimed to have been lost to Uniplate by Steenekamp included Dirlo and Top Parts.[47] According to Naicker, neither of these agreements (concluded in 2008 and 2010 respectively), contained exclusivity. Dirlo's was a cash sale for an Acrylic embossing machine without exclusivity over all blanks as Dirlo was only obliged to purchase Acrylic blanks from Uniplate consistent with the Acrylic machine it purchased.

122. Similarly, Top Parts concluded a rental agreement with Uniplate in 201O for a Type B embossing machine, which according to Naicker did not oblige Top Parts to purchase any other blanks but Type B blanks.

123. In our view, these isolated instances do not detract from the undisputed evidence that the majority of the contracts in the

relevant period contained exclusivity over all blanks regardless of machine type. Given that the "bait and hook" model was precisely to ensure exclusivity, these isolated instances are exceptions from the rule. What they do demonstrate is that without an exclusivity provision, customers would be able to switch away from NNPR to Uniplate, a feature of a competitive market.

124. Turning to Uniplate's claim that it has lost customers to NNPR, Naicker listed BB Group, Bidvest and McCarthy as customers it had lost to NNPR.[48]

125. However, Steenekamp explained the switch by these customers as follows.[49]

ADV QUILLIAM: A/right, let's move to next topic, which is customer switching in your experience with such customers. If you could move to page 133 of your bound pleadings bundle, just as a bit of background, this is a Uniplate submission on paragraph 70. In this paragraph Uniplate lists a number of firms that they are alleging that have switched or have changed from Uniplate to NNPR for the supply of what they call embossing systems. Could you just explain to us who these firms are, starting from the BB Group?

MR STEENEKAMP: The BB Group, Bidvest and McCarthy Motor Groups are not NNPR clients. These are clients of an embosser. So, the embosser was successful in persuading these companies to switch. The embosser is just using NNPR's equipment, but it is an independent company completely independent from NNPR.

126. The above demonstrates that the embosser could switch away from Uniplate to NNPR by using NNPR equipment. In other words, NNPR could only win embossers as customers only if it also offered a machine to them. Large embossers would then possibly have to carry two machines, one from Uniplate and one from NNPR.

127. It shows that NNPR could only enter the blanks market if it also provided the relevant machine. Thus, the Uniplate exclusivity

provision has the effect of raising rival's costs since a competitor seeking to compete in the blanks market would only be able to do so if they also supplied the relevant machine. A rival would not be able to win a customer for blanks only; they would have to enter both markets to compete effectively.

128. Furthermore, the few isolated incidents of customer switching that Naicker relies on as evidence of effective rivalry in the blanks market is not supported by the evidence of others.

de Lange

129. de Lange testified that from 2002-2009 JJ Plates sourced itsblanks from numerous blank suppliers including NNPR, Arga and Uniplate. In 2009 in order to conform to the change in Gauteng regulations JJ Plates entered into a rental agreement with Teqplate, a Uniplate

distributor for Type B embossing machine.

130. According to de Lange, the main reason for choosing Teqplate was due to his personal relationship with its previous owner, Mr Dawie

Pretorius.[50] However, it was not long after signing this agreement with Teqplate that JJ Plates began to fully understand the terms of this exclusive agreement. According to him the trouble came after he had purchased number plate blanks from Baleka (Uniplate's other exclusive distributor).

131. de Lange's testimony was that during one routine service a Teqplate technician noticed that a Baleka blank was being used on a Teqplate embossing machine. TeqPlate threatened him with legal action.[51] His oral evidence was as follows:

"MR DE LANGE: Okay, they have people, reps or people that service their machines for them and they would come to your shop regularly, especially when_they find out that your purchases have dropped and that's when they found out that I'm still purchasing plates from Baleka. ADV QUILLIAM: And after discovering this, how did TechPlate react?

MR DE LANGE: TechP/ate started sending me threatening letters. ADV QUILLIAM: A/right, and in reaction to these letters, what did you do?

MR DE LANGE: The letters were so threatening at that stage that I actually stopped purchasing immediately from Baleka."

132. de Lange's defence for purchasing number plate blanks from Baleka was simply that they were cheaper.[52]

133. As mentioned above, de Lange tried on numerous occasions to cancel the contract but was told he was bound for the 10-year period.

134. This threat of litigation by TeqPtate against de Lange for using blanks supplied by another Uniplate distributor is not the only instance of threats of legal action against de Lange. de Lange was threatened in at least two other instances.

135. In or about 2011, de Lange claims he started sourcing Type B number plate blanks from NNPR (to be used in a Uniplate machine). Upon learning of this, Teqplate served court papers on JJ Plates for defamation and for breaching its agreement with Teqplate. Following this, de Lange only sourced number plate blanks from Teqplate which he claims were 20% more costly than alternative suppliers.

136.

At the same time, de Lange also entered into a rental agreement with NNPR for a Type B embossing machine. Once again Teqplate threated legal action, JJ Plates then cancelled the contract with NNPR despite thefact that an

NNPR machine was being used to emboss NNPR blanks.[53]

137. de Lange submitted that based on his experience in the market, Uniplate's conduct had precluded him from sourcing cheaper number

plate blanks which were available from alternative suppliers in the market.

138. de Lange's evidence is borne out by a notice sent by Uniplate to embossers dated 5 November 2012[54] advising them that NNPR has been ordered by the South Gauteng High Court to stop supplying embossers with their requirements and .. .If NNPR does not abide by the court order, NNPR will be in contempt of court which is a criminal offence. Any embossers with contracts with Uniplate and who entertain NNPRs unlawful conduct will be assisting NNPR to commit a criminal offense and thereby subject themselves to possible legal action. We therefore strongly recommend that you do not assist NNPR to commit any criminal offenses and we inform you that Uniplate will not tolerate any embossers who breach their agreements with us by purchasing blanks from NNPR".[55]

Other evidence of Uniplate's alleged aggressive enforcement of the contracts against embossers

139. The record also contained evidence of other threats of legal action instituted against embossers in the market who breached Uniplate's exclusive agreements. The Commission alleged that Uniplate's litigious stance against embossers to enforce exclusivity further foreclosed the market to competitors.

140. The Commission addressed some of this litigation with Naicker in oral evidence, as discussed below.

141. One was a "Notice" sent by Uniplate [dated 2012, to embossers reminding them to adhere to the conditions of their agreements. This was prior to the High Court decision referred to above. The subject of the notice wasu NNPRoffering Non-paint blanks to our customers which have Uniplate Non-Paint Machines". Of significance was paragraph 6 of this notice which reads as follows[56]:

"6. Should you purchase the Non Paint Blanks from NNPR or any other competitor and use them on our equipment, this will be a direct and material breach of the agreement with Uniplate and you could face the following actions: Uniplate will proceed through the legal system for specific performance, damages, and or cancellation of our agreements whereby we will repossess the equipment and still demand the losses we have incurred through your breach. So you risk being out of business and facing additional costs in legal fees.”

142. Another instance involved Auto Number Plates which had in its possession equipment from both Uniplate and NNPR and had been purchasing its Acrylic blank requirements from NNPR. When Uniplate became aware of this, Auto Number Plates was litigated against in order to enforce Uniplate's exclusive contracts.[57]

143. Although these incidents relate to a period before Naicker joined NNPR, Naicker acknowledged these incidents and confirmed that Uniplate is serious regarding enforcing exclusivity through the courts.[58]

144. Additional incidents of threats of litigation in the record include an extract from a letter dated 12 January 2012, sent by Uniplate to a company called Q No More, which reads as follows[59]:

3. We will not tolerate any such loss occasioned by your wilful failure to honour the agreement, and at this stage, we demand that you within seven (7) days:

Immediately cease purchasing your requirements from opposition entities and furnish us with your written undertaking to that effect; Furnish us with a further written undertaking that you will honour the terms and conditions of the agreement in that you will specifically resume all your purchases solely and exclusively from us in accordance with the terms and condition of the agreement.

4. Should you fail to comply with the above, an application will be made to Court to interdict you and compel you, in which event the costs incurred will be for your account."

145. In a letter dated 14 March 2014, Doreen Wadsworth, the owner of Vos Plates an embosser which exited the blanks market in 2012 also speaks of her experience of litigation with Uniplate[60]:

[…]

146. These extracts serve to support the pattern that was already alluded to by de Lange, Steenekamp and Naicker himself, namely that Uniplate would resort to legal action to enforce its exclusivity, thereby putting paid to the claim by Uniplate itself that customers could terminate their contracts easily and switch to another supplier. The ability of customers to switch was hindered by Uniplate's exclusivity requirements. However, customer switching was rendered nigh impossible by Uniplate's litigious stance.

147. This is not to say a dominant firm may not legitimately enforce its rights through the courts, but there is authority for the proposition that in certain instances it may be abusive for a dominant firm to engage in litigation against its customers and competitors.[61]

148. Uniplate's litigious approach constituted sufficient harassment for embossers to desist with dealing with competitors, as testified to by de Lange and as the record shows.

Did Uniplate's contracts raise barriers to entry?

149. There was no disagreement between the parties' expert economists that barriers to entry were high. The contention was whether the exclusive arrangements made entry more difficult or not. As has already been shown by the evidence of Steenekamp, NNPR could only effectively compete against Uniplate in the blanks market if it also incurred the additional costs of producing and supplying embossing machines. Thus, barriers to entry in the blanks market were increased for rivals.

Potential entry

150. There has been no effective and sustained entry into the South African number plate industry.

151. In the relevant period three international firms allegedly explored the possibility of entering and attempted to enter the South African the number plate industry. However, according to the Commission, these plans were abandoned as a result of Uniplate's position in the market and the insufficient demand created as a result of Uniplate's exclusive contracts. The three firms that attempted to enter were Utsch, Utal sp and Smart.

152. The first potential partnership was in the form of Utsch in 2007. Utsch is a German based manufacturer of blanks who, according to Steenekamp wanted to acquire a stake in NNPR. However, after negotiations failed Utsch subsequently entered the blanks market on their own in 2009. This was short lived and Utsch ceased its operations in South Africa in 2010/2011.

153. Uniplate submitted that there were no facts or direct evidence from Utsch led by the Commission confirming that Utsch exited due to the exclusive arrangements in place between Uniplate and embossers. In a letter dated 1 February 2011[62] seemingly to its customers, Utsch stated that it exited due to the Department of Transport's delay in implementing a change in a certain number plate system and the economic downturn.

154. While Steenekamp's evidence about his involvement with Utsch is based on his personal knowledge, we agree that his evidence does not take the matter any further as to the reasons for Utsch's exit. Nevertheless, the evidence confirms Utsch's failed entry in the blanks market.

155. The second potential partnership took place in or about 2011. Steenekamp testified that during this time NNPR had been contacted by Utal sp to potentially partner to produce number plate blanks in South Africa. However, after Utal sp became aware of the insufficient demand for number plate blanks in South Africa, this business venture was abandoned.[63]

156. Uniplate submitted that Steenekamp's testimony was not direct evidence from Utal sp itself and there was no other evidence

corroborating Steenekamp's evidence. This objection is without merit because Steenekamp was not just a bystander to this situation, he was the other party in the potential NNPR/Utal sp relationship.

157. In any event de Lange corroborated Steenekamp's evidence regarding Utal sp but for a different period. According to de Lange, he was approached in 2014 by Utal sp to potentially enter into a partnership to produce licence plates in South Africa. However once Utal sp learnt about the exclusive contracts and the insufficient demand in the market, this potential business venture was abandoned.

158. This is confirmed in a letter from Utal to the Commission dated 28 September 2016 in which Utal confirms its visit to South Africa in 2014 to explore the possibility of entering the number plates market. Following the visit, Utal stated[64]:

"The current structure of the South African market for license plates, however, would predestine any investment from Utal for failure. As far as Utal is concerned, the main reason for this resides in Uniplate's dominating position on the market that clearly seems to extend to a great number of embossers, thus hampering other companies to access the market".

159. Steenekamp also testified that Smart had discussions with NNPR who were interested in setting up a number plate blank business but decided not to pursue this on learning of the exclusivity provisions that tied-up customers in long term contracts.

160. In conclusion, while there is no direct evidence from the abovementioned international firms, the evidence of failed entry into the relevant markets cannot be disputed. As the case law discussed above shows, it is not necessary to show that competitors have exited the market or that they lost market share, it is sufficient to show a prevention or lessening of competition, including impeding competition.

Conclusion on effect

161. Based on the evidence discussed above we conclude that Uniplate's exclusive agreements have had the effect of increasing rivals' costs (as in the case of NNPR) because they could only compete effectively in the blanks market by incurring the additional cost of entering the machines market.

162. Rivals in the blanks only market remained small as demonstrated by the market shares.

163. New entry was discouraged because demand for blanks was tied up in contracts enduring for 10 years or longer.

164. Customers were prevented from switching to rival suppliers by the wide exclusivity requirements (on all types of blanks and all embossing materials) and threats of litigation. Customers were even prevented from carrying rival's machines in conjunction with Uniplate's.

165. In our view the Commission has discharged the onus of showing that Uniplate's exclusive agreements had the likely effect of foreclosing rivals in the number plates market. The foreclosure was significant for the reasons mentioned above, resulting in higher prices for blanks and lack of choice for customers has been demonstrated.

166. We now turn to consider Uniplate's defences to determine whether these outweigh the anti-competitive effects mentioned above.

Efficiencies (Uniplate's Defences)

167. In terms of section 8(d)(i) the dominant firm, if found to have engaged in an exclusionary act must show whether there are any technological, efficiency or other pro-competitive gains which outweigh the anticompetitive effects of its conduct. This is often referred to as the "efficiency defence". Recall that in this instance, the evidential onus to establish this defence rests on the dominant firm i.e. Uniplate.

168. Uniplate identified three main efficiencies arising from the exclusivity provision i.e. (i) the exclusivity provision allows for the embossing machine to be offered at a reduced price (at a subsidised price); (ii) prevents free riding and ensures that the number plates are of the requisite quality and adhere to regulations; and (iii) finally that it enables Uniplate to offer free servicing and maintenance of its embossing machines.

Recoupment of Research & Development ("R&D") Costs

169. Naicker's testimony was that the cost of manufacturing an embossing machine is high. Since embossers are generally small and often family run businesses with low capital resources, they often are unable to purchase embossing machines.

170. As such, in order to provide the embossing machine at an affordable price, Uniplate subsidises the cost of the embossing machine and its blanking line by generating revenues through the sale of its blanks.

171. Naicker contended that without such cross subsidisation brought about by the exclusivity clause, Uniplate would be forced to recover all of its costs of the embossing machine (including the initial research and development costs) through significantly higher prices for its embossing machines to the detriment of embossers.[65]

172. As to the actual R&D costs Naicker said when Uniplate started developing the technology for Type B embossing machines in 2007 (in anticipation of the change in the number plate requirements in Gauteng) it incurred R&D costs in the amount of approximately R15 million. Exclusivity allows for the recoupment of these R&D costs over time.

173. However, as we discuss below Uniplate was unable to provide a convincing account of the R&D spend.

173.1 Firstly, the claim that Uniplate machines were provided at a subsidised cost or at no cost is not supported by the evidence. In this regard, Chagan (Naicker's predecessor) stated in a letter from Uniplate to the Commission dated 3 March 2014[66] that "/ state that the [Uniplate] makes only a limited profit, if any, from the sale it receives from its customers in respect of equipment."

173.2 Chagan's statement above was made in support of a statement made earlier by Mr Vishnu Arjun ("Arjun"),[67] the Chief Financial Officer of Uniplate in the relevant period, to the Commission in 2015 who confirmed that Uniplate makes a small

profit or at least covers the cost of the machine.

173.3 Naicker tried to water down Chagan and Arjun's submissions to the Commission that Uniplate was making a small profit or at least covering its costs. He queried their methodology by suggesting that Chagan and Arjun didn't take into account all costs of, for example delivery, installation and training.[68] However, he did not provide supporting documents to substantiate why their evidence should be disregarded.

173.4 To the contrary when the Commission tested whether Uniplate had, in 2014 made a profit on what Uniplate called its systems and dies, Naicker was reluctant to concede that Uniplate had been selling its systems at a profit but conceded that Uniplate's selling price had been higher than its cost. For example, when probed about the painted Type A embossing system, he gave the following response[69]:

ADV QUILLIAM: Alright, would you extend that to the second table that says "the painted type A embossing system" would on Mr Arjun's understanding as financial director of Uniplate at that time, before you joined, the total cost of a type A embossing machine, as he calls it a system, results in as you say, a margin or as I say a profit, on the selling price? So in other words, the selling price is higher than the cost? So let me put it in a question, according to Mr Arjun, for the type A embossing equipment, is the selling price higher than his understanding of the costs?

MR NAICKER: Selling price in terms of the numbers quoted here from him, it is higher than the cost.

174. A similar conclusion can be drawn from the non-paint Type B embossing machine:

ADV QUILLIAM: A/right, now if we take all of these elements and we cross-reference them to page 170 of Mr Arjun's attachment to his e-mail in 2014, do you see that each element that Mr Arjun lists, his cost ... well let's put it this way, in his understanding, the selling price of that item is higher than the cost in his understanding of that item in 2014, do you see that?

MR NAICKER: So we're looking at 170, we're back at 170 now? ADV QUILLIAM: Yes 170, non-paint type B embossing machines. MR NAICKER: Okay.

ADV QUILLIAM: Now I'm asking you is the cost price in terms of Mr Arjun's understanding lower than the selling price of each item listed in

that table?

MR NAICKER: Yes.

175. In cross-examining Mncube, Wesley suggested that if all these alleged additional costs mentioned by Naicker in paragraph 176.3 above were taken into account, the machines would be sold at below cost. However, no actual data was provided to substantiate this claim. As correctly pointed out by Mncube an inference of below cost could not be drawn without seeing the actual extent of those costs.[70]

176. In our view the evidence in totality confirms that Uniplate was covering its costs or breaking even from sale of its machines. Both Arjun and Chagan confirmed this to the Commission and Naicker himself conceded this under cross examination. Furthermore, no data was provided by Naicker to support his claim that machines were being sold below cost.

177. In the face of this evidence Murgatroyd argued that a closer analysis of margins earned on embossing machines compared to blanks showed that the latter cross­ subsidises the former. This is to the extent that Uniplate actually has earned negative margins on its embossing machines.

178. Murgatroyd put up the following table in defence of this issue[71]:

179. In estimating these figures, Murgatroyd claimed that he did not include investment costs. According to him if investment costs were included then the margins would decrease further.

180. But this exercise is flawed firstly because the margin analysis provided was outside of the complaint period; secondly it covered a very short period being 4 months; and thirdly we do not know what the underlying costs and prices of machines used in this analysis are. For all we can tell, Uniplate might have been discounting heavily in this short period of time to meet competition.

181. We agree with the Commission, the table is not helpful in supporting the claim by Uniplate that it required a 10-year exclusivity period in all types of blanks to recoup the tosses it was making on the machines.

182. Moreover, the recoupment justification for cash sales since ownership of the machine passes to the embosser on purchase is not sustainable. As discussed above, the requirement for a ten-year exclusivity on dies in cash sales when dies expire long before the lifespan of the machine is not justifiable. Uniplate provided no basis for requiring exclusivity for a further ten years for dies.

183. In respect of rentals, Uniplate submitted that while it maintained ownership of the machine, exclusivity was still required in order to prevent rivals from free riding off of its investment by supplying blanks to its customer to be used on a machine provided at a reduced price. We will consider this under the second defence.

184. While Steenekamp acknowledged that recoupment was necessary he did not accept the general proposition that exclusivity was the answer. As he testified, NNPR operated its business for 10 years without exclusivity. It only entered into exclusive arrangements with its customers as a defensive strategy against Uniplate. And then only for a 3-5-year period.[72]

185. Even if we were to assume that exclusivity was necessary to recover some R&D costs, exclusivity in respect of all blank requirements when the embosser may have purchased e.g. a Type B machine only could not be justified.

186. In oral evidence Steenekamp testified that it was not necessary to require exclusivity over all types of number plate blanks when only a Type B machine has been purchased, is not justified. He testified that the rationale of recouping costs for a player that is already in the market cannot justified since some of the investment would have been recouped as the innovation of the new machine (Type B) would have been on the back of a machine (Type A) for which investment costs would have already been recovered.

187. When pressed in cross-examination about whether it did not make commercial sense to spread the risk of recoupment over different sales of blanks, he said "It would be a nice-to-have."[73] This suggests that it is not necessary to have exclusivity over everything since the recoupment of costs can still be realised

without tying in all products unrelated to the subject of the contract. It only serves to unjustifiably increase profits for the

dominant firm.

188. Steenekamp's evidence in this regard, although demonstrating that the exclusivity was not related to the costs incurred in the R&D of the type of machine being sold, also highlights the "free riding" that Uniplate itself is engaging in. By requiring exclusivity on all blank types irrespective of the type of machine being sold, it was happily allowing its own customers to "free ride" on another supplier's machine.[74]

To prevent free-riding and Quality Control

189. The second claimed efficiency was that Uniplate required exclusivity in order to prevent free riding by competitors who have not invested in R&D to manufacture embossing machines and simply free ride "by making their blanks on Uniplate's machine. Uniplate could suffer reputational harm from such ''free riding". Exclusivity was therefore required to ensure that number plates were produced at the requisite quality and in compliance with the necessary regulations.

190. In Naicker's witness statement he explained that the exclusivity was in the supplier's interests so as to ensure that the embosser

maintained the necessary SASS certification and avoided the risk of any reputational damage to Uniplate.

191. However, neither the free riding nor quality control defence can be sustained.

192. Recall that Uniplate's exclusive agreements required embossers to purchase all their number plate blank requirements from Uniplate, regardless of the type of embossing machine purchased.

193. What this implies firstly as we have discussed, is that the claim that the market is for embossing systems is not borne out by the facts since Uniplate was willing to sell the full range of the types of blanks to embossers regardless of the type of embossing machine they have bought. Secondly, had free riding been a concern for Uniplate, it would not have been able to allow the customer to use its type A blanks (and free ride) on some other unknown suppliers' type A machine.

194. Furthermore, Uniplate also made a blanket claim of exclusivity being necessary to avoid free riding by embossers who use a competitor's blanks on its machines to make number plates. However, the concern of free riding does not arise in cash sales since ownership passes immediately to the embosser. What the evidence shows instead is Uniplate locks in cash customers into 10-year contracts by requiring them to buy dies (consumables) which wear out long before the 10-year period exclusively from them and signing them up for further 10-year periods when the dies wear out.

195. Turning to quality control and the maintenance of SASS standards, the evidence before us in this regard is de Lange's evidence. He testified that in the relevant period he has never had SASS certification revoked even when he used NNPR's blanks on Uniplate's

machine. According to Naicker the SASS does audit checks every six months. He testified that although on paper this is what the

SASS should be doing, it does not always happen. This suggests that there may be under­ enforcement by the SASS but there is no evidence of SASS having a difficulty with JJ Plates number plates.

196. To the contrary the SASS letter to the Commission dated 6 March 2014 confirms that there is no legislation or SASS requirement preventing

an embosser from using a different blank plate that differs from its embossing machine. The SASS' accreditation focuses on the output, which is the number plate rather than whether a Uniplate machine was used to make an NNPR plate.[75] Even Naicker testified that to a lay person, it would be difficult to distinguish between a compliant and non­ compliant number

plate.[76]

Covering maintenance services provided as part of Uniplate 's Warranty in the contracts

197. The third efficiency defence was the maintenance services offered by Uniplate. Naicker submitted that exclusivity allowed Uniplate to cover the costs of ongoing maintenance through the sale of its blanks.

198. The Commission rejected this claim, stating that while the maintenance of the machine was customer specific it was not non-contractible. In other words, the maintenance costs could be specified and listed in the contract. Uniplate as Mncube points out would be able to charge embossers for any maintenance that is done and contract on the optimal amount of maintenance.

199. Murgatroyd did not disagree with the Commission's analysis of maintenance being contractible but stated that one cannot disregard

the fact that the exclusivity allows this expense to be cross subsidised through the supply of its blanks. In closing argument

Wesley appeared to abandon reliance on maintenance costs as a stand­ alone justification but said that it formed part of recouping

investment costs.

200. Steenekamp testified that NNPR does not include maintenance costs in its pricing to its customers as a justification for exclusivity and the duration of the agreements. NNPR contracts out of the maintenance and invoices its customers for maintenance costs whenever the customer requires it. Based on the evidence, it does not seem that this has made the machines unaffordable to the customers.

201. Moreover, the warranty clause in Uniplate's contracts applies for six months. Naicker testified that although the warranty period is contractually for six months, in practice Uniplate provides maintenance services for the subsistence of the embosser. In essence, he said the practice was more favourable than what is contained in the agreement.

202. When asked if there was a reason this was not expressly stated in the agreements Naicker said it was to prevent embossers abusing the machine in the knowledge that it will be fixed free of charge.[77] There appears to be no basis for such a claim. No other evidence was put up in support of this contention which seems to be a belated

justification to bolster the rationale for exclusivity.

Conclusion

203. We have found that there is a primary market for the manufacture and supply of embossing machines and a secondary market for the supply of number plate blanks. For purposes of this decision, it is not necessary to further delineate the relevant markets for embossing machines according to the type of embossing machine since the economic experts agreed that this does not affect the analysis of the complaint before us.

204. Since the economic experts also agreed that whichever way the market is defined, Uniplate would be dominant under the Act, it is not necessary for us to determine the issue of dominance.

205. We have also found that Uniplate's exclusive contracts have foreclosed the market to Uniplate's competitors and has raised

barriers to entry.

206. Uniplate's contracts locked in customers, meaning that competitors such as NNPR was unable to access the market. Customers were compelled to exclusively purchase from Uniplate and were threatened with legal action in the event that they did not comply. Our legislation places a special duty on dominant firms, which Uniplate has conceded it is, to not engage in exclusionary conduct without any pro­ competitive justifications.

207. While market shares on their own cannot be relied upon as conclusive evidence of effects, in the context of the other overwhelming

evidence in this case, the market share evidence appears to support the Commission's contentions that rivals were excluded. There are only two effective competitors in the market for blanks, NNPR and Uniplate with Arga trailing behind with a very small market share. There is no question that Uniplate has retained its dominance over the relevant period and NNPR has only been able to grow its market shares between 17 and 21% over the relevant period.

208. Customers were also harmed by Uniplate's exclusive contracts in the form of higher prices and reduced choice.

209. Since this is not a 'systems market', there can be no reason for insisting embossers to exclusively purchase blanks from Uniplate when purchasing a Uniplate embossing machine. Any exclusivity would need to be justified. We are of the view that Uniplate has not sufficiently justified the tying of the blanks and blanking requirements to the embossing machine. In defence, it was suggested

that Uniplate had subsidised the price of machine through the sale of blanks. There was simply no evidence of this.

210. We are therefore of the view that Uniplate has contravened section 8(d)(i) of the Act penalty.

Administrative penalty

211. The Commission submitted that if we find that Uniplate contravened the Act, we should impose a penalty of R23 131 879 (being the maximum statutory penalty based on Uniplate's audited financial statements in 2016). Uniplate on the other hand submitted that the penalty should be no more than R9 905 465.

212. Both the Commission and Uniplate relied on the methodology developed by the Tribunalin Aveng.[78]

213. The relevant provisions of the Act are sections 58 and 59 of the Act. In particular, section 59(3) provides that:

"when considering an administrative penalty, the Competition Tribunal must consider the following factors:

(i) the nature, duration, gravity and extent of the contravention;

(ii) any loss or damage suffered as a result of the contravention;

(iii) the behavior of the respondent;

(iv) the market circumstances in which the contravention took place;

(v) the level of profit delivered from the contravention;

(vi) the degree to which the respondent has co-operated with the Commission and the Competition Tribunal; and

(vii) whether the respondent has previously been found in contravention of this Act."

214. Section 59(2) states: "An administrative penalty... may not exceed 10 percent of the firm's turnover in the Republic and its exports from the Republic during the firm's preceding financial year."

The Commission and Uniplate's submissions on methodology

215. The Aveng decision sets out a six-step approach to determining an appropriate penalty. This methodology together with the parties' submissions are discussed below.

215.1 Step one: determination of the affected turnover in the relevant year of assessment. It is common cause between the Commission and Uniplate

that the relevant financial year for the assessment of the affected turnover is 2014.

215.2 However the parties were in disagreement about what proportion of Uniplate's turnover should be considered to be affected turnover. While the Commission recognised that Uniplate's main business was the supply of number plates and the supply of consumables to sign manufacturers, it was unable to separate the sale of blanks from the other revenue. The Commission therefore used Uniplate's total turnover for this step which was R448 921 860.

215.3 Uniplate submitted that this figure significantly overstated the affected turnover as it included revenue earned from the Falcon business (mainly a signage business) and other small business units, from blanks sold to embossers located from outside of South Africa and from the supply of unrelated products.

215.4 Uniplate submitted that once these revenue streams were deducted from the total revenue of R448 921 860, the affected turnover was R108 871 419. From this figure Uniplate further submitted that the revenue for Type A blanks should be deducted from the affected turnover since Uniplate was not dominant in Type A blanks . In addition, as mentioned, Uniplate sought to also exclude turnover derived from exports.

215.5 In our view, while section 59(2) expressly provides for exports to be included, the evidence suggests that Uniplate's conduct would have been limited to its South African blanking business and is unlikely to have affected exports. This is because number plates are unique to the region in which they are provided and are generally subject to local regulations. We have therefore excluded exports from the affected turnover.

215.6 Regarding Uniplate's submission to exclude Type A blanks because Uniplate is not dominant in this market, we see no basis for this since in terms of section 59(2), the penalty is based on the affected turnover regardless of dominance. Moreover, Uniplate's exclusivity also included type A blanks.

215.7 We have therefore used the figure of R108 871 419 being the affected turnover excluding exports.

215.8 Step two: calculation of the "base amount" being that proportion of the relevant turnover relied upon. This base amount can range between 0-30% depending, inter alia, on the factors set out in section 59(3). The Commission proposed a base amount of 30% in line with its previous submissions in recent abuse of dominance matters. By applying this figure to the affected turnover figure, the Commission arrived at a figure of R134 676 558.

215.9 Uniplate disputed the Commission's claims that the appropriate base amount was consistent with its previous decisions. According to Uniplate this figure should be at the lower end of the scale. In particular, it submitted that applying a figure similar to that applied in Competition Commission vs Giuricich[79] was a more appropriate figure, since this base was used in a cartel case, the most egregious contravention. Applying a figure of 3.5%, Uniplate arrived at a base amount of R3 537 666.[80]

215.10 ln our view Uniplate fails to take into account that in Guiricich although the contravention was a cartel, it involved a once off incident of collusive tendering whereas the exclusivity of Uniplate's contracts was pervasive. We have determined the base amount to be 5% This gives a figure of R5 443 571.

215.11 Step three: where the contravention exceeds one year, multiply the amount contained in step two by the duration of the contravention. Since the period of the complaint is 5 years both the Commission and Uniplate used this multiplier. The Commission arrived at a figure of R673 382 790 (R134 676 558 x 5) and Uniplate at a figure of R14 150 664 (R3 537 666 x4).[81]

215.12 Based on our calculations we arrived at a figure of R27 217 855 (R5 443 571 x 5).

215.13 Step four: rounding off the figure obtained in step three, if it exceeds the cap provided for by section 59(2). Based on Uniplate's total turnover

in its 2016 financial year, both the Commission and Uniplate calculated the 10% statutory cap to be R23 131 879.

215.14 We accept this figure as the statutory cap.

215.15 Step five: considering factors that might mitigate or aggravate the amount reached in step four, by way of a discount or premium expressed as a percentage of that amount that is either subtracted from or added to it. The Commission was of the view that while this was Uniplate's

first contravention of the Competition Act that this mitigating factor matched the aggravating factors and that step five should be neutral.

215.16 Uniplate argued to thecontrary submitting that there were several mitigating factors which warranted a discount. Briefly this included (i) Uniplate did not engage in this conduct to deliberately create or enhance its market power; (ii) Uniplate's conduct benefited embossers; (iii) Uniplate did not earn additional profit from its conduct; (iv) Uniplate cooperated fully in the investigation and prosecution of the complaint, and (v) this was Uniplate's first contravention of the Act. On this basis, Uniplate recommended a discount of 30%.

215.17 In our view, the factors listed by Uniplate are at best neutral. We have decided on a discount of 30%. This gives a figure of R16 192 315 (R23 131 879 - 30%).

215.18 Step six: rounding off the amount in step five if it exceeds the cap provided for in section 59(2). If it does, it must be adjusted downwards so that it does not exceed the cap. After taking the above factors into account, the Commission arrived at an administrative penalty of R23 131 879 which is the statutory cap. Uniplate arrived at a figure of R9 905 465.

215.19 Since the figure we arrive at in paragraph 215.17 above does not exceed the statutory cap it is not necessary to round it off.

215.20 The penalty amount is thus R16 192 315.

ORDER

[1] Uniplate has contravened section 8(d)(i) of the Act in the period 2010-2014.

[2] Uniplate must pay an administrative penalty of R16 192 315.

[3] Uniplate must make payment of the administrative penalty within 90 business days of this order.

[4] There is no order as to costs.

27 June 2019

Date

_____

Ms Mondo Mazwai

Ms Yasmin Carrim and Mr Enver Daniels concurring

Tribunal Researcher: Bustsiwe Masina

Tribunal In-house Economist: Karissa Moothoo Padayachie

For the Commission: Layne Quilliam and Anisa Kessery

For Unlplate: Adv Mark Wesley Instructed by

Chris Charter of Cliffe Dekker Hofmeyr

[1] See Commission's Heads of Argument, para 186.

[2] Dies are used to produce the alphanumeric characters on a number plate. To produce such figures, the blank is placed between the dies 'male' and 'female' components which imprint the registration number onto the number plate. There are differences between the dies used toemboss Type A blanks, Type B blanks and Acrylic blanks - the dies cannot therefore be used interchangeably. See Witness Statement of Steenekamp, paragraph 18.2. Dies are consumables and are worn out with volume usage requiring them to be replaced.

[3] Witness statement of Naicker, pleadings bundle, page 126, paragraph 35.

[4] There was a dispute during the hearing as to whether Arga had exited this market or not, see Trial Bundle page 65-66; Transcript page 87, line 12-22, page 88, page 89, page 90 lines 1-22; Uniplate's Heads of Argument at page 36, paragraph 90.4. However, nothing much turns on this for purposes of this decision. What is clear from the evidence is that Arga last sold embossing machines in 2010 and not in the subsequent period of the complaint period (2011-2014).

[5] An Acrylic or plastic number plate blank consists of an acrylic window and reflective sheeting which includes the provincial emblem.

[6] The main difference between the Type A and B aluminium number plate blanks is that the Type B number plate blank consists of an aluminium base plate that is manufactured from pre-coated aluminium of various colours, which dispenses with the need to paint the number plate manually post­ manufacture. Type A requires the blank to be manually painted.

[7] Firms that only manufacture blanks are colloquially referred to as blankers.

[8] The standards only regulate the manufacture of number plate blanks and the final number plate and do not regulate the manufacture of embossing machines.

[9] Interim audits are conducted on a bi-annual basis by SABS in order to ensure compliance.

[10] The provision was also implemented in the supply agreements of Uniplate's third party distributors, Teqplate and Baleka.

[11] Transcript page 695, lines 18-22 and page 696, lines 1-2.

[12] Transcript page 706, lines 3-7.

[13] Transcript page 304, lines 2-12.

[14] Trial Bundle, pages 171- 172

[15] Witness statement of Naicker, pleadings bundle, page 121, paragraph 8.

[16] Witness statement of Steenekamp, pleadings bundle, page 113, paragraph 35.

[17] Witness statement of Devandran Naicker, page 15, paragraph 69.4.1. and 69.4.2.

[18] Uniplate's Expert Report, page 311 of the pleadings bundle, paragraph 170.

[19] DG Competition Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, 2005, paragraph 247, page 205 of the Economics Authorities Bundle.

[20] Tying describes a situation when a firm sells one product but only on condition that the buyer also purchases a different product. The tying product (the embossing machine in this case) is the product that is sold only if the tied product (the blank) is purchased. As such in substance the anticompetitive effects of tying would be similar to exclusive dealing, especially in instances where the incumbent starts from a dominant position

[21] Competition Commission vs South African Airways (18/CR/Mar01)

[22] South African Airways (Pty) Limited v Comair Limited and Another (92/CAC/Mar10)

[23] Nationwide Airlines and Others v South African Airways (Pty) Ltd and Others [2001] ZACT 1 (5 January 2001).

[24] Competition Commission v Telkom SA Limited (11/CRFeb04) para 99 on page 635 of the Authorities Bundle.

[25] While ownership of the embossing machine transfers Immediately with a cash sale, ownership only transfers once the agreement is paid in full at the end of the term for an instalment sale agreement.

[26] In terms of these agreements, embossers pay a set rental fee each month until the end of the contract, with ownership of the machine remaining with Uniplate throughout the standard term of the agreement which is typically 120 months.

[27] See Table 9 of the Competition Commission Expert Report.

[28] Competition Commission Expert Report, Table 11.

[29] Transcript page 141, lines 7-11.

[30] Transcript page 51,8 lines 4-8.

[31] Witness Statement of Naicker pleadings bundle page 132 paragraph 65.

[32] Transcript page 415, lines 9-18.

[33] Transcript page 558 lines 18-22 and page 559 lines 1-4. See also Transcript page 585 lines 15-21 and page 586 lines 1-12.

[34] See Exhibit D read with Exhibit E.

[35] […]

[36] For example Uniplate Group and Maseke Business Enterprise which contains the following cancellation clause: "A RS000.00 deposit will be paid if this contract is cancelled before 2 years. RS000.00 will not be refundable. After 2 years the embosser will be refunded the deposit.· See Trial bundle page 1788 clause 24.

[37] Transcript page 537, lines 1-11.

[38] Transcript page 538, lines 1-9.

[39] Trial Bundle, page 1039.

[40] Witness statement of Devandran Naicker paragraph 61 on page 131 of the pleadings bundle. See also Transcript page 650-651.

[41] Transcript page 1210, lines 5-15.

[42] Transcript page 1009, lines 11-22; Transcript page 1010, lines 1-7; Transcript page 1011, lines 1-22; and Transcript page 1012, lines 1-2.

[43] Transcript page 103, lines 11-17.

[44] Transcript page 1015, lines 6-16.

[45] See Steenkamp's factual witness statement paragraphs 51.2 and 51.3.

[46] Transcript page 223 lines 15- 21 and page 224 lines 1-14.

[47] Transcript page 83 lines 9-15.

[48] Transcript page 86, lines 1-13.

[49] Transcript page 86 lines 1-13.

[50] Witness statement of de Lange, paragraph 17, page 97 of the pleadings bundle. Transcript page 282 lines 6-17.

[51] Witness statement of de Lange, paragraph 19, page 98 of the pleadings bundle.

[52] See de Lange's Witness Statement paragraph 19, page 98 of the pleadings bundle.

[53] Witness statement of de Lange, paragraph 20-21, page 98 of the pleadings bundle.

[54] Trial bundle page 154-155.

[55] Trial bundle page 154-155.

[56] See also Transcript page 668.

[57] Witness statement of Steenekamp, pleadings bundle page 117, paragraph 49.

[58] Transcript page 667, lines 6-18.

[59] Trial bundle page 152-153.

[60] […]

[61] Whish & Bailey, Competition Law, B'h ed. (Oxford University Press, Oxford 2015) page 754; Case T- 111/96 ITT Promedia NV v Commission [1998] 11-2937; Competition Commission v. Telkom Case No. 11/CR/Feb04 (decision 7 August 2012).

[62] See Exhibit F Letter dated 1 February 2011.

[63] Witness Statement of Steenekamp, pleadings bundle page 114-115, paragraphs 40 1, page 114- 115 of the pleadings bundle.

[64] Letter by Utal dated 28 September 2016, Trial bundle page 775.

[65] Witness Statement of Devandran Naicker paragraph 50.2, page 129 of the pleadings bundle.

[66] Trial bundle, page 1.

[67] Trial Bundle, page 168.

[68] Transcript page 737, lines 9-21 and page 738 lines 1-4.

[69] Transcript page 744 lines 5-15.

[70] Transcript page 934 lines 1-8.

[71] […]

[72] Transcript page 547 lines 13-21 and page 138 lines 1-6.

[73] Transcript page 150 line 3.

[74] Transcript page 148 lines 4-22 and page 149 lines 1-20.

[75] Trial Bundle page 171-172

[76] Transcript page 460, line 22 and page 461, lines 1-3.

[77] Transcript page 515, lines 5-18.

[78] Competition Commission vs Aveng (Africa) Limited t/a Steeledale and others (84/CR/Dec09).

[79] Competition Commission vs Giuricich Coastal Projects (Pty) Ltd and Another (CR162Dec14)

[80] Uniplate's calculation was based on its view that the affected turnover excluded revenue earned from the Falcon business (mainly a signage business) and other small business units, exports and Type Its affected turnover figure was R101 076180.

[81] This was calculated in Uniplate's Heads of Argument on a period on 4 years instead of 5 years which Uniplate ultimately accepted was the correct duration.