Twin City Trading 2 (Pty) Ltd v Castle Gate (Pty) Ltd and Another (LM004Apr23) [2023] ZACT 28; [2023] 3 CPLR 45 (CT) (17 July 2023)

COMPETITION TRIBUNAL OF SOUTH AFRICA

Case No: LM004Apr23

In the matter between:

Twin City Trading 2 (Pty) Ltd Acquiring Firm

and

The commercial letting enterprise conducted by Castle Gate (Pty) Ltd

and the commercial letting enterprise conducted by Club Retail Park (Pty) Ltd Target Firms

Panel: Jerome Wilson (Presiding Member)

Thando Vilakazi (Tribunal Member)

Andreas Wessels (Tribunal Member)

Heard on: 26

June 2023

Order issued on: 26 June 2023

Reasons issued on: 17 July 2023

REASONS FOR DECISION

Approval

[1] On 26 June 2023, the Competition Tribunal (“Tribunal”) unconditionally approved the large merger in which Twin City Trading 2 (Pty) Ltd (“Twin City Trading”) will acquire an undivided 30% share interest in two rental enterprises known as Castle Gate Property and Club Retail Property (the “Target Properties”) from Castle Gate (Pty) Ltd and Club Retail Park (Pty) Ltd, respectively.

Parties to the transaction and their activities

Primary acquiring firm

[2] Twin City Trading is controlled by Twin City Development (Pty) Ltd (“Twin City Development”) which is, in turn, controlled by the L’MAE Trust.

[3] Twin City Trading, and all the firms that control it directly or indirectly, are referred to collectively below as the “Acquiring Group”.

[4] The Acquiring Group owns properties in various categories (residential, commercial, and agricultural) in South Africa and internationally. Relevant to the present transaction is the Acquiring Group’s property portfolio in Pretoria, which is comprised of office property and retail property.

Primary target firms

[5] Castle Gate Property is wholly owned by Castle Gate (Pty) Ltd (“Castle Gate”), and Club Retail Property is wholly owned by Club Retail (Pty) Ltd (“Club Retail”).

[6] Club Retail is wholly owned by Atterbury Property Fund (Pty) Ltd (“Atterbury Property Fund”), while Castle Gate is controlled by Atterbury Property Fund with a shareholding of 65%[1].

[7] Atterbury Property Fund is wholly controlled by Atterbury Property (Pty) Ltd (“Atterbury Property”). Atterbury Property is in turn, controlled by Atterbury Property Holdings (Pty) Ltd (“Atterbury Property Holdings”), with a shareholding of 75.35%. Of the remaining shares in Atterbury Property, a total of 10.06% is held by Talis Investment Partners (Pty) Ltd and Talis Property Fund (Pty) Ltd (referred to together below as “Talis”). Both of the Talis companies are 100% owned by Mr. Tebogo Mogashoa. Atterbury Property Holdings is, in turn, controlled by Atterbury Manfou (Pty) Ltd, which does not have any (direct or indirect) controlling shareholders.

[8] The Target Properties comprise rentable retail, hospitality and office space in Pretoria. The Castle Gate Property has a Gross Lettable Area (“GLA”) of 21 800 m² for retail property and a GLA of 3 200 m² for office property. The Club Retail Property has a GLA of 16 784 m² for retail property and a GLA of 12 698 m² for office property.

Proposed transaction

[9] The proposed transaction involves various different agreements that are intended to be implemented simultaneously, including:

(i) an agreement between Twin City Trading and Castle Gate in terms of which Twin City Trading will acquire a 30% undivided share in the rental enterprise conducted as a going concern from the Castle Gate Property (with Castle Gate retaining the remaining 70% undivided share therein); and

(ii) an agreement between Twin City Trading and Club Retail in terms of which Twin City Trading will acquire (i) certain sectional title units equal to 30% of the sectional title units owned by Club Retail, together with the rental enterprise carried on as a going concern by Club Retail therefrom and (ii) a 30% undivided share in the full title properties and exclusive use areas owned by Club Retail together with the rental enterprise carried on as a going concern therefrom (with Club Retail retaining the remaining 70% of the sectional title units, and the remaining 70% undivided share, respectively).

[10] In terms of a co-ownership agreement between the merging parties, Twin City Trading will acquire “negative control” of the Target Properties for purposes of section 12(2)(g) of the Competition Act, 89 of 1998, by virtue of veto rights over (i) the adoption of important policies within the business; (ii) material strategic decisions such as the adoption of budgets and business plans; (iii) major investments; and (iv) the appointment and removal of directors and senior management.

Indivisibility assessment

[11] The Competition Commission (the “Commission”) considered whether, in light of the fact that Twin City Trading intends to acquire two Target Properties in terms of different agreements, the proposed transactions should be considered together or separately for merger control purposes.

[12] The Commission concluded that the different agreements should be regarded as reflecting a single indivisible transaction on the basis that the target firms are controlled by the same shareholder, namely Atterbury Property; the Target Properties are all involved in the letting of retail property and office space; and the different agreements are interlinked and will be implemented simultaneously.

[13] Having regard to the above factors, we agree with the approach followed by the Commission in this regard.

Rationale

[14] The Acquiring Group submitted that the Target Properties present a low-risk investment opportunity that will generate an attractive yield or return on investment in the event that the rental enterprises are sold in the future.

[15] Castle Gate and Club Retail submitted that the proposed transaction is an opportunity to partially realise a return on its investment in the Target Properties whilst still retaining the majority stake therein, and thereby allowing them the opportunity to share in the future revenue generated by the Target Properties, and in the return on investment thereon in the event that the rental enterprises are sold in the future.

Competition assessment

[16] The Commission considered the activities of the merging parties and found that they are both active in the provision of lettable office space and lettable retail property.

Office space

[17] As regards lettable office space, the Commission followed its approach in Redefine/Leaf[2] where the Tribunal accepted that lettable office space can be divided into different grades, namely grades P[3], A[4], B[5] and C[6].

[18] In this regard, the Commission found that the merging parties offer different grades of lettable office space – the Acquiring Group offers grade A lettable office space, while the Target Properties offer grade P lettable office space. In addition, the respective office spaces are in different nodes, with the Acquiring Group’s office space being situated in the Waterkloof node, while the Target Properties are situated in the Menlyn and Hazelwood nodes. The Commission therefore concluded that there is no product or geographic overlap for lettable office space amongst the merging parties.

Retail property

[19] As regards the provision of lettable retail property, the Commission followed the classification of retail properties according to the Investment Property Databank Index (“IPD”)—

Table 1: Classification of Retail Property

Retail Property Type Definition (by GLA range) Small Freestanding Centres 500m² – 1,000m² Local Convenience Centres 1000m² – 5000m² Neighbourhood Shopping Centres 5 000m² – 12 000m² Community Shopping Centres 12 000m² – 25 000m² Small Regional Shopping Centres 25 000m² – 50 000m² Regional Shopping Centres 50 000m² – 100 000m² Super Regional Shopping Centres More than 100 000m²

Source: Investment Property Databank South Africa (Pty) Ltd

[20] Based on this classification, the Commission found that the merging parties’ activities overlap in the narrow market for lettable

retail space in small regional shopping centres. The Commission found that the Acquiring Group owns a small regional shopping centre with a GLA of 48 000 m² whilst the Target Properties comprise[7] a small regional shopping centre with a total GLA of 38 584 m².

[21] As regards the relevant geographic market, the Commission evaluated what it referred to as “comparative centres” within a 15km radius of the Target Properties.

[22] The Commission found that, according to the Shopping Centre Directory (2019), there are approximately 18 shopping centres within a 15km radius of the Target Properties with a combined GLA of 726 785 m².

[23] The Commission concluded that the merged entity will have a post-merger market share of approximately 12%, with an accretion of approximately 5%, within a 15km radius of the Target Properties.

[24] The Commission also contacted various tenants of the Target Properties to obtain their views on the transaction. These tenants did not express any concerns regarding the transaction, and they indicated that there were various alternative retail properties available to them in reasonably close proximity to the Target Properties.

[25] Based on its findings that the merged entity would have a low post-merger market share, with a small accretion, and that there were various alternatives in close proximity to the Target Properties, the Commission concluded that the proposed transaction is unlikely to result in a substantial lessening or prevention of competition in the market for lettable retail property.

The Tribunal’s assessment

[26] It was not clear to the Tribunal on what basis the Commission had calculated the market shares of the merging parties in the market for lettable retail property, and, in particular, whether that calculation included only small regional shopping centres or also other categories of shopping centres according to the IPD retail property classification. The Tribunal therefore requested the Commission to provide further information on the properties included in its market definition, including their distance from the Target Properties, size and IPD classifications.

[27] In its response, the Commission clarified that it regarded comparative centres for purposes of its analysis as comprising not only small regional shopping centres, but also community shopping centres, small regional shopping centres, regional shopping centres and super-regional shopping centres. The Commission relied in this regard on the Tribunal’s decision in Hyprop/Attfund[8]. The Commission also identified an error in its market share calculation, which reduced the combined market share of the merging parties (on its definition) to approximately 8%, with an accretion of less than 4%.

[28] In our view, the Tribunal’s decision in Hyprop/Attfund does not constitute authority for the approach adopted by the Commission in this case. It cannot be assumed, without any further analysis, that all the different categories of shopping centres within a 15km radius of the target firm in a merger constitute effective competitive constraints upon the target firm. More particularly in this case, it cannot simply be assumed that all the different categories of shopping centres included in the Commission’s market share calculations constitute effective constraints on the Target Properties. Merely by way of example, there does not appear to be any a priori basis to conclude that a super- regional shopping centre such as Menlyn Mall constitutes an effective competitive constraint on the small regional shopping centres comprising the Target Properties.

[29] In order to reach such a conclusion, the Commission would have to conduct a substantive competition analysis of all relevant factors, including (for example) the size, characteristics, rentals and proximity of each of the shopping centres in question, and any evidence of switching between them. Each case must ultimately be determined on its own facts.

[30] However, based on the information available in this case, it does not appear to us that the merger gives rise to competition concerns irrespective of the (product or geographic) market definition that is applied. Based on the information provided by the Commission, if super-regional shopping centres are excluded from the market, the Acquiring Group would have a post-merger market share of approximately 10%, with a market accretion of less than 5%. There would be at least 16 shopping centres located within a 15km radius of the Target Properties, including four shopping centres located less than 4km away.

[31] If the product market is defined even more narrowly to include only small regional shopping centres, the merged entity would have a market share of just under 20%, with at least 10 alternative small regional shopping centres within a 15km radius of the Target Properties, and two shopping centres located less than 5km away.

[32] In the particular circumstances of this case, therefore, we agree with the Commission’s ultimate conclusion that the proposed transaction is unlikely to raise any negative horizontal effects in the market for lettable retail property, having regard to the low market share of the merged entity, and number of alternatives in close proximity to the Target Properties irrespective of the market definition applied. We therefore reach this conclusion without needing to make a determination on the scope of the relevant product or geographic market in this case. The dimensions of the relevant market may however require more detailed investigation by the Commission in future retail property mergers.

Public interest

Employment

[33] The merging parties submitted that the proposed transaction will not have any negative impact on employment as they have specifically undertaken that there will be no job losses or retrenchments resulting from the proposed transaction.

[34] The Commission also engaged with the employee representatives of the Acquiring Group and the Target Properties. Neither employee representative raised any concerns regarding the proposed transaction, and the representative of the Target Properties confirmed that the Target Properties have no employees.

[35] The Commission also found that the management and administration of the Target Properties is currently outsourced to Atterbury Management Services (Pty) Ltd (“Atterbury Management Services”), and that this will remain the case post-merger, without any job losses at Atterbury Management Services.

[36] Based on the above, the Commission concluded that the proposed transaction is unlikely to have a negative impact on employment. We agree with this conclusion.

Spread of ownership

[37] As regards spread of ownership, the Commission found that, while neither Castle Gate nor Club Retail has any direct shareholding by historically disadvantaged persons (“HDPs”), Atterbury Property has an HDP shareholder, Mr Mogashoa, who (as set out above) indirectly owns 10% of the shares in Atterbury Property through Talis.

[38] Accordingly, the Commission found that, on a flow-through basis, a 10% HDP shareholding can be attributed to the Club Retail Property and a 6.5% HDP shareholding can be attributed to the Castle Gate Property.

[39] The Acquiring Group, on the other hand, does not have any HDP shareholders. Therefore, given that Twin City Trading will acquire a 30% interest in each of the Target Properties, the Commission found that the HDP shareholding in Club Retail property will, on a flow-through basis, reduce to 7% from a pre-merger HDP shareholding of 10%, whilst the HDP shareholding in the Castle Gate Property will decrease from 6.5% to 4.55%.

[40] Given this dilution in HDP ownership, the Commission requested the merging parties to remedy the dilution by way of an employee share ownership plan or HDP transaction.

[41] The merging parties responded that conditions of the sort requested by the Commission were not warranted in this case. They submitted that the HDP dilution that would be brought about by the proposed transaction is not substantial. They also pointed out that the assets held by Atterbury Property post-merger will remain unchanged, as will the HDP shareholding of 10% in Atterbury Property. As such, the HDP shareholder in Atterbury Property will benefit from the funds generated by the proposed transaction, and the forecasted growth of the Target Properties post-merger.

[42] The Commission also found that the purchase price for the proposed transaction constituted only 5% of the overall asset portfolio of Atterbury Property, and that the HDP shareholder in Atterbury Property supported the proposed transaction for the reasons outlined above.

[43] The Commission further noted that only 30% of the Target Properties is being sold and that Atterbury Property will continue to have an indirect interest of 70% in each of the Target Properties post-merger.

[44] On this basis, the Commission concluded that conditions were not warranted to address the HDP dilution in this case. Having regard to the facts in this particular transaction, the Tribunal agrees with this conclusion.

Conclusion

[45] For the reasons set out above, the Tribunal concludes that the proposed transaction is unlikely to substantially prevent or lessen competition in any relevant market and does not raise any substantial public interest concerns. We therefore approve the proposed transaction without conditions.

Date: 17July 2023

Jerome Wilson SC

Concurring: Dr Thando Vilakazi and Mr Andreas Wessels

Tribunal case manager:

Leila Raffee

For the merging parties:

Oscar Kolbé of Weavind & Weavind

For the Commission:

Yolanda Okharedia and Themba Mahlangu

[1] The remaining 35% shareholding in Castle Gate is owned by the Carl Jacobus Erasmus Trust (a non-controlling shareholder).

[2] Redefine Properties Limited/ Leaf Property Fund (Pty) Ltd (Tribunal Case Number: LM195Feb15)

[3] Grade P refers to top quality, modern space which is generally a pace-setter in establishing rentals and includes the latest or a recent generation of building services, ample parking, a high quality lobby finish and good views or a good environment.

[4] Grade A refers to offices which are generally not older than 15 years or which have undergone major renovations, and have high quality modern finishes, air conditioning, adequate on-site parking, and market rentals near the top of the range in the metropolitan area in which the building is located.

[5] Grade B refers to offices that are generally older buildings, but which have accommodation and finishes that are close to modern standards as a result of refurbishment and renovation from time to time, and have air conditioning and on-site parking (unless special circumstances apply).

[6] Grade C refers to buildings with older style finishes, services and building systems. They may or may not have air conditioning or have on-site parking.

[7] The Commission assessed the Target Properties as a whole. The total GLA 38 584 m² is comprised of the Castle Gate Property constituting a GLA of 21 800 m² and the Club Retail Property constituting a GLA of 16 784 m².

[8] Hyprop Investments Limited/ Attfund Retail Limited (Tribunal Case Number: LM092Jan11).