Fuel Retailers Association v Minister Of Energy and Others (28818/2014) [2023] ZAGPJHC 1067; [2023] 4 All SA 739 (GJ) (22 September 2023)

FLYNOTES: ADMINISTRATIVE – Regulatory accounting system – Adoption and implementation – Lawfulness –Wholesalers may not indirectly operate retail site or exercise substantial control over it – Implemented RAS allows wholesalers and oil companies to do exactly that – Rationale underlying prohibition on vertical integration is undermined – Object of government’s policy is prevented from being achieved by manner of implementation –

RAS at odds with legislative regime and amenable to review – Petroleum Products Act 120 of 1977, s 2A(5)(a).

REPUBLIC OF SOUTH AFRICA

IN THE HIGH COURT OF SOUTH AFRICA

GAUTENG DIVISION, JOHANNESBURG

Case No: 28818/2014

REPORTABLE

OF INTEREST TO OTHER JUDGES

22.09.23

In the matter between:

THE FUEL RETAILERS’ ASSOCIATION Applicant

and

THE MINISTER OF ENERGY First Respondent

THE CONTROLLER OF PETROLEUM PRODUCTS Second Respondent

THE SOUTH AFRICAN PETROLEUM INDUSTRY

ASSOCIATION Third Respondent

PETROSA (SOC) LTD Fourth Respondent

THE RETAIL MOTOR INDUSTRY ORGANISATION Fifth Respondent

AMISTEC (PTY) LTD T/A LIQUID FUELS WHOLESALERS Sixth Respondent

PETROLEUM RETAILERS ALIGNMENT FORUM Seventh Respondent

ROYALE ENERGY

Eighth Respondent

NATIONAL ENERGY REGULATOR OF SOUTH AFRICA Ninth Respondent

Coram: Ingrid Opperman J

Heard: 27 and 28 October 2023

Delivered: This judgment was handed down electronically by circulation to the parties’ legal representatives by email. The date and time for hand-down is deemed to be 14h00 on 22 September 2023

Summary: The Fuel Retailers Association applied to review and set aside a decision to adopt and implement an accounting system for the Petroleum Sector (the RAS) – the RAS does not make provision in its retail margin for Entrepreneurial Compensation (EC) for retailers in Company Owned and Retailer Operated (CORO) sites. The Department of Mineral Resources and Energy (the DoE) contended that the EC should be negotiated from the CAPEX portion of the retail margin by the oil companies forfeiting a portion of the CAPEX. SAPIA contended that the EC should be negotiated from the notional EC published in the BSS Matrix annually which notional EC is carved out from the CAPEX - the court held that the retailers are as of right entitled to EC and should not have to negotiate with the oil companies to ‘forfeit’ a part of the CAPEX -– court held further that a decision which requires the parties to negotiate is not per se irrational but a decision which requires parties to negotiate where i) the negotiating power is unequal; ii) the default position is that the oil company (the more powerful contracting party) must relinquish something; and iii) no guidelines or factors have been put into place to consider during this bargaining process, is irrational – court found that the model is fundamentally flawed for CORO sites as the RAS Matrix’s Benchmark Service Station is a Retailer Owned Retailer Operated (RORO) site – majority of service stations in SA are CORO sites – Benchmark Service Station should cater for this reality.

Order

1. Condonation for the applicant’s failure to bring this review application within 180 days of the original decision referred to paragraph 2 hereof alternatively within a reasonable time, is granted.

2. The original decision of the first respondent or her delegates [the Minister], taken in November 2013, to implement the RAS without providing for a ring-fenced Entrepreneurial Compensation (EC) to be claimed exclusively by the retailers in Company Owned Retailer Operated (CORO) sites and/or specifying the items to be claimed under the EC by retailers in CORO sites, is reviewed and set aside.

3. The determination of the treatment and calculation of the EC for retailers in CORO sites as an allocation within the retail margin of the RAS (the determination) is referred back to the first respondent [the Minister] to decide in accordance with this Court’s judgment, within a period of 9 months from the date of this order.

4. Pending the determination, the 2020 RAS Benchmark Service Station Matrix and any subsequently issued (or yet to be issued) Matrices are to remain in force and effect.

5. Pending the determination, the status quo of the outcome of each CORO site fuel retailer’s negotiation with its fuel supplier/landlord is to be maintained and all new agreements still to be concluded between CORO site retailer’s and fuel suppliers/landlords are to be on the basis that the Minister’s decision has not been reviewed or set aside.

6. The first, second and third respondents are to pay the costs of this application, jointly and severally, the one paying the other to be absolved such costs to include the costs of two counsel where so employed.

JUDGMENT

INGRID OPPERMAN J

Introduction

[1] The Applicant (the Fuel Retailers Association) applies to review and set aside the Minister of Energy’s (the Minister) decision to adopt and implement the Regulatory Accounting System for the Petroleum Sector (the RAS) to the extent that it establishes a retail margin for the activity of selling petrol to end-users without providing a trading margin, also referred to as an Entrepreneurial Compensation (EC), for retailers to compensate them for the dispensing services they provide. The Minister and the Controller of Petroleum Products (the Controller) - the Second Respondent who also holds the position of Deputy Director-General in the Department of Mineral Resources and Energy – (the DoE) and the South African Petroleum Industry Association (SAPIA), the Third Respondent, oppose the application.

[2] The Fuel Retailers Association represents the small business-owners and entrepreneurs within the retail sector who operate fuel service stations across the country.

Key activities in the fuel supply chain

[3] The procurement of crude oil in South Africa is not regulated, nor is the refining of crude oil. However, government, through the DoE, amongst others, regulates the price of petrol at service stations. In this manner, government determines the allowable returns on investments in all petrol activities once it leaves the refinery gate.

[4] The key activities in the fuel supply chain are: (a) refining which represents the large and established oil companies; (b) primary storage, which is the storage of fuel within the refinery boundaries; (c) wholesale operations which include the primary distribution and marketing of petroleum products; (d) secondary storage facilities which is the storage of petroleum products at storage facilities throughout the country, (refined petroleum products are transported from refineries to storage facilities in bulk); (e) secondary distribution, which is the distribution of fuel from secondary storage facilities to points of sale; (f) retailing which is mostly the sale of petrol and diesel through a service station network.

[5] This application concerns, in the main, this last stage i.e. retailing. Retailing is a licensed activity in South Africa. Retail stations typically also offer other petroleum products like paraffin, lubricants and other related services such as a convenience store and car wash. Retail stations may be both owned and operated by a retailer (a retailer owned retailer operator or RORO site) but are far more commonly owned by one of the oil companies and operated by a retailer (a company owned retailer operated or CORO site). The distinction is important because the retailer of a CORO site has additional costs that a RORO retailer does not have, including those associated with the lease and franchise agreements concluded with the oil company that owns or holds a long lease over the site.

Background

[6] Prior to December 2011, the fuel industry was regulated based on the Marketing of Petroleum Activities Return (‘MPAR’) system and the guidelines thereto. The MPAR system employed a methodology that permitted retailers to claim operating expenses and EC (entrepreneurial compensation) to compensate them for the service they provide through the fuel retailing business. There were several difficulties associated with the MPAR system. This then led to the appointment of Bates White, a company based in Washington DC in the USA, at the behest of the DoE to undertake a review of the petrol price accounting system.

[7] To implement certain of the Bates White recommendations, the DoE (Department of Mineral Resources and Energy) appointed the Institute for Petroleum and Research (‘the IPSR’) to develop recommendations on the RAS (Regulatory Accounting System for the Petroleum Sector) to enable the Minister to set the activity-based margins for each level on an annual basis. The report of the IPSR (Institute for Petroleum and Research) recognised that company owned retailer operated or CORO sites are the majority in South Africa – both in number and in terms of volume of petrol sold. The ‘company’ in this context is the oil company. Despite this, the report based its Benchmark Service Station (BSS) on a retailer owned retailer operator or RORO site.

[8] The DoE engaged in informal consultation with stakeholders between 12 April 2011 and 21 November 2013. The DoE established a RAS technical Committee chaired by DoE representatives comprising representatives of most stakeholders including retailers.

[9] Between the introduction of RAS in December 2011 and its final implementation in December 2013, stakeholders were afforded the opportunity to align their contracts with the margins set out in the RAS Matrix which would come into effect in December 2013.

[10] Since no agreement on the need for entrepreneurial compensation or EC had been reached by December 2011, it was agreed in the RAS Technical Committee that during the transitional stage (the so-called Rapid RAS stage), the RAS would provide for entrepreneurial compensation or EC by carving out a portion of the small stock premium and marketability adjustment margins. Throughout this transitional period, an amount of 17,9 cents per litre had been identified and allocated in the retail margin of the RAS as an entrepreneurial compensation or EC amount. This arrangement was agreed to, to ensure that entrepreneurial compensation or EC could be recovered, even though none had expressly been provided for in the RAS model.

[11] In December 2013, the Minister decided to finally approve the RAS.

The Litigation History

[12] In August of 2014, the Fuel Retailers Association launched the current application in order to obtain clarity on whether the EC was accommodated within the RAS and to compel the Minister to undertake the consultation process required to properly regulate the EC and setting aside the Minister’s failure to decide whether to prohibit any business practice, method of trading, agreement, arrangement, scheme or understanding which has the effect of allowing oil companies to recover from the EC allocation of the fuel retail margin. The debate at that stage, as the Fuel Retailers Association understood it, was about the purpose and proper implementation of the retail margin and how the EC was to be calculated and claimed by the retailers.

[13] In answer to that application the DoE contended that the RAS does not provide for an EC allocation at all and that the RAS structure leaves it to the investor (that is, the owner of the asset at retail level – which is usually an oil company) and the retailer to negotiate what proportion of the investor’s return on investment can be forfeited to the retailer. The DoE explained that the EC was only included in the retail margin for the Rapid Ras stage to align commercial agreements before the full RAS implementation in December 2013. This resulted in the need for the relief sought in prayer 1 of the notice of motion falling away. Thereafter, the stakeholders embarked on a further consultation process which resulted in the relief sought in prayer 2 of the notice of motion also falling away. After some interlocutory skirmishes, the notice of motion was amended and currently the Fuel Retailers Association is seeking the substantive relief formulated in paragraphs 5 and 7 of the amended notice of motion only, such relief being:

‘5. Reviewing and setting aside the original decision of the first respondent or her delegates [the Minister], taken in November 2013, to implement the RAS without providing for a ring-fenced EC to be claimed exclusively by the retailers and/or specifying the items to be claimed under the EC by retailers;

……..

7. Directing that the determination of the treatment and calculation of the EC as an allocation within the retail margin of the RAS is referred back to the first respondent [the Minister] to decide, in accordance with this Court’s judgment, within a period of three months from the date of this order;’

[14] It was made plain that the Fuel Retailers Association were not pursuing a freedom of trade case and were thus not persisting in seeking the relief in paragraph 9 of the amended notice of motion i.e. that the implementation of the RAS without an allocated and ring-fenced EC (or trading margin) to be claimed exclusively by the retailers is a breach of section 9(3) of the Constitution and thus unlawful and invalid. Condonation insofar it was necessary for the failure to have brought the review within 180 days of the Minister’s decision alternatively a reasonable time, was not opposed. I will endorse the agreement relating to condonation.

The RAS

[15] The RAS is a model and methodology that stipulates the ‘margin’ (in cents per litre) that accrues to each identified activity along the retail petroleum value chain. Regulated firms may not increase their prices in order to achieve more than this allowable profit margin. In this way, the Government not only controls and determines the end-price of petrol (which is directly regulated through the publication of a regulated pump price), but also the profit that can be earned by each sector within the petroleum retailing industry.

[16] The RAS is contained in three documents: (1) the Regulatory Margin Model Guidelines (the Guidelines), (2) the Benchmark Service Station RAS Matrix (the RAS Matrix) and (3) the 15 principles governing the RAS (the Principles).

[17] The Guidelines set out the accounting principles, policies, methods and definitions that stakeholders must use when completing their regulated returns for submission to the DoE. The Controller contends that the RAS was introduced to participants in the fuel industry through the distribution of the Guidelines.

[18] Each year, the DoE publishes the RAS Matrix which sets out the composition of the retail margin in real terms (i.e. cents per litre).

[19] The DoE has described the objects of RAS as i) eliminating all possible cross-subsidisation between products and between the different activities in the value chain of fuel supply and ii) promoting the investments in assets in the said value chain by ring-fencing these activities, applying an appropriate return on investments, and providing for the recovery of allowable operating costs.

[20] The 15 RAS Principles are quoted in full and are:

‘1. Economic principle: The RAS is based on petrol only and therefore eliminates cross-subsidisation between fuels and market channels.

2. Regulatory Rule. The RAS retail margin is based on a RORO site and the full retail·margin will accrue to the service station operators. The allocation of the retail margin between operators and investors is not regulated.

3. The value of Opex and Capex of a BSS will be added into one line item, "Retail margin" in the petrol price structures.

4. Regulatory principle: Secondary Storage and Secondary Distribution will be indicated as separate line items in the petrol price structures.

5. The DoE will have the prerogative to request financial and other data in line with the provisions of the Regulatory Accounts Margin Guidelines from any licensed wholesaler with the DoE and any secondary storage owner licensed with the National Energy Regulator

of South Africa in order to determine RAS margins.

6. RAS should be relevant, robust and transparent;

7. RAS should promote efficiency through benchmarking;

8. RAS should promote investment and growth through greater regulatory certainty;

9. RAS should minimise costs and market distortions;

10. RAS should be clear, simple and practical for users, minimizing regulatory burden.

11. RAS methodology should be consistent, transparent and provide a fair remuneration that promotes sustainable operations and investment

across the activities.

12. Commercial arrangements between asset owners and asset operators are not regulated from a price perspective.

13. RAS should harmonise the methodology for calculating margins with the methodology utilised by NERSA to calculate tariffs for petroleum pipelines and storage facilities.

14. Wholesale margin will be indicated as a separate line item in the petrol price structures.

15. These principles are intended to provide insight and guidance into the RAS methodology. They do not represent regulation and do not replace or supersede the various Acts and Regulations governing the lndustry.

[21] The RAS margins, together with several other regulatory costs, inform the Minister’s determination of the pump price of petrol for the end consumer.

[22] As mentioned, the RAS stipulates a margin for each activity. This application is concerned only with the margin allocated to the retailing sector within the petroleum industry. The retail margin includes a return on the capital for the owner of the assets portion (CAPEX) through which they can achieve a return on their investment or capital expenditure in the assets required along the petroleum value chain (tanks, storage units, pumps, land and the like), and an operating costs portion (OPEX) which is a straight cost-recovery facet to compensate parties for the costs involved in providing the services required along the petroleum value chain. The OPEX recovery is not a profit-making margin. It covers costs such as labour and overheads incurred by the service station operator (the retailer) in distributing petrol without adding a return or margin.

[23] The RAS does not differentiate between RORO and CORO sites but the margins it provides for are based on and benchmarked against a RORO site.

[24] In theory, both the OPEX recovery and CAPEX return will be recovered by a retailer in a RORO site. In a CORO site, the retailer will recover the OPEX and the oil company the CAPEX return.

[25] Business practices dictate that an entity would not retail for cost recovery only. In the normal course, a company or entity is rewarded a profit commensurate with the risk involved in that business.

[26] The retailers’ margin on petrol can never be more than the difference between the cost of the petrol after all other stakeholders have taken their margins, and the regulated pump price. Retailers cannot increase their margin by increasing the price of the petrol that they sell because the consumer petrol price is set by regulation. The fuel retailers consequently have an extremely limited scope to determine the trading margin (and subsequent profit) earned per litre of petrol sold at their fuel station.

What is an EC?

[27] An entrepreneurial compensation allocation (or trading margin) is a portion of the retail margin, over and above the OPEX recovery, that would be allocated to retailers to compensate them for the provision of services associated with operating service stations. They ensure a return on a retailer’s risk and investment in providing services.

[28] The Fuel Retailers Association argues that the RAS makes no provision for an EC for parties that do not own any assets, and whose business is purely a service-business. As stated, most fuel retailers in the country fall into this category: they do not own the assets required to dispense fuel – they lease them from wholesale oil companies under franchise and lease agreements. Under the RAS, these asset-less retailers can retain only the OPEX portion of the retail margin which covers their operating expenses.

[29] There appears to be a disconnect between the formal RAS methodology recorded in the documents and the practical way the DoE expects the model to operate.

[30] The RAS methodology provides only for a CAPEX and OPEX portion in the retail margin. The EC does not exist as an element of the retail margin at all. The RAS Matrix consequently does not include any allocation that recognises and compensates retailers for the investment they have made in the business of operating a retail station. In practice, however, the DoE accommodates the EC within the CAPEX portion of the retail margin. It describes this as a ‘split’ of the CAPEX margin into the EC and the margin for the investor in the service station assets. Each year, the DoE publishes the split in the full return on CAPEX in the RAS Matrix.

[31] The DoE published a 2014, 2015, 2016, 2017 and 2018 RAS Matrix. Each of these Matrices includes an EC referred to herein as the notional EC.

[32] With reference to the 2015 RAS Matrix, the notional EC can be identified as follows: The total retail margin is 161.7 cents per litre. This comprises of i) the line items constituting the OPEX margin contributions with a total of 90.3 cents per litre; ii) the line items constituting the CAPEX with a total of 48 cents per litre. In the final sum, the CAPEX portion is recorded as 71.4 cents per litre. This is because the notional EC return of 23.4 cents per litre is included in the CAPEX portion of the retail margin to comprise the total CAPEX margin distribution.

[33] In short: a portion of the CAPEX is carved out and is labelled ‘EC’. This is reflected in the RAS Matrix published by the DoE. This is a recommendation/guideline published by the DoE but one which appears to be adhered to by all stakeholders. It would also seem that this notional EC was brought across from the transitional stage (the Rapid RAS stage).

[34] The DoE has left it to the retailers and the oil companies to negotiate as there are different business models in the petroleum space. The Controller explained this in October 2013 as follows: ‘…..Both parties would know what is due to them when they engage in discussions to align the agreement with RAS as expected…..’.

[35] The issue in part is whether the EC is to be left to the parties to negotiate or whether the Minister should regulate how much is to go to either party or to, at the very least, provide the factors which ought to be considered when negotiating this split.

[36] In sum:

36.1 The RAS does not provide for an EC.

36.2 RAS leaves it to the investor and the retailer to decide what portion of the investor’s return on investment (CAPEX) it can afford to forfeit to the retailer.

36.3 There is no express provision in any of the Guidelines, Matrices or Principles which entitles the retailers to claim an EC from the CAPEX margin.

36.4 A notional EC is carved out from the pure CAPEX by the DoE and published annually in the RAS Matrix.

36.5 This notional EC published by the DoE is, in practice, split between oil company and retailer in negotiation.

Nature of the decision

[37] The decision to implement the RAS was taken on 25 November 2013, when the Minister approved and signed the following recommendation:

‘The implementation of the RAS margins into the retail price structure of petrol; and

The margins applicable to the wholesale price structures of diesel and IP.’

[38] The first question which falls for determination is whether the Minister’s decision is executive action or a policy decision that is exempt from judicial review under PAJA.

[39] In SARFU,[1] the Constitutional Court held that in determining whether a particular act constitutes administrative action, the inquiry should

focus on the nature of the power exercised and not the identity of the actor. The Constitutional Court stressed that the mere fact that the decision-maker is part of the executive arm of government does not mean that the action is executive. The relevant question is whether the task itself is administrative. In this regard, the focus of the enquiry must be the “nature of the power” the decision-maker is exercising.[2] The Court went on to note a number of other considerations that may be relevant to determining “which side of the line a particular action falls”:

‘The source of the power, though not necessarily decisive, is a relevant factor. So, too, is the nature of the power, its subject-matter, whether it involves the exercise of a public duty and how closely it is related on the one hand to policy matters, which are not administrative, and on the other to the implementation of legislation, which is.’ (emphasis provided)

[40] The Court held that when a senior member of the executive is engaged in the implementation of legislation, that will ordinarily constitute administrative action. The jurisprudence following from the SARFU decision has established that the implementation of legislation by the Executive is an administrative function.[3]

[41] In Permanent Secretary of the Department of Education of the Government of the Eastern Cape Province and Another v Ed-U-College[4] the Constitutional Court distinguished between the essentially political functions of formulating policy and initiating legislation, on the one hand, with the implementation of legislation, which is typically administrative, on the other.[5]

[42] O’Regan J explained the difference between policy formulation in the broad (political) sense and in the narrower (administrative) sense. The Court held that the Provincial Government’s decision to adopt a particular subsidy formula and the mechanism for allocations was “policy formulation in the narrow sense or within the framework of legislation” and was thus administrative action.

[43] The mere fact that a decision is underpinned by policy does not exclude it from the realm of administrative action. O’Regan J in Ed-U-College noted that it is quite possible for action to be administrative even when it has political implications.[6] Our courts have also accepted that certain types of policy decisions – although not having the force of law – will constitute administrative action and be susceptible to review under PAJA. In Greys Marine Hout Bay (Pty) Ltd and Others v Minister of Public Works and Others[7] the SCA rejected the argument that the Minister’s decision to let waterfront property was a policy decision. Nugent JA found that it was a case of policy execution rather than policy formation and was thus administrative action. Nugent JA stated “there will be few administrative acts that are devoid of underlying policy – indeed, administrative action is most often the

implementation of policy that has been given legal effect.”[8]

[44] The Minister’s power to approve the RAS is sourced in the Petroleum Products Act 120 of 1977 (Petroleum Products Act), as amended, which empowers the Minister to regulate various aspects of the liquid fuels industry. Section 2(1)(c) and (d) provide:

‘The Minister may by regulation or by notice in writing served on any person, whether personally or by post, and any person authorized

thereto by the Minister may by such notice so served—

…

(c) prescribe the price, or a maximum or minimum price, or a maximum and minimum price, at which any petroleum product may be sold or bought by any person, and conditions under which the selling or buying of petroleum products other than in accordance with the prescribed, maximum or minimum price may take place.

(d) regulate in such manner as he may deem fit, or prohibit, any business practice, method of trading, agreement, arrangement, scheme or understanding which, in the opinion of the Minister, is calculated—

(i) to influence, or which may have the effect of influencing, directly or indirectly, the purchase or selling price of petroleum products at any outlet; or

(ii) to cause, or which may have the effect of causing, directly or indirectly, an increase in the price referred to in paragraph (c).”

[45] The RAS Matrix, Guidelines and Principles are documents that are produced and published by the Minister under that provision of the Petroleum Products Act.

[47] It is common cause from the papers that the idea was that something would be created which is binding. So, although it is not a regulation, it is clear that the Minister expected everyone to run their operations in accordance with the RAS model and one asks if this were not binding and constituted advice only, why would everyone get involved? Of great significance is the fact that the RAS model is used to determine the regulated retail price. RAS was a way to achieve a range of policy considerations which preceded it. It was preceded by a white paper and a host of other policy building blocks which were already in place. This was not executive action; it was the fulfilling of a statutory role.

[48] But even if I were wrong on finding that this decision is administrative action, there is no dispute that a legality review would still be available to the Fuel Retailers Association. By virtue of my findings herein, it matters not whether the Minister’s decision is labelled as executive action which is exempt from judicial review under PAJA or whether the decision may be reviewed under the principle of legality. In this latter regard (the legality route) I am conscious of being constrained in the following respects: This court may only evaluate whether the RAS is rationally connected to its objective of enabling cost recovery and profit sharing for both the owner of fuel station assets (mostly the oil companies) and the retailers. This court may not interfere with that decision merely because there may be another way to achieve cost recovery and profit sharing – namely by including the EC in the RAS expressly and allocating it in fixed proportions as between the fuel retailer and the oil company. This court also may not interfere with the decision because it would prefer the alternative if it were in the position of the Minister.[9] This court will thus endeavour to ascertain whether the means employed are rationally related to the purpose for which the power was conferred.

[49] The Minister’s decision falls to be reviewed and set aside on any one or more of the following grounds, all of which I find impugn the decision taken:

Errors – Bates White Report

[50] The DoE explains that the RAS was initially conceived through the Bates White report. The key recommendation was for the DoE to implement a revised pricing system based on ‘revenue requirement principles’ and which would allocate costs to ring-fenced activities within the petroleum supply chain. Within the context of the broader recommendation, Bates White made specific recommendations on the calculation of the retail margin. Most importantly, it recommended that the DoE consider the possibility of different retail margins for differently situated retailers, and recommended further assessment of the impacts of the proposed model on the viability of service stations.

[51] Recommendation 9 of the Bates White report provided:

‘Retail margin – calculate the retail margin based on a benchmark service station:

a Initiate a retail service station benchmarking analysis to establish appropriate compensation for the return on fuel-related retail assets.

b Consider the possibility of different retail margins e.g. for rural versus urban service stations.

c Include in the study an assessment of industry impacts, including effects of the aggregate retail asset return, the viability of service stations and implications for provision of services in different areas.’

[52] Bates White provided these specific recommendations regarding the determination of the retail margin because it had identified several risks with the new model. At paragraph 4.2.1 of the Report, Bates White warned that the new methodology may have uncertain impacts on retailers, and particularly CORO retailers (who, it seems obvious, would be in an unequal bargaining position with the oil company):

‘The retail margin will include a return component for fuel-related retail assets. ... We propose that the retail asset return component be determined through a benchmarking survey of retail service stations. Rather than provide for a return on total industry assets, which could exacerbate investment incentive problems, the methodology would aim to provide an appropriate return for an ‘efficient’ benchmark service station. The supporting analysis would have to assess what level of efficiency should be supported, and estimate the impacts of the methodology on the viability of service stations, the provision of retail services in non-urban areas, and the likely effects on owners and operators.

As part of the impact assessment, it should be recognised that by shifting the retail asset return from the wholesale margin (received by the oil company) to the retail margin (received by the retail operators) the new approach will cause the oil companies to extract their returns from increased franchise and rental fees. If the aggregate return to oil companies implied by the return component of the retail margin is less than the return they have achieved under the current system (and which presumably was factored into their investment decisions), some retail operators might find themselves at a disadvantage.’ (emphasis provided)

[53] The DoE contracted the IPSR to conduct the analysis and establish appropriate compensation for the return on fuel-related retail assets. There is no evidence that the DoE considered the possibility of different retail margins or conducted the recommended study on the industry impacts and retailer viability. There is nothing before this Court that explains why the DoE decided to reject or disregard these recommendations by Bates White.

[54] The warnings raised by Bates White were never put before the Minister for her consideration and decision. Instead, the Minister was provided with – and thus relied on – an inaccurate summary of the Bates White report in the October 2011 Recommendation. The Recommendation document excluded the very important provisos and warnings contained in the Bates White report and did not explain that Bates White had recommended further work in respect of the retail margin.

[55] If the Minister and the Controller had properly applied their minds to the Bates White report, they would have realised that Bates White recommended a model that sought to ensure that “regulated firms must be able to recover their costs, including a fair return on their capital investment.” It made no provision at all for a fair return on the business of operating a fuel retail service station. Bates White assumed that in all cases, the retailer would be in a position to claim the full operating margin, as well as the margin providing a return on capital investment. But this is the case only for RORO service stations, where the same entity owns the assets, and incurs the operating costs arising from the retailing activity. The model makes no provision for CORO service stations – where retail operators do not own any capital assets. These retailers cannot secure a return on their investment in operating the retail business. This is crucial because CORO sites predominate in South Africa – both in number and in terms of volume of petrol sold.

[56] The model adopted by the DoE and the Minister was mismatched to the realities of the South African fuel retail sector. Its implications were that CORO site operators would inevitably be undercompensated for their operations.

[57] It means, in short, that the DoE proceeded with the implementation of a pricing model based on only a selective reading of the recommendations of Bates White. It disregarded the clear warnings raised by Bates White, failed to undertake the assessments of retailer viability proposed by Bates White, and failed to consider the rationality of a regulatory scheme that assumed all retailers owned the assets required for the retailing activity – when this was not the case.

The Minister was also not informed of these cautionary factors in recommending the final adoption and implementation of RAS.

That, in turn, meant that the Minister could not take them into account at all. The Minister’s decision was accordingly

procedurally unfair[10] and irrational[11] as a result.

Errors – IPSR Report

[58] The DoE alleges that following the Bates White Report, it appointed IPSR to conduct investigations into the pricing methodologies. IPSR investigated and developed the RAS margin models. The DoE states that it drafted the terms of reference based on the Bates White Report which the Minister accepted.

[59] The Controller stated that the DoE and the Minister were satisfied that IPSR’s analysis and the recommendations were “appropriate, practical and relevant to the requirements of the fuel retail sector”.

[60] This statement is difficult to comprehend considering the serious concerns raised by the IPSR and the issues flagged as requiring additional research and analysis.

[61] The Controller is mistaken when he summarizes the IPSR findings and recommendations as follows:

‘Although the IPSR has referred to the fact that the retail margin should also have a third component, namely the EC, its methodology does not include the EC in the final retail calculation as the retail margin is based on a RORO site. The IPSR, taking cognisance of CORO sites, indicated that stakeholders should determine that portion of the margins to be treated as an EC.’

[62] The implication that the IPSR made a decision not to include the EC in the retail margin is incorrect. The IPSR was not mandated to consider whether or not to include the EC in the retail margin. The DoE only proceeded with the first of the three Bates White

recommendations.[12] The Terms of Reference required IPSR to “formulate the appropriate rate of return for service station assets based on a Capital Asset Pricing Model.” In other words, the IPSR mandate was limited to the question of determining the characteristics of the benchmark service station, and the appropriate rate of return for service station assets. The report determined the methodology for calculating CAPEX. When stakeholders raised concerns about the failure of the model to include an entrepreneurial portion in the retail margin, the IPSR repeatedly stated that it was limited by its terms of reference.

[63] It is also incorrect that the IPSR recommended that the trading margin be negotiated between retailers and wholesalers. The IPSR made it quite clear that the proposal that the CAPEX portion of the retail margin be ‘split’ between wholesalers and retailers (to accommodate an EC) would require additional work beyond its mandate. It also emphasised that the split would have to be based on ‘sound economic logic’. The report notes that ‘there is scope for the investment margin to be reasonably split between investor and entrepreneur. This would require further work and would have to be studied separately as it is outside the scope of the Terms of Reference for Project ME 686.’[13]

[64] In response to concerns that the model ‘makes no provision for an appropriate return on the retail activity’ and ‘would challenge the prohibition on vertical integration’, the IPSR noted:

‘The proposal to split the investment margin into investor and operator elements is outside the scope of the terms of reference of ME 686.’

[65] Despite its limited mandate, the IPSR noted at several places in the report that the model with only an investment margin and an operating cost margin based on a RORO station, would be flawed. The IPSR recorded that most retail sites were CORO sites, and that retailers had proposed that the retail margin be split into three components.

‘In consultative meetings between the oil companies and dealers there have been suggestions that the retail margin be split into three components, namely, retail investment margin, entrepreneurial margin and a dealer margin. The rationale for the split in the retail margin is that the bulk of the existing retail sites are leased sites and not dealer owned sites and the Benchmark Service Station is premised on the fact that the site is a new dealer owned site.’

[66] The IPSR accepted that for the benchmark service station to be implemented to account for the reality of the retail sector in South Africa, the margin would have to be split into three components – but it considered that split to be outside the scope of its project.

‘The split in the BSS margin between the investor margin and the dealer / operator margin covers the investment and recovery of costs margin for a BSS site. It therefore constitutes two components of the BSS margin. In order for the BSS to be implemented within the reality of the retail sector in South Africa where the majority of service stations are leased sites whose volume throughput is 70% of total retail volume, the margin of the BSS would have to be split into three components, namely, investor, operator and operations (costs). This would only be able to occur if the investment margin were split into two components to reward investors and operators to reward entrepreneurship of the dealers on leased sites. There is scope for the investment margin to be reasonably split between the investor and the entrepreneur. This would require further work and would have to be studied separately as it is outside the scope of the terms of reference for Project ME 686.

In the final analysis the methodology employed to split the investor related margin of the BSS between the investor and operator would have to be based on sound economic logic for the BSS model to be successfully implemented as part of the regulatory dispensation in the future.’ (emphasis provided)

[67] The IPSR also warned that the RAS would lead to de facto vertical integration – the very issue that the new model sought to resolve:

‘there is a compelling case to be made that dealers who are entrepreneurs should receive some form of entrepreneurial compensation over and above the normal salary compensation for their investment and efficient functioning of the service station. If the dealers are only granted normal salary compensation then they would become de facto managers of the respective oil companies leading to vertical integration of the retail service station sector. Moreover, there would be no business incentive for entrepreneurs to enter the service station industry as dealers if there is no entrepreneurial incentive in comparison to other small franchised businesses whose return could possibly be greater. (emphasis provided)

[68] The Report continued at page 12:

‘This new system if implemented without further work on an appropriate split of margin between investors and dealers of leased sites would mean that the investment portion of the margin would be retained by the investor and operational cost margin would be retained by the dealer of a leased service station. This would mean that dealers would effectively become managers of leased retail service stations and only recover their costs thereby commencing a process of de facto vertical integration of the vast majority of service stations. ….

This means that from an economic perspective the proposed split accommodates the oil companies that develop service stations as well as the dealer owned dealer operated service stations. The BSS is premised on the fact that a single set of assets that facilitates the sale of regulated petroleum products would make up the site, but in reality, the bulk of the assets on the leased sites are not owned by the dealer and hence the proposed split in the retail margin. The split in the retail margin would mean that those who invest would receive a commensurate return on that investment while those who operate the site should receive entrepreneurial compensation and cost recovery for operational viability.’ (emphasis provided)

[69] The reasoning cannot be faulted. Why a retailer owned retailer operated (RORO) service station was ever chosen as a benchmark when it is the less common form of business in the South African industry is not explained. Ultimately, the IPSR left the issue of entrepreneurial compensation open. In conclusion, it said:

‘there are a number of ways of dealing with the problem, ranging from allowing market forces to settle the matter to a regulatory

solution. The former would mean that the oil companies would not find suitable dealers for leased sites if all that is being

offered by the oil company to a retail entrepreneur is a salary and cost related margin. On the other hand, if it is not an issue which can be amicably resolved commercially between the oil companies and the dealers therefore resulting in the new margin mechanism not being implemented then a regulatory intervention may be required to assist the parties to find a way forward.’

[70] The Controller says that it noted the concerns raised by Bates White and the IPSR. But the DoE disregarded these concerns because “the majority of refining wholesalers gave the full EC to its appointed retail service operators, as reflected in the annual benchmark

service station CAPEX matrix concluded by the Department.” This is not the case.

A model based on a RORO retail site is irrational

[71] The IPSR made it quite clear that there were serious risks associated with a model based on the assumption that all retailers were RORO operators, when this was not the reality. The current model meant that CORO retailers would only be able to recover operational costs. The reality of the retail sector meant that the retail margin should be split into three components. The methodology for calculating this split, IPSR opined, should be based on sound economic logic and would require additional research and analysis and further work was required to ascertain the consequences of the new model on asset-less fuel retailers. This was the information before the DoE when the decision to implement the final RAS was recommended. There is nothing in the papers before this Court that demonstrates that the DoE properly applied its mind to these warnings – or even communicated this information fully to the Minister. By necessary implication, she was precluded from taking these materially relevant considerations into account before deciding to implement the final RAS[14].

[72] The Minister was not informed of the consequences of a regulatory model that did not include a trading margin for CORO retailers. She was not informed of the fact that on this model, CORO retailers were only able to recover the costs of operations, and would earn no profit margin at all (as of right) from the risk and investment into the business of operating the fuel station.

[73] The Recommendation stated:

‘The partial implementation phase was not without any challenges. The retail margin model does not make provision for a split in what accrues to the service station operator nor the investor into the retail assets. The disagreement between the oil companies and retailers on this matter took longer than expected as retailers wanted the Department to regulate the component of the Retail margin. However, the split of the retail margin is not regulated by the Department and that position has been communicated in writing to all stakeholders.’

[74] The Recommendation did not disclose any of the Fuel Retailers Association’s concerns about the EC and the need for a ring-fenced trading margin. The Minister was not provided with the documents or minutes of the RAS Technical meeting disclosing the extent of the dispute regarding the EC and its consequences.

[75] There was a failure to properly consider the impact of the RAS on CORO retailers and to deal with the specific concerns raised by the Fuel Retailers Association.[15]

[76] The Minister was bound to act procedurally fairly. A decision taken in ignorance of, or without sufficient regard to, materially relevant considerations is procedurally unfair and procedurally irrational.[16]

[77] It is clear from the Rule 53 Record and the Controllers’ affidavits that the Minister took the decision to approve the full implementation of the RAS in ignorance of materially relevant considerations regarding the trading margin, and expressly raised by Bates White, IPSR and the Fuel Retailers Association. The decision falls to be set aside on this basis alone.

[78] In other regulated contexts, where retailers within the supply chain do not own the assets, the regulatory scheme provides for a trading margin to compensate traders for the services they provide. In the case of piped gas, NERSA regulates the piped-gas value chain. The methodology adopted by NERSA includes a trading margin as compensation for the trader. This is because traders are involved in the purchase and sale of gas, and do not own any of the related upstream infrastructure. In the case of liquefied petroleum gas, the Minister regulates the value chain in accordance with the Petroleum Products Act. The maximum retail price comprises of a number of components but includes a retail margin equal to 15% of the purchase price. The Fuel Retailers Association procured an analysis of the impact of the implementation of the RAS from an independent expert, Genesis Analytics. (Genesis’ expertise has not been challenged, nor has expert evidence been put up to refute its views.) Genesis notes that this is an example of where the retail price includes both a return on assets to provide the service, as well as margin on the fuel sold by the retailing entity. Genesis notes: “This ensures that traders of LPG that do not own cylinder filling assets required for retail distribution are still able to earn a margin on the sale of fuel.”

[79] By approving the RAS model without the proper allocation of the EC, the DoE has created a regulatory framework that strictly licences and regulates retailers, but does not provide for a return on their investment in the business of operating a fuel station and puts them at the mercy of the negotiating position of the oil companies. This position is inconsistent with the DoE’s general approach to regulation of the retail sector. The DoE’s licencing regime recognises and regulates two forms of investment in the retail sector: the capital investment in the assets required to operate a fuel station and the investment required to operate the business. Retailers must acquire a retail licence, and the owners of retail assets must acquire a site licence. Despite this, the RAS only provides a profit margin for the owners of the retail assets. It makes no provision for a trading margin to accrue to the operators of service stations – despite the fact that they are bound by the licencing and regulatory scheme. Retailers are taken into account in some parts of the regulatory framework, but completely disregarded in others without good reason.

[80] The focus on assets - to the exclusion of other services provided within the supply chain - means that CORO retail station operators are completely disregarded.

[81] The DoE accepts that CORO retailers should be compensated in some form. The DoE’s RAS fact sheet says:

‘The RAS technical team agreed during the development of the BSS model that an entrepreneurial compensation should be predetermined as a reward for the operator of the assets.’

[82] The DoE has adopted a methodology that is mismatched to the realities on the ground and which cannot secure that reward which its technical team agreed should be predetermined as a reward for the operator (not the owner alone) of the asset, which is consistent with the general policy of promoting small and medium sized enterprises and enabling those with no capital but their capacity to work hard and build a business through dedication and entrepreneurial energy, and the decision to adopt a model at odds with the technical team’s agreement accordingly falls to be set aside.

Does RAS achieve its objectives

[83] The Department has adopted a pricing methodology that includes only CAPEX and OPEX recoveries in the retail margin, and excludes a trading margin for retailers over and above those allocations. It then ‘tacks on’ a notional trading margin which it publishes in the Matrices. However, it tells industry stakeholders that retailers must negotiate with oil companies to carve out an entrepreneurial compensation amount out of the CAPEX portion of the retail margin. It expressly disavows the existence of an EC and even a notional EC. The implications of leaving this to market forces where those forces are so heavily tilted in favour of the oil company are understandably seen as disastrous by the retailers.

[84] A regulatory scheme without an allocated trading margin means that retailers are undercompensated for the risk they assume when running the business of a petrol station, and for the fuel-dispensing service they provide.

[85] Retailers cannot increase their margin by increasing the price of the petrol that they sell because the consumer petrol price (that is, the pump price) is set by regulation.

[86] The IPSR explains the economic pressures on fuel retailers and the crucial role played by the retail margin in ensuring that retailers can earn a profit and sustain their business:

“There is a clear distinction between the owners of service station (the investors) where the industry is capital intensive due to low margins and high equipment and building standards, and the operator or dealer where the industry is labour intensive as full service is required. Its profitability is a function of high volume sales. Hence the dealer’s margin is crucial in determining the profitability of the retail service station.

Secondly the retail margin on petrol, the main product, is prescribed by government within which the dealer needs to cover all costs of running the service station and ensuring an optimum level of service, product availability and safety. As the price of petrol is fixed there is no scope for the dealer to increase prices should costs rise. The dealer or retailer needs to keep operating costs down in order to maximise profitability and here the dealer margin is adjusted according to the Retail Margin Model.

However, equally no one is able to start a price war and only the price of diesel which is not fixed by government is able to be

discounted on the forecourt. On-going economic developments of the retail service station of other non-petroleum activities such as convenience stores, eateries and the sale of non-controlled products have also contributed to the profitability of the retail service station.”[17]

[87] In its first report, Genesis concludes that the ‘current framing of the RAS with respect to the retail margin does not allocate a specific return for CORO service stations, resulting in regulatory uncertainty and lack of sustainability in this part of the value chain.’ The RAS methodology allocates no return for the risk that they take on the business, precisely because it does not include a trading margin or EC.

[88] The DoE adopts the position that although there is no formal provision for the EC, retailers can negotiate an EC from the CAPEX portion of the retail margin that accrues to the asset-owner.

[89] The Department provides a cents per litre figure on the RAS Matrix each year, but accepts that the margin received by each retailer notionally depends on their commercial negotiation with the asset investor who must decide which portion of the CAPEX it will ‘forfeit’.

[90] The Department had decided that retailers must fend for themselves within this strictly regulated industry, and persuade the asset-owners to forfeit a portion of the CAPEX margin. The Department is content to allow the ‘split’ in the CAPEX portion of the retail margin to be subject to the “commercial negotiation” or market forces between oil companies and retailers. It contends that it has done enough because the EC is published in the matrix to assist retailers to negotiate with oil companies and to increase their bargaining power. Quite how this assistance is achieved without the force of law is a mystery to me.

[91] The DoE’s attitude displays some naivety. The commercial realties of asset-less retailers operating within a highly regulated environment negotiating for a share of the CAPEX margin with their backs against the wall and no legal force behind their demands is an unenviable position in which to be.

[92] The structure of the retail margin means that the oil companies are entitled - on paper - to the full CAPEX portion of the retail margin. There is no obligation on oil companies to “forfeit” any of the CAPEX portion of the margin to retailers. In fact, any decision by oil companies to “forfeit” a portion of the CAPEX margin means that the oil companies will be unable to fully recover their return on investment in the assets. Genesis concludes that “… it is likely that neither party in a CORO arrangement is given an adequate regulated return.”

[93] The commercial reality and power relations within the industry mean that the fate of the retailers is entirely in the hands of the oil companies. The DoE recognises this power imbalance and the fact that retailers have very little leverage in these negotiations. The oil companies can refuse to recognise the EC and claim the entire CAPEX for themselves. Oil companies can also recognise the EC, but - since the ‘forfeit’ of the CAPEX portion of the margin leads to under recovery - take steps to claim (directly or indirectly) a portion of the notional EC through additional line items for their own benefit.

[94] This is what transpired during the transitional period. In the RAS technical meeting of 14 August 2013 the Department acknowledged that the oil companies were under-recovering under the RAS methodology, and that they were using the entrepreneurial compensation element to “recover costs”. These concerns were not communicated to the Minister. The meeting of October 2013 recorded that three of the major oil companies refused to accept that the notional EC would accrue to the retailers.

[95] In the face of this evidence, the DoE’s statements that CORO retailers are accommodated within the RAS by the notional EC is simply not correct. The retailers are forced to operate within a regulatory scheme that makes no provision at all for a secured profit margin on their investment in the business of operating a service station.

[96] The DoE itself seems to live in hope that the disputes will simply go away. It has said that it has expected industry associations to “resolve areas of dispute and disagreement amongst its membership alternatively, rely on the dispute resolution mechanisms to keep the operations of this industry within the policy framework that Government has set”

[97] Unfortunately, the Genesis analysis reveals that the very structure of the RAS - the framework itself – is the source of the problem. Without a change to the RAS, there can be no fair resolution to these disputes, either generally or on an individual basis because the absence of recognition of a right to EC (as opposed to a mere hope that an oil company will part with some of its CAPEX recovery) leaves the one party without any power with which to negotiate and in circumstances like that to expect negotiation to resolve the problem is irrational.

[98] Simply put, the RAS is incapable of giving effect to its objects of providing a fair return to the participants in the retail sector, and avoiding cross-subsidisation of activities. It is irrational to the extent that it excludes a trading margin from the retail margin, it leaves one category of participant without a fair return.

Commercial negotiation of the EC

[99] To recap: the RAS model does not include an EC, a trading margin allocated to operators to compensate them for the business risk of running a service station. The retail margin includes only a CAPEX and an OPEX. There is nothing in the BSS RAS Matrix, principles or guidelines that entitles the retailers to claim a portion of the CAPEX margin (at least insofar as CORO sites are concerned). Indeed, the RAS Guidelines explain that the calculation of the retail margin “… comprises two components: the investment margin [capex] and the petrol operating margin [opex].” The Minister’s and the Controller’s position is quite clear. It states that:

“… RAS is made up of capex and opex. Capex includes the investors’ return on assets. There is no provision in RAS for an EC (or a notional EC). The current RAS structure leaves it to the investor and retailer to decide what portion of the investor’s return on investment it can afford to forfeit to the retailer.”

[100] It thus comes down to it having to be negotiated under circumstances where, in terms of the RAS methodology, the entire CAPEX is supposed to go to the investor (oil company). This is to be achieved in the following manner:

“What remains is for investors and retailers to negotiate the terms of their commercial arrangements in line with the approved RAS

methodology and the Sector Code as opposed to the Minister over-regulating the industry.”

[101] The hesitancy to over-regulate when the situation cries out for regulation is inexplicable. The RAS methodology does not regulate the Entrepreneurial Compensation at all.

“… The expectation is that the apportionment of recovery of a rate of return must be aligned to the quantum of investment and other factors taken into account in the commercial agreement between the investor and the retailer.” (emphasis added)

[102] There is no guidance of what these other factors must be and oil companies are, in the present situation, at large to simply refuse to forfeit (to use the DoE’s terminology) any part of their CAPEX recovery.

[103] The Fuel Retailers Association argue that the aforegoing is contrary to the objectives of RAS; that there is no certainty, no transparency and that the model does not prevent vertical integration, which it should. This leads to the conclusion that the RAS is premised on a fundamentally irrational decision because it is incapable of achieving the stated objectives of RAS.

[104] The Controller now alleges that the Fuel Retailers Association agreed to the final implementation of the RAS - including the fact that the EC would be subject to negotiation between the investors and retailers.

[105] This is not correct. Despite what the Controller now says in the affidavits, the Fuel Retailers Association repeatedly[18] and persistently stressed that the regulatory model without a trading margin would have catastrophic consequences for retailers.

[106] At the time the Minister took the decision many retailers took the view that they had a right to the full notional EC; certain of the wholesalers were refusing to acknowledge the retailers’ right to any portion of the CAPEX margin as a notional EC; SAPIA had taken the position that there was no EC component in the retail margin; the retailers continued to insist that an EC was required. At best for the Respondents, a dispute exists about whether there was consensus on excluding the EC from the RAS. I will thus accept, as I must, the version of the Respondents.[19] This does, however, not preclude the current challenge for a host of reasons not least of which is that this agreement did not form part of the Minister’s decision-making process. It is the decision not to regulate the split of the retail margin between fuel retailers and the fuel companies in CORO sites that is at the heart of this review application.

[107] The assertion of the existence of the agreement by the DoE, an agreement not drawn to the Minister’s decision, exposes the fact that the Minister failed to properly consider the impact of the RAS on CORO retailers and to deal with the specific concerns raised by the Fuel Retailers Association.[20] According to the DoE there was no issue left as it had been resolved by agreement between the relevant stakeholders. Thus nothing for the Minister to consider on that front.

[108] The Minister was bound to act procedurally fairly. As I have observed and as is well supported by authority and the wording of PAJA, a decision taken in ignorance of, or without sufficient regard to, materially relevant considerations is procedurally unfair and procedurally irrational.[21]

[109] Ms le Roux SC representing SAPIA argued that what is missing in the Fuel Retailers Association’s case before this Court is evidence to show that these commercial negotiations have failed to give rise to adequate compensation for the retail operators and the owners of retail assets respectively. SAPIA argues that precisely what that relationship entails and what costs must be recovered will vary from site to site. They submit that the parties will negotiate and strike a commercial deal to allocate income as needed between them considering their individual circumstances and that the DoE’s decision to enable these bespoke outcomes is thus rationally connected to the cost recovery objective of the RAS. They contend that failing to prescribe an EC for fuel retailers in the RAS is rationally connected to these objectives where there is a diversity of different commercial arrangements between retailers and asset owners, and a range of different levels and types of asset investment that must be recovered. This, they contend, satisfies the test in Albutt, Pharmaceutical Manufacturers and Democratic Alliance[22].

[110] The Fuel Retailers Association contends that the use of RORO sites as the model operating fuel station in the RAS excludes from consideration the lease agreements, branding agreements and service level agreements entered into by fuel retailers who operate CORO sites. It contends that these increase fuel retailers’ costs, which are then under-recovered in the commercial negotiation required under the RAS.

[111] Ms le Roux SC, representing SAPIA, emphasized the inadequacy of the evidence of this under-recovery before the Court. She placed much emphasis on the fact that retailers have multiple income streams, including the sales from the forecourt shop, diesel, car wash, to name but a few. The current complaint relates to the sale of petrol alone. She pointed out that not a single set of financial statements has been included in these papers evidencing the contention that the failure to regulate the EC, threatens the viability of retailers. It was contended that no evidence has been placed before this court that retailers and company owners cannot negotiate. That being so, this court cannot conclude that every CORO site is unviable or that ‘the regulatory scheme imposed by RAS exacerbates existing barriers to entry for historically disadvantaged South Africans and, as a result, serves to exclude (rather than promote) new retailers entering the market.’ To support this argument SAPIA referred to the outcome of 6 negotiations (out of some 3000 retailers comprising the membership of the Fuel Retailers Association). These do not show that retailers are excluded and evidences a range of R0,05 to R0,16 of the available R0,203 notional EC[23].

[112] Ms le Roux argued that not a shred of evidence had been placed before this court that a retailer was unable to reach agreement or that a retailer had no option but to agree on R0,0 of the notional EC and that the oil company was unwilling to ‘forfeit’ any of the R0,203 of the notional EC. There was thus no evidence that the negotiations always worked against the retailers. Of the 6 affidavits before me, the lowest is 5 cents of the 20 cents available (25% of the notional EC) and the highest is 16 cents of the 20 cents (80% of the notional EC) available. She submitted that negotiations certainly did not always go in one direction. These submissions fail to persuade.

[113] The DoE expressly disavows the existence of an EC and a notional EC. It contends that the stakeholders had agreed that there would be no EC in the RAS . The retailers are thus forced to work within a regulatory scheme that makes no provision at all for a secured profit margin on their investment in the business of operating a service station. The DoE contends that an EC should be negotiated by the retailers and the oil companies from the entire CAPEX margin. Yet what the oil companies start their negotiations with, is only the notional EC and not the entire CAPEX margin as suggested by the DoE. The notional EC should of course be for the benefit of the retailers in its entirety (if all the line items for a CORO site had been properly allocated and factored in which we know

is not the case as the RAS has been premised on a RORO site). What 5 of the 6 affidavits demonstrate is that the oil companies ‘clawed back’ and recovered their costs from what might otherwise have been the profits of the retailers.

[114] The Petroleum Products Act acknowledges an unequal bargaining position. Recently, the Constitutional Court emphasised the importance of transforming the petroleum industry:

‘to ensure that unequal bargaining power in the industry was addressed for those doing business in that industry, as well as empowering

historically disadvantaged South Africans in the petroleum and liquid fuels industry.’ [24]

[115] Ms le Roux contended that the demands of the retailers were unrealistic. In support of this argument she referred to the evidence of Mr Mbonambi stating that he had anticipated making a reasonable profit and that he would be able to make a return of 32%. She pointed out that no evidence has been placed before me as to what would constitute a reasonable return but that as a rule of thumb and in her view, it would be fair to suggest that a return of between 10 and 15% would be considered a good return and 32% unreasonably high.

[116] In my view, the grievance of the retailers is more fundamental. They contend that the existence of any profit they earn is at the whim of the oil companies and that a model which permits this is at odds with the objectives sought to be achieved by RAS. The retailers are forced to operate within a regulatory scheme that makes no provision at all for a secured profit margin on their investment in the business of operating a service station. In my view, the retailers should, as of right, be entitled to a profit on their sales of petrol only.

[117] The understanding of the DoE and the understanding of SAPIA are at odds with one another. The DoE contends that the retailers and the oil companies should negotiate an EC from the entire CAPEX portion of the retail margin whereas SAPIA contend that an EC needs to be negotiated from the notional EC published in the RAS matrices. This confusion appears to have been brought about because the DoE seeks to avoid any conflict regarding the EC. A ring-fenced EC within the retail margin in the formal RAS documents is avoided. On that approach, the full CAPEX of the retail margin accrues to the oil companies (in CORO sites). At the same time the

DoE appears to want to appease the retailers by acknowledging the need for an EC as part of the regulatory scheme, suggesting that it should be derived from the CAPEX portion of the retail margin. It does this by taking the margin allocated to two unrelated elements within the CAPEX margin: the small stock premium and the marketability adjustment each of which in fact relates to, and

compensates asset owners for particular capital investments. This results in the oil companies contending that they are entitled to ‘claw back’ some of the EC, as evidenced in the 5 supporting affidavits of Messrs Khoza, Nkosi, Nonkwebo, Mbonambi and Mbatha.

[118] Even if retailers were in a reasonable position to negotiate with the oil companies that supply them (which they are not), the inevitable outcome of the DoE’s decision to allow the EC to be determined by the commercial negotiations is that either the retailers are under-compensated (because they cannot procure an adequate return for their retail activities), or the asset-owners are under-compensated (because they have allowed retailers to take an EC out of the CAPEX that they would otherwise be entitled to). This flaw in the RAS means that it is arbitrary and incapable of giving effect to the objectives that it was designed to achieve.

[119] As mentioned, the Fuel Retailers Association commissioned reports from Genesis. The second report (which was abandoned by the Fuel Retailers Association) was commissioned to report on the impact of RAS. A survey was done and of the 3000 members of the Fuel Retailers Association only 53 responded and of those, only 30 were CORO sites. So, although it was abandoned, SAPIA argued that it should be considered for the following reasons: It was commissioned for the reason of supporting the application; If it were a problem that retailers were unable to reach agreement or that negotiations always favoured the oil companies, this survey would have revealed this or if there were a problem in practice, the retailers would have queued to voice their grievances. Although this argument on the face of it seems attractive it ignores the realities of the retailers being that they are, at least, breaking even because

of the OPEX component. But that is not what they are entitled to. The retailers are entitled (as of right) to an EC on their petrol sales.

[120] It is correct that of the 6 affidavits (of the potential 3000 members of the Fuel Retailers Association) not one suggests that they need certainty on the EC component failing which they will become unviable. It is unlikely that on the evidence before this court, this can be contended because this court knows that petrol is not the only stream of income and the overall performance of a station might present a positive picture.

[121] The DoE has adopted a margin-setting scheme that excludes any EC for retailers over and above the CAPEX and OPEX portions of the retail margin. It ‘tacks on’ a notional trading margin by telling industry stakeholders that retailers can negotiate with oil companies to carve out an EC amount out of the CAPEX portion of the retail margin. This scheme is flawed in that it is arbitrary. The oil companies are expected to ‘forfeit’ a portion of the CAPEX margin under circumstances where it is

recognised that there is unequal bargaining power. With what, one asks, must the retailers bargain?

[122] Confronted with this dilemma and with the proposition that the default position is that the oil companies take the entire notional EC, Ms le Roux argued that there was ‘no default’ position as the notional EC is the sum total of what is available and if the parties cannot agree, no agreement is concluded. I am not convinced by that construction of the model: the position seems to be that it is a ‘take it or leave it’ situation. The retailer either accepts what is on offer from the oil

company, or it must walk away from the transaction. That is how the model is designed.

[123] To look at what the negotiations actually yielded is of limited assistance if one accepts, as I do, that the negotiation power is unequal, for the reasons that I have set out herein.

[124] Mr Quixley representing the Fuel Retailers Association, argued that a decision which requires the parties to negotiate, where no guidelines or factors have been put in place as to how the negotiation should be conducted, is fundamentally irrational. In my view, the distinction he draws is crucial. The argument is not that negotiation per se is irrational. Clearly it is not. A decision which requires parties to negotiate where i) the negotiating power is unequal; ii) the default position is that the oil company (the more powerful contracting party) must relinquish something; and iii) no guidelines or factors have been put into place to consider during this bargaining process, is naïve to the point of irrationality. The fact that the retailers in practice manage to limp along despite this constraint cannot transform the decision from an irrational one to a rational one.

[125] The DoE’s position is that the industry should negotiate the EC and if it fails, it will regulate it. Mr Bokaba SC, representing the DoE drew attention to the following feature in the IPRS report that ‘…if all that is being offered by the oil company to a retail entrepreneur is a salary and cost related margin’, oil companies would not find suitable dealers. That may be so, but the model cannot be designed to, in principle, not provide an EC to the retailer at all, particularly as the retailer is up against the statutory price maximum, the retailer can pass no increase on to the consumer, being the last one in the line, the last link in the value chain. The retailer is the most vulnerable to being negotiated downward, the need for regulatory protection in the form of a regulated EC could not be more self-evident, and to deny the structural flaw by pointing to instances where the structural flaw has not manifested itself particularly harshly is to deny the inherent and logical flaw in the model on anecdotal evidence.

[126] The jackal should be entitled to its share in the kill because to ask the lion whether it will share its kill with the jackal without some protection of its minimum position, will result the jackal in having to be satisfied with the scraps or walk away hungry.

[127] Both Mr Bokaba and Ms le Roux emphasied that this model has been implemented now for 9 years and that after 9 years of implementation, only 6 affidavits were attached (not one of which shows that the retailers were incapable of negotiating an EC) and no financials were attached to support a 0 margin of profit. I do not accept that that is a necessary element of a cause of action based on irrationality or a failure to take account of relevant matter. Ms Le Roux argued most strenuously that without evidence this court cannot conclude that the lack of a regulated EC, leads to an unviable service station. She submitted that there is no evidence that market forces are incapable of managing the viability of ensuring reasonable compensation or of commercial failure due to an inadequate EC. She argued that there is no evidence that on the sale of petrol alone, retailers are unviable and finally, that there is no evidence that if the EC is regulated, it will affect viability.

[128] The Fuel Retailers Association contends that negotiation is fundamentally unworkable and an anathema to the model in respect of a CORO site. For all the reasons traversed thus far, I agree.

[129] Returning to the point made by the Constitutional Court,[25] the Petroleum Products Act acknowledges an unequal bargaining position in the industry. The model should allow for the circumstances of the majority of sites. To leave it to the investor and the retailer to decide what portion of the CAPEX the investor can afford to ‘forfeit’ to the retailer under circumstances where it is accepted that the bargaining power is unequal, is irrational.

[130] Ms le Roux argued that additional compensation may always be desirable but submitted that that is not the test that this Court must apply. I disagree. Retailers are, as of right, entitled to be adequately compensated. As things presently stand, the default position is that either the retailer takes what is on offer, or walks away and no agreement is concluded. When pressed on the ‘default’

position during argument, Ms le Roux argued that there was no ‘default’ position. That is incorrect. The starting point

is that the investor (oil company) must forfeit part of the CAPEX. The RAS does not provide for an EC at all. That is the default. There is no express provision in any of the Guidelines, Matrices or Principles which entitles the retailers to claim an EC from the CAPEX margin. There is not even provision for the notional EC which is published annually which does not have any legal force. No-one is bound by the notional EC published in the RAS Matrix annually. The stakeholders appear to adhere to it but that does not entitle the retailers to any portion of it.

[131] The legal position is: The EC is said by respondents to be part of the CAPEX, but the retailers have no entitlement to it. No considerations have been identified which can guide the industry in this regard. It does not avail SAPIA to argue that the insufficiency of evidence demonstrates that negotiation works in practice. In my view, it shows no more than that the retailers ‘make do’. What else would they do? The retailers are entitled to be compensated and are entitled to an EC based on a CORO model where the considerations informing the EC are identified.

The RAS with a notional EC enables vertical integration

[132] One of the key tenets of the government’s policy and legislation in petroleum products sector is that licensed wholesalers may not operate at the retail level of the fuel supply chain. Section 2A(5)(a) of the Petroleum Products Act consequently prohibits licensed wholesalers, among others, from engaging in a business practice, method of trading, agreement, arrangement, scheme or understanding that is aimed at or would result in their holding a retail licence.

[133] The policy is aimed at limiting vertical integration and creating opportunities for small business in the industry, especially for historically disadvantaged South Africans. Indeed, the Controller describes one of the objectives underlying the RAS as follows:

“A related policy objective is that of providing certainty to investors with regard to the return on assets throughout the petroleum value chain, whilst pursuing the objective of delinking the vertical integration of the industry to reflect a separation between wholesale operations and the retail services aspect of the industry.”

[134] Despite the prohibition on vertical integration, wholesalers are permitted to own retail assets – in order to lease them to retailers - and to earn a return on those assets. But they may not indirectly operate the retail site or exercise substantial control over it.

[135] Yet, RAS in its current form allows wholesalers and the oil companies to do exactly that: The inclusion of only a notional EC as a part of the CAPEX allocation of the retail margin means that the oil companies can either refuse to recognise the EC (and claim the entire CAPEX for themselves) or take steps to claim (directly or indirectly) a portion of it through additional line items for their own benefit. Because of the power differential between oil companies and fuel retailers, they are able to insist on the margin that retailers will recover from the notional EC. In effect, they determine the margins and profits that CORO stations are able to make, and thus exercise substantial control over or interfere with their operations. That, in turn, contributes to greater vertical integration, rather than reducing it. In effect, licenced wholesalers who own the retail site and assets, can determine the profits earned by retailers. They accordingly exercise substantial control over the operations of retailers. They are also able to secure a profit from the retailing activity. Thus, the rationale underlying the prohibition on vertical integration is undermined, the object of government’s policy s prevented from being achieved by the manner in which it is implemented. The implementation therefore is irrational.

[136] It means that the RAS undermines one of the purposes that it sought to achieve, or enables that which the legislation prohibits. On either basis, it is at odds with the legislative regime and amenable to review.

Remedy

[137] This court may grant any order that is just and equitable.[26] PAJA empowers this court with a discretion.

[138] The applicant did not persist in seeking a declaration of invalidity even though the default position is ordinarily to set aside unlawful administrative action.[27] This is so as the applicant conceded, correctly in my view, as submitted by Ms le Roux, that if this court were minded to remit the issue of the EC back to the Minister, the appropriate order would be one which maintains the status quo of the outcome of each fuel retailer’s negotiation with its fuel supplier/landlord. Clearly such a holding pattern while the DoE reconsiders the position is preferable as it will ensure contractual stability and will eliminate commercial uncertainty. This is what I seek to achieve in the order that I intend making.

[139] I do not agree that the new model would only need to make changes in respect of the retail margin, in order to introduce a trading margin which is a discreet amendment that can quickly and relatively easily be affected. It may well ultimately be so but as I see things what would be required is that the IPSR (if they are going to be employed again) will have to be given a wider mandate to report on a model for CORO sites. That recommendation will have to be traversed with all stakeholders and the Minister will have to decide whether the RAS in its amended form is to be implemented. I thus disagree with the applicant that it would require approximately three months’ work. It is for this reason that I have afforded the process a period of 9 months.

[140] In preserving the status quo pending the finalisation of the process, little uncertainty is caused in the industry thereby minimising prejudice. Retailers of CORO sites entering the market for the first time and concluding new agreements will be required to negotiate an EC with the oil company/investor/fuel supplier/ landlord from the notional EC published in the RAS Matrix annually.

I accordingly grant the following order:

140.1 Condonation for the applicant’s failure to bring this review application within 180 days of the original decision referred to in paragraph 140.2 hereof alternatively within a reasonable time, is granted.

140.2 The original decision of the first respondent or her delegates [the Minister], taken in November 2013, to implement the RAS without providing for a ring-fenced Entrepreneurial Compensation (EC) to be claimed exclusively by the retailers in Company Owned Retailer Operated (CORO) sites and/or specifying the items to be claimed under the EC by retailers in CORO sites, is reviewed and set aside.

140.3 The determination of the treatment and calculation of the EC for retailers in CORO sites as an allocation within the retail margin of the RAS (the determination) is referred back to the first respondent [the Minister] to decide in accordance with this Court’s judgment, within a period of 9 months from the date of this order.

140.4 Pending the determination, the 2020 RAS Benchmark Service Station Matrix and any subsequently issued (or yet to be issued) Matrices are to remain in force and effect.

140.5 Pending the determination, the status quo of the outcome of each CORO site fuel retailer’s negotiation with its fuel supplier/landlord is to be maintained and all new agreements still to be concluded between CORO site retailer’s and fuel suppliers/landlords are to be on the basis that the Minister’s decision has not been reviewed or set aside.

140.6 The first, second and third respondents are to pay the costs of this application, jointly and severally, the one paying the other to be absolved such costs to include the costs of two counsel where so employed.

___________________________

I OPPERMAN

Judge of the High Court

Gauteng Division, Johannesburg

Counsel for the applicant: Adv G Quixley and Adv F Hobden

(Heads of argument prepared by Adv R Bhana SC, Adv I Goodman and Adv F Hobden)

Instructed by: Seton Smith & Associates

Counsel for the 1st and 2nd respondents: Adv TJB Bokaba SC, Adv B Morris and Adv T Pooe

Instructed by: The State Attorney

Counsel for the 3rd respondent: Adv MM le Roux SC

Instructed by: Fasken Attorneys

Date of hearing: 27 and 28 October 2022

Date of Judgment: 22 September 2023

[1] President of the Republic of South Africa v South African Rugby Football Union 2000 (1) SA 1 (CC).

[2] SARFU at para 141.

[3] In Janse van Rensburg NO v Minister of Trade and Industry NO 2001 (1) SA 29 (CC) the Court found that the Minister’s powers to suspend the activities of a company and to attach or freeze its assets was subject to section 33 and therefore administrative action. Similarly, in Premier, Mpumalanga v Executive Committee, Association of State-Aided Schools, Eastern Transvaal 1999 (2) SA 91 (CC) at para 38, the Constitutional Court held that the decision of the Premier of Mpumalanga Province to withdraw state bursaries from state-aided schools amounts to administrative action.

[4] 2001 (2) SA 1 (CC)

[5] See paragraph 18: “ Policy may be formulated by the executive outside of a legislative framework. For example, the executive may determine a policy on road and rail transportation, or on tertiary education. The formulation of such policy involves a political decision and will generally not constitute administrative action. However, policy may also be formulated in a narrower sense where a member of the executive is implementing legislation. The formulation of policy in the exercise of such powers may often constitute administrative action.”

[5] See paragraph 18:

“ Policy may be formulated by the executive outside of a legislative framework. For example, the executive may determine a policy on road and rail transportation, or on tertiary education. The formulation of such policy involves a political decision and will generally not constitute administrative action. However, policy may also be formulated in a narrower sense where a member of the executive is implementing legislation. The formulation of policy in the exercise of such powers may often constitute administrative action.”

[6] Ed-U-College at para 17.

[7] 2005 (6) SA 313 (SCA)

[8] Greys Marine at para 27.

[9] Albutt v Centre for the Study of Violence and Reconciliation 2010 (3) SA 293 (CC) at para [51]

[10] Violating section 6(2)(c) of PAJA

[11] Violating section 6(2)(f)(ii) of PAJA

[12] “Initiate a retail service station benchmarking analysis to establish appropriate compensation for the return on fuel-related retail

assets”.

[13] IPSR Report, p104.

[14] Violating Section 6(2)(e)(iii) of PAJA

[15] Minister of Health & another v New Clicks SA (Pty) Ltd & others[2005] ZACC 14; , 2006 (8) BCLR 872 (CC) (30 September 2005) at paras [191], [391], [393], [400] and [402]

[15] Minister of Health & another v New Clicks SA (Pty) Ltd & others[2005] ZACC 14; , 2006 (8) BCLR 872 (CC) (30 September 2005)

at paras [191], [391], [393], [400] and [402]

[16] Democratic Alliance v President of the Republic of South Africa and Others, 2013 (1) SA 248 (CC) at paras [36] and [39]

[17] IPRS Report, 020-343, p7.

[18] In a letter of 7 November 2012 to Mr Maake; Minutes of a RAS meeting on 13 March 2013 reflects the oil companies’ professed intention to claim a share of the EC, there allegedly not being grounds to substantiate the quantum thereof; The Draft Minutes of the meeting of 10 April 2013 record the clear disagreement between the oil companies and the retailers regarding

the manner in which the CAPEX margin and the notional EC should be shared. It also records the DoE’s position, at that stage, that the notional EC should accrue to the retailer and that any wholesaler seeking to claim a portion of the EC must provide an evidentiary basis for it;

[19] Plascon Evans Paints. There was no request that this dispute be referred to the hearing of oral evidence or that such determination in favour of the Fuel Retailers Association would be dispositive of this issue.

[20] Minister of Health & another v New Clicks SA (Pty) Ltd & others[2005] ZACC 14; , 2006 (8) BCLR 872 (CC) (30 September 2005) at paras [191], [391], [393], [400] and [402]

[21] Democratic Alliance v President of the Republic of South Africa and Others, 2013 (1) SA 248 (CC) at paras [36] and [39]

[22] Supra

[23] The Matrix as at December 2013 was referenced – ‘FA1’ – 001-23 in which the Capex margin distribution is reflected as R0,203 and the EC as R0,401

[24] Rissik Street One Stop CC t/a Rissik Street Engen and Another v Engen Petroleum Ltd [2023] ZACC 4; see too Business Zone 1010 CCt/a Emmarentia Convenience Ventre v Engen Petroleum Ltd 2017 (6) BCLR 773 (CC)

[25] Footnote 24

[26] Section 8 of PAJA read with section 172 of the Constitution

[27] Aquila Steel (S Africa) (Pty) Limited v Minister of Mineral Resources, 2019 (3) SA 621 (CC) at para 108