Sasol Chemical Industries Limited (appellant) was the only significant producer of purified propylene in South Africa and dominant in the polypropylene (PP) market (64-80% market share during 2004-2007). Appellant purchased feedstock propylene from a fellow subsidiary, Sasol Synfuels (Synfuels), at a low price due to Synfuels' unique coal-to-liquids Fischer-Tropsch process which created feedstock propylene as a by-product with low fuel alternative value (FAV). The Competition Commission (respondent) alleged appellant charged excessive prices for purified propylene and PP from 2004 to 2007, failing to pass on its unique feedstock cost advantage to customers. The Commission claimed appellant's domestic PP prices were up to 32% higher than export prices and 41-47% higher than Western European prices when adjusted for feedstock costs. For propylene sold to Safripol, the Commission alleged prices were 55% above economic value based on export PP prices, and 326% higher than appellant's own 2003 predictions for Project Turbo. The Competition Tribunal found contraventions of s 8(a) of the Competition Act 89 of 1998 and imposed administrative penalties totaling R534 million.
The appeal was upheld. The decision of the Competition Tribunal was set aside and replaced with an order dismissing the complaint referral. No administrative penalties were imposed.
The binding legal principles are: (1) 'Economic value' under s 8(a) of the Competition Act should be determined using a dominant firm's actual costs where those costs are justified on rational grounds and reflect arm's length pricing subject to independent regulatory scrutiny, rather than hypothetical costs in a notional competitive market. (2) A firm's special cost advantages need not be adjusted away at the economic value determination stage; such advantages may be considered at the reasonableness stage of the s 8(a) inquiry. (3) A price is only excessive under s 8(a) if it is both: (a) higher than economic value, AND (b) bears no reasonable relation to economic value. Both elements must be established. (4) For a price to bear 'no reasonable relation' to economic value, the margin above economic value must be substantial - significantly greater than approximately 20% above the defined economic value. (5) In evaluating economic value through price-cost tests: capital assets should be valued at replacement cost adjusted for inflation rather than historical book value; return on capital should be calculated using inception WACC on an after-tax basis including appropriate hurdle rates for risk; and cost allocation methodologies should reflect economic reality rather than arbitrary volume-based allocations. (6) Courts should exercise a measure of deference in excessive pricing cases to avoid becoming price regulators, particularly where pricing can be justified on cognizable commercial grounds.
The Court made several important non-binding observations: (1) The Tribunal's suggestion that patent holders may charge prices bearing no relation to economic value for the duration of the patent is manifestly incorrect and would destroy the possibility of excessive pricing cases in intellectual property - patent holders are not licensed to engage in excessive pricing. (2) Expert witnesses in competition proceedings must provide independent, objective, unbiased opinions uninfluenced by litigation imperatives. Experts should not assume the role of advocates, should state assumptions and limitations clearly, should acknowledge matters outside their expertise, and should not opine on legal questions. The Tribunal must exercise greater discipline over expert evidence. (3) The failure of the Commission to present cogent, well-reasoned expert evidence to rebut appellant's expert testimony was fatal to its case. Figures cited without clear justification do not constitute expert evidence. (4) The Court's decision should not be interpreted to mean excessive pricing cases can never succeed - had the Commission presented proper expert evidence contesting the various cost calculation methodologies, the outcome might have been different. (5) Using the cost of substitutes (like import parity pricing) as economic value would mean no firm could ever price excessively, as there is always some next best substitute - this would constitute a 'cellophane fallacy'. (6) Where a dominant firm's position results from state support rather than innovation or risk-taking, this is relevant to the reasonableness inquiry, but creates difficulties for firms in determining the source of their current dominance given the passage of time.
This is a landmark judgment on excessive pricing under s 8(a) of the Competition Act. It clarifies that: (1) economic value should be based on a dominant firm's actual costs where those costs reflect legitimate arm's length transactions, rather than hypothetical competitive market costs; (2) the Mittal precedent does not require adjusting away all firm-specific cost advantages at the economic value stage - special advantages may be considered at the reasonableness stage; (3) substantial margins above economic value are required before a price will be found to bear 'no reasonable relation' to economic value; (4) courts should exercise deference in pricing matters to avoid becoming price regulators; (5) the judgment provides detailed guidance on cost calculation methodologies including capital asset valuation, return on capital, and cost allocation. The case demonstrates the extreme difficulty of successfully prosecuting excessive pricing cases and the stringent evidentiary burden on competition authorities. It also contains important obiter on the role and duties of expert witnesses in competition proceedings, emphasizing the need for independence and proper scientific methodology.