For nearly forty years, Article L. 420-2, second paragraph, of the Commercial Code has raised more disappointed hopes than convictions. Invoked in almost every dispute over a broken relationship between a supplier and a major retailer, it was set aside in almost all of them, because it demands proof of harm to a market where the client brings only a file of a deteriorated bilateral relationship. Two events six days apart in May 2026 changed that. Here are the six points that now determine whether a claim under this text stands a chance.
1. A landmark ruling has finally validated a conviction under the text, at the highest level
On 13 May 2026, the Commercial Chamber of the Cour de cassation, in the Apple Premium Resellers case, handed down a ruling published in the Bulletin and the Annual Report that confirmed, for the first time at that level, a conviction based on Article L. 420-2, second paragraph — validating an “ecosystem” reasoning that is directly transposable to selective distribution networks, franchises and digital platforms (Cass. com., 13 May 2026, no. 22-22.623 and others, FS-B+R). Six days later, the French Senate’s inquiry committee on manufacturer and retailer margins recommended revising the legal definition of abuse of economic dependence, finding the interpretation applied until then too restrictive. The text is now simultaneously enshrined by the courts and challenged by the legislature.
2. Three cumulative conditions apply, and the third is the one that defeats nearly every claim
The text requires a state of economic dependence, its abusive exploitation, and an actual or potential effect on the functioning or structure of competition. Each is governed by a distinct evidentiary regime, and failure on any one defeats the claim. The state of dependence itself is now assessed, per paragraph 86 of the 13 May 2026 ruling, through four indicia read together: the reputation of the supplier’s brand, its share of the relevant market, its share of the reseller’s turnover, and — the only truly decisive one — the reseller’s inability to obtain within a reasonable time a technically and economically equivalent solution. None of the first three needs to clear any fixed threshold on its own; Apple’s own market share was a modest 16.3% by volume, yet the finding stood because the other indicia converged strongly. The third condition, the market effect, is structurally the hardest: the more real the dependence, and therefore the weaker the victim, the harder the harm to competition is to demonstrate, because the text demands macroeconomic proof where the injury is microeconomic. This internal tension is what produced three decades of near-total ineffectiveness, and it is what the Apple case overcame, in a configuration where the weakening of an entire distribution channel was objectively measurable — the relevant unit of analysis being the category of operators, not the individual undertaking.
3. A larger competitor is not automatically an alternative — market share is the most deceptive indicator
Apple argued that, since its own market share was lower than Samsung’s, resellers necessarily had an alternative supplier available. The Commercial Chamber rejected this: only the Apple group operated a network of specialized retailers dedicated to its own brand across the entire national territory, and no supplier of comparable size held a brand with a comparable portfolio, reputation and customer base, so Samsung “did not constitute an alternative supplier” despite its larger footprint (§ 88). Market share measures an operator’s power, not the equivalence of a solution to the dependent party — a distinction worth building into any economic-dependence file from the outset, on either side of the case.
4. Partial solutions do not add up, and a percentage of turnover proves nothing on its own
The Commercial Chamber has rejected two arguments claimants and defendants alike keep reaching for. First, that combining several imperfect alternatives can substitute for one equivalent solution: the Court dismissed the “mistaken premise drawn from the existence of concurrent options,” each alternative having to be assessed on its own terms (Cass. com., 29 January 2025, no. 23-16.526). Second, that a high share of turnover with a single partner is, by itself, sufficient to establish dependence: even a share as high as 86.38% was held insufficient on its own, because the state of dependence requires proof that the concentration was neither chosen nor unwindable within a reasonable time (Cass. com., 26 February 2025, no. 23-50.012).
5. This litigation is won or lost before the relationship ends, not after
The cardinal rule fits in one sentence: the state of dependence is assessed at the time of the facts, but it has to be documented beforehand. An undertaking that discovers its dependence on the day it is delisted has, in nearly every case, already lost access to the documents that would have proven it — product allocation tables, comparisons of terms granted to other network members, the real criteria behind rebate decisions, and above all proof that alternative suppliers were actually approached and turned the undertaking away. All of that sits in the opposing party’s files, or depends on evidence — a dated refusal letter, a costed reconversion study — that only exists if it was gathered while the relationship was still running. Article 145 of the Code of Civil Procedure, mobilized before any trial to preserve or obtain evidence, is often the only route left to reach it once a dispute looks likely; it reaches not only the partner itself but third parties, including other members of the same network whose converging statements carry recognized probative value.
6. The best defense attacks the market effect first, not the dependence itself
The most common defense mistake is spending most of the argument disputing the state of dependence — the condition most favorable to the claimant, since it rests on facts the claimant controls — while dealing with the harm to competition in a few lines, even though that is the hardest element for the claimant to prove. The stronger sequence reverses this: challenge the market effect first, then the abusive conduct, and only then the dependence itself, reminding the court from the outset that the text protects the market, not the counterparty. Where dependence genuinely cannot be defeated, contesting the market effect and demonstrating a deliberate strategic choice remain the two lines of argument worth more than a frontal challenge to dependence itself.
This text has changed in nature more than it has changed in wording. What used to be a near-automatic dismissal is now, in the right configuration, a viable claim — and, for network heads and platforms watching the Competition Authority’s recent statements on structural dependence, a live compliance risk. Our practical guide sets out the full evidentiary methodology for claimants and the complete defense framework, together with the competing grounds — significant imbalance, abrupt termination, economic duress — that often carry a case further than this text alone.