The right of withdrawal, linked agreements and vehicle purchase contracts- the CJEU in C-143/23

On the 30th of October 2025, the CJEU delivered a judgment in the joined cases of consumers KI and FA against Mercedes-Benz Bank AG and Volkswagen Bank GmbH in C-143/23 , advancing the interpretation of Articles 10 and 14 of Directive 2008/48/EC on Consumer Credit.

 

The consumers entered into a credit agreement with their respective banks to purchase a motor vehicle for private use; where the car dealers from whom the vehicles were purchased acted as credit intermediaries and to whom the loans were paid in directly. The credit agreements did not inlcude the interest rate applicable to late payments at the time the agreement was concluded. Months and years after the contract had been concluded, the consumers notified their banks that they wished to withdraw from their credit agreements, arguing that the standard 14-day period did not begin to run due to omissions in the mandatory information included in their contracts. The referring Landgericht Ravensburg asked several questions to the CJEU.

 

 

The first question addressed by the CJEU was whether the consumers' right of withdrawal began to run despite the absence of mandatory information in their contract. According to Article 14(1)(b) of Directive 2008/48 the 14-day withdrawal period begins to run only on the day on which the information provided for in Article 10 has been received by the consumer, if that day is later than the day on which the credit agreement was concluded. Article 10(2)(l) provided that a credit agreement must state, in a clear and concise manner, the interest rate applicable to late payments, arrangements for its adjustment and any charges payable for default.

 

In its analysis, the CJEU emphasises the importance of mandatory information for consumers' informed decision-making, aimed at helping consumers in a weaker position vis-à-vis the bank. The CJEU then underlined the importance of informing consumers of the specific interest rate for late payment to enable consumers to be aware of the consequences of any late payment, information which is likely to influence not only the consumer’s decision to enter into the agreement, but also their ability to organise the repayment of the loan.

 

In view of the these reasons, the CJEU ruled that Article 10(2)(l) and Article 14(1)(b) of Directive 2008/48 must be interpreted as meaning that the withdrawal period provided for in Article 14(1) does not begin to run until the credit agreement does not specify, in the form of a specific percentage, the interest rate applicable in the event of late payment at the time of conclusion of the agreement, and until such information has been duly communicated to the consumer. With this, the CJEU confirmed its earlier position in C-33/20 (see our analysis here).

 

 

The second question tackled by the CJEU is interesting. The referring court asked directly whether the consumers’ potential intention to abuse their right can be considered here, in particular, that the consumer continues to use the vehicle until the national courts have ruled on the validity of the withdrawal and that the consumer refuses to pay compensation for the loss of value of that vehicle. The CJEU emphasised that a general legal principle is that EU law cannot be relied on for abusive or fraudulent ends. However, in the particular situation, the creditor cannot claim that the exercise of the right of withdrawal is unfair, as the withdrawal period has not, in such a case, begun to run. Under the circumstances, the credit cannot reply on the consumer’s improper exercise of the right of withdrawal provided for in Article 14(1).

 

The third question was what compensation should be provided to the creditor for the use of the vehicle. The CJEU referred to the 14th recital of the Directive, which stated that it was for the Member States to determine the conditions and arrangements following exercise of the right of withdrawal. Directive 2008/48 therefore grants Member States a margin of discretion leaving them to regulate matters relating to the return of the goods financed by the credit, which must follow the principle of effectiveness requiring that national provisions governing the consequences of the exercise of the right of withdrawal do not undermine the effectiveness and efficiency of that right to such an extent that it becomes impossible or excessively difficult to practice to exercise the right.

 

The CJEU, however, emphasised that compensation must be proportionate to the vehicle’s depreciation and its condition at the time of its return. Subject to the verifications to be carried out by the referring court, the CJEU was of the opinion that a method of calculation based solely on the difference in price between the purchase and resale of the vehicle, which includes factors unrelated to the use of that vehicle, such as commercial margins and resale costs – determined unilaterally by the car dealer – as well as value added tax, does not allow for the assessment of the depreciation of that vehicle resulting from its use by the consumer. In particular, if these circumstances are considered regardless of whether the vehicle has not been registered or used before the right of withdrawal is exercised. The CJEU concludes that this method, therefore, appears to impose on the consumer a burden resulting exclusively from the exercise of his or her right of withdrawal, and is likely to result in compensation that is disproportionate to the purchase price of that vehicle, making the exercise of the right of withdrawal impossible or excessively difficult to use in practice.

 

The CJEU ruled that Article 14(1) of Directive 2008/48 must be interpreted as precluding national case-law to calculate the amount of compensation for loss of value owed by consumer to the creditor by deducting from the sale price charged by the dealer at the time of the vehicle’s purchase the purchase price paid by the dealer at the time of the return of that vehicle, provided that that the method of calculation includes factors unrelated to the consumer’s use of that vehicle.

 

Fourth, the CJEU also addressed the question of payment of the interest for the credit agreement, ruling that Article 14(1) of Directive 2008/48 must be interpreted as not precluding national legislation under which a consumer who, after withdrawing from a consumer credit agreement linked to a vehicle purchase agreement, is required to pay the interest provided for in that first agreement for the period between the payment of the loan funds to the seller of the financed vehicle and the date of return of the vehicle to the creditor or seller.

 

Finally, the CJEU explicitly confirmed that Directive 2008/48 must be interpreted as not harmonising completely the rules relating to the consequences of the consumer’s exercise of his or her right of withdrawal from a credit agreement linked to a vehicle purchase agreement.

Who is the average consumer? CJEU in Compass Banca (C-646/22)

On the 14th of November, the CJEU published its long-awaited decision on Compass Banca (Case C-646/22; we have previously discussed it here). In this case, the CJEU, for the first time, elaborated on who the ‘average consumer’ is, especially in light of the persistent critiques from behaviouralists, and further clarified the assessment of and the consequences for unfair commercial practices under the UCPD (Directive 2005/29/EC).

The case involves a commercial practice by the Italian company Compass Banca which the Court termed ‘framing’ – a term typically associated with a specific type of cognitive bias rather than a concrete commercial practice. Namely, Compass Banca presented an offer for a personal loan alongside an unrelated insurance product, leaving consumers with the impression that it was not possible to obtain the loan without taking out the insurance. In particular, there was no cooling-off period between the signing of the two contracts. Even though Compass Banca claimed that it was made clear to consumers that the loan was not contingent on the insurance, the Italian consumer authority requested a seven-day cooling-off period to be granted and, upon Compass Banca’s failure to comply, found the practice of framing an ‘aggressive’ and thus ‘unfair’ commercial practice under the UCPD. Compass Banca challenged this decision in court, which invited questions reaching the CJEU.

The CJEU’s ruling

The first question concerns the extent to which behavioural insights about individuals’ cognitive biases should inform the concept of the ‘average consumer’, a notion that lies at the heart of the UCPD as a benchmark for assessing the effects of a particular commercial practice on consumers’ decision-making processes. Such an explicit reference makes this (abstract and somewhat academic) question not merely ‘hypothetical’ and justifies its admissibility (paras 37-39). Probably unsurprisingly, however, the Court avoided adopting academic terms like ‘homo economicus’ and ‘bounded rationality’ which were used by the referring court. This, of course, does not really make a difference to the substantive reasoning.

With reference to recital 18 of the UCPD (which came from the Court in the first place), the Court restated that the ‘average consumer’ is an individual ‘who is reasonably well-informed and reasonably observant and circumspect, taking into account social, cultural and linguistic factors’. The CJEU highlighted the nature of the ‘average consumer’ as an objective criterion which is independent of any specific consumer’s knowledge, but ‘not statistical’, which nonetheless allows national courts to take into account ‘more realistic’ considerations when exercising their own faculty of judgment to determine the ‘typical reaction of the average consumer’ (paras 48-51, recital 18 UCPD). With this understanding, the Court continued to clarify the two prongs of the average consumer benchmark: ‘reasonably well-informed’ and ‘reasonably observant and circumspect’. (I read it as the former relates to obtaining information/its availability, while the latter to processing information/its effectiveness.) As to the former, in view of the trader’s mandatory information obligations, it should be understood as ‘referring to the information which can reasonably be presumed to be known to any consumer, taking into account the relevant social, cultural and linguistic factors, and not to the information which is specific to the transaction in question’ (para 52). The lack of information is thus not excluded from the assessment of the effects of a commercial practice. Here, I think the Court was indicating that being ‘reasonably well-informed’ does not require consumers to actively seek out material information that the trader is legally obliged to provide.

Similarly, the nature of being ‘reasonably observant and circumspect’ does not exclude considering the influence of cognitive biases, should such biases be likely to affect a reasonable average consumer to materially distort their behaviour (para 53). The Court then recalled its several cases which acknowledge that an average consumer may be deceived, may have varied levels of attention regarding different goods and services, may be subject to an erroneous perception of a piece of information and may be simply unable to understand the technical details in certain transactions (paras 54-56). While these cases were usually discussed as ‘deviations’ from the average consumer standard, the Court used them as ‘evidence’ to confirm that a ‘reasonably observant and circumspect’ consumer is not a perfectly or particularly observant and circumspect one (adverbs used by AG Emiliou in para 42). Nonetheless, the Court cautioned that the existence of constraints like cognitive biases does not automatically make them legally relevant and decisive in finding an unfair commercial practice: ‘it is still necessary for it be duly established that, in the particular circumstances of a specific situation, such a practice is of such a kind as to affect the consent of a person who is reasonably well-informed and reasonably observant and circumspect, to such an extent as to materially distort his or her behaviour’ (para 57). There seems to be a high bar for courts to apply behavioural insights. In all, the Court concluded by sticking to the classic definition of a rational consumer while accepting the possibility of constraints that can impair consumers’ decision-making capacity, such as cognitive biases.

The second question concerns whether the practice of ‘framing’ in this case is in all circumstances aggressive or at least unfair. First, the Court found that framing is not categorically blacklisted in all circumstances, since it does not correspond to any practice listed in the ‘complete and exhaustive list’ of Annex I (para 68). Second, the Court indicated that neither can framing be found aggressive, in most cases, when applying the general test under Art. 8 UCPD: there is no ‘harassment’ and ‘coercion’ in their usual meaning in daily language (para 72), and there is no ‘undue influence’ as framing ‘does not, as such, imply the existence of acts of pressure, even if that practice is likely to create a bias of framing’ (para 75). Third, it is still possible that a non-aggressive practice can be a misleading one in the sense of Arts. 6-7 UCPD. In this case, the Court noted that framing leaves consumers with the (misleading) impression that it was impossible to get a loan without taking out the insurance (para 80) – though Compass Banco has claimed otherwise (para 82). Ultimately, it is for national courts to assess the unfair nature of a commercial practice (para 83).

The third and fourth questions ask: Should framing be found unfair, do the UCPD and Art. 24(3) of Directive 2016/97 on insurance distribution preclude the national authority from requiring a cooling-off period to be granted in order to put an end to the unfair practice? Regarding Directive 2016/97 the question was answered negatively, as Art. 24(3) only requires the possibility of buying a good or service separately without the ancillary insurance (as a package). As to the UCPD, the Court held that while it precludes ‘a general or preventive obligation to comply with a certain cooling-off period’ in an ex-ante manner (para 91), it does not preclude national authorities’ ex-post ‘power to issue directions to that trader’ once there has been an established unfair commercial practice (para 92). However, the measure taken cannot restrict the freedom to provide services (per Art. 4 UCPD) and must respect the rights codified in the Charter of Fundamental Rights, in particular the freedom to conduct a business under its Art. 16 (paras 94-95). In this light, the principle of proportionality mandates that a measure is only acceptable when ‘there are no other equally effective means of putting an end to that practice which are less prejudicial to the freedom to provide services and the freedom of the trader concerned to conduct his or her business’ (para 96). In short, requiring a cooling-off period is fine, unless there are less intrusive alternatives. Here, the Court took a rather constitutionally informed approach to the enforcement of consumer protection, though one might wonder why consumer protection itself (Art. 38 of the Charter) was not brought to the balancing exercise.

Comments

This case adds an interesting (but definitely not conclusive) annotation to the controversial notion of the ‘average consumer’. The Court wants to keep the baby and the bathwater: the average consumer is indeed ‘observant and circumspect’ (which receives heavy criticism) but only ‘reasonably’ so (which allows considerable leeway and flexibility). While the Court stressed that its analysis was specifically made ‘within the meaning of [the UCPD]’, its reasoning would most likely have broader implications as the average consumer benchmark is creeping into other consumer instruments. 

To some, the Court has said nothing new in this case – nowhere in EU law has it ever committed to interpreting the average consumer as ‘homo economicus’. To others, this decision may be celebrated as a victory for behavioural law and economics. However, the wording of the decision suggests (‘an individual’s decision-making capacity may be impaired by constraints, such as cognitive biases’) that cognitive biases are not the decisive nor the only factors that can ‘impair’ a consumer’s decision-making capacity. (Here, I would prefer ‘influence’ over ‘impair’ as we don’t want to reinforce ‘observant and circumspect’ and marginalise ‘reasonably’.) Indeed, while behaviouralists have commendably challenged the predominant information paradigm for better consumer protection, it has been pointed out that their critique lacks a social dimension – how choices are shaped not only by our individual cognitive capacities but also by our interpersonal interactions, social practices, cultural preferences and institutional set-ups. In this regard, the Court made reference ‘to the fact that a loan applicant is normally in need, to the complexity of the contracts presented for signature by the consumer, to the concurrent nature of the combined offer and to the short period granted to take up the offer concerned’ (para 80) – which appears to be a list of factual and contextual factors that should be considered when ascertaining how ‘reasonably observant and circumspect’ the average consumer should be in this case. And this list clearly goes beyond cognitive biases (so does the list by AG Emiliou in para 40).

In any case, determining the average consumer’s typical reaction should not be reduced to an empirical exercise solely aiming for a realistic approximation of real-life consumer behaviour, even with the help of behavioural science and even AI. This is clear in recital 18 of the UCPD (‘in line with the principle of proportionality’, ‘taking into account social, cultural and linguistic factors’, ‘not a statistical test’) and from the Court. Instead, delegating national judges ‘to exercise their own faculty of judgement’ ultimately asks them the normative question of how much protection should be afforded to the consumers in our political economy. So if we take this normative dimension seriously, we can explore other ways to flesh out the benchmark beyond what was discussed in this case. But then, to what extent should realistic considerations inform this normative assessment? And what assumptions, insights, frameworks, theories or imaginaries should serve as the normative guideline for judges to fill in the definition? 

Like it or not, the average consumer is here to stay, and the debate is certain to persist. How ‘reasonable’ the average consumer should be expected to be, what and who should inform this definition, and therefore what the desirable level of consumer protection should be – these questions will continue to puzzle academic debates, judicial reasonings and even political processes. 

Airbnb hosts can (and should) be regulated locally, AG says

Earlier this month Advocate General Bobek delivered his opinion in joined cases C-724/18 and C-727/18 Cali Apartments. The cases involve two Parisian hosts who were sanctioned for letting their apartments via Airbnb without an authorisation required under national and municipal law. Unlike in the previous cases on the so-called platform economy (see, in particular:  Airbnb scores a victory before the Court of Justice), the questions referred did not revolve around the role of the platform operator. Rather, the Court was asked to define the limits of the freedom to provide services by the peer providers.

Background of the case

Under applicable French law letting apartments for short periods to a transient clientele is only allowed if a number of conditions are fulfilled. Firstly, it is necessary to declare this fact to the mayor of the relevant municipality. Secondly, and most importantly to the dispute, in municipalities with more than 200 000 inhabitants, change of use of residential premises is additionally subject to a prior authorisation. Detailed conditions for obtaining the authorisation are set out at the local level and may include offset requirements in the form of the concurrent conversion of non-residential premises into housing. The appellants, who were sanctioned for letting their Parisian properties to tourists without necessary authorisations, argued that requirements imposed under French law were contrary to Directive 2006/123/EC on services.

Opinion of AG Bobek

The Advocate General considered the answer to the first two questions asked by the referring court to be fairly straightforward. In his view, Services Directive does apply to the activities and rules in question and the national/municipal framework at issue should be qualified as an 'authorisation scheme' under Articles 9 to 13 of Services Directive. While the reasoning in this respect is rather convincing (note especially reference to Visser), it is hard to overlook a similarilty with another recent case, in which AG Bobek gave an opinion: C-393/17 Kirschstein. There as well, the AG argued for a broad reading of the act's scope - in that case Directive 2005/29/EC on unfair commercial practices - and centered his analysis on the substance of the act. As we reported on this blog, however, the Court did not share this opinion and considered the applicability question to be decisive (see: CJEU in Kirschstein...).

Having established the applicability of Service Directive to the Cali case, the AG moved to assessing the compatibility of the relevant authorisation scheme with Services Directive. Here, a distinction was made between the benchmark for evaluating the need for establishing such a scheme in the first place and more specific conditions of such a scheme. Following the opinion, the former should comply with Article 9 of Services Directive, while the more extensive list of Article 10 is relevant for the latter. Crucially, both provisions contain a common core, requiring the applicable measures to be non-discriminatory, justified by an overriding reason relating to the public interest and proportionate to the objective pursued.

As regards the public interest objective, para. 97 of the opinion lists the different reasons put forward by participants to the proceedings: combating a housing shortage; offering affordable and sufficient housing; social housing policy; the protection of the urban environment; resisting pressures on land; the protection of consumers; the efficiency of tax inspections; fair trading; and the protection of the recipients of housing services. The relevant French authorities focused, in particular, on considerations of housing shortage and protection of urban environment, both of which the AG  accepted without hesitation. Other reasons, including the more consumer-oriented ones, thus remained on the sidelines of the analysis, even if their role in other context has not been discounted.

As regards the remaining questions, the AG took the view that establishment of the authorisation scheme at issue can be seen proportionate under Article 9(1) of the Services Directive, but questioned the proportionality and the non-discriminatory nature of its specific requirements under Article 10(2). According to the AG,  offsetting required by the City of Paris can in many instances actually defeat the purpose of asking for an authorisation in the first place (para. 127). This is especially the case for 'non-professional owners' of an extra flat in which they personally do not reside. Hence, the analysed offsetting requirement eventually reserves the access to market for short-term accommodation rental only for the big players, who would typically be legal persons or property developers. Having said that, the AG made clear that offsetting requirements need not always be contrary to EU law. Proportionality of such requirements could potentially be achieved by limiting them to premises above a certain size or to owners with more residential properties. A system of temporary authorisations not subject to offset, which would be periodically reviewed and potentially redistributed, could be another option to consider (para. 134).

Concluding thought

Two elements transpiring from the AG's proportinality analysis are additionally worth highlighting: the role of evidence-based assessment and of local diversity. According to the AG, the conclusion on both Article 9 and 10 requires that account is taken of "specific data" concerning the housing market in the cities where an authorisation scheme is envisaged. Such evidence, collected at the local level, is needed to assess whether the scheme does not go beyond what is necessary to achieve the objectives pursued (paras. 112, 121). The opinion concludes with a thought, towards which - as the AG notes - the Commission showed some "intellectual unease" during the hearing: that local diversity as to the specific authorisation conditions is not only permissible; it is even desirable (para. 136). While this finding might be true for limitations justified by the reasons related to housing shortage and urban environment, this is not necessarily the case for ones related to consumer protection.

* The author carries out a research project on consumer protection in the collaborative economy, financed by the National Science Centre in Poland on the basis of decision no. DEC-2015/19/N/HS5/01557.