The Application of Mandatory Rules by Arbitral Tribunals – Three Salient Issues
The Application of Mandatory Rules by Arbitral Tribunals – Three Salient Issues
When arbitral tribunals have to decide whether to apply or give effect to certain mandatory rules, they often face serious challenges. First of all, they are likely to receive no, or only very limited, guidance under the governing arbitration law, the relevant institutional arbitration rules (if any), and any domestic choice-of-law rules that they may consider relevant. Second, they may be called upon to ponder over basic theoretical questions such as the role of arbitrators (must arbitrators seek to defend certain state or public interests and, if so, which ones?) and the legal significance of the place of arbitration (including the ‘classic’ question of whether, or to what extent, arbitrators can be regarded as having a forum).
Another factor that complicates the legal analysis and, arguably, casts doubts on whether the question asked (can/should arbitral tribunals apply mandatory rules and, if so, which ones?) is at all appropriately framed pertains to the considerable heterogeneity of the rules that are commonly considered as mandatory norms. Indeed, these rules range from general public law norms, such as antitrust and currency exchange regulations, to decisions of the executive branch of the government (e.g. trade sanctions) to private law norms, in particular those aimed at the protection of weaker parties (consumers, workers, commercial agents etc.). One might wonder whether it is a realistic aim to elaborate a single set of rules that would offer sensible solutions for all these categories of mandatory norms.
Despite the above-mentioned difficulties, scholarly commentators and arbitral tribunals have reached some degree of consensus on a number of fundamental issues:
a) arbitral tribunals can apply mandatory norms (i.e. that the applicability of such norms does not render the relevant dispute non-arbitrable);
b) arbitral tribunals should or even must apply such rules under certain circumstances;
c) the mandatory rules of the lex contractus are, in principle always applicable;
d) it is wise to apply the mandatory rules of the seat in the light of enforcement considerations and;
e) the mandatory rules of third countries (i.e. countries other than those of the lex contractus and of the seat) may also, under certain circumstances (in particular, when there is a special connection with the parties or the dispute), be applied.
The purpose of this article is not to reiterate any of these points of agreement. Rather, it is to contribute to the academic debate by addressing controversial or unresolved issues and highlighting common misconceptions. Specifically, this study will discuss (a) the still controversial question of whether arbitral tribunals called upon to apply certain mandatory norms have to resort to the conflict norms of the seat, (b) the very widespread, but inaccurate, view that arbitral tribunals necessarily have to apply the mandatory rules of the lex contractus, and (c) the legal significance of the territorial scope of application of mandatory rules, an issue frequently overlooked by commentators.
This contribution is divided into four main sections. Section I provides a brief overview of the relevant issues, describing the current state of the academic debate and arbitral case law. Section II argues that arbitral tribunals do not have to apply the conflict norms of the seat. Section III shows that the scope of the lex contractus should be construed as covering only rules of contract law and that only the mandatory rules of contract law of the lex contractus are thus necessarily applicable. Section IV explains the legal significance of the territorial scope of application of mandatory norms for the purposes of determining whether such norms may or should be applied by arbitral tribunals.
I Overview of Issues
1 Application of Conflict Norms of the Seat
When an arbitral tribunal has to decide on the application of certain mandatory rules, it may ask itself whether it must or should decide this issue on the basis of the conflict of laws (choice-of-law) norms of the seat. The expression ‘conflict of laws norms of the seat’ (or simply, ‘conflict norms of the seat’) refers to those conflict norms that the courts of the relevant country are obliged to apply. For example, an arbitral tribunal seated in an EU Member State may ask itself whether it has to apply the Rome I Regulation on the law applicable to contractual obligations (which must be applied by EU courts) and, more particularly, its Art. 9 dealing with the application of overriding mandatory norms.
Scholars generally provide a negative answer to this question. Gary Born, for example, when discussing what he perceives as the ‘historic view that international arbitral tribunals were mandatorily required to apply the arbitral seat’s choice-of-law rules’, observes that ‘virtually all developed jurisdictions have abandoned this approach’. A similar view is expressed by other authoritative writers such as the authors of Redfern and Hunter on International Arbitration and those of Fouchard Gaillard Goldman on International Commercial Arbitration. The present author has expressed an opinion that is largely identical to such views.
To a very large extent, arbitral case law confirms the view that arbitral tribunals do not have to apply the choice-of-law norms of the seat. While there are of course arbitral tribunals that have applied the conflict norms of the seat (which does not necessarily imply that they considered themselves obliged to apply those rules), many arbitral tribunals have followed different approaches. For example, a number of tribunals have chosen to apply the conflict norms involved cumulatively, i.e. the conflict rules of the parties’ respective legal systems and, at times, those of the seat. Other tribunals have resorted to what they perceive as general principles of choice of law, again refusing to consider themselves bound by the seat’s conflict norms.
However, the opposite view nevertheless remains alive. In the specific context of the question of the application of mandatory rules, several authors either consider the conflict norms of the seat to be necessarily applicable or advocate their application. Calavros, for example, observing that the Rome I Regulation requires the application of the mandatory rules of the forum, concludes that those mandatory rules must thus be applied by arbitral tribunals, ‘provided that the Rome I Regulation is effective in the State of the place of arbitration’. Even Born, who is otherwise critical of the idea that arbitral tribunals should be bound by the conflict norms of the seat, argues that, when determining the applicability of mandatory rules, arbitral tribunals ‘should presumptively apply the choice-of-law rules of the arbitral seat’.
In arbitral practice, it is not uncommon for arbitral tribunals to apply the conflict norms of the seat. It is true that it is not always clear whether arbitral tribunals apply those rules because they are the rules of the seat or for some other reason. However, in the absence of any contrary indication, it is reasonable to consider that such determinations are based on the assumption that arbitral tribunals must or should apply the choice-of-law rules of the seat. Thus, decisions of arbitral tribunals seated in the Netherlands and Sweden applying Art. 7 of the 1980 Rome Convention can be regarded as illustrative of this approach.
2 Application of the Mandatory Rules of the Lex Contractus
The clearly predominant view in both scholarship and case law is that arbitral tribunals must apply the mandatory rules of the lex contractus, i.e. of the law governing the contract. Neither scholars, nor arbitral tribunals seem to have subjected this proposition to any restrictions arising from (a) the specific choice-of-law norms that may be held to be applicable or (b) the nature or type of the particular mandatory rule in question (in particular, no distinction is made based on the public or private law nature of the relevant rule).
It can be assumed that this understanding originates, at least in part, from the language used in the relevant provisions of domestic conflict of laws statutes, and the generally accepted interpretation of those provisions. In Switzerland, for example, Art. 19(1) of the PILA provides that ‘a mandatory provision of another law than the one referred to by this Act [i.e. in contractual matters, the lex contractus] may be taken into consideration, provided that the situation dealt with has a close connection with such other law’. Although this is not an inevitable conclusion, this sentence may reasonably be understood as implying that the mandatory provisions of the lex contractus are by definition applicable.
A very similar rule can be found in the 1980 Rome Convention. Art. 7 provides that ‘[w]hen applying under this Convention the law of a country, effect may be given to the mandatory rules of the law of another country with which the situation has a close connection…’. Like Art. 19(1) Swiss PILA, this provision can be understood as implying that the mandatory rules of the lex contractus are necessarily applicable. Additional support for such an interpretation is provided by the semi-official commentary to the Convention, the well-known Guiliano-Lagarde report. In their commentary, the authors explain that Art. 7 of the Convention reflects the established ‘principle that national courts can give effect under certain conditions to mandatory provisions other than those applicable to the contract’, clearly suggesting that the mandatory provisions of the lex contractus are applicable as such.
Art. 9 of the Rome I Regulation (which transforms the Convention into a piece of EU legislation, while introducing some changes) can similarly be read as supporting the view that the mandatory rules of the lex contractus are necessarily applicable. Like its predecessor, it singles out the mandatory rules of the forum (which courts can always apply). As regards the mandatory rules of other countries, it does not use language similar to that employed in the Convention, given that it only condones the application of the mandatory rules of the place of performance. However, the mere fact that Art. 9 specifically addresses the application of the mandatory rules of both the forum and the place of performance, while omitting to deal with the mandatory norms of the lex contractus, suggests that the applicability of the latter was considered to be self-evident.
Scholarly writers almost unanimously acknowledge the general recognition of the principle that the mandatory rules of the lex contractus must be applied. However, most of them, including Mayer, Zhilsov, Barraclough/Waincymer, and Calavros, fail to provide a specific justification for this principle, seemingly taking it for granted. Amongst the few authors who have sought to explain the applicability of the mandatory norms of the lex contractus, two basic approaches can be distinguished. Under one approach, the applicability of those mandatory norms results from the fact that they fall, presumably as a purely legal matter, within the scope of the lex contractus. Under another approach, the mandatory norms of the lex contractus are not applicable per se; whether or not they are will depend on the will of the parties and, more specifically, on the wording of the choice-of-law provision contained in the contract, if any.
Unsurprisingly, numerous arbitral tribunals have recognised the applicability of the mandatory rules of the lex contractus. Illustrative examples notably include the award in ICC case No. 4237 (affirming, in general terms, that arbitrators can apply trade usages and contractual stipulations only to the extent that ‘they do not deviate from the mandatory rules of the applicable law’); the award in ICC case No. 14046 (examining compliance of a contractual provision contained in a contract governed by Italian law with Italian antitrust law); and the award in ICC case No. 12127 (applying French competition law to a contract subjected to French law).
3 Territorial Scope of Application of Mandatory Rules
Like all legal norms, mandatory rules necessarily have a particular scope of application. They have, first of all, a material or substantive scope of application, i.e. they will only apply to certain categories of contracts or certain types of actions or conduct. Mandatory rules protecting commercial agents, for example, only apply to commercial agency agreements. Likewise, mandatory rules of competition law only apply to conduct that has the effect of excluding or distorting competition in a particular market.
Mandatory rules also have a territorial scope of application. Broadly speaking, one can distinguish between so-called ‘domestic mandatory rules’ and ‘international mandatory rules’. The former consist of norms of a particular legal system that are mandatory in a domestic context only, i.e. in connection with contracts that do not present any connection with countries other than the country that has enacted the relevant mandatory rule. The latter comprise norms that are mandatory in an international context, i.e. when the contract is connected to more than one country. It has to be noted that certain mandatory norms are by definition (only) applicable in an international context because they are not concerned with purely domestic conduct (e.g. currency exchange regulations, import/export restrictions, etc.). In some cases, the determination of the specific territorial scope of application of a mandatory norm may be difficult.
When a particular situation or contract falls within both the material and the territorial scope of application of a specific mandatory rule, then – to use the customary expression – that rule is applicable ‘on its own terms’. Applicability on its own terms (which can also be referred to as ‘internal’ applicability) is of course not generally viewed as sufficient for a court or arbitral tribunal to apply the mandatory norm at stake (with the exception of the mandatory norms of the forum, which are necessarily applied by the courts concerned). The crucial issue – which renders the analysis so complex and controversial – is in fact whether, or under what circumstances, certain mandatory norms that are applicable on their own terms must or should be applied by courts or arbitral tribunals (i.e. their ‘external’ applicability).
It is due to this focus on the external applicability of mandatory rules that legal scholars have largely neglected the analysis of the scope of application of such rules and, more particularly, of their territorial scope of application. This has led to some confusion and created misunderstandings. In fact, as has been explained above, many authors consider the mandatory norms of the lex contractus to be necessarily applicable, apparently implying that those rules are applicable regardless of whether their own territorial applicability requirements are met. Thus, Swiss competition law would in principle be applicable to a contract between a US and a Japanese party subjected to Swiss law, even though the contract is very unlikely to fall within its territorial scope of application (given that it most probably cannot have any effect on the Swiss market) – a solution which defies common sense.
As will be explained in Section IV, the determination of the territorial scope of application of mandatory norms constitutes a significant aspect of the broader question of the application of mandatory norms by arbitral tribunals and thus deserves closer examination. As will be shown, arbitral tribunals should never apply mandatory rules if the contract or situation does not fall within their territorial scope of application. In addition, the territorial scope of application of a mandatory norm may be relevant to determine whether such norm qualifies at all as an ‘overriding’ or ‘international’ mandatory norm, i.e. whether it meets the threshold generally required in the context of international litigation and arbitration.
II Arbitral Tribunals neither Must, nor Should Apply the Conflict Norms of the Seat
As has been explained above, the question of whether arbitral tribunals must or should apply the conflict norms of the seat is still rather controversial. Tackling this controversy, this Section answers both questions in the negative, i.e. it argues first that arbitral tribunals are not obliged to apply the seat’s conflict norms and, second, that it is generally not advisable for them to do so. The first argument finds support in the absence of any relevant statutory or institutional requirements (1). The second argument is based on the observation that the application of the conflict norms of the seat would be incompatible with established choice-of-law principles in the field of international commercial arbitration (2) and the fact that domestic conflict rules governing the application of mandatory rules are likely to be unsuitable in the context of international arbitration (3).
1 Absence of Statutory or Institutional Requirements to Apply the Conflict Rules of the Seat
It would, in theory, not be impossible for arbitration laws and institutional arbitration rules to require arbitral tribunals to apply the choice-of-law norms of the seat. The arbitration statute of country X could indeed provide for the mandatory application of X’s domestic conflict norms, or a specific part thereof (e.g. the part dealing with contractual obligations). A similar approach could, in principle, also be adopted by the drafters of institutional arbitration rules.
However, as a matter of positive law, no such provisions are contained in either arbitration statutes or institutional arbitration rules. Most arbitration statutes contain one single article dealing with the law applicable to the merits, typically laying down two basic principles: (1) that the parties are free to choose the applicable law (party autonomy) and (2) that, absent a choice of law by the parties, arbitral tribunals enjoy broad discretion in the determination of the applicable law. No reference whatsoever is generally made to the domestic conflict norms of the seat (or any other country, for that matter).
The absence of any obligation to follow the conflict rules of the seat also characterises institutional arbitration rules. In fact, the rules of leading institutions such as the ICC, LCIA, ICDR, and VIAC do not provide for any such duty. Like arbitration statutes, they are typically limited to a statement of the two basic principles referred to above, i.e. of choice-of-law party autonomy and arbitral discretion.
2 Incompatibility of the Application of the Conflict Norms of the Seat with Established Choice-of-Law Principles in the Field of International Arbitration
As has been explained above, the two basic choice-of-law principles in the field of international commercial arbitration are party autonomy and arbitral discretion. As far as party autonomy is concerned, it is characterised by a particularly high level of such autonomy, insofar as the parties are generally entitled not only to choose whichever domestic law they deem suitable, but also to opt for the application of so-called non-national or transnational law (for the sake of simplicity, reference will be made hereinafter to ‘non-national’ law only). In fact, most arbitration statutes do not allow the parties to select the applicable ‘law’, but ‘rules of law’, a concept that is generally understood as encompassing non-national law. The ability of the parties to choose non-national law is, today, almost universally recognised.
The second principle, arbitral discretion, means that arbitral tribunals have considerable power in determining the applicable law. The specific level of arbitral discretion may vary from one jurisdiction to another. Under the laws of a few countries, arbitral tribunals enjoy particularly broad discretion, given that they are empowered not only to select a particular domestic law, but also to choose non-national law. In most legal systems, however, arbitral tribunals are free to choose the applicable law, but that law must be a particular domestic law. In fact, under the relevant statutes, arbitral tribunals must apply conflict norms (which they can freely choose), which implies that they will ultimately apply a particular country’s domestic law.
The application of the domestic conflict norms of the seat (or any other country) is likely to conflict with the well-established principles of party autonomy and arbitral discretion. In fact, as regards the former principle, it has to be observed that domestic choice-of-law rules in matters of contract do not generally allow parties to opt for non-national law. The Rome I Regulation, for example, recognises the principle of choice-of-law party autonomy, enabling parties to a contract to choose the applicable law freely. However, as the language used in the relevant provision suggests (the parties are free to choose the ‘law’ governing the contract), the parties must choose a particular domestic law, i.e. they cannot select non-national rules.
The application of the conflict norms of the seat would also undermine arbitral choice-of-law discretion. In fact, domestic conflict of laws statutes do not generally grant any discretionary powers to courts. Instead, they lay down rather detailed choice-of-law rules using various connecting factors. Under the Rome I Regulation, for example, there are specific conflict norms for different categories of contracts. For example, a contract for the sale of goods is, in the absence of a choice of law by the parties, in principle governed by the law of the country where the seller has his habitual residence, while a distribution agreement would in principle be governed by the law of the country where the distributor has his habitual residence. It is true that the Regulation contains a so-called escape clause, under which a different law may be determined to be applicable. However, this escape clause is not intended to confer broad discretionary powers upon courts (or arbitrators); rather, it is intended to apply in those (presumably very rare) cases in which the relevant connecting factors fail to lead to the application of the law of the country presenting the closest connection to the contract.
The incompatibility of domestic conflict norms with the principles of choice-of-law party autonomy and arbitral choice-of-law discretion is not merely a formal problem; it affects the substance of the parties’ interests. In fact, the principles concerned have not been elaborated randomly or without reason; rather, they serve specific underlying purposes. The parties’ ability to select non-national law (such as, for example, general principles of law, the UNIDROIT Principles of International Commercial Contracts, or rules common to the parties’ respective legal systems) is intended to address the perceived deficiencies of domestic laws and to allow the parties to select rules that they consider as neutral and/or well-balanced. Arbitral choice-of-law discretion has developed in response to the perceived rigidity and obscurity of domestic conflict norms and seeks to promote the fairness and predictability of arbitral choice-of-law determinations. The achievement of both of these objectives would be compromised if arbitral tribunals had to apply the conflict norms of the seat (or any other domestic conflict norms, for that matter).
3 Questionable Suitability of Domestic Conflict Norms Governing the Application of Mandatory Rules
As has been shown above, the application of the conflict norms of the seat may conflict with established principles of choice-of-law in international commercial arbitration. It is thus preferable if arbitral tribunals do not apply such rules or, at the very least, ensure that their application does not threaten the parties’ legitimate interests. Now, it could of course be argued that arbitral tribunals should only apply those conflict norms of the seat that deal with the application of mandatory rules, a suggestion that would entail some form of ‘dépeçage’ at the level of choice-of-law rules. However, this would not be a suitable approach. For one, such dépeçage could be a source of potential complications. Moreover, and more importantly, it is questionable whether domestic conflict norms addressing the applicability of mandatory rules are at all suitable for use by arbitral tribunals.
First of all, those conflict norms almost inevitably allow (or require?) domestic courts to apply the mandatory rules of the forum. An Italian court, for example, will always be entitled (and will generally make use of this right) to apply mandatory rules of Italian law. The well-known reason for this is that the Italian court is an organ of the Italian state, charged with the application of the law and the protection of the basic public interests of the Italian legal order. Quite clearly, the same cannot be said about an arbitral tribunal seated in Italy. It might be argued that such a tribunal, due to the dual contractual and judicial nature of arbitration, should take into account and protect certain State interests, but there appears to be no reason why the interests of the place of the seat should receive priority over those of any other country. It is true that enforcement considerations may ultimately prompt an arbitral tribunal to apply the mandatory rules of the seat, but it is also true that the internationally binding nature of annulment decisions of the seat is questioned with increasing frequency, i.e. the courts of a number of countries have on occasion enforced arbitral awards that had been set aside at the seat. Thus, the non-application by an arbitral tribunal of the mandatory rules of the seat does not necessarily preclude the enforcement of the award in third countries.
Second, as regards the mandatory rules of third countries (i.e. countries other than the forum), two basic justifications are generally provided for their application by domestic courts: first, the actual interest in their application of the State concerned and, second, comity, a principle that requires respect for the rules and decisions of other jurisdictions, notably to ensure peaceful and harmonious relations between the countries concerned. None of these underlying rationales is necessarily applicable in the context of international arbitration. In fact, if one views arbitration primarily (or exclusively) as a creature of contract, there is no compelling reason why arbitral tribunals should take State interests into account, whether those of the country of the seat or any other country. As to comity, this is plainly not applicable because arbitral tribunals are not charged with the task of preserving healthy relationships between the country of the seat and other countries.
Last, the relevant domestic conflict norms may not be suitable inasmuch as they might simply unduly restrict the right of arbitral tribunals to assess freely which mandatory rules, if any, to apply. Under the Rome I Regulation, for example, mandatory norms other than those of the forum can only be applied if (a) the rules are rules of the country of the place of performance and (b) those rules render the performance of the contract unlawful. In other words, the Rome I Regulation does not authorise the application of rules of countries other than those of the place of performance, even though they may be closely connected with the dispute. Moreover, the Regulation only allows the application of those mandatory rules of the place of performance that invalidate the underlying contract, i.e. it does not, for instance, authorise the application of protective provisions such as, for example, consumer protection laws or laws protecting commercial agents, which do not have an invalidating effect. Arguably, such restrictions may unduly interfere with the ability of arbitral tribunals to render appropriate decisions.
III Only the Mandatory Rules of Contract Law of the Lex Contractus Are Necessarily Applicable
As has been explained above, the vast majority of commentators are of the opinion that the mandatory rules of the lex contractus are necessarily applicable. They affirm the applicability of those rules in general terms, without drawing any distinctions between different categories or types of mandatory rules. In particular, they fail to draw a distinction between mandatory norms of public and private law. Thus, according to the prevailing opinion, if Indian law applies to the contract (either because it has been chosen by the parties or because it is otherwise deemed to be applicable), all Indian mandatory norms (for example, currency exchange regulations, trade restrictions and competition law rules) will necessarily be applicable.
While most scholars provide little to no justification for the application of the mandatory rules of the lex contractus, those that do offer either of two explanations. One explanation consists of the idea that, as a matter of legal principle, the applicable law (and this is not necessarily confined to the lex contractus, but also applies to other areas) extends to all rules of the relevant legal system, including its mandatory rules. The alternative explanation focuses on the intent of the parties and posits that the application of the mandatory rules of the lex contractus depends on such intent and, more particularly, on the wording of the choice-of-law provision contained in the contract (if any).
As will be shown below, both approaches are flawed or, at the very least, highly questionable. In fact, it is much more sensible to hold that the scope of the lex contractus is confined to rules of (private) contract law, thus excluding public law rules and private law rules outside the field of contract law (e.g. tort law). Also, considering that the application of mandatory rules rests upon the importance of the State interests concerned, the intent of the parties should not be a relevant factor, i.e. it should not be possible for the parties to either choose or exclude the application of such rules.
1 The Scope of the Lex Contractus Is Confined to Rules of (Private) Contract Law
It is true that domestic conflict of laws statutes do not generally specify that the applicable law refers only to the relevant area of private law (in contractual matters, to the relevant contract law). The Rome I Regulation, for example, specifies which issues fall within the scope of the lex contractus, without however expressly restricting such scope to rules of private law. Under Art. 12 of the Rome I Regulation, the lex contractus governs: (a) contract interpretation, (b) performance, (c) remedies, (d) the extinction of obligations, as well as prescription and limitation of actions, and (e) the consequences of nullity of the contract. A separate provision of the Regulation essentially extends the scope of the lex contractus to the issue of substantive validity.
While the Rome I Regulation does not expressly state that the rules falling within the scope of the lex contractus are only rules of (private) contract law, the nature of the issues governed by the lex contractus rather clearly suggests that the relevant rules are extremely likely to be rules of contract law. Indeed, one will, for example, not be able to find rules governing remedies for breach of contract outside the boundaries of the relevant contract law. One may, however, hesitate when it comes to the issue of substantive validity. Here, one may indeed consider that competition law rules that prohibit certain categories of restrictive agreements are in fact rules that govern the validity of a contract in the sense of the relevant provision of the Regulation. In principle, such rules could thus be considered as falling within the scope of the lex contractus.
This is, however, not a sensible approach. There are three – somewhat connected – reasons for understanding the lex contractus as exclusively comprising rules of contract law. The first pertains to the very object of the choice of law as a legal area or discipline. In fact, although this may not always be expressly stated, choice of law deals with the determination of the law governing legal relationships or situations of private law, which suggests that choice of law is merely concerned with designating the applicable private law. This is also implicit in the terminology used in a number of countries to refer to the broader field of conflict of laws (which covers jurisdiction and enforcement of judgments in addition to choice of law), namely the commonly used terms ‘private international law’ and ‘international private law’. In short, choice of law is not, as a matter of principle, concerned with the determination of applicable rules of public law.
The second reason pertains to choice-of-law methodology. It is based on the observation that the determination of the applicable law and the determination of applicable mandatory rules are two separate, entirely independent, issues. They are separate in the sense that different conflict rules or ‘standards’ are applied to determine the application of a particular governing law, on the hand, and the application of certain mandatory rules, on the other. They are independent insofar as the applicable law is determined without having regard to possibly applicable mandatory rules, and vice versa. Thus, the fact that, say, Chinese law is determined to govern a particular contract has no implications for the determination of possibly applicable mandatory norms other than those that are part of the lex contractus.
The third and last reason, which is connected to the second, relates to the fact that the determination of the applicable law and the determination of possibly applicable mandatory norms are based on entirely different considerations. The former focuses on the private interests of the parties and, more particularly, on the need to ensure predictability of the applicable law, which explains the quasi-universal recognition of choice-of-law party autonomy in matters of contract. The latter, on the other hand, focuses on the (public) interests of States in the application of specific rules that are considered instrumental for the preservation of vital public interests.
The present author is not the only, or first, person to express the view that only the mandatory rules of contract law of the lex contractus are necessarily applicable. Such a view appears to be implicit in the reservations that a few scholars have expressed vis-à-vis the principled application of all mandatory norms of the lex contractus. The approach advocated in this article has also been adopted by the arbitral tribunal in ICC case No. 7528. In that case, the tribunal refused to apply mandatory provisions of French law governing sub-contracts, arguing that the scope of the lex contractus was confined to general contract law.
2 The Parties Can neither Choose, nor Exclude the Application of Mandatory Rules
Some authors argue that whether the mandatory norms of the lex contractus apply or not essentially depends on the parties’ intent. Thus, they argue, it is primarily a matter of interpretation of the choice-of-law clause contained in the contract. Born, for example, maintains that a choice-of-law clause providing that the contract ‘shall be interpreted in accordance with the laws of State X’ does not in principle cover non-contractual mandatory provisions. He reaches the same conclusion in relation to a broader (standard) choice-of-law clause under which ‘[t]his contract shall be governed by the laws of State X’, arguing that non-contractual mandatory provisions (he specifically lists competition law, trade controls, and intellectual property rules) would not in principle fall within the scope of such a clause.
Other authors also attach importance to the parties’ actual or presumed intent. Derains, for example, claims that whether the mandatory norms of the lex contractus apply notably depends on whether it (i.e. the lex contractus) has been chosen by the parties (in which case they must be applied) or determined by the arbitral tribunal (in which case they may or may not be applied), thus attaching significance to the presumed intent of the parties. Even Mayer considers (negative) party intent to be relevant, insofar as he considers that the absence of an express exclusion of the mandatory norms of the lex contractus by the parties strongly militates in favour of their application.
The views of these authors are not correct. In fact, they disregard what has been explained above, namely that mandatory rules, by their very nature (and by reason of the public interests involved in their application), are outside the sphere of party autonomy. Thus, the parties should not be allowed to exclude the application of mandatory rules. Conversely, they should not be allowed to choose specific mandatory rules either. More generally, the parties’ intent or preference should not be a factor having any impact on whether certain mandatory rules are held to be applicable or not.
IV The Territorial Scope of Application of Mandatory Rules
1 Mandatory Rules Only Apply if the Contract Falls within Their Territorial Scope of Application
It may seem obvious that mandatory rules should only be applied if they are applicable on their own terms and, more particularly, if the territorial requirements for their application are met. However, as has been mentioned above, the categorical affirmation of the applicability of the mandatory rules of certain countries (in particular, of the seat and the lex contractus) by a number of authors appears to imply that mandatory norms could be applied even if those requirements are not met. It is thus important to clarify that satisfaction of the territorial requirements of a particular mandatory norm is a necessary prerequisite for its application and should constitute the first step in the analysis performed by arbitral tribunals.
In practice, this first step may often be the end of the analysis, rendering any further inquiry into the appropriateness of applying the mandatory norms of certain jurisdictions (which, as has been seen, may be a very complex exercise) unnecessary. For example, as Born suggests, it is very unlikely for a purely domestic Zambian transaction to fall within the territorial scope of application of either Swiss or US competition law (which may be the law of the seat). Also, an export ban adopted by the country of the seat is obviously not territorially applicable to a transaction that does not involve any exports originating from that country.
One may of course argue that by choosing a particular lex contractus, the parties have implicitly agreed on the application of all mandatory rules of such law, regardless of whether those rules are applicable on their own terms. For example, one could argue that, by choosing German law to govern their contract, a US and a Japanese party have agreed on the application of EU competition law. This is not, however, a sensible approach. Indeed, as has already been explained, the application of mandatory norms is outside the sphere of party autonomy and parties should thus not be allowed to affirmatively choose certain mandatory norms.
In practice, it may often be difficult to determine the exact territorial scope of application of a given mandatory norm, in particular as far as mandatory rules of private law are concerned. The mandatory French rule prohibiting parties from excluding judicial adaptation of contractual penalties, for example, does not specify its territorial scope of application. Thus, it is only certain that such a prohibition necessarily applies in connection with purely French contracts, i.e. contracts that do not present a connection with any other country, but whether it also applies (and whether a French court would thus necessarily apply such prohibition) in relation to a contract involving a foreign party (or two foreign parties) is a question that is not answered by the provision itself.
Uncertainty may also arise as to the extent to which a particular mandatory norm is applicable in an international context. Under the EU Agency Directive and domestic implementing legislation, for example, provisions protecting commercial agents in the event of contract termination by principals are expressly stated to be of a mandatory nature. This means that they are unquestionably mandatory in a purely EU context, i.e. in connection with contracts involving an EU principal and an EU agent. But what if the principal is not established in the EU? What if the agent is not established in the EU? And what if neither party is established in the EU? None of these questions finds an express answer in the Directive and it is thus for the CJEU to clarify its territorial scope of application.
2 Relevance of the Territorial Scope of Application for the Determination of the ‘Overriding’ or ‘International’ Character of Mandatory Norms
As has been explained above, one habitually distinguishes mandatory norms that are mandatory in a domestic context (only) from those that are mandatory in an international context. The former are generally referred to as ‘domestic mandatory norms’, while the latter are labelled ‘international mandatory rules’. Since this terminology might suggest that the second category of norms are norms of international law (in particular treaty law), it would in fact be more appropriate to use the terms ‘domestically’ and ‘internationally’ mandatory norms instead, emphasising the specific context in which the rules concerned possess a mandatory character (those expressions will be preferred for the remainder of this article).
It should be noted that the distinction between domestically and internationally mandatory norms is not entirely compelling, given that the latter concept is not particularly precise. In fact, there is no such thing as an ‘international context’; rather, there are different ways in which a particular transaction may be connected to a plurality of countries. For example, as has been explained above, it does not make much sense to affirm, in general terms, that the EU Agency Directive is applicable in an international context. Indeed, it may apply in some such contexts, but not in others, depending on the specific connections that exist with EU countries.
In arbitral practice, it is well-established that only ‘internationally’ or ‘overriding’ mandatory rules should be applied – domestically mandatory norms are essentially irrelevant. In light of the close connection between this particular threshold (overriding mandatory rules) and the territorial scope of application of mandatory rules, it is clear that the territorial scope will generally be decisive as regards whether particular norms meet the relevant threshold. Where this threshold is not met, the application of the mandatory rule concerned is excluded, and there is no need for any further inquiry or assessment.
This article has aimed to contribute to the academic discussion of the application of (substantive) mandatory rules by arbitral tribunals by examining selected controversial or neglected issues, namely: (a) the duty of arbitral tribunals to apply the conflict norms of the seat, (b) the applicability of the mandatory rules of the lex contractus, and (c) the relevance of the territorial scope of application of mandatory rules.
Specifically, this contribution has argued that (a) arbitral tribunals are neither bound by the conflict rules of the seat, nor generally well-advised to apply those rules, (b) it is incorrect to affirm in general terms that all mandatory rules of the lex contractus are necessarily applicable, the better view being that only the mandatory norms of contract law of the lex contractus are applicable per se, and (c) the territorial scope of application of mandatory rules is relevant, not only because it constitutes a necessary prerequisite for their application (contrary to what a number of authors suggest), but also because it frequently proves decisive in resolving any threshold issue that may arise.
* Markus Petsche (PhD) is associate professor at Central European University, Department of Legal Studies, and of counsel at petsche pollak attorneys-at-law (e-mail: firstname.lastname@example.org).
 Two clarifications need to be made with regard to the meaning of the term ‘mandatory rules’ (or ‘mandatory norms’) used in this article. First, this article only deals with the application of mandatory substantive rules, i.e. it is not concerned with the application of mandatory procedural rules (an issue that is far less controversial than the one discussed in this contribution). Second, unless the opposite is expressly stated, the terms ‘mandatory rules’ or ‘mandatory norms’ must be understood as referring to ‘international’ or ‘overriding’ mandatory norms, not to ‘domestic’ mandatory norms. For an explanation of this distinction, see s I(3).
 There does not appear to be a single arbitration statute that contains provisions addressing the application of mandatory rules by arbitral tribunals. For illustrative examples of legislation that is silent on this question, see, for example, UNCITRAL Model Law on International Commercial Arbitration art 28, English Arbitration Act s 46, Swiss Private International Law Statute art 187 and French Code of Civil Procedure art 1511.
 The rules of arbitration institutions are generally silent as regards the application of mandatory rules. See, for example, ICC Rules art 21, LCIA Rules arts 22(3) and (4), ICDR Rules art 31, VIAC Rules art 24.
 While domestic private international law statutes generally deal with the issue of the application of mandatory rules, the relevant provisions are typically rather ambiguous, providing limited practical guidance to decision-makers with regard to (a) the identification of mandatory rules and (b) the determination of whether a particular mandatory rule is applicable or should be applied in a given case. See, for example, Swiss Private International Law Statute art 19, which defines mandatory rules as those that reflect ‘interests that are legitimate and clearly preponderant according to the Swiss conception of law’. It provides for the application of the mandatory rules of the governing law and also allows for the application of the mandatory rules of a ‘foreign law’ that is closely connected to the case, provided that their application is appropriate in light of the aims pursued by such rules and the consequences of their application. By any standard, art 19 fails to lay down simple and user-friendly rules governing the application of mandatory norms.
 See, for example, Adeline Chong, ‘The Public Policy and Mandatory Rules of Third Countries in International Contracts’ (2006) 2 Journal of Private International Law 27–70, 31 (mentioning that typical examples of ‘international’ mandatory rules notably include ‘rules concerning exchange regulations, antitrust laws, and import and export prohibitions’, while observing that s 27(2) of the English Unfair Contract Terms Act also constitutes such a mandatory norm. See also, with respect to the specific context of the 1980 Rome Convention, Council Report on the Convention on the law applicable to contractual obligations OJ  C282/1 (the ‘Guiliano–Lagarde report’). Commenting on Art. 7 of the Rome Convention – which deals with the application of mandatory rules – the authors of the report mention ‘rules on cartels, competition and restrictive practices, consumer protection and certain rules concerning carriage’ as examples of mandatory norms.
 One argument presented in this article, namely the idea that one needs to distinguish between public and private law mandatory norms (and that, as far as the lex contractus is concerned, only the mandatory rules of contract law are per se applicable) suggests that the answer to this question is in the negative.
 See, for example, Pierre Mayer, ‘Mandatory rules of law in international arbitration’ (1986) 2 Arbitration International 274–293, 277 (observing that arbitrators have ‘nearly unanimously’ answered the question of the arbitrability of disputes involving the application of mandatory rules in the affirmative); Marc Blessing, ‘Mandatory Rules of Law versus Party Autonomy in International Commercial Arbitration’ (1997) 14 (4) Journal of International Arbitration, 23–40, 29–31 (focusing on Swiss arbitration law); Donald Francis Donovan, Alexander K.A Greenawalt, ‘Mitsubishi After Twenty Years: Mandatory Rules Before Courts and International Arbitrators’ in Loukas A. Mistelis, Julian David Mathew Lew (eds), Pervasive Problems in International Arbitration (Kluwer Law International 2006, 11–60) (examining the arbitrability of antitrust disputes under US law).
 See Mayer (n 7), 280–86; Constantin Calavros, ‘The application of substantive mandatory rules in International Commercial Arbitration from the perspective of an EU UNCITRAL Model Law jurisdiction’ (2018) 34 Arbitration International 219–240, 224 (apparently arguing that where the lex contractus is the law of the seat, the mandatory rules of the lex contractus/seat must be applied).
 See s I.B.
10 See s I.B. For example, Marcus Commandeur, Sebastian Gößling, ‘The determination of mandatory rules of law in International Arbitration – An attempt to set out criteria’ (2014) 12 SchiedsVZ 12–20, 16 (observing that, ‘[g]iven the chance of vacating an award in its forum, mandatory rules are almost considered as a matter of fact by many tribunals’); Andrew Barraclough, Jeff Waincymer, ‘Mandatory Rules of Law in International Commercial Arbitration’ (2005) 6 Melbourne Journal of International Law 205–244, 223–24 (‘Enforceability concerns should therefore give mandatory rules of the seat a strong claim to be applied, at least insofar as they reflect the relevant public policy’).
 See Barraclough, Waincymer (n 10) 227–235 (arguing that arbitral tribunals possess discretion in determining whether to apply the mandatory rules of third countries); Commandeur, Gößling (n 10) 16–17 (discussing the application of the so-called ‘special-connection’ test).
 See s I.C.
 Gary Born, International Commercial Arbitration (2nd edn, Kluwer Law International 2014) 2630.
 Ibid 2631.
 Nigel Blackaby and others, Redfern and Hunter on International Arbitration (6th edn, Kluwer Law International, Oxford University Press 2015) 222: ‘The seat of the arbitration is invariably chosen for reasons that have nothing to do with the conflict rules of the law of the place of arbitration. This has led to the formulation of a doctrine that has found support in both the rules of arbitral institutions and the practice of international arbitration – namely, that, unlike the judge of a national court, an international arbitral tribunal is not bound to follow the conflict-of-law rules of the country in which it has its seat.’
 Emmanuel Gaillard, John Savage, Fouchard Gaillard Goldman on International Commercial Arbitration (Kluwer Law International 1999, The Hague – Boston – London) 867–68.
 Markus Petsche, ‘Choice of Law in International Commercial Arbitration’ in Sai Ramani Garimella, Stellina Jolly (eds) Private International Law – South Asian States’ Practice (Springer 2017, Singapore, 19–37) 21–26 (discussing arbitral discretion in choice-of-law determinations and several approaches followed by arbitral tribunals in the exercise of such discretion).
 See, for example, Egyptian Company (buyer) v Yugoslav Company (seller) (Final Award, ICC Case No. 6281, 26 August 1989) (1990) 15 Yearbook Commercial Arbitration 96, 97–98; Manufacturer v Distributor (Partial Award, ICC Case No. 7319, 1992) (1999) 24 Yearbook Commercial Arbitration 141, 145; Licensor and buyer v Manufacturer (Interim Award and Final Award, 17 July 1992 and 13 July 1993) (1997) 22 Yearbook Commercial Arbitration 197, 203. Evidently, such an approach is only feasible where the different conflict norms all refer to the same law.
 See, for example, Indian Company v Pakistani Bank (Award, ICC Case No. 1512, 1971) (1976) 1 Yearbook Commercial Arbitration 128, 130; Seller v Buyer (Interim Award, ICC Case No. 6149, 1990) (1995) 20 Yearbook Commercial Arbitration 41, 54.
 Calavros (n 8) 225.
 Born (n 13) 2707.
 See Austrian Company (Seller) v Dutch Company (Buyer) (Award, 11 January 1982) (1983) 8 Yearbook Commercial Arbitration 158, 160 (applying, without expressly referring to it, Art. 7 of the Rome Convention); Distributor (EU country) v Manufacturer (EU country) (Final Award, SCC Case No. 158/2011) (2013) 38 Yearbook Commercial Arbitration 253, 269–70.
 See n 5.
 Ibid 26.
 Regulation (EC) No. 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I)  OJ L177/6 (Rome I Regulation) art 9(2).
 Rome I Regulation art 9(3).
 Mayer (n 7) 280 (stating, without any particular explanation, that the mandatory rules of the lex contractus are unquestionably applicable when (a) they have not been excluded by the parties and (b) one party has invoked them before the tribunal).
 A.N. Zhilsov, ‘Mandatory and Public Policy Rules in International Commercial Arbitration’ (1995) 42 Netherlands International Law Review 81–119, 109 (observing, without offering a personal opinion or analysis, that ‘most commentators agree that an arbitrator should apply the law as it stands, including its mandatory rules’).
 Barraclough, Waincymer (n 10) 219–223 (discussing in some detail – but without examining its foundation – the view that the mandatory rules of the lex contractus should in general be applied and, what is more, regardless of whether the lex contractus was chosen by the parties or determined by the arbitral tribunal).
 Calavros (n 8) 227 (‘It follows from the foregoing considerations [which are unfortunately very difficult to comprehend] that the mandatory rules of law of the lex contractus are in any event applicable for the resolution of the underlying dispute submitted to international arbitration’).
 See Commandeur, Gößling (n 10) 15 (noting that ‘the mandatory rules of that substantive law [the lex contractus] must be applied… [because] [t]he selection of the law of a particular jurisdiction is considered to naturally include the selection of mandatory rules and public policies of that jurisdiction’).
 See Born (n 13) 2708 (noting that ‘the question whether a particular mandatory rule is within the scope of the choice-of-law agreement will require interpretation of the choice-of-law agreement’).
 Syrian State trading organization, buyer v Ghanaian State enterprise, seller (Award, ICC Case No. 4237, 17 February 1984) (1985) 10 Yearbook Commercial Arbitration 52, 55.
 Company A (Italy) v 6 Respondents (Italy) (Final Award, ICC Case No. 14046) (2010) 35 Yearbook Commercial Arbitration 241, 243.
 Licensor (France) v Licensee (US) (Award, ICC Case No. 12127, 2003) (2008) 33 Yearbook Commercial Arbitration 82, 84.
 See, for example, Chong (n 5) 31.
 See s IV.A.
 This terminology is commonly used by authors from common law jurisdictions. See, for example, Chong (n 5) 48; Born (n 13) 2711 (noting that, ‘by their own terms, the public policies and mandatory laws of the seat may not be applicable to the parties’ dispute’).
 See Chong (n 5) 48 (observing that ‘[i]t is obvious that the fact that the foreign rule may be applicable on its own terms to the contract in question is not sufficient to render it applicable’).
 See ns 27–32.
 See s IV.A.
 See s IV.B.
 See, for example, Model Law art 28, English Arbitration Act s 46, Swiss Private International Law Statute art 187, French Code of Civil Procedure art 1511.
 See, for example, Model Law art 28(1): ‘The arbitral tribunal shall decide the dispute in accordance with such rules of law as are chosen by the parties as applicable to the substance of the dispute.’
 See, for example, Model Law art 28(2): ‘Failing any designation by the parties, the arbitral tribunal shall apply the law determined by the conflict of laws rules which it considers applicable.’
 ICC Rules art 21, LCIA Rules arts 22(3) and (4), ICDR Rules art 31, VIAC rules art 24.
 See, for example, Model Law art 28(1).
 See, for example, Blackaby (n 15) 206–11 [discussing the use of the lex mercatoria (a particular form of non-national law) in international arbitration].
 An example is French arbitration law. See French Code of Civil Procedure art 1511 which provides that, in the absence of a choice of law by the parties, an arbitral tribunal shall apply ‘the rules of law it considers appropriate’ (emphasis added).
 See, for example, Model Law art 28(2).
 See Rome I Regulation art 3, which provides that: ‘A contract shall be governed by the law chosen by the parties.’
 For recent scholarly commentary on this particular feature of the Rome I Regulation, see, for example, Markus Petsche, ‘The Application of Transnational Law (Lex Mercatoria) by Domestic Courts’ (2014) 10 Journal of Private International law 489–515, 493–94.
 Rome I Regulation art 4(1)(a).
 Rome I Regulation art 4(1)(f).
 Rome I Regulation art 4(3).
 Rome I Regulation art 4(3) provides that: ‘Where it is clear from all the circumstances of the case that the contract is manifestly more closely connected with a country other than that indicated in paragraphs 1 [which contains specific choice-of-law rules for different categories of contracts] or 2 [which provides for the application of the characteristic-performance test in the event that the applicable law cannot be determined under paragraph 1], the law of that other country shall apply.’
 See, for example, Blackaby (n 15) 201 (referring to the search for a neutral law in the specific context of State contracts and beyond).
 See, for example, Petsche (n 17) 26–28 (explaining arbitral choice-of-law discretion on the basis of the need to ensure the predictability and legitimacy of choice-of-law determinations). Based on a concrete example, the author explains the paradoxical idea that increased autonomy in the determination of the applicable law may, in individual cases, improve the predictability of such determinations.
 See Rome I Regulation art 9(2) (‘Nothing in this Regulation shall restrict the application of the overriding mandatory provisions of the law of the forum.’); Swiss Private International Law Statute art 18 (‘This Act is subject to those mandatory provisions of Swiss law which, by reason of their special aim, are applicable regardless of the law referred to by this Act.’).
 On this dual nature of arbitration, see, for example, Barraclough, Waincymer (n 10) 209–11 (discussing contractual, jurisdictional, and hybrid theories of arbitration).
 See Mayer (n 7) 283: ‘Such a distinction [between the lex fori and foreign law] is clearly not to be made by arbitrators, because they do not have a forum. They are not faced with domestic mandatory rules as opposed to foreign mandatory rules, but simply with mandatory rules external to the lex contractus… For this reason, the arbitrator’s position is more favourable to an objective weighing of the interests that militate in favour of applying the mandatory rules and the lex contractus, respectively: he views all laws as being of equal dignity.’
 For discussion of the relevant case law, see Blackaby (n 15) 634–638.
 See Chong (n 5) 35–40.
 Rome I Regulation art 9(3): ‘Effect may be given to the overriding mandatory provisions of the law of the country where the obligations arising out of the contract have to be or have been performed, in so far as those overriding mandatory provisions render the performance of the contract unlawful.’
 This approach is different from the seemingly more flexible one followed in the Swiss Private International Law Statute, art 19 of which merely requires the existence of a close connection between the situation and the law concerned.
 See s I.B.
 Rome I Regulation art 12.
 Rome I Regulation art 10(1): ‘The existence and validity of a contract, or of any term of a contract, shall be determined by the law which would govern it under this Regulation if the contract or term were valid.’
 See, for example, Gilles Cuniberti, Conflict of Laws: A Comparative Approach (Edward Elgar 2017, Cheltenham-Northampton) 1 (defining the object of choice of law as the determination of ‘the law governing private relationships’).
 There is, of course, an exception to this principle. That exception is precisely that choice-of-law rules also address the issue of the application of mandatory rules, many of which are rules of public law. However, arguably, such provisions are largely superfluous because the relevant norms may be applied even without an express legal basis in the relevant choice-of-law principles.
 The term ‘standard’ is used here because it is debatable whether a provision such as Rome I Regulation art 9 can properly be considered as a conflict norm. See also Cuniberti (n 70) 19 (explaining that mandatory rules ‘are applied directly, as opposed to other substantive rules which only apply if designated by the relevant multilateral choice of law rule’).
 See, for example, Cuniberti (n 70) 18 (explaining that choice-of-law rules do not in principle take into account State interests because ‘most private law disputes are only concerned with private interests’).
 Ibid 19.
 See, for example, Nathalie Voser, ‘Mandatory Rules of Law as a Limitation on the Law Applicable in International Commercial Arbitration’ (1996) 7 American Review of International Arbitration 319–357, 339–340 (‘[T]here is no justification for assuming that the mandatory rules of the lex contractus have a special and paramount position and that therefore the interests of the state that provides the lex contractus have to be respected with less scrutiny than the interests of other states, simply because they are part of the chosen law.’).
 Sub-contractor v Contractor (Partial Award, ICC Case No. 7528, 1993) (1997) 22 Yearbook Commercial Arbitration 125.
 Ibid 128.
 Born (n 13) 2708.
 Yves Derains, ‘Public Policy and the Law Applicable to the Dispute in International Arbitration’ in Pieter Sanders (ed), Comparative Arbitration Practice and Public Policy in Arbitration, ICCA Congress Series vol 3 (Kluwer Law International 1987, 227–256) 244 (stating that, ‘when the parties have chosen a law to apply to their contract, it will be concluded from the outset that the arbitrator has to apply the mandatory rules of that law’).
 Ibid 245 (noting that in such a case the mandatory provisions of the lex contractus do not enjoy any special status when compared to the mandatory rules of other countries).
 Mayer (n 7) 280.
 That one cannot derogate from, or exclude the application of, mandatory rules is actually part of the very definition of the concept of mandatory rules. See, for example, Rome I Regulation art 9(1), which defines overriding mandatory norms as ‘provisions the respect for which is regarded as crucial by a country for safeguarding its public interests, such as its political, social or economic organisation, to such an extent that they are applicable to any situation falling within their scope, irrespective of the law otherwise applicable to the contract under this Regulation.’ (emphasis added).
 Born (n 13) 2711.
 French Civil Code art 1231-5 provides:
‘Where a contract stipulates that the person who fails to perform shall pay a certain sum of money by way of damages, the other party may be awarded neither a higher nor a lower sum.
Nevertheless, a court may, even of its own initiative, moderate or increase the penalty so agreed if it is manifestly excessive or derisory. Where an undertaking has been performed in part, the agreed penalty may be reduced by a court, even of its own initiative, in proportion to the advantage which partial performance has procured for the creditor, without prejudice to the application of the preceding paragraph. Any stipulation contrary to the preceding two paragraphs is deemed not written.’
 Council Directive of 18 December 1986 on the coordination of the laws of the Member States relating to self-employed commercial agents (86/653/EEC) (Agency Directive)  OJ L382/17.
 See Agency Directive art 19: ‘The parties may not derogate from Articles 17 and 18 [pertaining to the payment of an indemnity or compensation upon termination by the principal of the agency contract] to the detriment of the commercial agent before the agency contract expires.’
 The Court has answered the first question in the affirmative, while providing a negative answer to the second one. See Case C-381/98 Ingmar GB Ltd v Eaton Leonard Technologies Inc  ECR I-9325 (first question), Case C-507/15 Agro Foreign Trade & Agency Ltd v Petersime NV (second question).
 See n 88.
 See, for example, Grantor of exclusive distributorship v (Former) exclusive distributors (Final Award, ICC Case No. 6752, 1991) (1993) 18 Yearbook Commercial Arbitration 54, 56.