Stabilisation Provisions in Long-Term Mining Agreements


The mining industry is well acquainted with the risks of investing in foreign countries. Traditionally, the most acute form of political risk attending a foreign mining project has been nationalisation – the risk of the host government taking ownership of the project and evicting its foreign owner without fair compensation. However, the risk of nationalisation (a form of direct expropriation) has declined in recent decades, in no small part because of the growth of the global system of investor–state dispute settlement and the fact that most governments are now aware they risk being brought before an international tribunal if they seize foreign investments in their territory. Instead, the greatest political risk for a foreign mining project today is that, rather than take it over, the host state will destroy the project or render it economically non-viable by changing the laws and regulations applicable to it. Common examples include the sudden introduction of in-country beneficiation or ‘value-adding’ requirements, increases in royalties and taxes, and drastic changes to labour and environmental regulations. In extreme cases, the host state may even pass laws that purport to void certain provisions of the investor’s contract or subject the contract to unilateral variation by the host government.

Drawing on experiences of international oil companies during the twentieth century, mining companies have for some years now sought to manage the risks of legislative change by including special provisions – known as stabilisation clauses – in their long-term agreements with host governments.[2] There is considerable variation in how these clauses are drafted and the results they produce: at one end of the scale are stabilisation clauses that purport to ‘freeze’ the law entirely;[3] at the other end are clauses that expressly contemplate or permit change, subject to requirements of consultation and compensation where the change has a material adverse impact on the foreign party. Whatever form they take, it is clear that stabilisation clauses can provide meaningful protection to mining companies, provided they are backed up by a right to refer disputes with the counterparty government to international arbitration.

There is a considerable body of literature concerning the historical, theoretical and socio-economic dimensions of stabilisation clauses, particularly in the petroleum context.[4] This chapter has a different vantage: it is about the practice of stabilisation clauses – how to draft them and how to operate them in the real world of international mining projects.[5]

How to draft an effective stabilisation clause

Types of stabilisation clause

To understand how to draft a stabilisation clause, it is first necessary to identify certain key words and expressions used by stabilisation specialists, as they signal important distinctions in the procedural and substantive rights that a stabilisation clause creates.

First, and at the highest level of classification, a distinction must be drawn between stabilisation clauses of a fiscal nature and stabilisation clauses of a legal (or regulatory) nature. Fiscal stabilisation clauses relate to government revenue: taxes, royalties, duties and other imposts.[6], [7] In contrast, legal stabilisation clauses are broader, covering laws and regulations of a non-fiscal character, such as the statutes that govern operations at the project site on a day-to-day basis (mining laws, labour laws, environmental laws, etc.). Fiscal stabilisation clauses are the more common of the two varieties – fiscal stability terms are widely granted for major industrial projects, including mines, across a wide range of polities (from developed countries to least developed countries). This is because governments understand that tax and royalties are levers they can easily pull to encourage foreign investment in their territory.[8] Legal stabilisation clauses are less common, both because the link between regulation and economic viability is less obvious from a policy perspective and because governments are generally reluctant to contract with foreigners on terms that limit (or could be seen to limit) their sovereign prerogatives. But legal stabilisation clauses are often included in agreements for major, long-term mining projects, as we shall see.

Second, there are two main types of stabilisation clause: freezing clauses and equilibrium clauses. The term ‘freezing clause’ is used to describe any stabilisation clause that directly or indirectly purports to create a rule that the law applicable to the contract cannot be changed by the host state for an agreed period.[9] By way of further classification, the direct form of a freezing clause is where the contract contains a provision that expressly states that the law applicable to the contract will be the law as it is when the contract is executed and will remain the same for the term of the contract. The contract will either itemise the applicable statutes and regulations or contain a general statement of the lex lata (such as ‘the laws in force at the date of this Agreement’). As Burnett and Bret note:

[b]y incorporating a specific law by reference as a term of the contract, these clauses transform it into a lex specialis governing the parties’ contractual relationship for the life of the contract. By doing so, the parties agree that they cannot invoke a change to that law to excuse a breach of contract.[10]

The indirect form of a freezing clause may be observed where the contract specifies that it may only be varied or amended with the consent of all contracting parties. This type of provision – known as an ‘intangibility clause’ – can have a freezing effect insofar as it may limit the state’s power to effect a unilateral change of contract terms through the exercise of legislative power (though this result should not be assumed, for reasons explained below). The body of practice on intangibility clauses extends back over a century, to the seminal Lena Goldfields case,[11] and is an important point of reference for lawyers advising in this field.

Conversely, the term ‘equilibrium clause’ is used to describe a stabilisation clause that does not purport to prevent the host state from changing the law, but rather only to require that if the host state does change the law and the change adversely affects the foreign investor’s position under the contract, the parties will be bound to negotiate a solution that restores the foreign investor to the economic position it was in before the change in law took effect. Sometimes, the clause will specify compensation as the sole remedy, but the more common approach is to bind the parties (company and host government) to negotiate adjustments to the wider contract, or to other legal instruments, to restore the overall economic position of the company. Hence, clauses of this kind are sometimes also referred to as ‘adaptation’ or ‘renegotiation’ clauses,[12] the emphasis being more on their procedural character than the substantive result they produce. The natural question to ask is whether a clause of this nature is enforceable: clearly, if all it amounts to is an agreement to agree, an equilibrium clause will fail under many legal systems. But, as the tribunal noted in Aminoil v. Kuwait, ‘[a]n obligation to negotiate is not an obligation to agree.’[13] That is why it is essential that the mechanics of an equilibrium clause – the trigger, the negotiation procedure, and the subject-matter parameters of the negotiations – be clearly stipulated and that the clause is linked to a binding dispute resolution procedure, which does not depend upon party consent for its result (i.e., arbitration).

As with all aspects of international contracting, there are observable trends in the use of these types of stabilisation clause. While there are variations across markets and countries, it is fair to say that the equilibrium clause is prevalent today. Though freezing clauses are arguably the ‘most efficient form of contractual stabilisation’,[14] they are in decline for various reasons, including constitutional frailty and the perception that they are difficult to enforce because they impose too obvious a fetter on the host state’s ‘right to regulate’ (concerns about which are a common theme in the scholarship and public discourse on foreign investment).[15] It is more likely, therefore, that readers of this chapter will be called upon to draft or operate equilibrium-type stabilisation clauses than freezing clauses, and it is for this reason that we will focus on them.

Basic elements of a stabilisation regime

The basic elements of a stabilisation regime are:

  • the stability undertaking;
  • the trigger;
  • the notice;
  • the procedure for negotiations;
  • the procedure for enforcement; and
  • the remedies available.

A stabilisation clause of the equilibrium variety is meant to have dynamic operation: to absorb change, rather than prevent it. For the clause to work in a dynamic fashion, these basic elements must all be present and the links between them must be carefully constructed. If these elements are properly assembled, the clause should serve not only to protect the company from the adverse effects of changes in law, but to incentivise the host state to refrain from changing the law in the first place. It is often said that the incentivising effect of a stabilisation clause is its main virtue.

In the context of an oil and gas project, the appropriate place for the stabilisation regime is within the concession agreement or production-sharing contract (PSC) with the host state. In contrast, in the mining sector, where production-sharing arrangements are less common, the stabilisation clause will normally be included in a host government agreement (HGA), mining development agreement (MDA) or other state agreement.

Stability undertaking

The first element is a substantive promise. Whatever the intention of the clause – be it to freeze the law or establish a system for the restoration of economic equilibrium – it must contain a clear undertaking from the host government to provide stability to the company. Structurally, the contractual purpose of a stabilisation undertaking is to serve as a long-stop to the negotiation process provided for under the provision: if the negotiations are unsuccessful, and the government therefore fails to meet its undertaking not to cause the company to suffer the burden of a change in law, the change in law will then crystallise as a breach of contract. This, in turn, provides the company with a contractual cause of action for the purposes of arbitration.

In most contracts, stability undertakings are given in a sub-clause of the wider stabilisation provision (often the first sub-clause), the effect of which is to make the undertaking a term or condition of the contract; occasionally, the undertaking may be given in the form of a warranty.[16] If the stabilisation clause is of the ‘freezing’ variety, the undertaking will normally be narrow, and confined to tax law (in the case of a fiscal stability clause) or ‘laws applicable to the project’ (in the case of a legal stability clause), a limitation on subject matter may also be incorporated through a narrowed definition of ‘laws’ (such as a definition that refers to a list of existing statutes and regulations). Freezing undertakings also usually have limited time-scales: for example, the undertaking to refrain changing the ‘laws applicable to the project’ might be limited to the first five years, to provide the company with legal certainty for the initial pay-back period; alternatively, it might refer to a ‘stabilisation period’ which corresponds to the project’s debt schedule. Sometimes, the freezing clause will cover the entire period of a long-term contract, but this is unusual purely because it is not normally legally or politically feasible for the host government.

If the stabilisation clause is of the equilibrium variety, the undertaking will be focused on consequence rather than change. Undertakings of this kind typically contain some form of materiality threshold to ensure that the company does not have a stabilisation claim for de minimis impacts. In the mining context, materiality thresholds may be specific and linked to changes in laws relating to companies, mining operations, imports and exports, and employment. This approach may be observed in the IBA Model MDA:

GOVERNMENT undertakes that, during the Stability Period, it shall not in the case of:

13.1.1 legislation or regulations governing the regulation and management of companies, effect any changes thereto or to their application which would impose a requirement that the directors of COMPANY comprise a higher number of Country residents than that presently required by Section [x] of the Companies Act;

13.1.2 legislation or regulations governing the operation of mines or related activities but subject to Clause 12, effect any changes thereto or to their application which, individually or cumulatively, would have a Material Adverse Economic Effect;

13.1.3 regulations and procedures governing imports and exports within Country, effect any changes thereto or to their application which, individually or cumulatively, would have a Material Adverse Economic Effect;

13.1.4 legislation or regulations governing the terms and conditions of employment within Country, effect any changes thereto or to their application which would prevent COMPANY from:

(a) operating on a seven (7) days a week, twenty-four (24) hours a day, three hundred and sixty five (365) days a year basis; or

(b) negotiating with employees or relevant unions or engaging employees or terminating their contracts of employment in such a manner which would be likely to have a Material Adverse Economic Effect, individually or cumulatively.[17]

What is also notable about the stabilisation clause of the IBA Model MDA is that it treats different classes of legislation and regulation in different ways, imposing an effective freeze on some classes but allowing change in other classes, provided it does not have a material adverse economic effect.

A stability undertaking has considerable value outside the stabilisation clause itself. For example, if a contract contains a clear stability undertaking, it may give the company a basis for claiming reliance-based damages if the law does change and the company suffers loss not adequately addressed by the stabilisation mechanism. Equally, if the company and its investments are covered by a bilateral investment treaty (BIT) or treaty with investment protection provisions (coverage all mining companies should seek for foreign projects), the company will be able to point to the stability undertaking in the contract as evidence in a claim for frustration of legitimate expectations under the fair and equitable treatment (FET) standard of the treaty (legitimate expectations-based FET claims being one of the most common causes of action in mining-related BIT cases). The objective of securing a clear stability undertaking from the host government should therefore rank high in the priorities of the lawyers advising the company in negotiations with the host government.

Trigger clause

The trigger is the part of the clause that controls the key issue of whether a qualifying event has occurred. It is the gateway to relief and, as such, the way it is constructed is critical to the protection the clause ultimately gives the company. It should be assumed that the trigger will be heavily negotiated with the host government and its final wording will reflect a compromise. Most stabilisation clauses have triggers based on a definition of ‘change in law’.

Definitions of change in law are sometimes considerably more elaborate than this, covering changes to the way existing laws are interpreted or applied by the authorities of the host state. Express coverage for ‘interpretive change’ is important because it is often the case that stabilisation disputes arise not out of the promulgation of new laws but rather out of changes in the way the courts and other authorities construe and apply existing legislation. When the trigger is crafted around the existence of a legislative event, the clause will not respond to this kind of administrative or judicial act – even though it may be just as damaging as the enactment of a new statute. In drafting terms, there are two ways to cover changes resulting from administrative and judicial acts. The first is direct and involves expanding the definition of change in law to stipulate that it includes ‘any change in the way a Law is interpreted, assessed, applied or administered by any Government Authority’ (or words to this effect). The second is indirect, through the definition of law. Below is an example of the indirect approach, taken from a contract in Africa:

‘Law’ means any act, code, regulation or other written law promulgated by the State or other Government Authority (including with respect to taxation and fiscal matters), or any decision, judgment, pronouncement, order, decree, assessment or direction made or issued by the State or other Government Authority or other competent executive, judicial or administrative body.

Due regard must be had to the nature of the host state’s legal system and system of government, to ensure that the main forms of legislation are expressly identified. If the host state has a unique form of law, such as a special decree or executive ordinance, it should be included in the definition of law. The more closely the clause reflects the legal order of the host state, the better it will function as a control on change.

As noted above, the host state will often ask the company to accept a materiality threshold, such that a change in law will only trigger the stabilisation regime if it rises to certain level of materiality or results in a cost or loss above a stated monetary threshold. Drafting approaches vary considerably in this area: sometimes the threshold is built into the definition of ‘change in law’; in other contracts, the threshold is established by a separate definition (such as ‘material adverse economic effect’, ‘material adverse impact’ or ‘burden’), which is then deployed in the trigger provision. The principle risk in the monetary threshold approach is that it can expose the company to a situation where multiple changes in law occur, but no single one of them is sufficient to constitute a material change for the purposes of the contract. The only way to manage this risk is to draft in a rule of aggregation, whereby if two or more changes occur during a stated period (say, one year) and collectively result in a cost or loss above a certain threshold, the company will be entitled to make a stabilisation claim under the provision. Of course, it is still possible for the host state to manipulate the clause when it is written this way, but the risk of the company suffering loss through a series of insufficiently material measures is much reduced, and, ultimately, stabilisation clauses are about mitigating risk, not eliminating it entirely.

Materiality provisions sometimes also contain proximity requirements, such as a specification that the impact (loss or additional cost) be a ‘direct result of the change in law’. Care needs to be taken here as the effect of such requirements is to narrow the trigger. Generally speaking, the impact of proximity requirements is greater in the case of legal stabilisation clauses than it is in the case of fiscal stabilisation clauses: taxes and other revenue collection orders tend to be in personam measures (applying to the company as a class of tax payer) with an inherently direct monetary impact, with the result that a proximity requirement is more likely to be satisfied; legal stabilisation clauses, in contrast, cover a much wider range of instruments, which may be general (rather than personal) in their application and consequential in terms of their monetary impact. Thus, the imposition of a direct impact requirement in the materiality threshold should ordinarily be resisted by the investor where legal stabilisation is the objective, lest it narrow the trigger excessively or, at the very least, lead to unnecessary debate if a dispute arises.

Before moving on to discuss the next basic element (notice), it is worth emphasising that a properly crafted stabilisation clause will normally contain various other defined terms. The expression ‘government authority’ is one example: it is important that this expression (or an equivalent) be defined in a way that properly reflects the way the legal system of the host state is administered. If, for example, a state-owned enterprise has the power to make laws or regulations for mining, this body must be included in the definition of ‘government authority’. Other defined terms will need to be used to make the stabilisation clause compatible with the wider contract.

Notice requirement

As a stabilisation clause will ordinarily establish a procedure for consultations or negotiations – the object of which is dispute avoidance, rather than dispute resolution – it must clearly state how this procedure is initiated. The standard approach is to require that the company issue a formal notice, the purpose of which is to provide the host government with sufficient information to enable it to meaningfully engage in the consultation process that follows. Terminology varies but ‘request for negotiations’ and ‘request for consultations’ are common appellations; the title ‘notice of change in law’ is sometimes used. In crafting the contractual framework for such notices, the two main issues to consider are timing and content. While it may be tempting to see the notice provision as a mere formality, it must be understood that the notice provision does in fact have an important substantive dimension, because it will be linked to the trigger. The notice provision must therefore be given proper attention in the drafting process.

In terms of timing, the principal issue is how much time the company will have to give notice after a ‘change in law’ occurs: if the time frame is too short, the company may not have enough time to identify the change, analyse or model its impact, and, if the trigger contains a materiality threshold, determine whether the resulting cost or loss is sufficient to activate the stabilisation clause.[18] A time frame of months, rather than days, will usually be preferable. As to the content of the notice, the standard approach is to require that the notice identify the ‘change in law’ giving rise to the request for negotiations (and describe the ‘material adverse impact’ or ‘burden’, if a materiality threshold applies). In this sense, the notice provision will often serve as the functional trigger in the wider stabilisation clause (though the true trigger will normally be contained in the definition of ‘change in law’, as discussed above).

From the company’s perspective, it is preferable not to have a hard deadline for the notice and instead specify that it ‘shall be issued as soon as reasonably practicable after becoming aware’ of the change in law. It is also preferable for the notice to be preliminary in its quantification of impact – for example, ‘to the extent possible, estimate or quantify the Material Adverse Impact’ – and not to include a statement of the remedy it seeks in relation to the change, which is an additional feature of some notice provisions.

It must be kept in mind that it is the content of the notice that fixes the parameters of the negotiation and, in turn, the subject matter of any subsequent dispute: if no agreement can be reached and it is necessary to refer the change to arbitration, it should be assumed that the government will challenge the admissibility to arbitration of any claim or issue that was not included in the original notice. In this sense, a more detailed notice provision may pose more risk to the company than a clause requiring only the bare minimum of information.

Procedure for negotiations

The next basic element of the clause is the procedure for negotiations. Again, drafting approaches vary but the uniform objective is to bind the parties to a process for finding ways to address the impact of the change in law that has been notified by the company. The purpose is not to bind the parties to reaching an agreement (because agreements to agree are generally unenforceable). As the focus is on process rather than product, it is important that the clause name the branch of the host government that will participate in the negotiations and specify a time frame within which the discussions must commence (the time limit for completion of negotiations will ordinarily be specified in another clause, which gives the parties the right to refer the matter to dispute resolution if the negotiations are unsuccessful).

Some clauses go further and specify the subject matter of the negotiations, such as by stipulating the solutions or remedies that the parties shall consider. These more elaborate provisions are normally found in equilibrium clauses, where the solutions available to the parties will include adjusting the project agreement (or other legal instruments) to neutralise the economic impact of the change in law. The approach developed in the petroleum sector, where PSCs are common, is to negotiate changes to the parties’ production allocations, such that the contractor receives a greater share of production (and the government receives less) to compensate for the loss or cost caused by the change in law. The topic of remedies is discussed further below. For present purposes, the key point is that parties must take care not to confine the scope of the negotiable subject matter too narrowly – if, for example, the clause speaks only of negotiating adjustments to the HGA, and the HGA is unable to absorb an impact of the scale that has been caused by the change in law, the negotiation mechanism may fail to achieve its objective. This is especially relevant for mining companies, as the production-sharing arrangements that are commonly found in the petroleum sector (and that are well suited to adjustment for equilibrium purposes) are less often used in mining. The contract in question must be examined to determine whether its provisions are suitable to an adaption or adjustment solution before this remedy is built into the stabilisation clause.

A final element to consider is good faith. Generally, good faith means being reasonable. The negotiation provisions of stabilisation clauses often expressly require the parties to act in good faith during consultations. Approaches to good faith vary across legal systems and jurisdictions: in countries of the civil law tradition, a rule of good faith may apply automatically through the law of obligations (e.g., the Civil Code); in common law countries, there is generally no automatic duty of good faith, meaning it is necessary for the parties to contract into a duty of good faith if they wish to place themselves under this positive obligation. Even where the applicable law imposes a duty of good faith, it is often considered prudent to expressly refer to good faith in the negotiation provision of a stabilisation clause because, for the clause to be effective as a dispute-avoidance mechanism, a greater than usual degree of cooperation is required of the parties. This is especially so where the clause is of the adaptation variety and exhorts the parties to restore the economic equilibrium by agreeing amendments to the economic provisions of their contract. Best practice is, therefore, to include an express obligation of good faith in the negotiation provision of a stabilisation clause.


The basic remedy that underpins all stabilisation clauses is damages: monetary compensation is the fall-back solution for the host government in negotiations with the company (if other solutions are not feasible) and it is ultimately only a damages award that a company can enforce against a country. As damages are recognised as a remedy for breach of contract under every system of private law, it is strictly not necessary to specify the availability of damages in a stabilisation clause (provided the clause includes a clear stability undertaking). Nonetheless, it is not unusual for stabilisation clauses to mention damages specifically, if only to ensure that the option of last resort is clear to the parties and the arbitrators. However, where other remedies are to be available, it may be necessary as a matter of law to specify what they are and when they will be available. The main examples are specific performance, declarations and contractual adaptation.

In the analysis of whether it is necessary to specify the availability of these or other remedies in the stabilisation clause, the key factor is the governing law – not just the law that governs the contract (the lex contractus), but the law that governs the arbitration process that will be used if the parties cannot reach a negotiated solution (the lex arbitri). To take the example of a declaration, some legal systems do not allow arbitrators to issue declarations, or the position is at best unclear. But a declaration might be valuable in a stabilisation context as it may provide the company with the ability to explain its non-compliance with a new law to shareholders and other interested parties; it might also put pressure on the host government. Similarly, in the case of specific performance, if parties wish to empower their arbitrators to make an order compelling a party to do a certain thing (rather than simply pay damages), the law may well require that the parties expressly record their agreement to this remedy. Specific performance may be necessary to support solutions such as the grant of exemptions from new laws or the conferral of additional benefits to neutralise the impact of the change (though when such remedies are sought from arbitrators, damages should be sought in the alternative to cover a situation where the host government cannot or will not specifically perform).

The remedy where the need for explicit agreement is greatest is adaptation: few legal systems empower arbitrators to modify or rewrite contracts without a clear agreement to that effect, yet it is often the power to modify the economic provisions of a contract (such as a PSC or joint venture agreement) that underpins the operation of an equilibrium-type stabilisation clause. There is wide variation in the way parties build the remedy of adaptation into stabilisation clauses. However, as Burnett and Bret note:

Adaptation and renegotiation clauses rest on three basic elements: (i) a triggering event; (ii) a method to address the impact of the event (either through automatic adjustment based on third-party information, renegotiation of the contract terms, or international arbitration); and (iii) standard to be achieved by renegotiation.[19]

If adaptation is desired as a solution, the clause should include these core components. Even outside the context of adaptation, best practice is to specify the remedies that will be available to the parties and their arbitrators in the event the stabilisation clause is activated, especially where adaptation is to be available. How this is done depends on the overall structure of the clause, but a common approach is to specify the remedies within the negotiation provision, and then cross-refer to this provision in the clause that specifies the remedies available in arbitration (such that the arbitrators have the same remedies as solutions available to the parties). Below is an example from a contract in Asia, in which the stabilisation clause is Clause 5 and the arbitration clause is Clause 7:

5.1 The Government shall within 45 days of the receipt of the Request for Consultations meet with the Contractor and the Parties shall attempt to agree on what action is required to relieve the Contractor of the Burden which has been caused, or which is likely to be caused, by the Change in Law. The solutions that the Parties shall discuss in good faith include the Government taking steps to exempt the Contractor from the Change in Law, granting the Contractor additional rights or benefits of equivalent value to the Burden, modifying the terms of this Contract to off-set the Burden, and paying the Contractor monetary compensation in an amount equal to the Burden.


7.1 Without limiting the scope of remedies generally available in an arbitration under this Contract, the remedies that may be granted by an arbitral tribunal in a dispute arising under Clause 5 [the stabilisation clause] shall include a declaration that the Contractor is exempt from a Change in Law, an order of specific performance that compels the Government to take any action under Clause 5.1 and an award of monetary damages […]

The following is an example from a contract for a project in Africa:

(b) If at any time or from time to time there should be a change in legislation or regulations which in the opinion of Contractor, materially and negatively effects the commercial and fiscal benefits afforded to the Contractor under this Agreement, the Parties will consult each other and shall agree to such amendments to this Agreement as are necessary to restore as near as practicable such commercial and fiscal benefits which existed under the Agreement as of the Effective Date.

(c) If the Parties fail to agree on such amendments within one hundred and eighty (180) days following the date on which the change in question took effect, the matter may thereafter be referred at the option of either Party to arbitration under article 28 hereof. Following final determination by arbitration, this Agreement shall be deemed forthwith amended in accordance with such determination.

Though the two examples above are quite different, they both contain express links between the stabilisation clause and the arbitration clause. The advantage of this linkage approach is that it avoids debates regarding the remedies the arbitral tribunal can grant in a stabilisation dispute. Without such specifications, the scope of available remedies will be determined by reference to the lex contractus and lex arbitri. Although this default approach has the attraction of simplicity, it does result in some uncertainty, particularly where unusual remedies such as adaptation are concerned.

Sister provisions

Before turning to discuss how a stabilisation clause is operated and enforced, it must be emphasised that the enforceability of a stabilisation clause depends not only on the clause itself, but also upon certain other provisions. It is the combination of these provisions that forms the ‘stabilisation architecture’[20] of the contract. Fundamentally, these sister provisions address two key points that apply to every stabilisation clause: first, the best stabilisation clause in the world will be worth nothing if the company’s only way to enforce it is against the host government in its own courts, subject to its own law; second, as noted above, it is ultimately only a monetary judgment that can be enforced against a country. It is only where the contract (HGA, MDA, state agreement, etc.) provides the foreign company with a clear route to a neutral forum having the power to issue an enforceable award of money damages, that the stabilisation regime in the contract will be effective.


The use of international arbitration is widespread in the mining industry, at all levels of the value chain. Arbitration has many advantages but ultimately it is enforceability that gives it the edge over all other forms of dispute resolution in international contexts. Under the Convention on the Recognition and Enforcement of Foreign Arbitral Awards 1958 (the New York Convention) – the leading multilateral enforcement convention for international arbitration – a foreign arbitral award may be enforced in over 150 countries, including against government property (subject to the sovereign immunity rules of the place where enforcement is sought). No other dispute resolution process can offer anything close to this level of coverage or portability of outcome. An international arbitration clause therefore gives a cross-border contract ‘teeth’ in a way no other clause can. It is critical, therefore, that whatever stabilisation mechanism is included in the mining company’s contract with the host government, it must be accompanied by a clause that requires any disputes arising out of the contract to be resolved by arbitration in another country (as is the assumption in the example provisions above).

The arbitration clause need not be elaborate; quite the contrary, the best approach is to use one of the standard (model) clauses published by the leading international arbitration institutions, such as the International Chamber of Commerce (ICC) in Paris or the London Court of International Arbitration. Parties often select the International Centre for Settlement of Investment Disputes (ICSID), a World Bank-affiliated institution based in Washington, DC, though the ICSID system is unique and specialist input is needed to ensure an ICSID clause is effective.[21] With the exception of ICSID clauses, provided the clause clearly selects an appropriate set of rules and a seat outside the host state, it will likely be fit for purpose. The importance of the seat cannot be overstated, as it determines which courts will have ultimate control over the arbitration process and its product (the award).[22] Thus, it is imperative that another country is selected as the seat in a contract with a foreign government – otherwise, the host government may be able to use its ‘home ground advantage’ to influence the local courts to set aside the award if it loses in a stabilisation dispute with the foreign company, or even change the law to thwart the arbitral process.

If, as is usually the case in contracts for major mining projects, the law of the host state is expressly selected as the governing law of the contract, it is prudent to specify that the law of the seat governs the arbitration clause of the contract. This will ensure the host state cannot use its legislative power to void its agreement to arbitrate. The practice of including specific governing law clauses within arbitration clauses grew out of the Sulamérica decision of 2012, in which the English Court of Appeal held for the first time that the law of the seat (London) governed the validity of an arbitration clause in a contract that would otherwise generally be governed by the law of Brazil (which, if applicable, would have rendered the arbitration clause invalid).[23] The importance of this issue was brought into focus in 2017, when Tanzania enacted legislation that purports to expunge international arbitration clauses in mining contracts that the government has entered into with foreign parties.[24]

Governing law

Governing law is a critical factor in any cross-border agreement, as it supplies the rules that determine the validity of the written terms of the agreement and therefore constitute the primary control on its enforceability. For international disputes practitioners, the idea that the governing law clause is ‘boilerplate’ is anathema. In major natural resources projects, the law of the host state will normally be the lex contractus, because of the nature and location of the activity concerned (it is rare for a government to accept foreign law as the governing law for a contract that confers rights over natural resources in its territory). In the specific context of stabilisation, this creates a risk of circularity: when host state law governs the contract that contains the stabilisation clause, host state law can be changed to invalidate the stabilisation clause. Put another way, where host state law governs, the law that the clause is meant to stabilise can destabilise the clause itself. The solution to this problem was developed in the petroleum industry. For some time now, it has been standard practice in the petroleum industry to incorporate international law in the governing law clauses of concessions and other state agreements subject to host state law. These are sometimes referred to as ‘Libyan clauses’, as oil concessions granted by Libya were some of the first contracts to contain governing law clauses incorporating international law (and, when Libya nationalised its oil industry in the 1970s, claims under the contracts led to some of the most important arbitral awards of the twentieth century). A modern example of a Libyan clause can be found in the AIPN Model Joint Operating Agreement, the governing law clause of which reads as follows:

The laws of [the host State], to the extent consistent with international law, shall govern this Agreement. To the extent the laws of [the host State] are not consistent with international law, then international law shall prevail.

This language ensures that, if there is a dispute, the state cannot use its own laws to excuse its breaches of the contract, as Libya sought (unsuccessfully) to do in the seminal case of TOPCO v. Libya.[25] With a Libyan clause, the main principle of international law that is imported is pacta sunt servanda (the ‘sanctity of contracts’ principle). If the host state changes its law in a way that voids its contract with the investor, or purports to excuse a breach of that contract, a conflict with international law arises and the pacta sunt servanda principle overrides the inconsistent host state law, with the result that the host state is liable to pay damages for breach of contract. In the case of a stabilisation clause, if the host state passes a law that purports to invalidate or expunge stabilisation clauses in contracts that the government has entered into with foreign companies, the presence of a Libyan clause will mean that this change in law does not defeat the stabilisation clause of the contract. The foreign company will be entitled to claim damages for breach of the stabilisation clause, through the principle of pacta sunt servanda – the mechanics of the stabilisation clause might be compromised or rendered unenforceable by the change in host state law but the company’s basic right to compensation will stand.[26]

Sovereign immunity

Turning to the final sister provision, it is critical that, in a contract that contains a stabilisation clause, there be a waiver of sovereign immunity clause. This is because, unless the host state expressly waives its sovereign immunity, it may not be possible for the foreign party to collect if it receives an arbitral award in its favour under the contract. The waiver must be explicit, and it must cover enforcement and execution (many legal systems distinguish between the two steps and if both are not covered by the waiver, it may not be possible to collect on the judgment).

How to operate a stabilisation clause

There are two main phases in the operation of a stabilisation clause: negotiation and arbitration. The two phases are linked and so it is essential that when the negotiation phase is opened, careful consideration be given to the arbitration phase, should it be necessary. Of course, the precise steps that must be or may be taken in each phase will depend on the way the contract is written, but it is possible to identify certain themes and issues common to all stabilisation cases.

Negotiation phase

The negotiation phase has three main components: initiation, engagement and evaluation. There will be a fourth phase – documenting the agreement – if the negotiations are successful, but for present purposes we will focus on the first three phases, which are relevant regardless of outcome.

Initiating negotiations: giving notice to the host government

As noted above, the way the negotiation phase is initiated is through a formal notice, issued by the company to the host government (or the branch of it nominated in the contract). In preparing the notice, the first task is to ensure that all the content requirements in the contract are satisfied – if they are not, the host government may challenge the validity of the notice (it bears emphasising here that, in stabilisation cases, it is not unusual for the host government to take a strict approach to contractual interpretation and raise all possible objections to the procedure). Of these requirements, the trigger is the most important. All notice provisions require that the trigger event (the change in law) be identified, which means that, at a minimum, the legal instrument (or instruments) that constitute the change must be properly named. The laws should be described in accordance with the conventions of the host state’s legal system.

Assuming the trigger contains a materiality threshold, the notice will need to explain how this threshold has been met in the circumstances. While it may not be strictly necessary for the company to provide supporting documentation with the notice, there will be situations where it is desirable to do so. Thus, to establish materiality, the company may attach financial documentation to show how the financial impact of the change has been calculated. However, care should be taken to ensure that, in providing such information with the notice, the company does not inadvertently confine its claim. The notice should therefore expressly state that the assessments it contains are preliminary only, and that the company reserves its right to modify or update them following further analysis.

If the notice requires an indication of the remedy sought, the company must similarly take care not to confine the scope of the negotiations that follow. Before any remedy is proposed in the notice, a careful analysis must be conducted internally, with assistance from external counsel and other advisers. This is especially so where the remedy is an adaptation of the project agreement. For example, if the company and the government are in a joint venture with a production-sharing element, any proposal to reduce the government’s share of production must be fully modelled, including from a tax perspective. If the company is under time pressure, it is prudent to keep any statement of remedies as broad as possible and frame it in inclusive terms (e.g.,‘the company is willing to discuss remedies including . . . ’).

As the notice sets the tone for the negotiations that follow, it should be drafted in a way that is not unnecessarily adversarial. It needs to be remembered that, in most cases (except for cases where a true freezing clause is in place), the fact that the host government has changed the law will not mean that the host government has breached the contract – rather, it means only that the host government’s obligations under the stabilisation clause are activated. Ordinarily, a breach of contract will only occur if the host government fails to negotiate or does negotiate but fails to reach an agreement that makes good on its obligation not to require the company to suffer the adverse impact or burden of the change in law.

Engaging with the host government

To be effective in a stabilisation negotiation with a foreign government, a company must set out to strike the right balance between asserting its contractual rights and being sensitive to the position of the host government. The dynamic is quite different to a normal business negotiation between two companies, as the parties to a stabilisation negotiation have radically different legal personalities and systems of accountability.

The natural inclination of many companies is to ask the government to reverse the law. But it is no good to simply demand that the host government reverse the law, because doing so fails to acknowledge that law-making and law-changing are core functions of every sovereign and being seen to be exercising this prerogative at the direction of a foreigner may expose the host government to severe criticism by its opponents. Further, if the political cost of the solution is too high, it may not be durable (e.g., a subsequent government may overturn it, meaning the company ends up back at square one). The basic principle of engagement in a stabilisation negotiation is that the optimal solution is the one that has the highest commercial value and lowest political cost. This means the company must craft any solutions it proposes in a way that is sensitive to the political situation of the host government and reflective of its legal system. For example, rather than demand that the law be reversed, the company should ask the government whether it would be possible to create an exception for the project in another legal instrument – such as a ministerial regulation or executive order – which would avoid the need for the government to take the matter before parliament, the highest-cost venue for the government of a democratic polity.

In an equilibrium-type stabilisation context, the company should be aware that it will often be significantly more sophisticated than the host government in its understanding of the economics of the project and the contracts that govern it. As many of the officials who become involved in a stabilisation negotiation will not be familiar with the project, it may be necessary to educate them before engagement on commercial solutions can take place. The company should therefore come prepared with a presentation that provides an overview of the project and explains its operational, commercial and legal fundamentals. The company should compose its delegation in a way that ensures the right people from the company are present to deliver these messages and respond to any questions the host government may have. Care must be taken to avoid condescension but still provide the officials with the information they need to consider the solutions available and explain them to their masters.

In terms of the way the negotiations are conducted, there is a broad spectrum of possibilities. Certainly, a complex stabilisation negotiation will usually involve multiple rounds of negotiations, with exchanges of documentation in parallel. Often, the venue of these meetings will rotate between the host state and a neutral location, though it is often preferable to meet inside the host state’s territory so that the officials from other departments are on hand if their input is needed (as delegations are sometimes large, this also reduces the government’s costs).

Practice varies regarding whether stabilisation negotiations are conducted on a ‘without prejudice’ or open basis. Generally, the adoption of ‘without prejudice’ cover tends to enable a more constructive dialogue, and it is often used for stabilisation negotiations (even if, strictly speaking, no breach or dispute has yet arisen). However, it should be remembered that the parties are under a positive obligation to negotiate, so any ‘without prejudice’ arrangement should allow the parties to refer to the fact a meeting occurred, but not what was said and done at the meeting. In any case, it is usually wise to agree that the negotiations will be kept confidential, as this will make it easier for the government to engage with the company.

Evaluating the available solutions

The analysis of the available stabilisation solutions is obviously different for every case. But it is possible to identify certain considerations that bear on the evaluation of the available solutions in every stabilisation case. This section assumes that the negotiations have been successful to the extent that, although a final deal is yet to be done, the government has expressed a willingness to agree to one of the solutions proposed by the company. The key considerations are the following:

  • First, what is the objective value of the solution to the company? By this point, the company will have conducted a detailed internal evaluation of the solutions on offer, but it needs to be recognised that this evaluation will be affected by subjective factors, such as the company’s own view of the value of the assets and the way commodity prices are likely to behave, advice the company has received on legal aspects of the solution, and the rapport the company has built up with key members of the host government. These subjective factors need to be excluded to the extent possible. The objective question is what a hypothetical purchaser of the mine would consider that the solution is of equal value to the cost or loss the company suffered as a result of the change in law.
  • Second, what is the political cost of the solution to the host government? The company should have no interest in a solution that will be too costly from a political perspective, because if the political cost is too high for the host government the solution will be unlikely to endure. Rather, as noted above, the company should focus on the solution that strikes the right balance between commercial value and political cost, even if this would entail the company making some concessions.
  • Third, what would the company be entitled to if it rejected the solution and proceeded to arbitration claiming monetary compensation alone? This analysis must be as objective as possible and therefore requires the input of international counsel and valuation experts with experience testifying before arbitral tribunals. If the objective figure is significantly greater than the best solution the government will accept, serious thought may need to be given to commencing arbitration. As part of determining whether the company would be better off claiming damages, due regard must be had to the cost of arbitration (and subsequent enforcement proceedings), both in dollar terms and in terms of the impact it would have on the company’s relationship with the host government.

If the available solution passes through these filters, such that it is acceptable as a remedy for the change in law, the question becomes how to document the arrangement. While the approach will depend on the circumstances, care must be taken to ensure that all necessary provisions from the original agreement (including the sister provisions described above) are incorporated by express reference into the instrument that records the parties’ agreement to the stabilisation solution. The instrument that concludes the stabilisation negotiation must be at least as strong as the contract that contains the stabilisation regime.

Arbitration phase

If the negotiations with the host government are unsuccessful, the company will have three choices: commence international arbitration, commence litigation in the local courts, or do nothing. The case for doing nothing should always be given proper consideration, especially where the company is a long-term investor in the host state with no near-term plans for country exit. For the reasons outlined above, commencement of litigation in the host state’s courts is unlikely to yield the desired result. The choice will therefore ordinarily be between doing nothing and launching arbitration. However, before arbitration is commenced the company should ensure that there is no contractual requirement to pursue or exhaust local remedies before proceeding to arbitration. A well-drafted arbitration clause would state that the company is not bound to litigate in the local courts before it can activate the stabilisation mechanism in its contract with the host government. Many contracts are silent on this matter and careful consideration should be given to the effect local court proceedings could have on international remedies before any action is taken in response to the change in law. For present purposes, we will assume that the decision is taken to commence arbitration against the host state, without any action at the local level.

Notice of dispute

It is standard practice for contracts to require that, before any party may resort to arbitration, it must first issue a notice of dispute, and then negotiate with its counterparty for a minimum period. Where this requirement is not satisfied, the counterparty (the respondent to the claim) may raise objections to the effect that the claimants’ claims are not admissible because a pre-condition to arbitration was not satisfied. Stabilisation clauses sometimes contain express language to clarify that, where the dispute arises under the stabilisation clause, the general notice and negotiation requirements of the arbitration clause will be deemed satisfied. This reflects the fact that stabilisation disputes are different: by the time they get to arbitration, an extensive process of negotiation will normally have already taken place (and failed), making the usual step of giving notice redundant. If the stabilisation provision does not contain such language, the company will need to comply with this formality. Companies often issue notices of dispute towards the end of stabilisation negotiations, once it becomes apparent that no agreement will be reached (sometimes, the threat of arbitration does lead the government to reconsider its position and a compromise is reached at the eleventh hour).

Commencing arbitration

Under all sets of arbitration rules, the process must be initiated by the claimant issuing a notice of arbitration or request for arbitration. Usually, the rules do not require much detail on the substance of the matter – the notice or request is not meant to be a statement of claim (though, under some arbitration rules, the claimant may elect to treat it as such). It is, however, important to state in the notice that the contractual pre-conditions to arbitration have been satisfied: in a stabilisation dispute, this means the notice must identify the date on which negotiations with the host government were commenced and that they failed to produce an agreement. Basic information on the dispute must also be provided: the notice must refer to the contract (and normally attach a copy), extract the arbitration clause within it, cite the stabilisation clause and provide an overview of the events that triggered it (i.e., the change in law). In a stabilisation dispute, best practice is to attach the request for negotiations that was dispatched to the host government (ideally a copy with proof of receipt) and exhibit copies of the laws or regulations at issue. The notice of arbitration should also indicate the relief that the company seeks: damages, declarations, specific performance, etc. If the remedy of adaptation is sought, the notice should identify the provision of the contract or the lex arbitri that permits the arbitrators to grant this remedy. Some systems of arbitration require that the claimant include an indication of the amount in dispute. Any statement of the quantum of compensation sought should be framed as preliminary and subject to adjustment during the proceedings.

Commencing arbitration in a joint venture

Often, large-scale mining projects are undertaken by two or more companies in a joint venture arrangement. Unless the stabilisation clause is drafted appropriately at the outset, difficulties can arise in commencing a stabilisation arbitration in the joint venture context. This is because (1) not all parties to the joint venture may experience the same effects of a change in law (such as where the joint venturers have different tax structures); and (2) not all joint venture parties may have the same appetite for litigious action against their host government. Ideally, the stabilisation clause should allow for member of the joint venture to commence arbitration independently of the other members. The purpose is to ensure that the decision to commence arbitration is not subject to the will of one member of the joint venture. If the joint venture uses an incorporated joint venture vehicle, the joint venture agreement should similarly ensure the right to commence arbitration for stabilisation purposes does not need a unanimous decision. Finally, the clause may also include ‘drag-along’ or ‘tag-along’ provisions that bind other joint venturers to participate or give them the benefit of the outcome if they do not.

Interim measures

In international arbitration, it is possible for a party to seek urgent protection from the tribunal in the form of interim measures (also known as ‘provisional measures’).[27] Under some of the main institutional rules (such as the ICC Rules), interim measures may also be sought before the arbitral tribunal is constituted, from an ‘emergency arbitrator’, in truly exceptional cases.[28] In the context of a stabilisation dispute, as part of commencing arbitration a company should consider whether the circumstances require interim measures (for instance, to restrain the enforcement of a new law or regulation that could cause serious or irreparable harm to the company or its mining operations).

Although arbitrators vary in the way they evaluate requests for interim measures, most arbitrators require that the applicant for interim relief demonstrates that:

  • the arbitral tribunal has prima facie jurisdiction;
  • there is a prima facie basis to the claim;
  • the requested measures are necessary to protect the applicant’s rights (or to preserve the status quo and avoid aggravating the dispute);
  • the requested measures are urgent; and
  • the requested measures are proportional.

These requirements are usually applied ‘in the round’, rather than as a strict checklist.[29] In practice, the elements of necessity and urgency are given the most weight, and there is considerable overlap between the two.[30]

In preparing a request for interim measures in a stabilisation case, the emphasis should be on these two elements and how damages would not be an adequate remedy for the harm that the change in law will cause to the company (if the harm can readily be cured through an award of damages, the prospects of interim measures being granted are lower). Emphasis should also be placed on how interim measures would prevent the aggravation of the dispute and maintain the status quo. While it is normally difficult to obtain an interim order restraining a sovereign from enacting or enforcing a law, there have been cases where such orders have been made.[31]

Appointing arbitrators

Stabilisation disputes usually go before a tribunal of three arbitrators. The standard approach is for each party (or each side of parties) to appoint one arbitrator, with the third arbitrator (who serves as president of the tribunal) appointed by agreement of the parties or, failing that, by a neutral institution such as the ICC.

Many rules require, or at least permit, the party initiating arbitration to appoint its arbitrator in its notice of arbitration or request for arbitration. The appointment of arbitrators is an area where experienced international counsel can add real value, as they have the most current and direct exposure to the pool of men and women who serve on tribunals in international disputes. The selection of arbitrators for stabilisation cases requires that consideration be given to a range of factors, including the individual’s experience in stabilisation disputes, his or her knowledge of the relevant industry (mining), and the similarities and differences between his or her jurisdiction of origin and the legal system of the host state (a basic understanding of the host state’s legal system being an important requirement in a stabilisation case). The company should also consider the extent to which the candidate arbitrator has worked in government or has a record of being appointed by governments in other arbitrations: a person who places too high a premium on sovereign prerogatives may not be able to render an impartial decision on the application of a stabilisation clause. If unusual remedies such as adaptation are to be sought, thought needs to be given to how suitable the candidate would be for this specific task: mandating a stranger to rewrite a long-term contract is a major step for any business to take, and although the parties will have the opportunity to present their cases on how the contract should be changed, they still end up placing a great deal of trust in their arbitrators.

Written submissions

The first main written submission in an arbitration is the statement of claim or memorial of claim. This is where the company presents its detailed case on law and fact, articulating the remedies it seeks and why they are justified. In preparing the case, practitioners need to be aware of the unique nature of stabilisation disputes. As Burnett and Bret observe, ‘the success of a [stabilisation] claim will depend upon the specific factual matrix, including the terms of the commitment, related terms of the contract, and how it interacts with the applicable law more broadly.’[32] The statement of claim must address these key factors.

Ordinarily, the statement of claim must be accompanied by all evidence on which the company intends to rely, including expert evidence. In a stabilisation case, the input of local counsel is essential in this part of the process, as there will necessarily be a strong element of host state law in the dispute. It is for this reason that parties often file reports from legal experts (former judges, eminent professors) in stabilisation cases. Expert evidence will also be required to substantiate the economic aspects of the claim: how the change caused a material adverse impact on the project, the precise quantum of the cost or loss resulting from the change, and, if adaptation is sought, how the proposed modifications to the contract would remedy the situation. Experts play a crucial role in complex arbitrations and they must be selected with care.

As the statement of claim will normally be filed long after the change in law occurred, it must capture any developments in the interim. For example, if the law has changed further, and the adverse impact on the company has grown, the statement of claim should detail what occurred and seek appropriate relief. It bears noting here that, if a stabilisation dispute reaches arbitration, it is likely that other breaches of contract will have occurred as the relationship between the company and the host government has eroded. Stabilisation claims therefore often end up as one part of a larger dispute and it is important that the statement of claim cover all matters of contention between the parties.

As noted above, the remedies sought should be within the solutions available under the stabilisation clause. Damages are always available. In the analysis of remedies, the nuance is around specific performance and adaptation. If the clause provides for the host state to do certain things to relieve the burden caused by the change in law, the company may seek an order of specific performance to compel the government to do those things (with an alternative prayer for damages if the government cannot or will not perform as ordered). If the clause is silent, the availability of specific performance will need to be established by reference to the law of the contract and the law of the seat of arbitration. This applies equally to the remedy of adaptation: if the clause says the parties shall negotiate to agree changes to the contract that restore the economic equilibrium, the company may request that the arbitrators carry out this exercise; if the contract is silent in this regard, the availability of adaptation as a remedy will depend on the interaction of the lex contractus and the lex arbitri.

After the statement of claim is filed, the government will respond with a statement of defence. This document may include preliminary objections, such as challenges to the jurisdiction of the tribunal or objections to the admissibility of the company’s claim. It will also set out the government’s case on the merits. Each case is different, but it can be assumed the government will defend on the basis that no trigger event occurred (i.e., there was no change in law as defined by the contract, or it was not sufficiently material). Governments also often argue that the clause is invalid, as it constitutes an unlawful fetter on the state’s legislative power or sovereign discretion. The jurisprudence is generally against this argument and lawyers representing companies in stabilisation cases should familiarise themselves with the relevant authorities (namely AGIP v. Congo,[33] Sapphire Petroleum v. National Iranian Oil Co,[34] Aminoil v. Kuwait[35] and TOPCO v. Libya[36]).

Interpretation and analysis of stabilisation clauses by arbitrators

Given the political sensitivities surrounding stabilisation clauses (discussed above), arbitrators tend to approach them with a greater than usual level of interpretive rigour. Nevertheless, there is a presumption of validity that applies to stabilisation clauses.[37] The international law doctrine of permanent sovereignty over national resources does not operate to excuse states from their stabilisation commitments,[38] since it provides that ‘foreign investment agreements entered into by or between foreign states shall be observed in good faith’. In fact, by entering into a stabilisation provision, a state arguably exercises its sovereign prerogatives.[39] Therefore, it is likely that most arbitral tribunals would start from a presumption that the stabilisation provision is valid, meaning it is the respondent government’s burden to prove otherwise.

The sensitivities surrounding the enforcement of a stabilisation provision also mean that most proceedings are conducted confidentially, so there are few publicly available awards discussing an arbitral tribunal’s interpretation and application of stabilisation provisions in a commercial context, and even less in the mining space.[40] However, the available jurisprudence shows that one of the common hurdles claimants encounter is whether the scope of a change in law is interpreted widely by the arbitrator to include a change in the interpretation of a law, or the issuance of new or changed interpretive guidance by a government authority with respect to a law. This situation often arises in the context of taxation disputes, where the application of an existing tax law is clarified or changed by way of a subsequent tax ruling or guidance note. An arbitrator who adopts a narrow approach may exclude such events from the scope of the stabilisation clause. Therefore, from the company’s perspective, it is crucial that the stabilisation provision is crafted broadly at the outset to clearly include changes in interpretation and application of laws. As the stabilisation clause will be part of a contract between a state and an alien (foreigner), general principles of international law apply, including general principles of contractual interpretation, and practitioners should be prepared to invoke these norms when they argue for their preferred interpretation of the provision.

Another recurring interpretive theme is the weight that arbitral tribunals give to the state’s ‘right to regulate’. Normally, weight is given to this sovereign prerogative and this often translates into a tendency to construe the stabilisation clause narrowly. However, consistent with the general principle of pacta sunt servanda, most arbitrators will be reticent to allow a state to avoid an express commitment provided to an investor. Although not related to a mining agreement, the tribunal in Parkerings v. Lithuania held that:

A state has the right to enact, modify or cancel a law at its own discretion. Save for the existence of an agreement, in the form of a stabilisation clause or otherwise, there is nothing objectionable about the amendment brought to the regulatory framework existing at the time an investor made its investment.

Fact can be helpful to the interpretation of stabilisation clause. The weight given to the state’s written stability commitment will be greater where the company is able to show that it relied on the state’s stabilisation promise when it made its decision to invest. Similarly, if the company is covered by a BIT or other investment treaty, the existence of a stabilisation clause will also greatly enhance the likelihood that the tribunal will find that, by changing the law, the host state has frustrated the company’s legitimate expectations and thereby breached the FET standard.[41]

Conduct of the hearing and the award

Once the written phase of the proceedings is complete, a hearing will be held. Hearings of between one and two weeks are common in stabilisation cases. From a legal perspective, the focus tends to be on the text of the stabilisation clause, especially the wording of the trigger: what it covers and what it does not. Counsel should therefore come prepared for protracted debates on interpretation. From a factual perspective, the witnesses will mostly be questioned on the nature of the measure and its impact. As noted above, most proceedings will be conducted confidentially, and heightened confidentiality restrictions can be sought where sensitive commercial mining information is presented as part of the case, for example, in the context of proving loss.

After the hearing closes, the tribunal will retire to deliberate and draft its award. This often takes many months, and so there is a window of opportunity for the parties to hold settlement talks if they are inclined to do so. As hearings tend to give parties a sense of their strengths and weaknesses, the post-hearing window can be productive for settlement purposes. In a stabilisation case, where the dispute arises out of an inability to find a solution to an event the parties positively contemplated in their contract, the potential for settlement is higher than usual.

If the tribunal renders an award in favour of the company, it will be necessary to carefully consider how to proceed. If the company wishes to continue with its project, the usual approach of demanding payment of the award or commencing enforcement proceedings may not be appropriate. Instead, it may make sense to offer to reopen negotiations with the host government, using the award as leverage to secure the outcome that could not be achieved in earlier consultations. In considering this option, it must be remembered that, for some governments, the fact that an international tribunal has ruled in favour of the foreign investor may make it much easier to achieve a settlement, purely because the government will be able to explain the settlement as a legal necessity and present its performance of the contract in a favourable light to other foreign investors. If the company decides to enforce its entitlements under the award, a formal demand should be made to the government. However, it should not be expected that the government will simply write the company a cheque. If the award orders the government to pay monetary compensation to the company in US dollars, the government may simply not have the currency on hand. Thus, even where the decision is taken to demand payment, the company may find it has little choice but to negotiate. Common post-award settlements include a blend of local currency, foreign currency and debt instruments (such as government bonds). The company should keep an open mind as to these possibilities, as the reality is that, though it is possible to enforce an arbitral award in most countries that are party to the New York Convention, the enforcement process takes time as it may be that the host state does not have sufficient assets in enforcement-friendly jurisdictions to satisfy the amount due under the award. This is especially true of least-developed countries, who also happen to be the countries most likely to grant stabilisation terms to foreign investors.


Every sovereign has the exclusive right to make and modify laws within its territory. But, mining companies, like all foreign investors, need to know what the legal and fiscal regime for their long-term investment will be before they will commit capital abroad, especially to countries with high sovereign risk. Companies do not expect that the law will not change at all, they only expect that the law will not change in a way that seriously undermines the economics of their project. Stabilisation clauses are now an accepted means of striking the right balance between sovereign prerogative and commercial imperative. However, to be effective, stabilisation clauses must contain certain basic elements, and be accompanied by certain other provisions. It is only when the necessary combination of clauses is present in the contract that the stabilisation architecture will hold together in the event the law changes. The purpose of this chapter has been to explain how this architecture is put together and how to use it if a change in law occurs. Every contract is different, as is every case, and so the observations and recommendations made in this chapter are necessarily general in nature. But stabilisation is a special branch of cross-border legal practice, characterised by a sui generis confluence of legal theory and contract economics. It is hoped that, with the benefit of this chapter, practitioners will have an enhanced understanding of these unique features and will be better positioned to advise and represent their clients in the negotiation of stabilisation clauses and the resolution of stabilisation disputes.


[1] Sam Luttrell is a partner and Amanda Murphy is a senior associate at Clifford Chance.

[2] The International Council on Mining and Metals (ICMM) reported that companies ‘are most likely to require stability clauses in cases where the legislative framework for mining investment is not well established’, ICMM, ‘Minerals Taxation Regimes: A review of issues and challenges in their design and application’, February 2009, p. 56.

[3] Freezing clauses have been criticised as an undue restriction on sovereign regulatory power; see Ruggie, ‘Responsible Contracting’, Principle 4, which states that ‘contractual stabilisation clauses, if used, should be carefully drafted so that any protections for investors against future changes in law do not interfere with the State’s bona fide efforts to implement laws, regulations or policies, in a non-discriminatory manner, in order to meet its human rights obligations.’

[4] AFM Maniruzzaman, ‘Some reflections on stabilization techniques in international petroleum, gas and mineral agreements’ (2005) 4 International Energy Law and Taxation Review 96, citing inter alia, Margarita Coale, ‘Stabilisation Clauses in International Petroleum Transactions’ (2002) 30(2) Denver Journal of International Law and Policy 217; Paul Comeaux and N. Stephan Kinsella, ‘Reducing political risk in developing countries: Bilateral investment treaties, stabilisation clauses and MIGA and OPIC investment insurance’ (1994) 15 New York Law School Journal of International and Comparative Law 1.

[5] It is well recognised that since ‘mining projects are tied to the location of the natural resource and much of the infrastructure for extraction is immovable . . . investors who successfully explore for minerals or oil are vulnerable to unilateral changes in local rules once they have taken the initial risk of investing’, Ruggie, ‘Principles for Responsible Contracts: Guidance for negotiators’, UNHCR, 2015, p. 16.

[6] If the government or a government-owned entity is a participant in the project and has an entitlement to a share of production, this may also fall within the scope of fiscal stability. While production-sharing arrangements are more common in the petroleum industry, they are starting to take hold in the mining sector.

[7] ICMM note that ‘companies highlight stability and predictability as the most important aspects of taxation regimes’, ICMM, ‘Minerals Taxation Regimes: A review of issues and challenges in their design and application’, February 2009, p. 11.

[8] Another reason fiscal stability terms are more common is that they are often granted unilaterally by operation of host state law. For example, many countries have investment promotion laws that include a certification-based system of tax incentives, through which a degree of fiscal stability may be provided to the owners of a sponsored project. Legal stability is rarely granted this way and is, instead, almost always a matter of contract. The exception is least developed countries emerging from periods of political or economic transition, where legal stability undertakings may be made to investors in certain sectors. The Democratic Republic of Congo is an example: under the 2018 Mining Code, foreign companies that hold mining titles enjoy specified periods of stability in the laws and regulations applicable to their projects (see Loi n.18/001 du 9 mars 2018 modificant et completant la Loi n. 007/2002 du 11 juillet 2002 portant Code minier, Articles 276 al 2, 340, 342 bis and 31). Generally speaking, these kinds of unilateral stability undertakings are less effective than contractual stability clauses, because they can be unilaterally withdrawn by the host government (and the investor will only have a remedy in this situation if it has standing to dispute the revocation in an international forum, as is the case under the DRC Mining Code).

[9] We say ‘purports’ because, as discussed below, it is often a complex question of local (constitutional) law whether the host government can fetter its legislative power in this way: some countries have laws that allow the government to do so, subject to compliance with certain constitutional procedures (such as approval of the contract by an act of parliament); other countries do not.

[10] Henry G. Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 262.

[11] In this famous mining case, the Concession Agreement between Lena Goldfields and the Soviet government required that the concessionaire was to submit to all existing and future Soviet legislation ‘in so far as special conditions are not provided in this agreement’ (Article 75) and the Soviet Government undertook not to make any alteration to the Concession Agreement by order, decree, or other unilateral act or at all except with Lena’s consent (Article 76). Arthur Nussbaum, ‘The Arbitration Between the Lena Goldfields, Ltd. and the Soviet Government’ (1950) 36 Cornell Law Review 31, p. 45.

[12] Henry G. Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 263.

[13] Aminoil v. Kuwait, Award, 24 March 1982, para. 24.

[14] Henry G. Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 262.

[15] For example, according to the United Nations Human Rights Commission (UNHCR) negotiators of a stabilisation clause should ensure that it ‘does not create obstacles to the State’s bona fide efforts to introduce and implement laws, regulations or policies in a non-discriminatory manner to meet its human rights obligations’, Ruggie, ‘Principles for Responsible Contracts: Guidance for negotiators’, UNHCR, 2015, p. 16. These issues can be more prominent when negotiating with a developing country, as the UNCHR found according to its research that ‘compared to contracts agreed with Governments in developed countries, those negotiated with Governments in developing countries were: (a) typically much broader in their coverage; and (b) much more likely to include exemptions for or award compensation to business investors for compliance with future laws – even in areas that are directly related to human rights, such as health, environmental protection, labour and safety’, Principles for Responsible Contracts: Guidance for negotiators, UNHCR, 2015, p. 16.

[16] The distinction in form is most significant where the contract is governed by the law of a common law country, such as England. Under English law, breach of a condition (meaning a clause that goes to the heart of the bargain) gives the injured party a right to terminate the contract; breach of a warranty gives the injured party a right to claim damages.

[17] International Bar Association, Model Mining Development Agreement (MMDA 1.0), Article 8.5, State Guarantees, Example 3, 4 April 2011.

[18] The issue of timing must also recognise the fact that the analysis usually involves the company taking external legal advice (from local counsel and, potentially, international counsel) and accounting advice. Sufficient allowance needs to be made for the company to go through its internal decision-making processes and resolve on whether to ‘pull the trigger’. For most companies, the decision to initiate a formal consultation process with the host government will not be taken lightly.

[19] Henry G. Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 263.

[20] Henry G. Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 260.

[21] This is because the multilateral treaty that governs ICSID arbitration, the Convention on the Settlement of Investment Disputes between States and Nationals of Other States 1965, contains special jurisdictional requirements.

[22] In the New York Convention system, only the courts of the seat are competent to annul (set aside) an arbitral award – the role played by the courts of all other countries is limited to enforcing the award (or declining to do so, if a ground for refusing enforcement is made out).

[23] Sulamérica Cia Nacional De Seguros SA v. Enesa Engenharia SA [2012] EWCA Civ 638. England is not alone in this approach to the issue of what law governs an arbitration clause. In Singapore, there is case law that arguably goes further by presuming that the law of the seat should be the governing law of the arbitration agreement, rather than the law that the parties chose to govern their agreement generally: FirstLink Investments Corp Ltd v. GT Payment Pte Ltd and others [2014] SGHCR 12.

[24] See Resources Contracts (Review and Re-Negotiation of Unconscionable Terms) Act 2017. For an analysis of the Tanzanian legislation of 2017, see Sam Luttrell, ‘An international perspective on the Tanzanian Natural Wealth and Resources Acts’ (2018) 36 Australian Resources & Energy Law Journal 3.

[25] Texaco Overseas Petroleum Company (TOPCO) v. Libya (1977) 53 ILR 389. All lawyers engaged in cross-border natural resources contracting or dispute resolution should familiarise themselves with this seminal decision. By way of background, in 1955, the Libyan Petroleum Commission awarded concession agreements to 17 international oil companies. In 1969, Muammar Kaddafi overthrew the Libyan monarchy and commenced socialist reform of Libya. Part of Kaddafi’s reform agenda involved the nationalisation of the oil industry, the state’s main source of revenue. By 1971, most of the International Oil Companies (IOCs) engaged in Libya had been subject to the host government’s revision of terms of their agreements. In 1973, the Libyan government announced the nationalisation of 51 per cent of the interests of nine IOCs. In February 1974, the remaining interests of three of these foreign companies (TOPCO, California Asiatic Oil and the Libyan American Oil Company) were nationalised, such that their investments were expropriated in full. The affected IOCs had arbitration clauses in their concessions, and so they were able to bring claims directly against Libya without the need to rely on the remedy of diplomatic protection. The award handed down by the sole arbitrator in the TOPCO case, Professor Rene-Jean Dupuy, became a seminal decision in the construction of the modern international law of foreign investment. This was due mainly to Professor Dupuy’s decision that the concession agreement on which TOPCO relied was ‘within the domain of international law’, and that international law was part of the legal order that governed it. Professor Dupuy reached this conclusion on the basis that (1) the governing law clauses of the concessions called for the application of both Libyan law and ‘principles of international law’; (2) the contracts contained stabilisation clauses (in which Libya effectively undertook not to nationalise for a certain period); and (3) the contracts contained international arbitration clauses. Professor Dupuy deduced from these clauses that the concession contracts were linked to the international legal order – in this matrix, Libyan law applied unless it was inconsistent with international law. In substance, applying this body of rules, the arbitrator found that the nationalisation of TOPCO’s assets was an expropriation that violated one key principle of international law: pacta sunt servanda. Professor Dupuy held that although a host State has the right to nationalise foreign-owned property in the exercise of its sovereign prerogatives, this right must be exercised in a manner that conforms to any contractual obligations that the host State may owe to the relevant foreign owners at the time (the relevant obligation being, in this case, Libya’s undertaking not to nationalise, expressed in the stabilisation clauses of the concession contracts), as the contract was held to be subject to international law. Professor Dupuy ordered Libya to specifically perform its contractual obligations under the concession agreements. Eight months after Professor Dupuy’s decision, the Libyan government and TOPCO reached a settlement, under which Libya agreed to provide compensation in the form of oil valued at US$152 million. For a more detailed analysis of the TOPCO case, see Robert von Mehren and P Nicholas Kourides, ‘The Libyan Nationalizations: TOPCO/CALASIATIC v Libya Arbitration’ (1979) 12(2) Natural Resources Lawyer 419.

[26] Beyond insulating the contract against unilateral invalidation by change in host state law, the inclusion of an express reference to international law in the governing law clause gives the foreign party the ability to access a range of rules and principles of international law, including the general principle of good faith (which may not necessarily be part of the law of the host State), the principles of expropriation and compensation, the doctrine of acquired rights and the international minimum standard of treatment of aliens (i.e., foreigners). To be sure, there is a good argument that these principles of international law apply to any contract between a government and a foreign person or company, but with an express reference to international law in the governing law clause of the contract the foreign party’s prospects of successfully invoking these principles are enhanced significantly.

[27] ICSID Arbitration Rules, Article 39; UNCITRAL Arbitration Rules (2013), Article 26; ICC Arbitration Rules (2017), Article 28.

[28] ICC Arbitration Rules (2017), Article 29.

[29] Sam Luttrell, ICSID provisional measures ‘in the round’, Arbitration International, Volume 31, Issue 3, 1 September 2015, pp. 393–412.

[30] This overlap was described by Professor Klaus-Peter Berger (cited with approval by the tribunal in Paushok v. Mongolia), who said (in relation to the interim measures provision of the UNCITRAL Rules) that: ‘[t]o preserve the legitimate rights of the requesting party, the measures must be “necessary”. This requirement is satisfied if the delay in the adjudication of the main claim caused by the arbitral proceedings would lead to a “substantial” (but not necessarily “irreparable” as known in common law doctrine) prejudice for the requesting party.’ See Klaus-Peter Berger, International Economic Arbitration in Studies in Transnational Economic Law, Vol 9, p. 336 (Kluwer 1993); see also Sergei Paushok, CJSC Golden East Company and CJSC Vostokneftegaz Company v. The Government of Mongolia (UNCITRAL), Order on Interim Measures, 2 September 2008, para. 68.

[31] For example, Quiborax SA v. Non Metallic Minerals SA and Allan Fosk Kaplun v. Plurinational State of Bolivia, ICSID Case No. ARB/06/2, Decision on Provisional Measures, 26 February 2010.

[32] Henry G Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes (Oxford University Press, 2017), p. 265.

[33] AGIP v. Government of the Popular Republic of Congo (ICSID Case No. ARB/77/1), Award, 30 November 1979.

[34] Sapphire Petroleum v. National Iranian Oil Co, 35 ILR 136 (1967).

[35] Aminoil v. Kuwait, Award, 24 March 1982.

[36] Texaco Overseas Petroleum Company (TOPCO) v. Libya (1977) 53 ILR 389.

[37] Henry G Burnett and Louis-Alexis Bret, Arbitration of International Mining Disputes, (Oxford University Press, 2017), p. 264.

[38] UN General Assembly Resolution on ‘Permanent Sovereignty over Natural Resources’ (1962) UNGA Res 1803.

[39] AGIP v. Government of the Popular Republic of Congo (ICSID Case No. ARB/77/1), Award, 30 November 1979, para. 86; see also Nour Eddine Terki, ‘The Freezing of Law Applicable to Long-Term International Contracts’ (1991) 6 Journal of International Banking Law 43, who explains that ‘it would in any case be wrong to imply that the granting of such a guarantee to the foreign firm would impair the sovereignty of the State. For by acting in this manner, the public authority is doing nothing more than implementing gone of the prerogatives attached to this concept. It is precisely by placing itself in the context of manifesting its sovereignty that a State can legitimately grant a stabilization clause, either by enacting a law a priori, or by approving a contracting containing such clause a posteriori.’

[40] There are a number of well-known petroleum arbitrations which have upheld stabilisation commitments in petroleum production agreements with host states. These include Lena Goldfields Limited v. USSR, Award (1929-1930) 5 ADIL 3 (Case No. 1); Libyan American Oil Co (LIAMCO) v. Libya, Award on Jurisdiction, Merits and Damages, 20 ILM (1981) 1; Saudi Arabia v. Saudia Arabian Oil Co (Aramco) (1958) 27 ILP 117; Sapphire International Petroleum Co v. National Iranian Oil Co, 35 ILR 136 (1967); Kuwait v. American Independent Oil Co (AMINOIL), Award, 21 ILM 976 (1982), 24 March 1982.

[41] Occidental Petroleum Corporation and Occidental Exploration and Production Company v. The Republic of Ecuador, ICISD Case No. ARB/06/11, Award, 5 October 2012, para. 526.

Unlock unlimited access to all Global Arbitration Review content