May 11
Dolores Utrilla
Dolores Utrilla
3rd May 2021
External Relations & Trade Internal Market Justice & Litigation

Op-Ed: “PS Achmea: P L Holdings and the ‘individual arbitration agreement’” by Gillian Cahill

On 22 April 2021, Advocate General Kokott handed down her Opinion in the case of PL Holdings v Poland (C-109/20). This case follows hot on the heels of the Court of Justice’s Achmea decision (C-284/16) which prohibited an investor state dispute resolution provision contained in an intra EU Bilateral Investment Treaty that permitted an investor of a Member State (MS) to initiate arbitral proceedings against another MS, before an arbitral tribunal whose jurisdiction that MS has undertaken to accept. The fallout of the Achmea decision was swift and fast, effectively putting an end to investment arbitration arising out of intra-EU BITs in Europe. Or at least so it was thought until the Opinion in PL Holdings. In this case, AG Kokott determined that an ‘individual arbitration agreement’ might be able to avoid the fate of the investor state dispute resolution provisions impugned in Achmea and could therefore be compatible with the duty of sincere cooperation and autonomy of EU law. However, the AG’s finding was conditioned on a MS’s court being able to ‘comprehensively review the arbitration award for its compatibility with EU law’, which must include the possibility of making a preliminary reference. She also found that the principle of equal treatment would have to be respected. In a curious twist in this curious Opinion, the AG appears to have made these conclusions with considerable hesitancy, stating at the same time that the compatibility of clauses such as that at issue in PL Holdings is indeed doubtful (point 75). 

The final result is that the Opinion appears to try and give a little to all sides and in so doing risks falling between two stools. Particularly where its reasoning falls on the nebulous qualification of the arbitration clause in question as an ‘individual arbitration agreement’.

Facts and Background to the Case

The PL Holdings preliminary reference arrived at the Court of Justice from the Swedish Supreme Court who had been seised of an annulment action in respect of an arbitral award rendered in the context of an investor-state arbitration brought pursuant to the dispute resolution provisions of a BIT entered into by Belgium and Luxembourg (then making up the Belgium-Luxembourg Economic Union, ‘BLEU’) and Poland. The award had been rendered consequent on PL Holdings’ (a Luxembourgish company) claim that its investment in a Polish Bank had been expropriated as a result of the revocation by the Polish Financial Market Commission of PL Holdings’ voting rights in that bank and its order to PL Holdings to divest its shares in that bank. 

The Preliminary Question  

In the context of an annulment action of the arbitral award, rendered in PL Holdings’ favour for some 150 million euros, the Swedish Supreme Court, as permissible under Swedish Arbitration law, referred a preliminary question to the Court of Justice asking whether ‘Articles 267 and 344 TFEU, as interpreted in [the judgment in] Achmea, mean that – where an investment agreement contains an arbitration clause that is invalid as a result of the fact that the contract was concluded between two Member States – an arbitration agreement is invalid if it has been concluded between a Member State and an investor by virtue of the fact that the Member State, after arbitration proceedings were commenced by the investor, refrains, by the free will of the State, from raising objections as to jurisdiction’?

The manner in which the question was posed was the first curious twist in this tale. The Swedish Supreme Court appeared to make, in its framing of the question, a distinction between ‘an arbitration clause that is invalid because it is contained in an intra EU BIT’ that is, Achmea type provisions and the type of clause at issue in PL Holdings. This latter type of clause that it deemed to be an arbitration agreement concluded between a Member State and an investor ‘by virtue of the fact’ that the Member State participates in the arbitration proceedings and does not raise a jurisdictional objection to the proceedings. On this basis it wished to know: did this distinction, as the Swedish Supreme Court saw it, change the scope and application of the Achmea ruling? Although not explicitly addressed by AG Kokott, it is arguable from the Opinion that she was somewhat influenced by the distinction that the Swedish Supreme Court made. But such a distinction from the arbitral perspective is very surprising. The nature of consent to the arbitral process, be it investment or commercial arbitration, is contained within the underlying arbitral agreement. Where that agreement is contained in ‘investor to State’ dispute resolution provisions in an intra-EU BIT – as was the case in both Achmea and in PL Holdings, the nature of consent to the arbitral process is not fundamentally altered by a Respondent state not objecting to jurisdiction (as the Swedish Supreme Court appeared to suggest). Rather, the argument concerning participation in the proceedings without objection is arguably a question of waiver of the right to object. But in neither case can consent or waiver alter the nature of a type of clause that has been found by the Court of Justice to be fundamentally invalid and contrary to EU law. 

The key finding: the ‘individual arbitration agreement’

In fact, the BLEU-Poland BIT contains a fairly standard ‘state to investor’ dispute resolution clause in its Article 9(2)) which provides that ‘if the dispute cannot be resolved within six months of the written notification provided for in (1), then at the investor’s choice, it is submitted to arbitration at one of the following designated institutions (…)’. Such an asymmetric arbitration clause found in a BIT and providing for investment arbitration at the investor’s choice, is extremely common in investment arbitration. Moreover, and despite suggestions to the contrary in the Opinion, this clause is very similar to the clause that was at issue in Achmea which provided in Article 8(1) of the Czechoslovakia-Dutch BIT that ‘All disputes between one Contracting Party and an investor of the other Contracting Party concerning an investment of the latter shall if possible, be settled amicably’ and in its Article 8(2) that ‘If diplomatic resolution does not occur, one of the contracting parties may submit, six months after notification provided for in paragraph (1), the dispute to an “ad hoc” arbitral tribunal.’ Paragraphs (1) and (2) of Article 8 therefore provide another classic example of an ‘investor to state arbitration’ clause. Thus, both the Achmea dispute resolution provision ultimately impugned by the Court of Justice and Article 9(2) of the BLEU-Poland BIT for asymmetric ‘investor to state’ arbitration. 

Although not mentioned by AG Kokott, the only obvious differences between the two clauses appears to be that Article 8(2) of Czechoslovakia-Dutch BIT the provided that arbitration could be elected by either the investor or the State, whereas in PL Holdings, the reference to arbitration is arguably compulsory once the investor so elects for arbitration. AG Kokott did not enter into an analysis of these clauses and it is therefore unclear whether, why and where she perceived there to be such a significant difference that justifies the reasoning on which she bases her conclusions as to EU law compatibility of the clause at issue in PL Holdings. 

AG Kokott argues that in Achmea the issue was the agreement between two Member States to remove disputes which might concern the application or interpretation of EU law (point 30), whereas in PL Holdings the arbitration agreement is an ‘individual’ agreement between a ‘Member State and an investor’ to remove disputes in the same way (paragraph 31). But this conclusion is hard to follow given that both clauses are similar ‘investor to state’ arbitration provisions contained in an intra-EU BIT. Similarly, given that both BITs provide for State to State arbitration clauses (Article 8 of the BLEU BIT and Article 10 of the Czechoslovakia-Dutch BIT) it is difficult to understand how AG Kokott can so easily divorce the arbitration agreement in PL Holdings from its underlying intra-EU BIT.

In this regard, AG Kokott also appears to suggest that an individual arbitration agreement is not based on an investment treaty (point 69). This conclusion is extremely difficult to understand, unless AG Kokott is implicitly referring to and perhaps even accepting the type of argument suggested by the Swedish Supreme Court in the formulation of its question. No discussion by AG Kokott of the Swedish Supreme Court’s argument is found elsewhere in the Opinion and therefore no neat answers to these questions are found. 

Surprisingly, having made such a finding, AG Kokott immediately appeared to row back on the importance of the distinction between the Achmea and PL Holdings type clauses, arguing that whether an individual case is removed from the EU judicial system (that is, would fall foul of the Achmea rationale) ‘depends on the specific dispute and not on whether the dispute is brought before an arbitration tribunal under a general investment treaty between Member States or under an individual arbitration agreement between an investor and a Member State’ (point 31). Not only is this statement questionable for the reasons set out above but it is also unclear why, if indeed that is the case, the distinction is relevant in the first place. 

Despite this confusion, AG Kokott goes on to find that it could not be ruled out that the ‘ arbitration tribunal misconceived the obligations of the Polish banking supervisory authority’ under Directive 2006/48/EC, at issue (point 39). For the AG, both the ‘risk of an infringement of EU law and the risk of divergent interpretation’ could be ‘limited or even eliminated if compliance with EU law by arbitration awards were comprehensively reviewed by the national courts – where appropriate, after having conducted a preliminary-ruling procedure’ (point 40).  Arguably, AG Kokott is here implicitly assimilating her ‘individual arbitration agreement’ to something akin to a commercial arbitration agreement, which can co-exist with the EU judicial system (in part) because of the possibility of back-end Member State review. 

Moreover, even accepting the ‘individual arbitration agreement’ distinction, it seems hard to ignore that given that any investment tribunal constituted on the basis of these BITs’ dispute resolution provisions will operate in the sphere of international law as envisaged and agreed to by the Member States who signed these treaties. Do both such BITs therefore not raise precisely the same risk to the autonomy of EU law that the Court of Justice was concerned with in Achmea? That is to say, do both arbitration provisions as agreed to by the Member States involved not offer equal possibility for the removal of disputes concerning the application and interpretation of EU law from the EU judicial system via an investor to state arbitration?

AG Kokott appears to recognise this, holding that recognition of individual arbitration agreements between Member States and investors from other Member States would create the risk of an infringement of EU law by the arbitration tribunals (point 42). But in so far as the national courts can ensure that such arbitration awards comply with EU law then these agreements may escape an Achmea fate (point 42). It is the reasoning for this latter significant finding that, for the reasons set out above, appears less than fully convincing. 

Finally, as regards the application of the principle of equality, AG Kokott recognised that the present investment arbitration did fundamentally institute a difference of treatment on the basis of nationality. Without entering into significant analysis, she further held that it was for the national court to determine whether such an individual arbitration agreement could be justified under Article 20 of the Charter (point 71).


AG Kokott’s ultimate findings in this case are two-fold. First, the Achmea type threat of autonomy of EU law is still present as regards an arbitration agreement at issue in PL Holdings or what she deems an ‘individual arbitration agreement’. Second, despite these risks, this threat may be overcome if MS review is significant enough, if necessary by the courts an MS courts using the preliminary reference procedure. Although AG Kokott herself at times suggests otherwise, it is hard not to determine that the entire premise of her findings are based on the qualification of the arbitration agreement at hand as an ‘individual arbitration agreement’. Yet for all the reasons set out above, such a qualification is far from being certain. Having thus started with a bang, as the follow on to Achmea the Opinion seems to end with more of a flutter. Given that precisely the same Achmea type concerns to the EU legal order are present in PL Holdings, whether or not the Court of Justice decides to follow AG Kokott’s line of reasoning will likely depend wholly on whether they are convinced by her new and novel notion of the ‘individual arbitration agreement’ in investment arbitration. 


Gillian Cahill is a Barrister specialised in EU law and international arbitration. She has particular expertise in the cross-over issues arising out of these two areas of law.


Your privacy is important for us

We use cookies to improve the user experience. Please review privacy preferences.

Accept all Settings

Check our privacy policy and cookies policy.