Publications & Events

Decision of the Paris Court of Appeal on "adverse inferences" doctrine

The Paris Court of Appeal has issued an important decision endorsing an arbitral award based, in part, on the "adverse inferences" principle in the IBA Rules on the Taking of Evidence in International Arbitration (the "IBA Rules").1

Article 9(5) of the IBA Rules states: "[i]f a Party fails without satisfactory explanation to produce any Document requested in a Request to Produce… the Arbitral Tribunal may infer that such document would be adverse to the interests of that Party".2

The benefit of the "adverse inferences" principle is two-fold: it acts as a deterrent to non-compliance with a document production request or order, and it provides a basis for a Party to assert a positive case in circumstances where important evidence has been withheld by the other Party.

Although traditionally seen as a doctrine supported more by common law than civil law jurisdictions (a corollary of the former's expansive approach to document production), the recent decision of the Paris Court of Appeal shows that thinking to be outdated.3


The Facts

Dresser-Rand Group Inc, a Denver registered affiliate of Siemens, entered into an agreement for the purchase of shares in the Spanish company Grupo Guascor, with the contract providing for a price adjustment at a future date. Dresser-Rand Group Inc subsequently transferred the shares to Dresser-Rand Holdings Spain. The original Spanish shareholders, unable to agree on the price adjustment with Dresser-Rand Group Inc, commenced arbitral proceedings in October 2012 against both Dresser-Rand entities (together, "Dresser").

The arbitral award, which was handed down in February 2015, went partially in favour of the Spanish shareholders. In coming to its decision, the tribunal drew an adverse inference against Dresser under Article 9(5) of the IBA Rules on the basis that it had failed to produce certain audit reports requested by the Spanish shareholders.

Dresser challenged that part of the award before the French courts. It argued that the IBA Rules should not have been relied on at all without prior consultation of the parties. In addition, Dresser alleged that drawing an adverse inference due to the non-production of documents when the tribunal had not even ordered their production, or asked Dresser to explain the non-production, breached the rules of due process.

The Paris Court of Appeal disagreed. It noted that the use of the IBA Rules had been explicitly referenced in the first procedural order, which itself was negotiated by the parties, and as such there was no need for further consultation before the IBA Rules could be relied on by the tribunal. It followed that, as the tribunal had acted within the language of Article 9(5), there was no breach of due process.



This decision may give pause to arbitrators, clients and counsel alike. Arbitrators should see it as a nod from domestic courts that their arbitral discretion, especially on procedural decisions, will only rarely be undermined. Tribunals, while still exercising caution to avoid precluding a party from presenting its case, should feel emboldened to grant requests for adverse inferences if appropriate. Indeed counsel may feel similarly emboldened to make those requests more often (although the importance of being very clear on exactly what they are asking a tribunal to infer should not be forgotten).4

For counsel and clients the decision also signals the importance of document production and the very real consequences that can arise from a failure to produce. Finally, and more generally, it highlights France's position as a leading arbitration-friendly jurisdiction.

Yet the decision also has its limits. The Paris Court of Appeal made it clear that the tribunal in fact reached its decision on the basis of exhibits submitted to it and not on the basis of any adverse inference, which was drawn more as a matter of completeness.

Whether a decision based more substantially on an adverse inference would be treated in the same way is unclear, although a subsequent decision of the Paris Court of Appeal may have given an indication.5 In this more recent case it was alleged that arbitrators had drawn adverse inferences following non-compliance with an order to produce documents which purportedly could not be disclosed due to national security requirements. The Court of Appeal found that adverse inferences had not been drawn yet also noted that the consequences of any adverse inference were not disproportionate. It remains to be seen whether this requirement for proportionality of an adverse inference is limited to the specific, national security issues in this case, or if it will have a wider application.


1 Cour d'Appel de Paris, Pôle 1 – Chambre 1, 28 February 2017, 15/06036.
2 Article 9(5), the IBA Rules on the Taking of Evidence in International Arbitration (2010).
3 Cour d'Appel de Paris, Pôle 1 – Chambre 1, 28 February 2017, 15/06036.
4 For more detailed analysis of the practice of, and rationale behind, the doctrine of adverse inferences see Michael Polkinghorne and Charles B. Rosenberg, The Adverse Inference in ICSID Practice, ICSID Review, Vol. 30, No. 3 (2015), pp. 741–751.
5 Cour d’Appel de Paris, Pôle 1 – Chambre 1, 21 March 2017, 15/17234.


Benjamin Ainsley Gill, a trainee solicitor at White & Case, assisted in the development of this publication.

This publication is provided for your convenience and does not constitute legal advice. This publication is protected by copyright.
© 2017 White & Case LLP