Kiri Industries Limited v Senda International Capital Limited and another

In earlier proceedings in Suit No 4 of 2017, the Singapore International Commercial Court found that Senda International Capital Ltd had acted oppressively towards Kiri Industries Ltd in relation to their joint venture vehicle, DyStar Global Holdings (Singapore) Pte Ltd, and ordered Senda to buy Kiri's 37.57% shareholding at a valuation as at 3 July 2018. The SICC later valued Kiri's shareholding at US$481.6m.

One element of that valuation was a notional licence fee for Zhejiang Longsheng Group Co, Ltd's unauthorised use of DyStar's patent over Orange 288 dyes. The SICC had assessed the fee using the method of Senda's expert, Mr Chan, partly because Senda had not given adequate discovery of how much of the infringing products Longsheng made. On appeal, the Court of Appeal held that this approach rewarded Senda for its under-disclosure and remitted the issue, directing that the fee be reassessed on the best available evidence of the quantity of the related products.

On remittal, the court had to fix the annual tonnage of related products Longsheng produced between 31 August 2010 to 23 March 2019. Kiri's expert, Ms Harfouche, put it at 120,000 tonnes a year; Senda relied on Mr Chan's later figure of 41,750 tonnes and an alternative of 20,875 tonnes.

1. Whether the Armory v Delamirie principle or an adverse inference required the court to accept the highest estimate against Senda.
2. Whether related products meant only products made using the patent or all navy and black disperse dyes.
3. Whether either expert's estimate of the annual tonnage could be accepted.
4. What annual tonnage should be used to calculate the notional licence fee.

The court preferred to apply the established rules on adverse inferences under illustration (g) of s 116 of the Evidence Act 1893 rather than Armory v Delamirie, and found neither appropriate here. The worst cannot be assumed against a wrongdoer where contrary evidence shows the assumption is wrong; the principle is not a licence for guesswork. It adopted a robust approach to quantifying loss on the available evidence.

On definition, DyStar's CEO used the term in a way showing related products were those made using the patent, and the earlier Valuation Judgment had accepted this, so Kiri's position was preferred.

Mr Chan's figures were rejected. They relied on a February 2020 model previously rejected because its underlying data were not disclosed, they were lower than Longsheng's sales of just three related products, which industry reports put at over 50,000 tonnes a year, and the revised proxy based on DyStar's own sales mix was untested. Ms Harfouche's main estimate rested on an unverified assumption that all navy and black disperse dyes made by Longsheng and another producer used the patent, which the evidence contradicted, since Longsheng could make many such dyes without it.

The court instead used Ms Harfouche's second table, based on broker reports showing that half of disperse dye output consisted of related products, which both experts had accepted. Correcting a mathematical error by applying Longsheng's 25% market share directly gave 56,750 tonnes, a figure consistent with the three-product sales evidence.

On the issue remitted by the Court of Appeal, the court determined the annual tonnage of related products produced by Longsheng to be 56,750 tonnes, and 53,550 tonnes after deducting the 3,200 tonnes produced for DyStar, for calculating the notional licence fee. The parties were invited to submit revised agreed calculations within 21 days of the licence fee and of the value of Kiri's shareholding, without a discount for lack of marketability.