The judgment of Phillips J in the English High Court in Mercuria v Citibank  EWHC 1481 (Comm) raises an important question about which documents ought to be treated as “documents of title” for the purpose of effecting delivery of goods.
Facts of the case
Citibank provided finance to Mercuria, on the security of metal which was believed to be in the possession and ownership of Mercuria at warehouses in certain Chinese ports. The finance was provided by way of “repo” agreements, pursuant to which Mercuria would sell metal to Citibank, and Citibank would repurchase it at a later date and for a higher price, albeit that the metal would never leave the warehouses, so that they were continuously held by a third party. The agreements conferred exclusive jurisdiction on English courts (see generally - for the terms of the agreements).
In May 2014, evidence emerged of fraud at Chinese ports where some of the metal had been warehoused. There was no suggestion that either Mercuria or Citibank was involved. However, the repo agreements entitled either party to terminate on the occurrence of an event which would affect the ability of the counterparty to perform (see ). In July 2014, Mercuria asserted that the events at the Chinese ports constituted a termination event, which obliged Citibank to deliver the metal to it (see ). In purported compliance with this obligation, Citibank delivered to Mercuria warehouse receipts for the metal, which evidenced its ownership and right to possession (see ).
The ‘central and largely determinative issue’ for Phillips J was whether, in so doing, Citibank had effected delivery (). Phillips J commenced this part of the judgment by considering the general law concerning delivery of goods in the possession of a third party ( to ). .
The Sales of Goods Act 1979 provides in s 29(4) that in order to effect delivery of goods held by a third party, the third party must attorn, or “acknowledge to the buyer that he holds the goods on his behalf”.
Phillips J described it as a “well established” that the mere transfer (from the seller to the buyer) of a warrant or receipt issued by a third party, in the absence of such an attornment, does not constitute delivery. His decision followed Farina v Home (1846) 16 M&W 119 (HL) and Dublin City Distillery v Doherty  AC 823 (HL) which unambiguously support this proposition (see  and ). Section 29(4) of the Sale of Goods Act was a codification of the existing common law. It is sensible, therefore, to read the first part of the Sale of Goods Act, section 29(4) as requiring an attornment (see ). Thus, Phillips J correctly held that under general sales of goods law, there had been no constructive delivery of the metals to Mercuria.
Section 29(4) of the Sale of Goods Act also provides that “nothing in this section affects the operation of the issue or transfer of any documents of title to goods”. Delivery could therefore be effective by transfer of a document of title (see ). However, it was conceded by Citibank that the receipts tendered did not constitute “documents of title”, and thus were not affected by this provision (see ).
Could the Factors Act 1889 have provided a different result?
Debattista et al have contended that Citi ought not to have conceded that the warehouse receipts did not constitute documents of title: see Debattista, Firth, and Selvaratnam QC, ‘Bankers Gambling in China: Down but not Out’ (2015) JIBFL 478, 479. In making that concession, they suggest that Citibank – and therefore Phillips J – proceeded on the basis that a bill of lading would be the only document of title which could transfer possession without attornment.
In particular, Debattista et al argue that the trial judge ought to have been referred to Factors Act 1889, s1(4), which provides in s 1(4) a wider definition of documents of title. The Sale of Goods Act uses the same definition of documents of title as the Factors Act.
Debattista et al argue that the warehouse receipts provided by Citibank fell within the definition contained in the Factors Act – and observe that that definition does not require any attornment. Accordingly, in their view, Phillips J “affect[ed] the operation of the issue or transfer of any document of title to goods”, contrary to the Sale of Goods Act, s 29(4), by requiring that there be an attornment for an effective delivery.
This argument ought to be rejected because it conflates two senses in which the expression ‘document of title’ is used. The definition of documents of title under the Factors Act (the broad sense) is generally understood to apply only for the purposes of the exceptions to the nemo dat rule. Bills of lading are the only true document of title, in terms of transferring constructive possession (the narrow sense): Lickbarrow v Mason (1794) 101 ER 380. As noted in Benjamin’s Sales of Goods (9th ed), “the wording of the qualification in section 29(4) is expressed in negative terms and does not define the role of a document of title in effecting delivery”.
The mere transfer of the receipts to Mercuria cannot have constituted a constructive delivery. Contrary to Debattista et al’s assertions, the failure to refer to the Factors Act, s1(4) does not cast Phillips J’s decision into doubt.
There is a further policy reason to reject this argument. If the broad sense of documents of title under the Factors Act is adopted, there will be multiple documents relating to any inventory of goods that can transfer constructive possession of them. This creates the potential for fraud, by creating the possibility that multiple parties will believe that they have constructive possession. It is thus preferable to maintain the current position, that only bills of lading are sufficient to transfer constructive possession, in order to reduce the potential for fraud. This is precisely why it is commercial practice to deal with bills of lading only if a complete set can be tendered.