CL-2018-000297, CL-2018-000404, CL-2018-000590, - [2025] EWHC 2364 (Comm)

Mr Horn did not understand or intend that any of the CANs, or tax refund claims supported thereby, would represent to SKAT that the named client was the Danish tax law owner of any shares when the dividend was declared (or at all), or was entitled as a matter of Danish tax law to the dividend referred to in the CAN or had received directly or indirectly from the Danish company a payment net of tax in respect of such an entitlement. He did not think at the time (and as it happens still does not think now) that anything like any of those representations was being or had been made to SKAT.

Mr Horn considered that the SCP CANs, and those issued by NCB, Indigo or Lindisfarne under the Maple Point Model, did no more than report to the custodian’s client an account credit that the custodian was treating as a ‘dividend’ item, in an amount calculated from the gross dividend declared and the standard WHT rate in Denmark (at the time, 27%). To Mr Horn at the time (and still today), the ‘dividend’ label did not mean or imply that a real dividend had been earned, because he was steeped in the habits of the market that extended to the use of that label for various types of dividend-related income, including what I am calling a dividend compensation payment (which, in his use of language at the time, Mr Horn would have said was a species of ‘market claim’). He did not think anything more was being communicated to SKAT by the provision of a CAN in support of a tax refund claim; and he regarded it then as a matter entirely for SKAT whether it took that as a sufficient basis for making payments.

Rajen Shah likewise did not understand or intend that anything similar to any of those pleaded representations would be or was being made to SKAT. From his experience of the market, including in particular seeing the CANs that Merrill Lynch issued in the context of the Broadgate fund activity, Mr Shah considered that CANs were generic credit advices that did not make any claim of share ownership at all, let alone of ownership for the purpose of any tax law. He did not expect them “[to] distinguish between a market claim or a compensation payment or a real dividend, as you describe it, which is traceable to the CSD. There is absolutely no distinction made.” Therefore, similarly to Mr Horn, he saw them as merely reporting a dividend-related credit and saying nothing as to the basis or legal characteristics of the payment.

Mr Dhorajiwala also did not understand or intend that any of those representations would be or was being made to SKAT. He held the mistaken, but honest, view that there was some sense in which a share purchaser was a shareholder from the trade date even where the trade settled only through the synthetic settlement method used in the Solo and Maple Point Models. He saw CANs as recording, as it was put to him, that the named client “was the legal and beneficial owner of the stated amount of the shares on the key date when the dividend would be declared”; but because of his confused view of share ownership, that was evidence (which I accept) to the effect only that, to his way of thinking, CANs recorded that the client had by the dividend declaration date purchased (i.e. contracted to acquire) the stated share volume. He was led to give some answers that SKAT argued were concessions that he thought CANs represented to the reader that the client had an entitlement to the gross dividend amount, but I do not accept that that was the tenor of his evidence taken as a whole. He had in mind that if the client was able to make a successful tax refund claim, it would ultimately receive the gross dividend amount, but not that the CANs made any relevant representation. I agree with the DWF Ds’ submission that when Mr Dhorajiwala clarified that he saw the ‘tax’ references in CANs as “just showing the calculation made in order to demonstrate the amount that was credited to [the client’s] account”, he was not retreating to a party line (as SKAT submitted in closing), but telling the truth as to how he saw it at the time.

Mr Horn did not understand or intend that the tax representation, or any statement essentially to its effect, would be or was being conveyed to SKAT. I agree with the DWF Ds’ submission that Mr Horn’s thinking at the time was accurately captured by the following answer given under cross-examination:

“Q. … So when you look at the gross dividend as a means of identifying the dividend that is being referred to, what is being said is that Acorn had the right to the dividend declared, tax has been deducted from it, and that is why the net dividend has been credited to the addressee’s account, yes?

A. No, you are going too far there. All it is saying is a custodian has credited somebody with a net dividend and that this is the method by which the net dividend is calculated. It cannot be characterising something for tax purposes which is a matter of local tax law.”

Similarly, and as with Mr Horn a finding bearing well in mind my concerns as to Rajen Shah’s credibility as a witness, Rajen Shah did not understand SKAT to be being told anything more than that the Danish withholding tax rate had been used to calculate the amount credited to the client: “… DCAs are produced on the basis of a calculation which requires the gross dividend less the maximum -- the highest rate of withholding tax in the relevant jurisdiction, and in Denmark that is 27%. It is a calculation methodology, it is not a reference to an actual amount of tax withheld. It is a calculation reference.”

The same goes for Mr Dhorajiwala (see paragraph 152(iv) above).

It was submitted that (a) the data, by reference to which the off-the-shelf WP2 software used by NCB operated, contained “the whole omnibus account”, and therefore (b) an NCB CAN “could not be making, or read as making, the statements in the terms SKAT alleges in these proceedings”. The conclusion would not follow, even if the premise were correct. Mr Horn, arguing the point from the witness box, said that the NCB CANs contained only “standard words that this system always uses. It is using them here in respect of a non-traceable market claim. It cannot have the meaning that you ascribe to it.” But if by its form and content the CAN issued by the WP2 system conveyed, on the face of things, ‘traceability’ (in other words, that there were real shareholdings with real dividend entitlements), it was a system flaw, rather than a reason to read the CANs differently, that CANs were generated for dividend compensation payments where the payee had not acquired any real shareholding. In any event, I am in no position on the evidence to accept the premise that the data ‘contained the whole omnibus account’, the key element there being how the coding of the software resulted in or recorded settlement rather than failure for a Maple Point Model share-less transaction loop.

It was submitted that even if an investor “has a settled balance of shares traceable to an account at VP Securities [i.e. is a shareholder], there is still no entitlement to the gross dividend amount from the issuer; the entitlement is only to the net dividend amount. Characterising this entitlement as a ‘gross entitlement which is then reduced by the WHT amount’ is entirely artificial: where there is a WHT regime in place, there is only ever an entitlement to the net dividend amount.” I disagree. The real shareholder’s dividend entitlement is to the gross dividend amount, but it carries with it a dividend tax liability that is discharged by the withholding. There is nothing artificial about that, just as it is not artificial, but rather it is natural and accurate, to say that an employee who is paid net of income tax under a PAYE scheme earns so as to be entitled against their employer to their gross wage.