Tax avoidance laid bare: the BNP Paribas case
Back in 2005, times were hard for investment banks. While the financial crash was still some years away, the “DOTAS” rules, introduced in 2004, required them to disclose tax avoidance schemes to the UK tax authority.
It put a serious dent in part of their business model: tax planning. No longer could they rely on creating a steady stream of structured avoidance schemes that might run for a few years before being detected by HMRC and closed down.
The new rules changed the dynamics: schemes would now be closed down much more quickly, on being disclosed to HMRC.
The initial response to this change, at least by some of the banks, was to make them more aggressive in seeking an immediate, one-off tax advantage. So they were more willing to take a punt on a risky scheme if it had a chance of generating a one-off deduction immediately before their December year-end.
Even if the scheme was closed down immediately after disclosure, they would have secured their tax benefit – which in the case of the London branch of a foreign bank might wipe out the whole of its tax capacity for the year.
This is the kind of aggressive tax planning laid bare in the recent First-tier Tribunal (FTT) case of BNP Paribas SA (London branch) v Commissioners of HMRC.
The case involved a complex series of financial transactions within the BNP group, followed by a sale of dividend rights (a 'dividend strip') for £150 million to an unrelated counter-party in the Alliance and Leicester (AL) group.
The effect was in substance a loan from AL to BNP: a fixed annual dividend akin to interest, with a final dividend of £150 million equivalent to the repayment of the loan.
The sale of dividend rights took place in mid-December 2005. The London branch of BNP first bought the rights from BNP’s Luxembourg subsidiary for around £149 million, and a day later sold them on to AL for £150 million.
The attempted trick was to crystallise an upfront tax deduction for the £149m payment against the London branch’s corporation tax liability for 2005, while claiming that the receipt of £150 million receipt from AL was not taxable due to the arcane provisions of section 730 ICTA 1988.
In other words, they hoped to translate a transaction producing a profit of just short of £1 million into a tax loss of £149 million, wiping out their UK tax liability for 2005.
The purchase and resale were claimed to be part of the normal course of financial trading by the London branch of BNP, and the arrangements as a whole between the BNP and AL groups were said to provide BNP with finance at a commercially attractive rate.
The smoking guns
The problem with the arrangements, however, was that – when viewed as a whole – they were not commercial for the BNP group. The FTT found that, contrary to BNP’s arguments, the “cheap financing” was merely window-dressing. In fact they obtained rather expensive financing in order to generate a large tax benefit.
The FTT based its findings on essentially two kinds of evidence: the contemporaneous documents and the underlying economics. The smoking guns included some of the internal BNP approvals papers, which were “candid” about the purpose of the transaction. As a prophetic note from a meeting put it: “the tax audit team have stated that they think the transaction is very aggressive and that it will go to court. If the transaction does go to court, they consider that we will lose."
But perhaps most damning was an unhelpful email exchange within the BNP team expressing concerns that it would be better to exit the arrangements immediately after the sale to AL because “we are losing money after day 2” and the continuing dividend payments would be a “drag on the group” – hardly the hall-marks of cheap financing.
The underlying economics bore out these concerns. While the arrangements looked commercially advantageous for BNP on a pre-tax basis, this advantage was more than cancelled out by the fact that the dividend payments to AL were not tax deductible (as, for example, interest on a simple loan would have been). The only basis on which the arrangements generated a benefit for the BNP group was by assuming they secured a huge tax deduction without a corresponding taxable receipt.
It’s interesting to note that the economic analysis by the FTT considers the transactions as a whole, on a group basis. Individually, each member of the BNP group involved in the transactions showed a profit. But the FTT wasn’t fooled by this: these individual profits were just “engineered ‘commercial’ incentives considered necessary for the planning to succeed”.
Based on these findings, the FTT decided the transaction was not part of the BNP London branch’s financial trade. Instead, it was a “tax recovery device”. It also found in HMRC’s favour on the alternative arguments.
Even if the transaction had been part of the London branch’s trade, the outlay of £149 million was not wholly and exclusively for the purposes of that trade but for the purpose of obtaining a tax benefit, and therefore not an allowable expense.
Finally, the whole premise of the scheme – that the receipt of £150 million wasn’t taxable due to section 730(3) – failed in any event because, on a broad, purposive reading of that provision, it didn’t apply in the context of a trading receipt. Nor did the FTT have any sympathy with a procedural point raised by BNP, aiming to preclude consideration of this final issue.
So this was a comprehensive victory for HMRC. BNP were left with a transaction that had made no commercial sense without the large tax benefit, and had even generated an additional tax liability due to the non-deductible dividends, albeit small in relation to the scheme as a whole.
The supposed loophole on which the scheme was based was closed – as BNP had feared it would be, once disclosed to HMRC – in January 2006. But in the end this proved to be a “belt and braces” approach as the scheme did not work.
It has subsequently been reported that BNP have decided not to appeal the decision, and indeed had already paid the tax in advance of last year’s tribunal hearing. The only question left in my mind is why they even litigated the case, rather than conceding without litigation, given the prejudicial nature of the evidence. I suspect that a UK-based retail bank, conscious of its reputation with consumers, would not have wished to have this kind of artifice aired in public.