Squirreled away somewhere on W*k*l**ks, you may come across some papers prepared by members of Barclays Structured Capital Markets team describing transactions submitted for approval. For the benefit of journalists, politicians and other readers who haven’t a clue about what all those boxes and numbers mean, here is the lowdown on each of the papers as summarised by The Financial Crimes:
This is just a large purchase of index-linked gilts. To Barclays, me and you the inflation linked uplift comes tax free, so what they are doing here is nothing new. They aren’t to keen on the inflation based uplift, so they shift some of the risk with a Total Return Swap.
Tax Aggression (1/10), Harm to the UK Exchequer (3/10)
This submission starts off with a red herring and an outright lie: “The benefit of the investment derives from the fact that the BBPLC group will generate pre-tax income at $R interest rates of c.19% but will not have any exposure to the $R:GBP exchange rate.” because in the detailed write-up it says “HoldCo hedges the expected $R-linked dividends on the RealCo Equity by entering into a series of nondeliverable currency forward sale agreements with the markets floor (the “Market Forwards”) under which HoldCo agrees to pay fixed $R-linked amounts in exchange for the GBP equivalents translated at the forward GBP:$R FX rates.”. In other words Barclays makes an investment in securities denominated in Brazilian Reals and then hedges the income return from its investment so that it has no net currency exposure, at least as far as the income is concerned.
So what is going on? The income is hedged but there still seems to be a currency risk associated with the capital. The giveaway is that although the investment in Brazilian Real securities is only for £300 million, as that money is flowed through the Barclays group structure, it is mixed with an extra £700 m from within the Barclays group, which is invested in an intermediate holding company which lends the £700 million back where it came from and passes the £300 million to a lower tier company which makes the £300 million Real investment. That lower company and the holding company enter into a currency swap, supposedly to hedge the position of each company, but curiously the swap is for the same principle amount as the investment in the holding company £1 billion, not the amount of the Real exposure in the company £300 billion The clue as to what is going on here is that the tax rate is 30%, the same as the ratio of the actual exposure to the hedged amount. We quickly deduce that what is going on is something to do with hedging the after tax value of the return to Barclays irrespective of the future exchange rate.
To understand it in a little more detail we have to know that the FX futures price (at which the Real income was sold) broadly reflect the interest rate differential between the Real and sterling, so that if the currencies actually remained the same there would be a benefit to Barclays because there would be a profit on the forward sale, but this is tax free because it is a hedging transaction on preference shares even though it only hedges the income and not the capital.
This is where the intra-group cross currency swap kick in. The swap is designed so that any profit or loss is only taxable or deductible in one company in the group, although Barclays doesn’t know at the outset whether it will depreciate or appreciate. If we assume that the value of the Real fells, the Barclays group would lose some real economic value in the value of the unhedged capital value of the Real securities (original cost £300 million), but it would have a proportional deductible loss on the value of the cross currency swap in one company in the group, although overall within the group there is no net loss but the company with the profit on the swap is not taxed on that profit again because it is a hedging transaction. But lo and behold the loss on the securities (X% of £300 million) is the same as the tax savings on the swap (30% of X% of £1,000 billion), so the group as a whole is perfectly hedged.
A bit complicated this one, but what is clear is that this is not much of an investment in Brazilian Reals as an exploitation of the rules of the taxation of foreign currency instruments. The trouble for the Revenue is that they cannot say definitively that this is a tax avoidance scheme because they have no idea what the future value of the Real will be. In the unlikely event that the Real appreciated, there would be substantial extra tax paid.
Tax Aggression (4/10), Harm to the UK Exchequer (4/10)
This is an old classic Luxembourg profit participating loan structure. At least one of the big 4 accounting firms was hawking this structure around in the late 1990’s so it is good to see it still in use. Maybe they stole it from Barclays, because they audit them. Anyway, this has nothing to do with the UK tax man and is a raid on the Italian treasury, courtesy of crummy Italian law and a ruling from the Luxembourg authorities.
Tax Aggression (3/10), Harm to the UK Exchequer (0/10), special bonus point for international business promotion
OK, this is a little more juicy. The idea here is that a Barclays sub in the Isle of Man manages to invest in a portfolio of government securities in such a way that they manage not to pay much in the way of tax. If we assume that Barclays would otherwise have held such a portfolio, even though the return is sub LIBOR (their normal benchmark), the deal doesn’t seem so bad because it is pretty much tax free. The deal needs a compliant unconnected Luxembourg bank, which in this case is the Luxembourg branch of Nordbank, a German bank owned by the City of Hamburg and a few others, who take little risk but get paid £2 million for their trouble. Barclays manage to get most of their income from the portfolio to the Isle of Man tax free, because £14 million of the income is taxed as a capital gain on shares, and £52 million of dividend income from redeemable preference shares is reduced by a £52 million credit for Luxembourg withholding taxes. Nordbank doesn’t mind paying the Luxembourg withholding taxes because they get to deduct the cost from their own tax liability and are left whole.
Tax Aggression (5/10), Harm to the UK Exchequer (5/10)
Another biggie. This one from Michael Keeley, who used to manage the son of an absconded murderer (and an aristo at that) when he worked at Kleinworts, not that that has anything to do with this story, but the numbers are big - £16 billion, so this is likely to be a story based on the deductibility of interest and the non- taxation of income – which it is, but contrary to what the “expert” in the Sunday Times said this was not all about money going round in a circle, but about Barclays borrowing $4 billion dollars in the UK (and claiming tax relief for doing so), routing that money through Luxembourg, back into a UK partnership and off to the US where it financed a portfolio of real Estate and other assets. The UK revenue that the income coming from the US to the UK partnership is mostly taxable in Luxembourg to the Luxembourg partner, and therefore outside the scope of UK tax, while the Luxembourg authorities think that the income is all exempt from tax because it is covered by the participation exemption regime. Uncle Sam does very well out of it because he claims a big chunk of withholding tax.
So the US is up by $80 million a year in tax and the UK loses $61 million a year in revenue, and Barclays think they are ahead, but according to Keeley’s proposal Barclays have a $45 million annual benefit based on $65 million of pretax margin. Where does that pretax margin come from? According to Keeley’s figures the American borrower is willing to pay an effective premium because of some US tax breaks, which he says gives the US borrower an effective borrowing rate of 0.40% under LIBOR, whilst simultaneously giving Barclays a return of 1.60% over LIBOR, the difference being a US tax saving equivalent to the missing 2% or $80 million. So Uncle Sam is out of pocket $80 million from BB&T, but he gets $80 million back from the withholding tax on the distribution to the UK, with $45 million a year ending up in Barclays, about $16 million a year in BB&T and a smidgin in Luxembourg and because it is always a zero sum game it is just the UK taxman who is out of pocket to the tune of $60 million a year, but then Keeley is Australian, so no surprises there.
But if I was a Barclays shareholder I would want to know why anybody thought $45 million was a good annual return on $4 billion of balance sheet exposure to a single counterparty, even if it was a 20% risk weighted asset, given all of the risks. What was the strategic reason behind it? In my view, not enough to justify the use of balance sheet and risks?
Tax Aggression (5/10), Harm to the UK Exchequer (8/10), Congressional Medal of Honour for $20 million annual contribution to the US economy.
“This memo sets out the minutes from the SCM Approvals Committee meeting on 13 July 2006 for the Establishment of a Luxembourg Office for SCM.”.
No idea what this is doing in here.
A valiha is a bamboo zither from Madagascar, but that has nothing to do with this transaction. This transaction is from Ian Abrahams and is not very complicated. Despite his reputation as a tax lawyer, Abrahams also another reputation as a frequent poacher of other people’s ideas. Many unsuspecting young bankers from other institutions were brought in to interview with Abrahams only to have him quiz them on their ideas and reject them. He had a similar reputation amongst advisers for demanding to know their best ideas. Maybe it is his own work, but on balance, even if it is, he has stoledn so much in the part that he doesn’t deserve any credit for it.
Barclays just happened to have two large interest rates swaps entered into between different parts of the group that were substantially in and out of the money for the two sides of the swap. There is nothing unusual in this. A special purpose company within might raise funding from a particular source which does not match the interest rate risk of the assets it holds, so it will enter into a swap with the treasury desk in the main bank company. One side pays a fixed amount of interest on a notional amount of principal and the other side pays a floating rate of interest. At the outset there is no implicit value in the contract, but over time as short term rates of interest move the fair value of the swap may become increasingly valuable to one party and correspondingly less valuable to the other party. Since this is all within the Barclays group, nobody is too bothered about it. Taken on its own the value of the more profitable side of the swap is reduced by the cost of the tax that would be payable on the profits, and similarly value of the losing side of the swap (which is not really mentioned in the paper) is increased by tax savings from the allowable losses.
Barclays figure that they can get some additional value from the swap if they can sell the swap to another bank (in this case CSFB) for a price which more closely reflects the pre-tax value of the swap (i.e. reducing the discount for tax). CSFB are most likely to be making losses in the UK. Foreign banks in the City have a nasty habit of making losses in the UK by paying their London based staff, while actually making their profits from deals booked in other countries, so they often have an appetite to buy in taxable profits, and in this case the price they set on their losses is only 20% of the face value of the tax. The problem for Barclays is that an outright sale would give rise to a capital gain, so they get around this by forming a partnership between two Barclays companies and and one of the partners contributing the swap into the partnership in return for its interest. CSFB is then admitted to the partnership for a substantial consideration, most of which is owed to the Barclays company that contributed the swap, The partnership agreement allocates 95% of the profits to CSFB, and though unstated in the paper, the swap would probably be unwound within a few months and the partnership some time thereafter.
Almost the same effect as a sale, but not quite and there is no capital gain because that’s the way the law works. Hey, I didn’t write the law so don’t look at me like that, but frankly, can you honestly say you wouldn’t do the same as Barclays?
Tax Aggression (4/10), Harm to the UK Exchequer (6/10)
One final footnote: The comments in the press about Barclays SCM being a nasty place to work ring true. The City can be a brutal place to work, but sine the 1980's I have known people from many other banks involved in similar tax based work at many banks (Goldman, Morgan Stanley, Merrill Lynch, JP Morgan, Chase, Citi, Bankers Trust, Deutsche, Kleinwort, AIG, Gen Re, Nat West, RBS Lloyds, Halifax and more), and they are generally intelligent, even geeky, but mostly personable, but the ones from Barclays have often have been without a doubt more unpleasant than most, theonly people coming close being Jenkins ex-staff at his former employer Kleinwort Benson.