How London bankers at Goldman Sachs and Nomura made millions from taxpayers

If you can’t spot the sucker at the poker table, better fold fast: the sucker is you.

That truism, taught to every investment bank intern, sadly rarely seems to filter out to civil servants when they’re sent in as cannon fodder to negotiate with the City’s deal makers.

But if you suspected that we British taxpayers have been fleeced at every turn by the bulge-bracket banks, you should see some of the deals the public sector was suckered into over in Italy, France and Portugal, not only in the run-up to the financial crisis, but also afterwards.

Documents filtering out of a parliamentary inquiry in Portugal have found their way to The Independent, with a particularly shocking example of what one financial expert describes as “possibly the most stupidly complicated financial transaction ever sold”. The trades were all perfectly legal, but provide a salutary lesson in why you should never buy something you don’t understand – particularly from a London investment banker in a thousand-dollar suit.

The paperwork describing them consists of page after page of algebra, representing an ever-growing, unpredictable combination of impenetrable risks, the impact of each moving part multiplying like a virus. As the career banker Moorad Choudhry, of Brunel University’s Department of Mathematical Sciences, said when we showed him: “It made me laugh and cry in equal measure. What on earth were they thinking? I’ve seen some bad products in my time but this is something else. This is the Apocalypse Now point of the banking industry, when everyone and everything is descending into madness.”

The sucker that bought the product was a Portuguese state-owned train and subway company, Metro do Porto. It was sold by a team of London-based bankers at Goldman Sachs at the start of 2008.

In effect, it was an attempt by Metro do Porto to offset and lower the amount of interest it was paying on a portion of its debt – €126m (£100m) of debt, to be precise.

Nothing wrong with that – a prudent effort to save the Portuguese taxpayer millions of euros.

However, the deal it agreed with Goldman was so toxic that, far from saving millions of euros, in less than a year Metro do Porto ended up owing the bank millions.

The documentation explaining the product is so fiendishly complicated, with so many moving parts, that according to a specialist banker The Independent showed it to, the extent of the risk would have been impossible for anybody but the most experienced of experts in the field to gauge. Sources close to Goldman say Metro do Porto knew what it was buying and that it was the result of a “request for proposals” – in effect, a tender inviting banks to pitch.

Whatever the process, the result was a disaster, which I’ll try to explain as simply as possible.

In effect, the deal was what is known in City parlance as a “swap” – a bet (although the bankers dispute that term) – lasting for 20 years. Every quarter, Goldman and Metro de Porto exchange payments depending on whose side is winning.

The Goldman end of the deal meant it would owe Metro do Porto if the 10-year swap rate (a type of interest rate) for sterling minus the 10-year swap rate for the euro was higher than a specific level at certain dates. It only involves a few mathematical formulae to calculate.

The Metro do Porto side, on the other hand, was fiendishly complicated. It consisted of more than 20 equations based on the relative movements of five different rates – some bizarrely related to the dollar and the British pound – nothing to do with Portugal whatsoever. The rates were: the 12-month Euribor rate for euros, the 12-month Libor rate for US dollars, the 10-year swap rate for euros, the two-year swap rate for euros and the 10-year swap rate for sterling.

Within the calculations is a function known as a “snowball” – a fitting metaphor given its appearance in a blizzard of algebra. A snowball is where losses at one stage of a bet are compounded into the next stage, the one after that and so on in such a way that losses spiral dramatically and fast.

In fairness, there was a cap to the amount of losses Metro do Porto could run up. The cap was, and I quote: “{SumDCF [n] x 6.33% - SumCpn[I,n-1]}/DCF[n], where I = 1,2,3”.

Got that? I doubt if the poor saps at Metro do Porto did either. Because, while that brought a small element of safety to the bet, it was not much, and it too was based on moving parts that meant the cap – and hence the maximum amount Metro could lose – increased every year.

So it was that, a few months later, a valuation of the contract turned up the fact that Metro do Porto was now down on the bet with Goldman to the tune of €120m. That’s a lot of signalmen’s salaries.

We can but imagine the panic this must have instilled in a sleepy provincial public service in an already-impoverished corner of southern Europe.

The train operator contacted another investment bank in London, Nomura, to try and come up with a scheme that would break it out of its disastrous position. Nomura – perhaps spotting that sucker in the room again – carved up a whole new bet in 2009. New, more opaque, and toxic.

This time, there were two parts to the deal. Part One was a mirror image of the Goldman bet, in effect cancelling the risk out. Fair enough.

But as well as this remedy, the folk at Nomura decided to make things a bit more interesting. They added a new bet involving payments based on a “black box” index invented by Nomura. “The black box index is a real piece of work,” said one swap expert. It is in effect an algorithm devised by Nomura’s whizz-kids that responds to an array of movements in the financial markets. An investment in their picks is supposed to perform well whatever the financial weather, although in the past other algorithms have blown up spectacularly when unexpected events, like the sub-prime mortgage disaster, have hit.

Despite those risks, Metro do Porto agreed to take bets based on the movement of this black box index, hoping they would fare better than its bet with Goldman.

This time, there was a fixed cap to the rail group’s exposure – 14.75 per cent a year – but, again, the losses on this Nomura bet were to grow rapidly from zero on day one to more than €80m in less than a year. This despite the fact that during this period the losses on the Goldman bet also increased – hardly what anyone could call a “remedy”.

To recap, then. A financial product aimed at keeping down the interest rate on a €126m loan was, just two years later, to cost the little transport utility more than the amount of the loan itself.

That, though, was only the half of it.

Metro do Porto, instead of continuing the costly Goldman bet and its mirror image (part one of the Nomura bet), decided, understandably, that it should cancel them – after all, why bother running a bet for and against Brazil winning the World Cup? To this effect, Metro do Porto asked the banks to work together and with Metro do Porto to cancel the identical components of the two bets.

The two banks had already made many millions on their sucker trades, so Metro do Porto’s executives presumably thought the pair would willingly agree to bring them to an end – perhaps even rather sheepishly.

How wrong they were.

Imagine the look on the subwaymen’s faces when the two banks said they would only cancel the supposedly identical components of the two bets, in return for €26m to reflect the difference in their valuations of these identical parts. €26m! And this on top of the losses the group was suffering from the other bets with Goldman and Nomura.

In 2010, Metro do Porto finally sought some independent financial advice and, as a result, confronted the banks about their behaviour. Perhaps fearing a public scandal, from demanding €26m to end the identical components of the two bets, the valuation models of the banks magically changed. The new valuations – funnily enough – meant Metro do Porto would get a windfall of about €20m to cancel the identical parts out.

The banks insist this was not a joint trade between them, although Nomura’s documentation from the time refers to both its and Goldman’s valuations, suggesting they must at least have been aware of each other’s changing positions. The banks also dispute our figures, which are based on paperwork from the time. They also declined to give their versions of what the amounts were, citing client confidentiality.

Nomura declined to comment for this article. A Goldman spokesperson said: “Goldman Sachs has a long track record of working with the Portuguese Government, with whom we continue to have a constructive relationship.”

As for the remaining tens of millions of euros that the two banks ended up winning on their various bets against Metro do Porto, we will never know the figures. This is because the Goldman-Nomura trade was far from being alone. In fact, there were many other disastrous swap deals done by Portuguese state entities with London bankers from Deutsche Bank, BNP and JPMorgan that also lost tens, and hundreds of millions of dollars. Portugal paid off its losses for most of them in one big negotiated settlement. There are no outstanding disputes between Metro do Porto, Nomura and Goldman.

Swaps like these have caused some of the biggest scandals in modern financial history. A bigger, but arguably less complex, case involving Santander and Metro do Porto is still rumbling through the London High Court. Politicians have resigned, parliamentary inquiries have found appalling levels of mismanagement. You could argue that the banks are not to blame: caveat emptor – “buyer beware” – being the City watchword. The banks involved say Metro do Porto’s staff were entirely aware of what they were signing up to. But the wave of scandals leave a nasty taste in the mouth of many in the City.

As one expert in swaps told The Independent: “These exotic products can incorporate such large profit margins for the banks who create and sell them that there is really only the tiniest chance that the buyers could ever win. But, because of their in-built complexity, the client never knows that until it is far, far too late.”

Which begs the question: is this really how we want our financial industry to behave?

Clear as mud: Details of the deal

Here is a taster of the deal Goldman Sachs sold Metro do Porto from documents obtained by The Independent.

This is the interest rate formula for the Metro do Porto side of the Goldman Sachs “swap”. For swaps officianados, bear in mind the buyer is a relatively small, Eurozone state-owned transport utility.

Here goes:

Interest: PayIndex[1]+PayIndex[2]+PayIndex[3]

PayIndex[1]: If Index 1 is less than or equal to 7.75%, 1.665%+Radial [1,n], subject to a maximum of AverageCap [1,n]

Otherwise,

7.75%+Radial[1,n], subject to a maximum of AverageCap[1,n]

PayIndex[2}: Radial[2,n]subject to a maximum of AverageCap[2,n]

PayIndex[3]: Radial[3,n], subject to a maximum of AverageCap[3,n]

Index1: Max (12m Euribor, 12m USD Libor), Reset in Arrears

Index2: 10 year EURSwap Rate - 2 year EUR Swap Rate, Reset in Arrears

Index3: 10 year GBP Swap Rate - 10 year EUR Swap Rate, Reset in Arrears

12m Euribor: EUR-EURIBOR-Reuters, with a Designated Maturity of 12 months as published on Reuters page [EURIBOR01]

12m USD Libor: USD-LIBOR-BBA, with a Designated Maturity of 12 months as published on Reuters page [LIBOR01}

10 year EUR Swap Rate: 10 year EUR Interest Rate Swap Rate

Source: Reuters Page ISDAFIX2, 11am London Time Fixing

2 year EUR Swap Rate: 2 year EUR Interest Rate Swap Rate

Source: Reuters Page ISDAFIX2, 11am London time fixing

10 year GBP Swap Rate: 10 year GBP Interest Rate Swap Rate

Source: Reuters Page ISDAFIX3, 11am London time Fixing

Radial[1,n]: Max (Radial[1,n-1]+Index1-7.75%,0),Radial[1,0]=0

Radial[2,n]: Max (Radial[2,n-1]-0.30%-Index2, 0), Radial[2,0]=0

Radial[3,n]: Max (Radial[3,n-1]-0.45%-Index3, 0), Radial[3,0]=0

n: Number of the Interest Calculation Period, n=1,...,80

For the avoidance of doubt, n=1 corresponds to the Interest Calculation Period from 15feb08 to 15May08

AverageCap[i,n]: {SumDCF[n]x6.33%-SumCpn[i,n-1]}/DCF[n], Where i=1,2,3

SumCpn[i,n] = SumCpn[i,n-1]+Cpn[i,n], SumCpn[i.0]=0, Where i=1,2,3

Cpn[i,n]:PayIndex[i]xDCF[n], Where i=1,2,3

DCF[n]: Daycount Fraction of Interest Calculation Period n, according to 30/360, adjusted modified following convention

SumDCF[n]: Daycount Fraction from the Effective Date to the Interest Payment Date of Interest Calculation Period n, according to 30/360, adjusted modified following convention.