The Default Line: THE INSIDE STORY OF PEOPLE, BANKS AND ENTIRE NATIONS ON THE EDGE (26 page)

These tales from the Default Line would appear to be firmly rooted in bubble-era America. But much of the ultimate source of this credit was to be found in Europe and – especially – in London.

Shadow banks were structures that walked, talked and acted like banks, doing many of the same things as banks, but which were not regulated as banks. In 2009 I spoke to Brad Setser, now a US Treasury official, just before he joined the Obama administration as a White House economic adviser. He told me about the shadow banks. ‘They were buying lots of securities, and issuing lots of short-term liabilities. They were taking on a lot of liquidity risk, and running a mismatch between the maturity of their assets and liabilities.’ In other words, they were lending, aggressively, taking in a type of sophisticated deposit, and performing the roles of conventional banks. ‘Yet they didn’t understand the risks they were taking, and a lot of these institutions were undercapitalised for what they were doing.’ The shadow banks were not single institutions but byzantine microstructures of hedge funds, insurers, investment-bank trading desks, credit-raters, off-balance-sheet conduits, special investment vehicles, master trusts, and parts of the normal banking system. The shadow banking system was self-unaware – each cog operated independently – but the end result was the bubble machine that fed Northern Rock, HBoS, Caja Madrid, US subprime providers, Anglo Irish Bank, and all their indebted customers. The billions whose life it affected didn’t even know the shadow banking system existed, but it was worth ten trillion dollars. In 2008 it collapsed, and we are beginning to notice now that it’s gone. Its impact was felt everywhere, from the letters landing on your doormat offering unsolicited credit and 125 per cent mortgages to private equity buyouts of the world’s biggest companies. The shadow banking system turned risky on-balance-sheet loans into supposedly indestructible financial instruments deemed as safe as a government bond by ratings agencies, and insured by the world’s biggest insurance scheme to stay that safe. The system worked well in an era of flexible credit ratings and even more flexible regulators. But it lacked the insurance against bank runs present in the conventional banking system. There was no backstop for liquidity in the event of a run on the shadow banks. London was its capital – and in 2008, we got the run.

Like the dark matter that makes up much of the mass of the universe, the shadow banking system is detectable only by deduction, rather than by direct observation. The US Treasury International Capital (TIC) reporting system offered some light into the financial shade. It was set up in 1934 to track international investment and capital flows, and requires most US financial entities to track foreign purchases of bonds and shares. All purchases are anonymised and given a numerical code corresponding to a nation, with the information returned to the New York Federal Reserve. The TIC is most famous for detailing the amassing of US Treasury bonds by China since 2000. This was a cause of great international angst and financial-market commentary. Some argue that it was the root cause of the build-up of global economic imbalances and thus of the financial crisis. By 2002, according to the TIC data, mainland China (country code 41408) had acquired long- and short-term US government debt totalling $93 billion. By 2007 this had ballooned to $476 billion. A flood of Chinese-earned dollars returned to the USA, so helping to keep Chinese exports cheap, and US lending rates low (see Chapter 4,
here
). Meanwhile, elsewhere on the TIC spreadsheet, another foreign nation, code number 13005, was seemingly hoovering up a broadly similar proportion of another type of US debt. In 2002 its holdings of long- and short-term ‘US corporate debt’ was $101 billion. By 2007 it was $423 billion. That country, the largest holder of US private debt, was the United Kingdom. In the run-up to the crisis, Britain had been buying up as much US debt as the much-feared China, albeit ‘corporate’ rather than government debt. How on earth was the UK managing to ingest American private debt worth one-seventh of the entire British economy?

In the US classification system, ‘corporate debt’ included, yes, an IBM or General Motors bond, as one might expect. But it also included non-Agency mortgage bonds. Traditional parcels of mortgage debt that conformed to Agency standards (and were therefore underwritten by the US state through the agencies Fannie Mae and Freddie Mac, etc.) were thirty-year loans given at a fixed rate. The market for non-conforming loans included subprime, interest-only mortgages, and as we know, this sector was booming at the time. The non-Agency mortgage-backed securities market showed up in the US TIC data as ‘purchases of corporate debt’. For Setser this points to purchases by securitisation vehicles, key parts of the shadow banking system, based in the UK, in the City of London.

‘A lot of the shadow banking system was operating through London,’ Setser told me. ‘You see it in the capital-flows data. You see it in 2005–6–7 when every risk, every mortgage was being repackaged and spliced and diced, and AIG Financial Products [based in London] was selling [credit default swap] insurance to banks. During this period there was a huge boom in US TIC data of US corporate debt sales to London. Then, in 2007, that stopped. These institutions were operating through London – both sides of their balance sheet were actively operating through London. European banks were borrowing a lot of dollars and using the funds to buy a lot of repackaged but still risky mortgages. London was a critical part of the US financial system for regulatory and tax reasons.’

No other G7 nation had acquired anything like the amount of US corporate debt as the UK. Japan, and China of course, amassed US government bonds. But the only other jurisdictions that saw such a rapid boom-time increase in purchases of US corporate bonds were the Cayman Islands ($80 billion in 2002, rising to $375 billion in 2007), Luxembourg ($78 billion to $370bn), Belgium ($92 billion to $323 billion) and Ireland ($43 billion to $216 billion).

The TIC data only gives a glimmer of the bigger picture. The UK figure does reflect the way that Wall Street washed its dirtiest laundry in the City of London. It also reflects the fact that big European banks operated their dollar trading through the City. The full reality would only be revealed if the UK published its own version of the TIC data. Although the chancellor has often told us that Britain owes a portion of its debts to China, a figure has never been published. Why?

In the realms of international financial diplomacy, fingers point towards the UK’s ‘appallingly bad’ data on financial flows, more suited to a secretive offshore financial centre than the world’s capital of finance. London’s wish to protect its role as manager of Gulf oil money and as a conduit for China’s surplus dollars might explain why. A portion of China’s purchases of US government bonds were originally scored as ‘UK’ purchases. ‘More attention is needed in the USA to flows through UK banks,’ Setser told me. ‘There are important lessons to be learned, and the UK needs transparent flow of money in and out. These would be important crisis warning signs.’ UK TIC-style data would be revelatory for global finance. But perhaps the City fears that would be a step too far out of the shadows.

The most tangible manifestation of the shadow banking system was, as Setser pointed out to me, an office in London’s smart Mayfair district. The people who worked there were told it was the most profitable office per capita in the world. This was the office of AIG Financial Products (AIG FP), run by Joe Cassano. The philosophical core of this triumph of high capitalism was, as Cassano triumphantly told AIG investors, ‘the bifurcation of credit from the host contract’. (This splitting of the credit atom is described in detail in the next chapter.) Such was the confidence that prevailed in Mayfair barely ten months before AIG’s collapse. The business written in London was primarily for ‘regulatory capital management’ by its banking clients.

AIG had begun selling so-called credit default swaps (CDSs) in 1998. J. P. Morgan were first in line to take advantage, having developed the underlying derivative technology. In the eyes of J. P. Morgan’s bankers in London, they had ‘stuffed our leftover positions into AIG, getting rid of risks’. The theory was that AIG, with its giant insurance-company balance sheet, could offer to protect the owners of parcels of loans – everything from subprime mortgages to corporate bonds to car loans – from the risk of default. AIG would charge a fee worth a fraction of a percentage of the security, and in return AIG would make good the loss in any default. J. P. Morgan began to arrange more and more such trades with AIG for its clients, beginning with an Italian bank. At first, these trades principally involved corporate debt, about which there was a wealth of historical default data, and well-tested modelling of losses (see Chapter 7,
here
).

In essence AIG’s clients paid a small amount to make their trading risks disappear. AIG itself collected the fees, but did not then further hedge these risks, and did not at first put aside any collateral. The financial system had, in theory, found a way to make credit risk pass along a chain of buyers, and then at the end, at AIG’s Mayfair office, it simply
disappeared
. It was magic. How so? AIG’s AAA credit rating was the bedrock, as it meant there was little or no need to post collateral, as was normally the case with these CDS trades. AIG also employed the services of Ivy League quantitative analysts, who used a broad number of models filled with data from banks, central banks and international think tanks, detailing losses in past recessions. They had special models for Dutch real estate, for example, and a different model for German small- and medium-scale enterprises. Mortgage securities would model the likelihood of default of each of the 80,000 individual mortgages. Even before the modelling stage, AIG staff would visit the banks writing the original loans, check their internal models, visit the repossession department, and identify lax credit standards. AIG would ‘positively select’ loan portfolios so they were neither too concentrated on a single industry nor too risky in general. They were replicating the credit-risk function of a bank. Except they were covering several banks at once, using the same model. There were two results.

Firstly, the trades were fabulously profitable. AIG had assumed away the loss statistically, so they had no need to post collateral or hold buffer funds for the contingency of a default, because the model said that default was impossible. The annual fees paid on credit default swaps could be booked as almost pure profit. Then, if modelled in the right way, a future stream of these fees over a period of years could be booked upfront as day-one profit. This is how you get fabulous profits: (1) offer an insurance policy against a risk that you prove will never happen; (2) assume it will never happen; (3) take all the fees as profit; (4) take all the future fees as upfront profit. This is how AIG FP ended up with what its employees were told was the most profitable office space in the world. In 2006 Joe Cassano received $44 million in salary and bonuses. Even in 2007, as things began to unravel, he got $24 million.

AIG’s major customers were the European banks. So the second result was that AIG was at the core of a machine that was decapitalising the basis of the European banking system. Andrew Forster of AIG explained how it worked on that fateful December 2007 conference call: ‘The majority of our trades are regulatory capital-motivated rather than for economic risk transfer purposes… And buying the Super Senior protection from us, they’re able to reduce their capital charges down from 8 per cent to just 1.6 per cent.’ In essence, European banks could trade away their credit risk with AIG, and hold dangerously low levels of capital, safe in the knowledge that AIG would cover their backs in a disaster. Prior to its takeover by RBS, the Dutch bank ABN Amro had $56 billion of these trades, which accounted for $3.5 billion of questionable ‘capital’, intended to satisfy the regulators. Danish, German and French banks were all involved in a quarter of a trillion dollars worth of these so-called ‘regulatory capital swaps’. It turned out that it was only a short step from decapitalisation to decapitation. Not surprisingly, Christine Lagarde, then French finance minister, was at the forefront of efforts to persuade US authorities to bail out AIG in the immediate aftermath of Lehman Brothers collapse. The net result was that European banks held much less capital than they had done, and could deliver even higher returns to shareholders and bankers.

AIG fell over in 2008. But it received the largest-ever injection of US government cash into a private entity – not because of defaults, but because the models underestimated the impact of collateral calls. Reductions in the value of the underlying loans, and reductions in AIG’s own credit rating, would result in AIG having to post collateral to its trading partners. At first, its AAA rating protected it, but that was lost, and AIG suffered a series of downgrades, ultimately requiring $50 billion of collateral. This spiral of doom accelerated because AIG also had to account for the fall in value of the CDS contracts on its accounts. It protested vehemently that the contracts were at no risk of actual default, and that its rude financial health meant that it could ride out any dip in the value of the contracts. Ultimately, though, both risks combined, and AIG’s trading partners demanded collateral it did not have, until it ran out of funds. It had suffered an invisible but epic cross-continental modern-day bank run. And it wasn’t even a bank.

What was allowed to happen at AIG said an awful lot about regulatory standards. No one in London really seemed to know what was going on. An insider told me that officially AIG FP was regulated by the French central bank, the Banque de France, even though trillions were traded from London. ‘Inspector Clouseau turned up twice a year and pretended to understand our accounts,’ he told me. Another former insider, Shirley Beglinger, told me: ‘The FSA should have noticed that AIG’s name popped up all over the place, and should, as a result, have started to take a more piercing look at AIG Financial Products. As long as it was going right, heaven help the regulator who tried to stop it, and they just weren’t willing to pick that fight, I suspect.’