Ben Bernanke, speaking before the Senate budget committee, said he was particularly angry that at AIG, “There was no oversight of the financial products division. This was a hedge fund basically that was attached to a large and stable insurance company.” What was going on?
The company’s net losses, reported as $62bn for the fourth quarter of 2008, were just shy of $100bn for the whole year. AIG’s SEC filing, the Form 10-K, tells us that the financial products division posted an operating loss of $41bn. This piece of AIG is a huge player in the derivatives market with $1.5 trillion of notional exposure to all manner of risks. It is now notorious for what they call – wait for it – its ‘super senior multi-sector collateralized debt obligation credit default swap portfolio.’
Typically, the principal and interest flows of a book of sub-prime mortgages were sliced and diced into asset sub-pools that would underpin a tiered series of securities. Each security would have a different credit quality, depending where it sat in the tier, what its asset backing was, and how it was structured. The rating agencies scored them, up to AAA, the highest investment grade, at the top. The ‘super senior’ stuff was even better than AAA and this was mostly where AIG played, providing insurance to the holders of these wonderful assets. They really were turning muck into brass.
But then the mortgages started going bad, so bad that even the super senior notes faced losses. So much for the financial structure. As the securities fell in value and were downgraded AIG had to put up increasing amounts of collateral and by August 2008 had already shipped out $20bn. And when AIG itself was downgraded the whole thing spiralled out of control. Within days the Fed had stepped in. Since then most of the AIG-backed CDOs - $62bn of face value - have been taken out of the market. On average, these super quality assets had lost over half their value. The investors, however, have received the whole of their capital back, since AIG, using Fed money, has paid the balance to settle and terminate its CDS contracts.
It is interesting to look at the counter-parties for these transactions. Contrary to one’s worst suspicions, most of these CDS transactions were not naked speculation with people who had no interest in the related debt instruments. AIG generally issued the CDSs to buyers of CDOs, and this made it possible for the rescue vehicle to buy the CDOs and tear up the attached CDS contracts with AIG. The lucky recipients were mostly investment banks, notably Société Générale, Goldman Sachs, and Deutsche Bank. So through a blend of loans to AIG and the CDO buyout, that $62bn should really be seen as a component of the bank bailout. The Fed is now a significant holder of securitised sub-prime mortgage debt.
Where else did AIG lose $100bn? By far the next biggest operating loss, over $37bn, came in the life insurance and pensions division. Life and pensions is a slow-moving, long term business, so this is puzzling. Here’s how they did it. They used a ‘US Securities Lending Program.’ Investors borrowed bits of AIG’s life and pensions investment portfolio in exchange for cash collateral, quite possibly to support short selling, presumably by investment banks and hedge funds. But AIG wasn’t happy with this little business. “To earn a spread,” as the 10-K says, AIG invested the cash collateral in – you’ll never guess - residential mortgage backed securities (RMBS). By the third quarter of 2008, $69bn of borrowings were outstanding (it had been $82bn at the start of the year) and people wanted their money back. But by then the RMBS market had seized up, which made things tricky. In December the Fed bought $39bn face value RMBS from AIG for $20bn.
What can one make of this shocking tale? Here are some brief thoughts.
Across the group, not only in financial products, there was a clear urge to pursue financial, that is, money returns over and above any trading profit that might accrue from the ostensible business. This is perhaps the real story of the whole financial crisis.
In exposing itself so massively to default and liquidity risks in the US housing market on both the asset and liability sides of its business, AIG ignored the basic principles of risk management. The management bet the whole company on the housing bubble.
The practice of lending stock for other people to short-sell is pretty shabby. Taken to extremes, it must be self-defeating, since the shorting activity will end up reducing the value of the lender’s investment portfolio. Did the growing demand for short-able stock not ring any bells? Had they held on to the cash collateral, they might have got away with it, but by investing in RMBS, AIG turned more than 10% of its life and pensions investments into junk. That seems little short of theft.
We’ll never know how far AIG or its counter-parties might have been influenced by the belief that AIG would not be allowed to collapse. Whether they were or not, what we know now is that whatever is done about new regulation, if you’re big enough anything goes. There must surely be moves to limit the size and scope of all kinds of financial institutions.
The basic concept of insurance is simple and ancient: to share the losses of the few amongst the many. The insurer or community pools risks that individuals could not afford to suffer. In a modern financial system, as the insurer’s aggregate exposures rise he’ll buy reinsurance on the same principle. The pooling and re-distribution of risk has an important social function. If there were no insurance I would do many things differently or not at all. When an insurer switches from averaging risk to concentrating it, we have a problem that calls for social intervention, either to require the insurer to change its behaviour or to take it out of the market. The question for regulators is whether they can credibly develop and deploy the tools to do the former. If they cannot, then the case for turning at least some insurance into a public utility starts to look attractive.
