Collapse of AIG

The Selected American Funds (SLADX) took a big hit when their investment in AIG blew up. Despite this mistake, in my opinion, Chris Davis and Ken Feinberg who run the Selected American Fund are very capable and honest managers. They eat their own cooking. The Davis Family, employees, and the directors have over 1.5 billion dollars invested side by side with the shareholders in the funds that their firm manages. They have very low fees for a actively managed funds. They are honest in their communication, and share very candidly their investment successes and more importantly their investment errors. In general, they set a great example for the rest of the mutual fund industry on how to live by their fudiciary responsibilities. So when their investment in AIG blew up, they shared what happened at AIG and why they continued to hang onto AIG as the situation got worse (which in hindsight was a bad idea). Here is an extract from 2009 Q1 manager's commentary:

"By far our largest mistake over the last five years was our investment in American International Group (AIG), which cumulatively detracted roughly 6% from our returns, almost three times as much as any other mistake. In essence, this mistake resulted from our incorrectly assessing three factors: the financial sophistication of management, the leverage of derivatives and the danger of collateral requirements tied to mark-to-market accounting.

Starting with management, the chief executive officer of a complex financial institution also serves as the de facto chief risk officer. He or she must understand the nature and extent of the risks being taken and must have the courage to forgo profits if the risks, however remote, could prove catastrophic. At AIG, the need for a highly capable CEO was accentuated by the diverse business models and risk profiles of its semiautonomous divisions. Unfortunately, the abrupt ouster of its longtime CEO and the promotion of a far less able successor came at the worst possible time. Although this successor had deep experience in the field of property and casualty insurance, the company’s financial operations and risks went far beyond this sector into areas in which new management had virtually no expertise.

The second factor that we–and this new management team–grossly underestimated lay in a relatively small division of AIG called AIG Financial Products. This unit contributed about 5% of total profits, only a small portion of which came from selling a type of highly complex and highly leveraged derivative known as a credit default swap, or CDS. At its heart, a CDS is a type of financial insurance in which a buyer would pay AIG a tiny premium in exchange for coverage against a highly unlikely financial event, such as the default of a triple-A security. In some cases, the premium could be as low as $1 for $1,000 worth of coverage. Unfortunately, many of the triple-A securities that AIG insured were tied to mortgages and presumed that residential real estate prices would never go down substantially. To make matters worse, unlike normal insurance contracts, credit default swaps are marked tomarket, meaning that they are priced not for what the actual losses are today but for what the market estimates the losses will be in the future. This mark-to-market accounting brings us to the third aspect of our mistake.

As the estimated losses on the contracts went up, AIG was required to post cash collateral for the buyers of the swaps. Worse still, the amount of cash collateral was also tied to AIG’s own credit rating. As is now apparent, all the factors were in place for a spiral. Because these contracts assumed virtually no losses, small increases in loss estimates led to huge losses. As these losses mounted, rating agencies became concerned and downgraded the company. Thus, AIG was required to post more collateral for both higher estimated losses and its lower credit rating. Adding to these liquidity demands, the company engaged in a massive securities lending program in which it lent out many of its investments. These loans were themselves collateralized with cash. Unfortunately, rather than simply hold this cash in short-term liquid instruments, the company invested much of it in illiquid, often mortgage-related securities. As nervous customers returned borrowed securities and immediately demanded their cash back, AIG was forced to try to sell these illiquid securities at a time when there were no buyers. In just a matter of months, these liquidity demands exceeded AIG’s available resources. Facing default, AIG asked for government intervention. In exchange for providing the necessary liquidity and guaranteeing or assuming responsibility for a number of assets and liabilities, the government took equity warrants for approximately 80% of the company.

All of the above analysis begs the question: “Why did we continue to hold the shares even as the situation got worse and worse?” The answer is twofold. First, we remained focused on the fact that the company had close to $100 billion of tangible equity, more than $1 trillion in investments and more than $20 billion of pretax earnings from global insurance, leasing and asset management operations. Furthermore, in contrast to a bank, it is difficult to have a run on an insurance company, as policyholders, unlike depositors, generally cannot suddenly take back their premiums. Thus, we thought the company’s powerful diversified earnings, tangible equity and assets would more than cover the losses over time. This thinking ignored three important facts. First, because of the collateral requirements discussed above, the company did not have the luxury of time but needed to come up with the cash immediately. Second, because there had been inadequate disclosure about the massive securities lending operation, the scale and immediacy of the cash requirements were far greater than anyone imagined. Finally, although the company’s financial statements showed a huge amount of liquid assets, most were held in the company’s insurance subsidiaries. As regulated entities, these subsidiaries were not permitted to release the assets to the parent company in order for AIG to meet these collateral calls. As a result, even if these mark-to-market losses proved temporary rather than permanent (as they still could), even if the company’s net worth remained substantial, or even if the company could have earned enough in the next five years to pay the losses as they came due, it still faced bankruptcy as a result of defaulting on the collateral calls.

We will end this discussion with an example of the dangers of this sort of collateral posting requirement. Imagine a homeowner has a $400,000 home with a $300,000 mortgage. Now imagine he earns $100,000 per year of which he uses $50,000 for living expenses, leaving $50,000 to service the mortgage. If we add to this example as a given that this person will never move and will never lose his job, it would seem that the mortgage is virtually risk free. But if we change one feature that on its face seems minor, we completely alter the risk profile. Specifically, imagine that the mortgage requires that if the estimated price at which the house could be sold on any given day falls below $300,000, the homeowner must put the shortfall in an escrow account or face eviction. Now, even if the homeowner could service this mortgage forever and even pay it down completely over time with no risk to the bank, he could still be bankrupted by having to mark his house to market and post collateral. It was this type of liquidity risk that, in a matter of months, brought down what had been the most profitable and highly valued insurance company in the world."