Tuesday, November 11, 2008

When Government Becomes An Insurer: The AIG Revised Bailout

Every major U.S. Newspaper covered the revised AIG deal (WSJ NYTimes CNN MarketWatch,  Bloomberg) yesterday, but none of the articles explain the need for the revision.  As the NY Attorney General continues to probe some very lose spending habits post bailout,  I find it very difficult to trust this company, so any plan that shifts even more cost and risk to the government disturbs me a bit. Ideally we would examine the new needs AIG since the original bailout plan and how this revision satisfies those needs, but since need has not been made public we are left with comparing the old plan to the new. The analysis will show this is a much sweeter deal for AIG. This deal clearly puts more cost onto the government, injects significantly more capital into AIG, and gives all the momentous risk of real estate backed securities onto the taxpayer. I am left wondering in what way this is a better policy decision. 

When the first AIG bailout was announced I thought it was the best concept that had been put forward at the time. It was an $85 billion loan, the term was 2 years at 8.5% plus three-month LIBOR interest rate. In today's market the rate would be approximately 11%. The rate was way above prime which gave the incentive for the government to be a lender of last resort, the position the government should take in a free market economy. Together all of the above articles give enough detail on the revised plan to make some interesting comparisons between the old and new AIG bailout. (I thought the WSJ article was the best, even though the real juicy details were buried in the continued portion of the article on page A15 of the paper.) The new program reduces the loan amount to $60 billion, extends the term to 5 years and reduces the interest rate to 3% plus LIBOR, a total of about 5.5% interest rate in today's market. Hardly a deterrent for a lender of last resort, the government is giving a prime rate for an "emergency loan."

Here is a simplified (rounded to the nearest billion) chart comparison of the original and revised loan scenarios:

 

Total Loan Amount

Total Interest Payments (nets out principle repayment)

Scheduled Monthly Payment

(assumes 12 pmts/yr)

Original Bailout

85,000,000,000

10,000,000,000

4,000,000,000

Revised Loan Structure

60,000,000,000

8,000,000,000

1,000,000,000

 

In total there is a nominal loss of $2 billion in loan interest payments over the life of the loan.1 But what I think is most important here is the scheduled payment. The original plan had an extremely high month-to-month cost, which was a strong deterrent to use the program in the future. This is an important incentive structure the revised plan is missing which may make the government the first source of money instead of the last. 

Additionally the government will purchase $40 billion in preferred shares which bumps up the total upfront bailout cost by $15 billion (total of $100 billion loan+shares now compared to $85 billion loan originally). These shares will carry a 10% annual interest payment. How this 10% is calculated is  not entirely clear to me so I cannot do comparison against the original bailout. A preferred stock option usually pays a dividend, a certain dollar amount for each outstanding share, the WSJ states these shares "carry 10% annual interest payments", it is not clear if the payments are calculated off the principle $40 billion, or the value of the shares at time of payment (which would be calculated as: number of shares*share price at time of calculation). These details could have a significant impact as AIG shares still have the potential of dropping further, which would expose the government to further loss.  

The most interesting of the new proposals are two new entities that will be created to remove the toxic assets off of AIG's balance sheets. These entities will be almost entirely funded by the government. 

The first entity will be capitalized with $30 billion from the government and $5 billion from AIG. The entity will purchase securities that AIG agreed to insure with credit default swaps2.  The entity will pay about 50 cents on the dollar for these securities. Unless I am viewing this wrong, this vehicle will pay AIG Parent 50 cents for each dollar of toxic debt it takes off the balance sheets. Essentially allowing another $12.5 billion (half of the government infusion minus AIG portion plus the return of half of the AIG portion) to be added to the parent company's bailout in exchange for its riskiest holdings. The separate entity that now holds all of this toxic debt, is essentially a government funded (and therefore controlled) company, with very little chance of surviving, let alone returning a profit to the taxpayers. This seems like an extremely risky bet with tax dollars. 

I saved the best for last. The second entity is designed to solve the liquidity problem in AIG's securities-lending business. This business lent out securities to short sellers in exchange for collateral, AIG then took this collateral and invested it for gains.  Unfortunately for AIG these gains turned into losses and now they owe a substantial amount of money to return the collateral.This business epitomizes risky behavior and the bailout plan is a wonderful example of a terrible incentive structure. To establish  this entity the government is expected to foot a $20 billion bill, while AIG contributes $1 billion. The government will inherit all of these bad debts and AIG will walk away risk free. I cannot imagine a plan that rewards a company better for risky behavior. 

These two entities will essentially remove all of AIG's high risk debt and hand it over to the taxpayer. Taxpayers may see a positive return if these assets, which are tied to the housing market, begin to recover. But these entities also open taxpayers up to what I think is the more realistic scenario of losing a substantial amount more as these assets continue to decline in value.  In theory the housing market will eventually come back, and hopefully every penny invested in AIG will be returned, however this new program seems extremely lucrative to the surviving AIG entity and gives taxpayers an extremely risky portfolio. Without knowing the whole circumstance of AIG's current situation it is hard to judge if they are taking advantage of the situation, but when all the numbers are worked out in this new deal you can't help but think they are greatly benefiting from shedding all risk at little cost. From a policy perspective this seems to be the worst incentive structure I have seen thus far. The program has changed from making the government a lender of last resort, to a prime lender that pays a risky company's debts at little or no cost... Is this still America

 

 

1.   These are nominal calculations because I did not think it was appropriate to compare by present value since the government is not weighing the opportunity cost of two investments but rather the nominal value of two costs it will put onto AIG to use the government bailout.

2.   A credit default swap is insurance against a bond default.  

No comments: