CORRESP
March 18, 2009
Mr. Jim B. Rosenberg
Senior Assistant Chief Accountant
United States
Securities and Exchange Commission
Division of Corporation Finance
100 F Street, NE
Mail Stop 6010
Washington, D. C. 20549
Re:   American International Group, Inc.
Form 10-K for Fiscal Year Ended December 31, 2007
Form 10-Q for Fiscal Quarter Ended September 30, 2008
File No. 1-8787
Dear Mr. Rosenberg:
We are in receipt of your letter dated February 25, 2009 and thank you for your comments concerning American International Group, Inc.’s (AIG) captioned filings. This letter sets forth AIG’s responses to each of the Staff’s comments contained in your letter.
AIG acknowledges that the adequacy and accuracy of the disclosure in AIG’s filings is the responsibility of AIG, that Staff comments or changes to disclosure in response to Staff comments do not foreclose the Commission from taking any action with respect to the filings and that Staff comments may not be asserted by AIG as a defense in any proceeding initiated by the Commission or any person under the Federal securities laws of the United States.
We have repeated your questions below to facilitate your review.
General
1.   Please refer to prior comment one. Your proposed disclosure primarily addresses your accounting for the Maiden Lane III agreement under SFAS 159, and does not appear to address the accounting for the Maiden Lane II agreement. Please revise your disclosure to explain how each payment made or received by AIGFP under the Maiden Lane II and III agreements was recorded, including whether any gain or loss was recognized. Please link this disclosure to the related amounts that were previously recorded in your financial statements, including the related derivative liabilities and any amounts of collateral posted and received.
 
    AIG Response:
 
    The accounting for and the payments made and received under the Maiden Lane II agreement are disclosed on pages 250 and 251 of AIG’s Form 10-K for Fiscal Year Ended December 31, 2008 (2008 Form 10-K). This disclosure also links the related amounts to previously recorded amounts in AIG’s financial statements. The Maiden Lane II transaction related to the sale of securities previously included in AIG’s U.S. securities lending program. AIGFP was not involved in the Maiden Lane II transaction; therefore payments made or received under the Maiden Lane II agreement had no effect on AIGFP.

1


 

    The payments made and received, including collateral amounts posted and received, under the Maiden Lane III agreement, are described on pages 136, 137 and 251. This disclosure links the related amounts to previously recorded amounts in AIG’s financial statements.
 
    AIG was exposed to changes in fair values of the underlying assets in the securities lending program as of October 31, 2008. AIG was also exposed to the derivative liabilities in the credit default swaps portfolio through October 31, 2008 after taking into account the provisions of the shortfall agreement between AIG and ML III. Accordingly, as disclosed in the first bullet point on page 64 of AIG’s 2008 Form 10-K, AIG recognized $1.6 billion of realized capital losses, after tax, during the fourth quarter of 2008 on the assets in the U.S. securities lending program. In addition, AIGFP recognized $2.6 billion of unrealized market valuation losses, after tax, during the fourth quarter of 2008 on the super senior credit default swaps related to the CDS terminated in connection with ML III. This amount is included in the $4.5 billion of unrealized market valuation losses, after tax, disclosed in the third bullet point on page 64.
10-Q for the Fiscal Quarter Ended September 30, 2008
Management’s Discussion and Analysis of Financial Condition and Results of operations
Critical Accounting Estimates
AIGFP’s Super Senior Credit Default Swap Portfolio, page 113
2.   We acknowledge your revised disclosure in response to prior comment seven. We continue to believe that you should provide additional quantified data conveying the nature and quality of the underlying securities and a comparison of the subordination at transaction inception to the subordination at the balance sheet date for your regulatory capital relief trades, similar to that which you have agreed to provide for your multi-sector CDOs. While you disclose that you do not expect to make payments pursuant to the contractual terms of the regulatory capital relief trades, the disclosure on page eight of your response does not enable an investor to quantify the potential exposure relating to each underlying transaction. Please revise your disclosure accordingly.
 
    AIG Response:
 
    In response to the Staff’s comment, AIG will provide additional disclosure to present a comparison of the subordination at transaction inception to the subordination at the balance sheet date for the regulatory capital relief trades in its Form 10-Q for the quarterly period ended March 31, 2009 (First Quarter 2009 Form 10-Q) as follows:

2


 

The following table presents, for each of the regulatory capital CDS transactions, the net notional amount at March 31, 2009, attachment points at inception and at March 31, 2009, and inception to date realized losses through March 31, 2009:
(Dollars in millions)
                                         
            Net Notional                     Realized Losses  
            Amount at     Attachment Point     Attachment Point     through March 31,  
CDS   Type     March 31, 2009     at Inception(a)     at March 31,2009(a)     2009(b)  
1
                                       
2
                                       
3
                                       
...
                                       
 
                         
Total
                                       
                         
 
(a)   Expressed as a percentage of gross transaction notional amount of the referenced obligations.
 
(b)   Represents realized losses incurred by the transaction (defaulted amounts less amounts recovered) from inception through March 31, 2009 expressed as a percentage of the initial gross transaction notional amount.
    AIG evaluated the data that is available to AIGFP in respect of these regulatory capital CDS transactions and assessed which of the available data would be helpful to users of financial information in understanding the credit quality of the assets (predominantly loans) underlying these CDS transactions. Based on available data, AIG does not believe that it can provide additional useful information as to the credit quality of the underlying assets. First, the nature of the information provided to AIGFP by the trustees for each transaction is not consistent across all transactions. For example, the vast majority of the assets underlying these transactions are not rated by nationally recognized statistical rating organizations. Thus, any rating information provided to AIG for these transactions represents internal ratings of the counterparties and AIG is unable to assess the comparability of these ratings either to those provided by nationally recognized statistical rating agencies or between counterparties. Second, many of the regulatory capital CDS transactions relate to assets actually held by the counterparties and, in a number of cases, AIG is precluded by confidentiality agreements from disclosing information with respect to these assets, including the internal ratings.
 
    As a result of the limited data available to AIGFP, AIG does not believe that it can provide accurate, comparable quantitative information with respect to the credit quality of the underlying assets. However, AIG will provide information with respect to realized losses, as an indication of the overall quality of the underlying assets.
 
3.   Please refer to the proposed disclosure on page 23 of your response. Please revise to provide similar disaggregated information with respect to CLOs that would enable an investor to evaluate the nature and quality of the assets underlying the insured obligations. Also, for credit default swaps related to corporate debt and CLOs, please provide quantified information depicting subordination at both transaction inception and at the balance sheet date. Given that the derivative liabilities relating to the corporate debt/CLO portfolios have increased significantly during the year ended December 31, 2008, please consider whether the corporate debt and CLO portfolios should be separately disclosed in the table on page 16 of your response and elsewhere, as appropriate.
 
    AIG Response:
 
    As indicated on page 139 of AIG’s 2008 Form 10-K, the net notional amount of the credit default swaps written on super senior tranches of CLOs at December 31, 2008 was $1.5 billion. Given the relatively small size of AIG’s exposure to these transactions, AIG believes that presenting separate tables would not provide meaningful information to users of financial information. However, AIG will include the CLO transactions with the corporate debt

3


 

    transactions in the First Quarter 2009 Form 10-Q table that corresponds to the table on page 139 of 2008 Form 10-K. Further, AIG will provide additional disclosure to present a comparison of the subordination at transaction inception to the subordination at the balance sheet date for the CDS transactions related to corporate debt and CLOs in its First Quarter 2009 Form 10-Q as follows:
The following table presents, for each of the corporate debt and CLO CDS transactions, the net notional amount at March 31, 2009, attachment points at inception and at March 31, 2009, and inception to date defaults through March 31, 2009:
(Dollars in millions)
                                         
CDS   Type     Net Notional
Amount at
March 31, 2009
    Attachment Point
at Inception(a)
    Attachment Point at
March 31, 2009(a)
    Defaults through
March 31, 2009(b)
 
1
                                       
2
                                       
3
                                       
...
                                       
 
                                       
                         
Total
                                       
                         
 
(a)   Expressed as a percentage of gross transaction notional amount of the referenced obligations.
 
(b)   Represents defaults (assets that are technically defaulted but the losses have not yet been realized) from inception through March 31, 2009 expressed as a percentage of the gross transaction notional amount at March 31, 2009.
4.   Please refer to the proposed disclosure on page 16 of your response. Please revise your proposed disclosure to clarify how the ML III transaction affects the information presented in the table. For example, the reference in footnote (f) to realized gains appears to be inconsistent with the tabular column heading depicting unrealized gains. Consider whether the format of this disclosure should be revised such that realized losses are shown in a separate column.
 
    AIG Response:
 
    AIG revised the proposed disclosure to clarify how the ML III transaction affects the information presented in the table on page 131 of its 2008 Form 10-K.
 
5.   Please refer to prior comment nine. We acknowledge your statement that you are unable to predict reasonably likely changes in the prices of the securities underlying a CDO. However, we continue to believe that your disclosure should be revised to explain whether the range of bond prices used in your sensitivity analysis is consistent with recent observed historical activity. If this is not the case, please disclose why you selected a range of prices that differs from recent historical activity, and why you believe the resulting analysis is meaningful to an investor. For each material input to the BET model, disclose the factors you used to determine the range used in your analysis.
 
    AIG Response:
 
    AIG revised the proposed disclosure to include the CDO price declines in 2008 on page 150 of its 2008 Form 10-K. While AIG believes that the ranges used in the sensitivity analyses are reasonable, given the current market conditions, AIG is unable to predict which of the scenarios is most likely to occur. As recent experience demonstrates, actual results in any

4


 

    period are likely to vary, perhaps materially, from the modeled scenarios, and there can be no assurance that the unrealized market valuation losses related to the AIGFP super senior credit default swap portfolio will be consistent with any of the sensitivity analyses. On average for any quarterly period during 2008, prices for CDOs declined between 6.14 percent and 11.93 percent of the notional amount outstanding. Further, it is difficult to extrapolate future experience based on current dislocated market conditions.
 
    The 5 point ranges used in the sensitivity analysis for the estimated price of the securities within the CDO collateral pools represent 15 percent variances from the actual input used at December 31, 2008. In future filings AIG will also quantify the percentage change of the current inputs that 5 points represents. The one-year changes in the weighted average life represent 20 percent variances from the actual input used. The weighted average life of the underlying securities has historically been between 4 and 6 years. In future filings AIG will note that the change in the weighted average life is consistent with AIGFP’s historical experience.
 
6.   Please refer to Appendix 2 of your response. Please disclose the actual input used at the balance sheet date in order to provide appropriate context for the sensitivities you are quantifying. Please also provide quantified sensitivity for each of the three primary factors effecting estimated future cash flows.
 
    AIG Response:
 
    Fair value estimates for AIG’s interests in Maiden Lane II and III are based on estimated future cash flows from the collateral and relevant discount rates derived from the entire asset pool. Under the terms of the arrangement through which AIG collects this data, non-disclosure agreements limit the ability to provide the information requested. Because the Maiden Lane Interests constituted less than one percent of total assets at December 31, 2008, AIG does not conduct analysis on performance data of the underlying mortgage loan portfolios, and the sensitivities. Therefore, AIG does not have or have ready access to quantitative data about the impact of changes in other variables such as underlying prepayments, housing price appreciation, or default assumptions. The sensitivities to changes in cash flows and the applicable discount rates have been provided, and we believe these effectively demonstrate the risk to adverse changes in the fair values in our Maiden Lane Interests.
     
Very truly yours,
   
 
   
/s/ Kathleen E. Shannon
   
 
Kathleen E. Shannon
   
Senior Vice President, Secretary & Deputy General Counsel
   
cc:     Frank Wyman, Staff Accountant
Carl Tartar, Accounting Branch Chief
(Securities and Exchange Commission)
 
    David Herzog

5