AMERICAN GENERAL FINANCE CP AGC1
January 22, 2009 - 2:59pm EST by
madmax989
2009 2010
Price: 50.00 EPS $1.00 $1.00
Shares Out. (in M): 1 P/E 1.0x 1.0x
Market Cap (in M): 1 P/FCF 1.0x 1.0x
Net Debt (in M): 23,500 EBIT 1 1
TEV: 1 TEV/EBIT 1.0x 1.0x

Sign up for free guest access to view investment idea with a 45 days delay.

 

Description

 

Does this sound interesting?

 

-          Strongly cash flowing entity

-          Senior unsecured debt trading at 50-60 cents

-          IRRs held to maturity of 35-45%

-          Assets are whole loan mortgages with 4% delinquencies

-          To be impaired dollar one at 50 cents would need 100% defaults and 57% severity on mortgage (66% severity on average home price)

-          Likely government support

-          Assets for sale

-          Liquid bonds - over $20 billion

-          No secured debt ahead of you

 

If so, read on...

 

American General Financial Corp. ("AGFC" or the "Company"), an indirect wholly-owned subsidiary of AIG, provides loans, retail financing and other credit related products to more than two million families through over 1,400 branches in 40 states, Puerto Rico and the U.S. Virgin Islands.  The main products are first lien mortgages, consumer lending and retail sales financing.  Mortgage loans are originated in the branches (primarily sub-prime) or purchased from others (primarily prime and near-prime).  Non-real estate loans are secured by consumer goods, automobiles, or other personal property (only 3% are unsecured) and generally have maximum original terms of 60 months.  The retail financing products are primarily revolving and installment financing provided to retailers.  AGFC was founded in 1920 and acquired by AIG in 2001. 

The Company has no secured debt and its senior unsecured bonds are currently trading in the 50-60 cent range (for the 2011-2012 bonds).  At these prices an investor is paying only 43-51% into the value of the mortgages, and only 34-41% into the value of the underlying homes, based on a simplified calculation.  In order to be impaired at these levels, for instance, we would need 100% of mortgagees to default and enter foreclosure with 59-66% severity (losses on the homes).  To have 66% severity would mean that a $130,000 home (roughly the average for an AGFC home) could only be sold for $44,200.  We just think this is an absurdly pessimistic outcome.  We believe these bonds are an attractive investment as they have very little risk of impairment and are worth par.  Our stress tests with worst case assumptions for asset performance still put fair value in the high 70s to low 80s. 

We have confidence in these assumptions because AGFC has proven to be a well disciplined underwriter even through the crazy years of the mortgage boom.  The current portfolio of mortgages is 91% fixed rate, 92% first lien, 94% owner-occupied and 97% fully documented.  The average LTV is 80%, with weaker credits at lower LTVs and stronger credits at higher ones.  The Company does not delegate underwriting to unrelated parties and has no option ARMs in the portfolio.  Our diligence has also led us to believe that most of the borrowers have been in their homes for some time prior to taking out their AGFC mortgages.  Basically, AGFC underwrote on the basis that borrowers with plain vanilla fixed rate mortgages and meaningful down payments or existing home equity are significantly less likely to encounter payment problems, delinquencies and defaults.  We agree - people with affordable monthly payments and skin in the game will fight to stay in their homes - and we are therefore willing to assume that losses on these mortgages are unlikely to reach the absurd level required to impair value on the bonds at current market prices.  These bonds traded in the 90s in September, well into the housing market crash.

Although most of our analysis focuses on asset valuation and balance sheet strength, it is important to realize that it is unlikely that asset valuation will even come into play at the end of the day.  Our analysis of AGFC's available sources of liquidity leads us to conclude that the Company should be able to meet its debt maturities for the foreseeable future.  Management's commentary in the most recent 10Q lends some support to our conclusions as they speak at several points about their ability and potential need to significantly curtail originations.  The incentives of the relevant parties - AIG, the federal government and AGFC management - reinforces our views as all are aligned towards the goal of maintaining going concern equity value in the business.  Furthermore, the small demand note AGFC recently pulled down from a subsidiary of AIG indicates that AIG (and the government by association) would likely intend to support AGFC.  Because the Company has meaningful liquidity options, we believe that the most likely outcome of our bond investment is timely payment of scheduled interest and repayment of full principal upon maturity.

 

Capital Structure

The chart below shows AGFC's capital structure as of September 30, 2008 ($000):

 

Capital Structure

 

% of Structure

Senior Unsecured Long Term Debt

20,953,057

 

Senior Unsecured Short Term Debt (1)

2,434,490

 

Senior Unsecured Total Debt

23,387,547

88.1%

Junior Unsecured Long Term Debt

349,209

1.3%

Total Debt

23,736,756

89.4%

 

 

 

Equity

2,823,340

10.6%

Total Capitalization

26,560,096

 

 

 

 

(1) Assumes $225 million of cash retained for working capital and all other cash

 used to reduce short-term debt.

 

 

 

Included in long-term debt above is $2.125 billion under AGFC's multi-year credit facility.   Included in short-term debt above is $1.7 billion of commercial paper, $2.05 billion under the Company's 364-day credit facility and $420 million under an AIG Funding demand note.  Short-term debt is shown net of $2.0 billion of excess cash as discussed in the footnote to the chart. 

This capital structure finances two main assets: finance receivables and a specialty insurance company that is related to the finance receivables business.  The asset side of the balance sheet at September 30, 2008 was as follows:

 

September 30, 2008

 

(dollars in thousands)

As Reported

Assets

 

Net finance receivables:

 

Real estate loans

19,108,608

Non-real estate loans

4,078,208

Retail sales finance

2,240,545

Gross finance receivables

25,427,361

Allowance for finance receivable losses

-959,192

Net finance receivables, less allowance for finance receivable losses

24,468,169

 

 

Real estate loans held for sale

25,771

Investment securities

1,218,085

Cash and cash equivalents

2,195,650

Notes receivable from parent

325,782

Other insurance investments

306,546

Other intangibles

87,894

Other

778,539

Total assets

29,406,436

 

Net finance receivables make up 89% of the total assets, with the insurance segment and other assets accounting for the remainder. 

 

Lending Portfolio

Before we discuss the Company's history, we would put forth the caveat that history is probably not a great guide in these unprecedented times.  However, understanding how AGFC came to be where it is and what types of losses it has experienced can put the problem in some context and is worth exploring.  Furthermore, it provides insight into the quality of management's underwriting and discipline. 

To put the question of asset quality into perspective, the worst charge-off experience AGFC has had in the last 16 years was 5.51%, which occurred in 1996.  The bulk of the charge-offs the Company experienced in 1996 came from its credit cards business, its non-real estate loans and its retail sales finance business.  The real estate business also peaked in terms of charge-offs, but at 1.21% it was still quite low relative to the other categories (in the most recent quarter, it surpassed the 1996 peak with a 1.23% charge-off rate).  Because 1996 was such a stand-out year for losses, it is worth exploring it in more detail to see how the current situation stacks up.

In 1994, AGFC attempted to expand its business by buying credit card receivables and riskier consumer loans.  This strategy of emphasizing higher-yielding finance receivables, which are characterized by higher credit risk, proved to be an unwise move and the Company got killed in this business.  Management used the experience to completely revamp the Company's underwriting standards and realign the compensation structure.  AGFC adopted an action program for improving credit quality that included raising underwriting standards, expanding the use of credit scoring, slowing branch expansion, stressing collections, improving branch office training and rebalancing the finance receivable portfolio credit risk.  Strategies for rebalancing the portfolio credit risk included slowing growth, de-emphasizing some higher risk portfolios and increasing the proportion of real estate secured receivables.  During 1996, the Company eliminated underperforming non-branch marketing programs, established higher underwriting standards, revised the field office incentive compensation system and increased use of credit scoring.  Also in 1996, management decided to exit the credit card business they had acquired in 1994 and instead focused on the best performing asset categories in real estate lending, mostly on a secured, first position, fixed rate basis. 

The following chart shows the historical charge-off performance:

 

[See Chart 1: http://tempcharts.blogspot.com/]

 

As we can see and as one might expect, much of the performance of the non-real estate and retail sales businesses can be explained by cyclical macroeconomic dynamics (we have included the U.S. unemployment rate to illustrate this relationship).  We can therefore assume that as the economy becomes mired more deeply in the current problems, this cyclical element will cause a marked increase in charge-offs (as we are already beginning to see). 

Since 1996, however, AGFC has grown its real estate business substantially while essentially maintaining but not growing its other lines of business.  The following chart makes this apparent ($ in millions):

 

[See Chart 2: http://tempcharts.blogspot.com/]

 

In 1996, real estate loans were 49% of total Company gross finance receivables.  By September 2008 this figure had grown to 75%.  This mix shift really makes a difference when thinking about historical charge-offs.  If we apply the charge-off rates that AGFC experienced in 1996 (which were the worst charge-off rates the Company had experienced since 1991) to the current book of receivables, we can calculate an "historical worst case" charge-off rate for today's book of business.  As the chart below shows, this "historical worst case" charge-off rate is 2.95 ($000).  This figure is actually below the 3.77% allowance currently on the books and is significantly lower than the 5.51% we mentioned earlier that the Company experienced in 1996.

 

"Worst Case" Charge-Off Rate

 

 

 

 

Worst

9/08

 

 

 

Charge-Off Rate

Gross Finance

Implied

 

 

1991 - Q3 2009

Receivables

9/08 Charge-Offs

Charge-Off Rate

Real Estate Loans

1.23%

19,108,608

235,036

 

Non-Real Estate Loans

8.96%

4,078,208

365,407

 

Retail Sales Finance

6.65%

2,240,545

148,996

 

Total

 

25,427,361

749,440

2.95%

 

Now clearly the current environment is likely to be quite a bit worse than this history shows, but still this means that in order for our bonds purchased at 50-60 cents to be impaired, total losses on the total book of business would have to be about 17-20 times as bad as any that AGFC has experienced in the last 16+ years (ignoring the benefit from any interim cash flow generation). 

 

Real Estate Lending

Given the size of the real estate loan portfolio relative to the rest, clearly the focus in terms of downside risk should be on the mortgage book.  As of September 2008, 48% of AGFC's mortgage loans were prime (FICO ≥ 660), 18% were near-prime (620 > FICO > 660), and 34% were subprime (FICO ≤ 620).  Current delinquency rates on AGFC's underlying sub-prime real estate loans are performing more than five times better than the current sub-prime index.  It is worth taking a moment to think about what this means.  The current mortgage environment is extreme.  Sub-prime mortgages industry-wide were showing 60+ day delinquencies of over 28% as of July 2008 and these have likely worsened considerably since then.  In this same environment, AGFC is running 60+ day delinquencies on sub-prime mortgages of only 6.0%.  When we include the near-prime and prime loans, the total mortgage portfolio is showing 60+ day delinquencies of only 4.2%. 

We believe there are structural reasons why AGFC has had superior performance to most of its peers.  AGFC underwrites primarily two types of mortgage loans: prime and sub-prime.  The Company originates and services most of its "riskier" sub-prime consumer and mortgage loans from its 1,400+-branch network.  These branch-originated mortgage loans account for 38% of AGFC's financing receivables portfolio and have a $53,000 average balance (as of March 31, 2008) with a 10.01% average yield.  The Company claims that each branch manager's compensation is linked to branch profitability and credit quality performance.  This connection of pay to performance caused AGFC's managers to stick to their knitting and maintain underwriting standards throughout the mortgage boom.  We think this intelligent incentive structure is the key to AGFC's loan performance

Prime loans are primarily purchased centrally from third parties after a thorough diligence process to ensure the loans conform to AGFC's underwriting standards.  These centralized loans account for 42% of AGFC's financing receivables portfolio.  The bulk of centralized real estate customers are generally described as either prime or near-prime, but for reasons such as elevated debt-to-income ratios, lack of income stability, or the level of income disclosure and verification required for a conforming mortgage, as well as credit repayment history or similar measurements, they have applied for a non-conforming mortgage.  These loans are 99% first lien and generally have 30 year terms.  The average loan balance in this portion of the portfolio was $194,000 (as of March 31, 2008) with an average yield of 6.08%.  These loans could probably be sold if necessary.

 

Insurance Segment

The Company offers various insurance products that serve to protect AGFC's receivables by insuring the borrowers for unforeseen events in amounts equal to their borrowings from AGFC.  The Company's loss ratio has averaged 48% in the past 17 years, and 44% in the last decade.  Applying the maximum losses (as a percentage of premium earned) that the Company has experienced since 1991, which was 60.8% in 1992, to 2007 earned premiums of $160 million would result in losses of $97 million.  Compare these "worst case" losses with the existing $1.05 billion in insurance segment equity and we can see that it would take over 11 years of such losses before the equity base was eroded.  There is clearly a good deal of excess capital in this subsidiary.  Some of this equity could be up-streamed to the parent (as the Company did in the last quarter).

 

Company for Sale

AIG announced that AGFC is one of its subsidiaries that is up for sale as part of its reorganization.  We think these assets might be of interest to a number of potential buyers, including Berkshire Hathaway, one of the better capitalized banks like Wells Fargo or JPMorganChase, or a foreign buyer with an interest in expanding in the U.S.  Buffett is a particularly intriguing potential buyer as he clearly has the balance sheet, is looking to put out money into large opportunities and has played out a similar thesis in the recent past with Clayton Homes in the manufactured housing industry.  We think an AGFC deal looks like a play right out of Buffett's playbook.  We do not expect nor require a sale of the Company to make these bonds an attractive investment; we merely highlight the possibility as it represents a significant imbedded option value for bondholders.  We do, however, feel that a sale could be the most expedient way for AIG and the government to realize positive equity value.

 

Liquidity

There are several sources of liquidity for AGFC, including cash on the balance sheet (which we already assumed is used to pay short-term debt), distributions from the insurance subsidiary (easily $500 million), cash profits from the business (annualizing the last quarter would provide over $700 million in cash earnings), proceeds from asset sales, and cash flow from principal collections, net of originations.  This last one is by far the biggest opportunity to generate cash as AGFC has historically generated a substantial amount of cash flow every year from principal amortization and early repayments.  Since about 2003, principal collections have run about $7-$8 billion per year, and collections as a percentage of net receivables have been very consistent within each loan category:

 

Principal Cash Collections

 

 

 

 

 

 

 

 

2001

2002

2003

2004

2005

2006

2007

$ in Millions

 

 

 

 

 

 

 

Real estate loans

2,392

2,894

3,708

3,985

4,715

4,157

3,243

Non-real estate loans

1,622

1,582

1,563

1,717

1,850

1,908

1,931

Retail sales finance

1,801

1,702

1,612

1,551

1,593

1,792

2,134

Total

5,815

6,178

6,883

7,254

8,158

7,856

7,309

 

 

 

 

 

 

 

 

% of Avg. Net Receivables

 

 

 

 

 

 

 

Real estate loans

33.53%

36.19%

38.27%

30.57%

26.73%

22.63%

17.51%

Non-real estate loans

55.99%

56.40%

55.21%

58.95%

60.71%

58.31%

52.93%

Retail sales finance

130.47%

127.46%

125.78%

123.09%

117.52%

110.91%

109.20%

 

Clearly, the current environment is highly unusual and we would not expect to see rapid pay down of some loans, but the interesting thing to note is that the average lives of the non-real estate and retail sales finance loans are only 1.7 and 0.9 years, respectively.  We would therefore expect to see reasonably consistent pay down out of these areas with much more muted pay down in real estate loans.  It is possible that the low mortgage rates available to homeowners today could lead to a refi boom, but we do not want to count on it.  Even so, imagining collections in the range of $2-$4 billion per year is pretty conservative given the historical level of collections.

We have also identified at least $5.3 billion of very high quality mortgage loans based on the Company's disclosures.  Borrowers of these loans have FICO scores greater than or equal to 660 and the loans also have the following characteristics as compared to the total real estate portfolio:

 

Very High Quality Loans

 

($ in billions)

Total

Very

 

Portfolio

High Quality

Loans outstanding

20.1

5.3

Loan-to-value

80.00%

73.50%

60+ day delinquencies

3.50%

0.70%

 

 

We think the Company could probably work with borrowers to churn these prime loans into conforming loans and then sell them off to the GSEs with no haircut to value.

            The debt maturity schedule is as follows ($000):

 

Debt Maturity Schedule

 

 

 

 

 

 

 

 

 

 

Bonds

 

Credit Facil.

 

Sub Debt

 

Other (1)

 

Total

Remainder of 2008

738,800

 

 

 

 

 

1,999,100

 

2,737,900

2009

2,397,900

 

2,050,000

 

 

 

356,000

 

4,803,900

2010

3,509,700

 

2,125,000

 

 

 

 

 

5,634,700

2011

2,420,700

 

 

 

 

 

 

 

2,420,700

2012

1,809,400

 

 

 

 

 

 

 

1,809,400

2013

2,397,800

 

 

 

 

 

 

 

2,397,800

2014

1,077,800

 

 

 

 

 

 

 

1,077,800

2015

801,000

 

 

 

 

 

 

 

801,000

2016

375,000

 

 

 

 

 

 

 

375,000

2017

3,300,000

 

 

 

 

 

 

 

3,300,000

Thereafter

0

 

 

 

349,200

 

 

 

349,200

Total, gross

18,828,100

 

4,175,000

 

349,200

 

2,355,100

 

25,707,400

Excess cash

 

 

 

 

 

 

 

 

-1,970,650

Total, net

 

 

 

 

 

 

 

 

23,736,750

 

 

 

 

 

 

 

 

 

 

(1) Includes $1.7 billion of commercial paper, $420 million in an AIG FP demand note and various other debt instruments.

 

 

Note that the $2.05 billion credit facility coming due in 2009 can be extended for an additional year at the Company's option.  With the sources of liquidity we mentioned, we think the Company will be able to meet its obligations.

 

Downside Scenario

            If we are wrong on our liquidity analysis and if the Company doesn't get sold and if the credit markets do not open for refinancing corporate debt and if the government / AIG doesn't step in to fund AGFC, then it is possible the Company will be forced to file for bankruptcy.  We think Geithner will want to avoid a black eye like that on the public stage, but we admit it could happen.  We have run rigorous analysis both on the debt agreements (see next section), and on the numbers, and we do not think we can be impaired under any reasonable scenario.  For instance, if we assume the following:

 

-          30% default and 50% severity on the real estate loans (remember current delinquencies are 4%)

-          40% default and 80% severity on the other loans

-          25% impairment on the equity in the insurance subsidiary

-          $1.5 billion up-streamed to AIG, which is half the current book value (very unlikely given the personal liability around an unlawful dividend)

-          $600 million of operating company liabilities given senior treatment

-          $200 million of bankruptcy costs

-          $3.5 billion of debt matures ahead of us (Q4 2009 filing)

 

We still come to nearly 80 cents recovery on the bonds in a run-off.  Leaving all the assumptions the same but assuming 80% frequency and 60% severity on the real estate loans, we break even at a 50 cent purchase price. 

 

Covenants and Other Legal Restrictions

            Please note, we do not purport to be lawyers, so take this analysis with a grain of salt and do your own homework.  Although our liquidity analysis shows that the Company should be able to meet its bond maturities without playing potentially value-destructive games, it is still important to understand the contractual protection that bondholders have.  To be clear right up front, we do not believe that AGFC's relationship with AIG as a wholly-owned subsidiary puts these bonds at any additional meaningful risk.  Most importantly, these bonds are issued by AGFC, not by AIG.  This means that our claims are on the assets of AGFC and AIG's lenders and other creditors are structurally subordinate to all of our claims.  AGFC does not guarantee any of AIG's debt nor the debt of any other AIG subsidiary (including AGFC's parent American General Finance Inc.).  AIG indirectly owns the common equity in AGFC.  Other than that AIG has no claim on AGFC's assets and has limited ability to upstream cash or other assets to the parent.  AIG (and any other entity, related or not) is also effectively prohibited by an "equitable and ratable" clause in its ability to provide funds to AGFC in exchange for secured positions.  It can therefore only lend to AGFC on an unsecured basis, as it has recently through a demand note, or inject equity.  We have also reviewed the agreement between AIG and the federal government and do not believe the agreement provides AIG with any additional powers that could impair value in our bonds.  Finally, we reviewed the publicly available information regarding all of the AIG entities in the AGFC chain of ownership to ensure there are no other areas of maneuverability that could impair value in the bonds.  We explain all of these issues in more detail below.

 

AIG Loan Agreement with the U.S. Federal Government

The Fed's loan agreement is worth dissecting in detail as it has strong covenants that serve to protect AGFC bondholders.  First, the Fed's loan was signed in conjunction with a Guarantee and Pledge Agreement (the "Guarantee Agreement").  In the Guarantee Agreement, AIG caused all domestic subsidiaries with aggregate assets of greater than $50 million, except Excluded Subsidiaries, to guarantee the obligations of AIG by granting Liens on certain of its assets. The Excluded Subsidiaries include: Regulated Subsidiaries, Securitization Subsidiaries, International Lease Finance Corp ("ILFC") and AGFC and AGFI. [Note: The exact amount of the security attained under this guarantee agreement is currently unknown because the Guarantee Agreement was limited in its scope due to section 3(a)(xiv)(ii).  The gist of this clause is that it limits the amount of any Guarantee to the amount permitted "without contravening" any existing financing agreements or being obligated under such financing agreements to "equally and ratably" secure the existing indebtedness.]  Therefore, AGFC was specifically excluded as a guarantor or pledgor of the government loan.

The covenants of the Fed's loan limit AIG's options.  First and most importantly, the loan's negative covenants apply to AIG as well as any "Restricted Subsidiary".  A Restricted Subsidiary is any "non-insurance" subsidiary with aggregate assets of greater than $50 million, or an equity investment in a subsidiary with aggregate assets of greater than $50 million.  As such, AGFC would be considered a Restricted Subsidiary under this definition and the covenants of the government loan to AIG therefore apply to AGFC.  To be clear, AGFC is excluded as a guarantor under the Guarantee Agreement, but is included with regards to the negative covenants.

The following material negative covenants of the Fed's loan prohibit AIG or any Restricted Subsidiary to (note we have left out some negative covenants that we believe are relatively immaterial in the larger picture of this analysis.  All capitalized terms as defined in the loan agreement):

 

  1. Indebtedness. Incur, create, assume or permit to exist any Indebtedness, except:
    1. existing Indebtedness and any extension, renewals, exchanges or replacements as long as it doesn't materially change the terms
    2. Indebtedness created by the Fed's loan itself
    3. intercompany Indebtedness (conditions applicable)
    4. Indebtedness of AIG or a Restricted Subsidiary incurred to finance the acquisition, construction or improvement of any fixed or capital asset (other than aircraft and aircraft-related equipment purchased by ILFC and its subsidiaries), and extensions, renewals and replacements of any such Indebtedness that do not increase the outstanding principal amount (conditions applicable)
    5. other Indebtedness incurred or assumed in the ordinary course of business to finance the business in the ordinary course
    6. other unsecured Indebtedness; provided that (i) the proceeds thereof are applied to prepay the Fed's Loan and (ii) such Indebtedness shall mature six months after the term of the Fed's Loan
  2. Liens. Create, incur, assume or permit to exist any Lien on any property or assets now owned or hereafter acquired by it or on any income or revenues or rights in respect of any thereof, except:
    1. Existing Liens
    2. Liens created via the Fed's Loan
    3. Liens existing on any property or asset prior to the acquisition so long as the Lien was not created in contemplation of the acquisition and pertains only to the asset acquired
    4. Tax Liens
    5. carriers', warehousemen's, mechanics', materialmen's, repairmen's or other like Liens created in the normal course
    6. other ordinary course liens
  3. Sale and Lease-back Transactions except:
    1. If it is allowed under the sale/ merger covenant
    2. Capital Lease Obligations allowed under (1) or (2) above
  4. Mergers, Consolidations, Sales of Assets and Acquisitions, except:
    1. AIG and Subsidiaries may purchase and sell assets in the ordinary course of business
    2. AIG and Subsidiaries are prohibited from any other Asset Sales unless:

                                                               i.      consideration at least 90% of which is cash,

                                                             ii.      such consideration is at least equal to the fair market value of the assets being sold

                                                            iii.      the Net Cash Proceeds of such Asset Sale are applied to pay down the Fed's loans

  1. Other Indebtedness and Agreements.
    1. any waiver, supplement, modification, amendment, termination or release of any indenture, instrument or agreement pursuant to which any Indebtedness of AIG or any of the Subsidiaries is outstanding if the effect of such waiver, supplement, modification, amendment, termination or release would materially increase the obligations of the obligor or confer additional material rights on the holder of such Indebtedness in a manner adverse to AIG, any of the Subsidiaries or the Fed

 

Essentially, the government loan precludes AGFC, as a restricted subsidiary, from incurring additional indebtedness, creating liens (priming our bonds), engaging in sale-leasebacks, and selling substantial amounts of assets (subject to certain conditions).  These covenants go a long way to protecting our bonds for so long as the government loan is outstanding.

There are further protections in the government loan document.  Under the Events of Default section, if AIG or any Subsidiaries were to fail to pay principal or interest on "Material Indebtedness" when due or payable it will result in an Event of Default under the Fed's Loan.  Furthermore, if any subsidiary were to file bankruptcy, either voluntary or involuntary, this would constitute an Event of Default under the Fed's Loan.  Under any Event of Default, the loan from the Fed would become immediately due and payable.

We should point out that there is language in the federal loan that states that the government could ask AGFC to guarantee the government loan so long as in so doing the government did not trigger the equitable and ratable clause.  We do not see a way for the government to get around the clause.  We therefore believe there is little risk of the government priming our loan by taking a secured position in the assets.

In fact, if anything the government loan appears to be a potential source of liquidity for AGFC.  In September, the Company borrowed $420 million from AIG Funding, Inc., an AIG-guaranteed funding affiliate, in the form of a demand note.  AIG Funding appears to be an entity that has access to the government credit facility funds, and therefore is an intermediary through which AGFC can access government funding.  The Company noted in its September 10Q that if necessary it would look to tap into the government funding facility, so this is consistent with the loan we saw in September.  It is also consistent with the government's incentive to keep AIG's subsidiaries out of bankruptcy.  In fact, one could argue that the demand note was not really necessary (given that AGFC finished the quarter with over $2 billion in cash on the balance sheet), and that the loan was more a symbolic gesture by the government and AIG that AGFC would be supported.

 

Support Agreement / AGFC Credit Facility

Another powerful document protecting AGFC lenders comes from the Support Agreement that was put in place concurrent with the renewal of the Company's 364-day credit facility on July 10, 2008.  Essentially, the banks that are party to the facility required that AIG sign the Support Agreement to protect the banks' investment.  The credit facility, which is unsecured, is pari passu with our bonds so we get the benefit of the arrangement the banks negotiated to provide for their own protection.  As part of the Support Agreement, AIG agreed to have AGFC maintain an Adjusted Tangible Leverage Ratio of no greater than 8-to-1 at the end of each calendar quarter.  The Support Agreement thereby effectively limits the amount of debt at AGFC, particularly if portfolio write-downs or discounted asset sales erode the Company's net worth.  At current debt levels, the Support Agreement requires that the Company maintain approximately $2.9 billion of tangible net worth.  With $3.2 billion of tangible equity as of the most recent quarter, this covenant also serves to put a limitation on how much the Company can dividend up to the parent, as dividends would reduce tangible book value.  As further protection, the Support Agreement states that AGFC must maintain minimum Consolidated Net Worth of at least $2.2 billion.  This last piece is not meaningful for now given the $2.9 billion requirement discussed above, but at some point it could come into play. 

AIG's commitment under the Support Agreement will terminate automatically upon the earlier of (a) the termination of the 364-day credit facility, (b) payment in full of the loans outstanding under the 364-day credit facility, (c) AIG providing an irrevocable and unconditional guarantee of the outstanding amounts under the 364-day credit facility, or (d) consent from the majority of the 364-day credit facility banks.  Because the Support Agreement provides the first layer of defense for bondholders, we need to understand the risk that any of these four events might come to pass. 

The termination of the facility considered in part (a) above is irrelevant because AGFC already drew down the facility in its entirety.  There is therefore no way to terminate it without paying off the debt, which is actually part (b).  As we will show, part (b) above is the only viable option, and as such we will address it after discussing parts (c) and (d).  Regarding (c) above, which contemplates AIG guaranteeing the credit facility, our read of AIG's loan with the federal government indicates that AIG is limited in its ability to guarantee any debt (including the credit facility) of AIG's subsidiaries.  This language can be found in Section 6.09 of the loan document.  Part (d), consent of the majority of the banks, is even more unlikely as there is no incentive for the banks to provide such consent in light of the current stress of the parent.    

This leaves (b) above, paying off the balance of the credit facility, as the only viable option.  If AGFC were to pay down the credit facility, it could conceivably work itself out of the restrictive covenants.  We think this is unlikely prior to maturity, however, because the credit facility provides two-year paper (it is extendable for an additional year at AGFC's option) and it would be irrational for the Company to repay two-year paper when it has nearer-term maturities coming due.  We think it is unlikely that the Company would put itself at risk for nearer-term maturities by removing its safety net in the two-year credit facility just to escape the covenants.  Furthermore, the optionality AGFC would create by relieving itself from the restrictive covenants in the credit facility would not be that great.  We therefore believe we will benefit from the credit facility's protection until its maturity in 2010.

 

AGFC Bond Indenture

The bond indenture is fairly light on covenants, but it does afford bondholders some protection beyond the government loan and the credit facility discussed above.  The indenture (and the credit facility for that matter) has an "equal and ratable" clause.  The clause specifically states that if the Company takes on any secured debt then the notes we would be purchasing would receive a lien that is "equal and ratable" to any new secured borrowings.  What this means is that any prospective secured bondholder would have to be comfortable with a security claim that is diluted by the existing unsecured debt.  This would probably be unpalatable to most any lender given the amount of senior unsecured debt ($25.4 billion unadjusted, or $23.4 billion net of the $2.0 billion of excess cash).  With that being said, there are a few loopholes in the clause that create some risk exposure provided that the protections afforded by the government loan on the limitation of additional indebtedness and liens are, for whatever reason, not in effect. 

One loophole is that the Company has a carve-out that allows it to secure up to 10% of its consolidated net worth.  This would work out to $282 million, which is a relatively small amount.  A bigger loophole risk is the potential ability of the Company to structure a repo transaction as a "sale" that would allow it to finance its mortgage receivables.  It is an important distinction that it be deemed a sale because as we have shown, AGFC is effectively prohibited from creating secured debt by the equitable and ratable clause.  The grey area is caused by language in the indenture (Section 1007(b)(12)) that could be interpreted to give AGFC the ability to pursue repos, provided that the transaction is limited only to the assets being sold (in other words, it must be non-recourse). 

We think that the Company's entering into a repo transaction would be challenged by bondholders, but that might not prevent the Company from attempting such a transaction.  If successful, AGFC could pay off loans in front of ours (or underwrite new loans, which would probably be value neutral to us at worst).  In this case we would be worse off because there would now be new "debt" in front of ours in a secured position (this would not be as big a problem if the repo arrangement matured after our bonds mature).  Given the multitude of liquidity options available to the Company, we think the risk of repos is very low.

Another related risk that we need to consider is the potential for the Company to sell assets outright.  The indenture does not specifically preclude asset sales.  There is therefore a risk that management might sell assets, potentially at a discount to fair value, in order to extend the life of the Company by paying off maturities as they came due.  While this is a risk, we believe it is manageable because even if the Company were to sell off all of its assets at 70 cents on the dollar, the bonds would still return 82 cents.  Even under this fairly absurd scenario, a recovery of 82 cents would be a perfectly acceptable investment result given our low purchase price.  Furthermore, large discounts on large sales of assets would significantly reduce management's tangible net worth and risk violating covenants in the Support Agreement for so long as the credit facility is outstanding.  Again, given the multitude of liquidity options available to the AGFC, we think the risk of heavily discounted asset sales is very low.

Finally, in the event that AGFC can not meet its maturities or otherwise would like to renegotiate the terms of the indenture, a 100% vote is required to pass changes to any of the important money terms.

 

Unlawful Dividends

As asset stripping is the primary risk we are concerned about, the main protection bondholders have is a restriction on the ability of the Company to pay dividends up to the parent.  Although there is no specific limitation on the ability to dividend up to the parent in the bond indenture, there is strong protection against "unlawful dividends," which is a dividend that would leave AGFC insolvent.  This effectively puts a limit on the value of assets that could be distributed out of AGFC.  Solvency is a simple test based upon the assets being worth more than the liabilities - if they are not the company is considered technically insolvent.  There is nothing in the document that would prevent AGFC from up-streaming all of its net worth (which was $2.8 billion as of September 30, 2008).  However, if the Company were deemed to be insolvent as a result of the dividend, we believe the directors of the Company would be held personally liable.  As such, we do not believe that a lawyer would let the Board dividend out more than half of AGFC's current net worth, nor do we think any Board member in his right mind would even consider putting himself personally at risk with such a transfer.  This puts an effective cap on the amount of a dividend to AIG at about $1.5 billion, if at all.  We would not be happy to see that happen but it certainly does not come close to impairing value on the bonds. A $1.5 billion dividend would first eat into equity value and if it did end up impairing the bonds, it would represent a 6.4% hit ($1.5 / $23.4 billion).  Again, such a dividend would not be possible for so long as the credit facility is outstanding.

 

Conclusion

We repeat, does this sound interesting?

 

-          Strongly cash flowing entity

-          Senior unsecured debt trading at 50-60 cents

-          IRRs held to maturity of 35-45%

-          Assets are whole loan mortgages with 4% delinquencies

-          To be impaired dollar one at 50 cents would need 100% defaults and 57% severity on mortgage (66% severity on average home price)

-          Likely government support

-          Assets for sale

-          Liquid bonds - over $20 billion

-          No secured debt ahead of you

 

We think these bonds are a no-brainer bet. 

 

 

Disclaimer:

 

We are neither a registered investment advisor nor a broker/dealer. 

 

Readers are advised that this report is issued solely for information purposes and should not to be construed as an offer to sell or the solicitation of an offer to buy any security. The opinions and analyses included herein are based from sources believed to be reliable and written in good faith, but no representation or warranty, expressed or implied is made as to their accuracy, completeness or correctness.

 

YOU SHOULD VERIFY ALL CLAIMS AND DO YOUR OWN RESEARCH BEFORE INVESTING IN ANY SECURITIES MENTIONED IN THIS REPORT.

 

We accept no liability whatsoever for any direct or consequential loss arising from any use of information in this report.

 

We may on occasion hold positions in the securities mentioned; however, these positions may change at any time.  We currently own AGFC bonds.

 

 

Catalyst

- Sale of the company.

- Liquidity picture crystallizes.

- Government support.

    sort by   Expand   New

    Description

     

    Does this sound interesting?

     

    -          Strongly cash flowing entity

    -          Senior unsecured debt trading at 50-60 cents

    -          IRRs held to maturity of 35-45%

    -          Assets are whole loan mortgages with 4% delinquencies

    -          To be impaired dollar one at 50 cents would need 100% defaults and 57% severity on mortgage (66% severity on average home price)

    -          Likely government support

    -          Assets for sale

    -          Liquid bonds - over $20 billion

    -          No secured debt ahead of you

     

    If so, read on...

     

    American General Financial Corp. ("AGFC" or the "Company"), an indirect wholly-owned subsidiary of AIG, provides loans, retail financing and other credit related products to more than two million families through over 1,400 branches in 40 states, Puerto Rico and the U.S. Virgin Islands.  The main products are first lien mortgages, consumer lending and retail sales financing.  Mortgage loans are originated in the branches (primarily sub-prime) or purchased from others (primarily prime and near-prime).  Non-real estate loans are secured by consumer goods, automobiles, or other personal property (only 3% are unsecured) and generally have maximum original terms of 60 months.  The retail financing products are primarily revolving and installment financing provided to retailers.  AGFC was founded in 1920 and acquired by AIG in 2001. 

    The Company has no secured debt and its senior unsecured bonds are currently trading in the 50-60 cent range (for the 2011-2012 bonds).  At these prices an investor is paying only 43-51% into the value of the mortgages, and only 34-41% into the value of the underlying homes, based on a simplified calculation.  In order to be impaired at these levels, for instance, we would need 100% of mortgagees to default and enter foreclosure with 59-66% severity (losses on the homes).  To have 66% severity would mean that a $130,000 home (roughly the average for an AGFC home) could only be sold for $44,200.  We just think this is an absurdly pessimistic outcome.  We believe these bonds are an attractive investment as they have very little risk of impairment and are worth par.  Our stress tests with worst case assumptions for asset performance still put fair value in the high 70s to low 80s. 

    We have confidence in these assumptions because AGFC has proven to be a well disciplined underwriter even through the crazy years of the mortgage boom.  The current portfolio of mortgages is 91% fixed rate, 92% first lien, 94% owner-occupied and 97% fully documented.  The average LTV is 80%, with weaker credits at lower LTVs and stronger credits at higher ones.  The Company does not delegate underwriting to unrelated parties and has no option ARMs in the portfolio.  Our diligence has also led us to believe that most of the borrowers have been in their homes for some time prior to taking out their AGFC mortgages.  Basically, AGFC underwrote on the basis that borrowers with plain vanilla fixed rate mortgages and meaningful down payments or existing home equity are significantly less likely to encounter payment problems, delinquencies and defaults.  We agree - people with affordable monthly payments and skin in the game will fight to stay in their homes - and we are therefore willing to assume that losses on these mortgages are unlikely to reach the absurd level required to impair value on the bonds at current market prices.  These bonds traded in the 90s in September, well into the housing market crash.

    Although most of our analysis focuses on asset valuation and balance sheet strength, it is important to realize that it is unlikely that asset valuation will even come into play at the end of the day.  Our analysis of AGFC's available sources of liquidity leads us to conclude that the Company should be able to meet its debt maturities for the foreseeable future.  Management's commentary in the most recent 10Q lends some support to our conclusions as they speak at several points about their ability and potential need to significantly curtail originations.  The incentives of the relevant parties - AIG, the federal government and AGFC management - reinforces our views as all are aligned towards the goal of maintaining going concern equity value in the business.  Furthermore, the small demand note AGFC recently pulled down from a subsidiary of AIG indicates that AIG (and the government by association) would likely intend to support AGFC.  Because the Company has meaningful liquidity options, we believe that the most likely outcome of our bond investment is timely payment of scheduled interest and repayment of full principal upon maturity.

     

    Capital Structure

    The chart below shows AGFC's capital structure as of September 30, 2008 ($000):

     

    Capital Structure

     

    % of Structure

    Senior Unsecured Long Term Debt

    20,953,057

     

    Senior Unsecured Short Term Debt (1)

    2,434,490

     

    Senior Unsecured Total Debt

    23,387,547

    88.1%

    Junior Unsecured Long Term Debt

    349,209

    1.3%

    Total Debt

    23,736,756

    89.4%

     

     

     

    Equity

    2,823,340

    10.6%

    Total Capitalization

    26,560,096

     

     

     

     

    (1) Assumes $225 million of cash retained for working capital and all other cash

     used to reduce short-term debt.

     

     

     

    Included in long-term debt above is $2.125 billion under AGFC's multi-year credit facility.   Included in short-term debt above is $1.7 billion of commercial paper, $2.05 billion under the Company's 364-day credit facility and $420 million under an AIG Funding demand note.  Short-term debt is shown net of $2.0 billion of excess cash as discussed in the footnote to the chart. 

    This capital structure finances two main assets: finance receivables and a specialty insurance company that is related to the finance receivables business.  The asset side of the balance sheet at September 30, 2008 was as follows:

     

    September 30, 2008

     

    (dollars in thousands)

    As Reported

    Assets

     

    Net finance receivables:

     

    Real estate loans

    19,108,608

    Non-real estate loans

    4,078,208

    Retail sales finance

    2,240,545

    Gross finance receivables

    25,427,361

    Allowance for finance receivable losses

    -959,192

    Net finance receivables, less allowance for finance receivable losses

    24,468,169

     

     

    Real estate loans held for sale

    25,771

    Investment securities

    1,218,085

    Cash and cash equivalents

    2,195,650

    Notes receivable from parent

    325,782

    Other insurance investments

    306,546

    Other intangibles

    87,894

    Other

    778,539

    Total assets

    29,406,436

     

    Net finance receivables make up 89% of the total assets, with the insurance segment and other assets accounting for the remainder. 

     

    Lending Portfolio

    Before we discuss the Company's history, we would put forth the caveat that history is probably not a great guide in these unprecedented times.  However, understanding how AGFC came to be where it is and what types of losses it has experienced can put the problem in some context and is worth exploring.  Furthermore, it provides insight into the quality of management's underwriting and discipline. 

    To put the question of asset quality into perspective, the worst charge-off experience AGFC has had in the last 16 years was 5.51%, which occurred in 1996.  The bulk of the charge-offs the Company experienced in 1996 came from its credit cards business, its non-real estate loans and its retail sales finance business.  The real estate business also peaked in terms of charge-offs, but at 1.21% it was still quite low relative to the other categories (in the most recent quarter, it surpassed the 1996 peak with a 1.23% charge-off rate).  Because 1996 was such a stand-out year for losses, it is worth exploring it in more detail to see how the current situation stacks up.

    In 1994, AGFC attempted to expand its business by buying credit card receivables and riskier consumer loans.  This strategy of emphasizing higher-yielding finance receivables, which are characterized by higher credit risk, proved to be an unwise move and the Company got killed in this business.  Management used the experience to completely revamp the Company's underwriting standards and realign the compensation structure.  AGFC adopted an action program for improving credit quality that included raising underwriting standards, expanding the use of credit scoring, slowing branch expansion, stressing collections, improving branch office training and rebalancing the finance receivable portfolio credit risk.  Strategies for rebalancing the portfolio credit risk included slowing growth, de-emphasizing some higher risk portfolios and increasing the proportion of real estate secured receivables.  During 1996, the Company eliminated underperforming non-branch marketing programs, established higher underwriting standards, revised the field office incentive compensation system and increased use of credit scoring.  Also in 1996, management decided to exit the credit card business they had acquired in 1994 and instead focused on the best performing asset categories in real estate lending, mostly on a secured, first position, fixed rate basis. 

    The following chart shows the historical charge-off performance:

     

    [See Chart 1: http://tempcharts.blogspot.com/]

     

    As we can see and as one might expect, much of the performance of the non-real estate and retail sales businesses can be explained by cyclical macroeconomic dynamics (we have included the U.S. unemployment rate to illustrate this relationship).  We can therefore assume that as the economy becomes mired more deeply in the current problems, this cyclical element will cause a marked increase in charge-offs (as we are already beginning to see). 

    Since 1996, however, AGFC has grown its real estate business substantially while essentially maintaining but not growing its other lines of business.  The following chart makes this apparent ($ in millions):

     

    [See Chart 2: http://tempcharts.blogspot.com/]

     

    In 1996, real estate loans were 49% of total Company gross finance receivables.  By September 2008 this figure had grown to 75%.  This mix shift really makes a difference when thinking about historical charge-offs.  If we apply the charge-off rates that AGFC experienced in 1996 (which were the worst charge-off rates the Company had experienced since 1991) to the current book of receivables, we can calculate an "historical worst case" charge-off rate for today's book of business.  As the chart below shows, this "historical worst case" charge-off rate is 2.95 ($000).  This figure is actually below the 3.77% allowance currently on the books and is significantly lower than the 5.51% we mentioned earlier that the Company experienced in 1996.

     

    "Worst Case" Charge-Off Rate

     

     

     

     

    Worst

    9/08

     

     

     

    Charge-Off Rate

    Gross Finance

    Implied

     

     

    1991 - Q3 2009

    Receivables

    9/08 Charge-Offs

    Charge-Off Rate

    Real Estate Loans

    1.23%

    19,108,608

    235,036

     

    Non-Real Estate Loans

    8.96%

    4,078,208

    365,407

     

    Retail Sales Finance

    6.65%

    2,240,545

    148,996

     

    Total

     

    25,427,361

    749,440

    2.95%

     

    Now clearly the current environment is likely to be quite a bit worse than this history shows, but still this means that in order for our bonds purchased at 50-60 cents to be impaired, total losses on the total book of business would have to be about 17-20 times as bad as any that AGFC has experienced in the last 16+ years (ignoring the benefit from any interim cash flow generation). 

     

    Real Estate Lending

    Given the size of the real estate loan portfolio relative to the rest, clearly the focus in terms of downside risk should be on the mortgage book.  As of September 2008, 48% of AGFC's mortgage loans were prime (FICO ≥ 660), 18% were near-prime (620 > FICO > 660), and 34% were subprime (FICO ≤ 620).  Current delinquency rates on AGFC's underlying sub-prime real estate loans are performing more than five times better than the current sub-prime index.  It is worth taking a moment to think about what this means.  The current mortgage environment is extreme.  Sub-prime mortgages industry-wide were showing 60+ day delinquencies of over 28% as of July 2008 and these have likely worsened considerably since then.  In this same environment, AGFC is running 60+ day delinquencies on sub-prime mortgages of only 6.0%.  When we include the near-prime and prime loans, the total mortgage portfolio is showing 60+ day delinquencies of only 4.2%. 

    We believe there are structural reasons why AGFC has had superior performance to most of its peers.  AGFC underwrites primarily two types of mortgage loans: prime and sub-prime.  The Company originates and services most of its "riskier" sub-prime consumer and mortgage loans from its 1,400+-branch network.  These branch-originated mortgage loans account for 38% of AGFC's financing receivables portfolio and have a $53,000 average balance (as of March 31, 2008) with a 10.01% average yield.  The Company claims that each branch manager's compensation is linked to branch profitability and credit quality performance.  This connection of pay to performance caused AGFC's managers to stick to their knitting and maintain underwriting standards throughout the mortgage boom.  We think this intelligent incentive structure is the key to AGFC's loan performance

    Prime loans are primarily purchased centrally from third parties after a thorough diligence process to ensure the loans conform to AGFC's underwriting standards.  These centralized loans account for 42% of AGFC's financing receivables portfolio.  The bulk of centralized real estate customers are generally described as either prime or near-prime, but for reasons such as elevated debt-to-income ratios, lack of income stability, or the level of income disclosure and verification required for a conforming mortgage, as well as credit repayment history or similar measurements, they have applied for a non-conforming mortgage.  These loans are 99% first lien and generally have 30 year terms.  The average loan balance in this portion of the portfolio was $194,000 (as of March 31, 2008) with an average yield of 6.08%.  These loans could probably be sold if necessary.

     

    Insurance Segment

    The Company offers various insurance products that serve to protect AGFC's receivables by insuring the borrowers for unforeseen events in amounts equal to their borrowings from AGFC.  The Company's loss ratio has averaged 48% in the past 17 years, and 44% in the last decade.  Applying the maximum losses (as a percentage of premium earned) that the Company has experienced since 1991, which was 60.8% in 1992, to 2007 earned premiums of $160 million would result in losses of $97 million.  Compare these "worst case" losses with the existing $1.05 billion in insurance segment equity and we can see that it would take over 11 years of such losses before the equity base was eroded.  There is clearly a good deal of excess capital in this subsidiary.  Some of this equity could be up-streamed to the parent (as the Company did in the last quarter).

     

    Company for Sale

    AIG announced that AGFC is one of its subsidiaries that is up for sale as part of its reorganization.  We think these assets might be of interest to a number of potential buyers, including Berkshire Hathaway, one of the better capitalized banks like Wells Fargo or JPMorganChase, or a foreign buyer with an interest in expanding in the U.S.  Buffett is a particularly intriguing potential buyer as he clearly has the balance sheet, is looking to put out money into large opportunities and has played out a similar thesis in the recent past with Clayton Homes in the manufactured housing industry.  We think an AGFC deal looks like a play right out of Buffett's playbook.  We do not expect nor require a sale of the Company to make these bonds an attractive investment; we merely highlight the possibility as it represents a significant imbedded option value for bondholders.  We do, however, feel that a sale could be the most expedient way for AIG and the government to realize positive equity value.

     

    Liquidity

    There are several sources of liquidity for AGFC, including cash on the balance sheet (which we already assumed is used to pay short-term debt), distributions from the insurance subsidiary (easily $500 million), cash profits from the business (annualizing the last quarter would provide over $700 million in cash earnings), proceeds from asset sales, and cash flow from principal collections, net of originations.  This last one is by far the biggest opportunity to generate cash as AGFC has historically generated a substantial amount of cash flow every year from principal amortization and early repayments.  Since about 2003, principal collections have run about $7-$8 billion per year, and collections as a percentage of net receivables have been very consistent within each loan category:

     

    Principal Cash Collections

     

     

     

     

     

     

     

     

    2001

    2002

    2003

    2004

    2005

    2006

    2007

    $ in Millions

     

     

     

     

     

     

     

    Real estate loans

    2,392

    2,894

    3,708

    3,985

    4,715

    4,157

    3,243

    Non-real estate loans

    1,622

    1,582

    1,563

    1,717

    1,850

    1,908

    1,931

    Retail sales finance

    1,801

    1,702

    1,612

    1,551

    1,593

    1,792

    2,134

    Total

    5,815

    6,178

    6,883

    7,254

    8,158

    7,856

    7,309

     

     

     

     

     

     

     

     

    % of Avg. Net Receivables

     

     

     

     

     

     

     

    Real estate loans

    33.53%

    36.19%

    38.27%

    30.57%

    26.73%

    22.63%

    17.51%

    Non-real estate loans

    55.99%

    56.40%

    55.21%

    58.95%

    60.71%

    58.31%

    52.93%

    Retail sales finance

    130.47%

    127.46%

    125.78%

    123.09%

    117.52%

    110.91%

    109.20%

     

    Clearly, the current environment is highly unusual and we would not expect to see rapid pay down of some loans, but the interesting thing to note is that the average lives of the non-real estate and retail sales finance loans are only 1.7 and 0.9 years, respectively.  We would therefore expect to see reasonably consistent pay down out of these areas with much more muted pay down in real estate loans.  It is possible that the low mortgage rates available to homeowners today could lead to a refi boom, but we do not want to count on it.  Even so, imagining collections in the range of $2-$4 billion per year is pretty conservative given the historical level of collections.

    We have also identified at least $5.3 billion of very high quality mortgage loans based on the Company's disclosures.  Borrowers of these loans have FICO scores greater than or equal to 660 and the loans also have the following characteristics as compared to the total real estate portfolio:

     

    Very High Quality Loans

     

    ($ in billions)

    Total

    Very

     

    Portfolio

    High Quality

    Loans outstanding

    20.1

    5.3

    Loan-to-value

    80.00%

    73.50%

    60+ day delinquencies

    3.50%

    0.70%

     

     

    We think the Company could probably work with borrowers to churn these prime loans into conforming loans and then sell them off to the GSEs with no haircut to value.

                The debt maturity schedule is as follows ($000):

     

    Debt Maturity Schedule

     

     

     

     

     

     

     

     

     

     

    Bonds

     

    Credit Facil.

     

    Sub Debt

     

    Other (1)

     

    Total

    Remainder of 2008

    738,800

     

     

     

     

     

    1,999,100

     

    2,737,900

    2009

    2,397,900

     

    2,050,000

     

     

     

    356,000

     

    4,803,900

    2010

    3,509,700

     

    2,125,000

     

     

     

     

     

    5,634,700

    2011

    2,420,700

     

     

     

     

     

     

     

    2,420,700

    2012

    1,809,400

     

     

     

     

     

     

     

    1,809,400

    2013

    2,397,800

     

     

     

     

     

     

     

    2,397,800

    2014

    1,077,800

     

     

     

     

     

     

     

    1,077,800

    2015

    801,000

     

     

     

     

     

     

     

    801,000

    2016

    375,000

     

     

     

     

     

     

     

    375,000

    2017

    3,300,000

     

     

     

     

     

     

     

    3,300,000

    Thereafter

    0

     

     

     

    349,200

     

     

     

    349,200

    Total, gross

    18,828,100

     

    4,175,000

     

    349,200

     

    2,355,100

     

    25,707,400

    Excess cash

     

     

     

     

     

     

     

     

    -1,970,650

    Total, net

     

     

     

     

     

     

     

     

    23,736,750

     

     

     

     

     

     

     

     

     

     

    (1) Includes $1.7 billion of commercial paper, $420 million in an AIG FP demand note and various other debt instruments.

     

     

    Note that the $2.05 billion credit facility coming due in 2009 can be extended for an additional year at the Company's option.  With the sources of liquidity we mentioned, we think the Company will be able to meet its obligations.

     

    Downside Scenario

                If we are wrong on our liquidity analysis and if the Company doesn't get sold and if the credit markets do not open for refinancing corporate debt and if the government / AIG doesn't step in to fund AGFC, then it is possible the Company will be forced to file for bankruptcy.  We think Geithner will want to avoid a black eye like that on the public stage, but we admit it could happen.  We have run rigorous analysis both on the debt agreements (see next section), and on the numbers, and we do not think we can be impaired under any reasonable scenario.  For instance, if we assume the following:

     

    -          30% default and 50% severity on the real estate loans (remember current delinquencies are 4%)

    -          40% default and 80% severity on the other loans

    -          25% impairment on the equity in the insurance subsidiary

    -          $1.5 billion up-streamed to AIG, which is half the current book value (very unlikely given the personal liability around an unlawful dividend)

    -          $600 million of operating company liabilities given senior treatment

    -          $200 million of bankruptcy costs

    -          $3.5 billion of debt matures ahead of us (Q4 2009 filing)

     

    We still come to nearly 80 cents recovery on the bonds in a run-off.  Leaving all the assumptions the same but assuming 80% frequency and 60% severity on the real estate loans, we break even at a 50 cent purchase price. 

     

    Covenants and Other Legal Restrictions

                Please note, we do not purport to be lawyers, so take this analysis with a grain of salt and do your own homework.  Although our liquidity analysis shows that the Company should be able to meet its bond maturities without playing potentially value-destructive games, it is still important to understand the contractual protection that bondholders have.  To be clear right up front, we do not believe that AGFC's relationship with AIG as a wholly-owned subsidiary puts these bonds at any additional meaningful risk.  Most importantly, these bonds are issued by AGFC, not by AIG.  This means that our claims are on the assets of AGFC and AIG's lenders and other creditors are structurally subordinate to all of our claims.  AGFC does not guarantee any of AIG's debt nor the debt of any other AIG subsidiary (including AGFC's parent American General Finance Inc.).  AIG indirectly owns the common equity in AGFC.  Other than that AIG has no claim on AGFC's assets and has limited ability to upstream cash or other assets to the parent.  AIG (and any other entity, related or not) is also effectively prohibited by an "equitable and ratable" clause in its ability to provide funds to AGFC in exchange for secured positions.  It can therefore only lend to AGFC on an unsecured basis, as it has recently through a demand note, or inject equity.  We have also reviewed the agreement between AIG and the federal government and do not believe the agreement provides AIG with any additional powers that could impair value in our bonds.  Finally, we reviewed the publicly available information regarding all of the AIG entities in the AGFC chain of ownership to ensure there are no other areas of maneuverability that could impair value in the bonds.  We explain all of these issues in more detail below.

     

    AIG Loan Agreement with the U.S. Federal Government

    The Fed's loan agreement is worth dissecting in detail as it has strong covenants that serve to protect AGFC bondholders.  First, the Fed's loan was signed in conjunction with a Guarantee and Pledge Agreement (the "Guarantee Agreement").  In the Guarantee Agreement, AIG caused all domestic subsidiaries with aggregate assets of greater than $50 million, except Excluded Subsidiaries, to guarantee the obligations of AIG by granting Liens on certain of its assets. The Excluded Subsidiaries include: Regulated Subsidiaries, Securitization Subsidiaries, International Lease Finance Corp ("ILFC") and AGFC and AGFI. [Note: The exact amount of the security attained under this guarantee agreement is currently unknown because the Guarantee Agreement was limited in its scope due to section 3(a)(xiv)(ii).  The gist of this clause is that it limits the amount of any Guarantee to the amount permitted "without contravening" any existing financing agreements or being obligated under such financing agreements to "equally and ratably" secure the existing indebtedness.]  Therefore, AGFC was specifically excluded as a guarantor or pledgor of the government loan.

    The covenants of the Fed's loan limit AIG's options.  First and most importantly, the loan's negative covenants apply to AIG as well as any "Restricted Subsidiary".  A Restricted Subsidiary is any "non-insurance" subsidiary with aggregate assets of greater than $50 million, or an equity investment in a subsidiary with aggregate assets of greater than $50 million.  As such, AGFC would be considered a Restricted Subsidiary under this definition and the covenants of the government loan to AIG therefore apply to AGFC.  To be clear, AGFC is excluded as a guarantor under the Guarantee Agreement, but is included with regards to the negative covenants.

    The following material negative covenants of the Fed's loan prohibit AIG or any Restricted Subsidiary to (note we have left out some negative covenants that we believe are relatively immaterial in the larger picture of this analysis.  All capitalized terms as defined in the loan agreement):

     

    1. Indebtedness. Incur, create, assume or permit to exist any Indebtedness, except:
      1. existing Indebtedness and any extension, renewals, exchanges or replacements as long as it doesn't materially change the terms
      2. Indebtedness created by the Fed's loan itself
      3. intercompany Indebtedness (conditions applicable)
      4. Indebtedness of AIG or a Restricted Subsidiary incurred to finance the acquisition, construction or improvement of any fixed or capital asset (other than aircraft and aircraft-related equipment purchased by ILFC and its subsidiaries), and extensions, renewals and replacements of any such Indebtedness that do not increase the outstanding principal amount (conditions applicable)
      5. other Indebtedness incurred or assumed in the ordinary course of business to finance the business in the ordinary course
      6. other unsecured Indebtedness; provided that (i) the proceeds thereof are applied to prepay the Fed's Loan and (ii) such Indebtedness shall mature six months after the term of the Fed's Loan
    2. Liens. Create, incur, assume or permit to exist any Lien on any property or assets now owned or hereafter acquired by it or on any income or revenues or rights in respect of any thereof, except:
      1. Existing Liens
      2. Liens created via the Fed's Loan
      3. Liens existing on any property or asset prior to the acquisition so long as the Lien was not created in contemplation of the acquisition and pertains only to the asset acquired
      4. Tax Liens
      5. carriers', warehousemen's, mechanics', materialmen's, repairmen's or other like Liens created in the normal course
      6. other ordinary course liens
    3. Sale and Lease-back Transactions except:
      1. If it is allowed under the sale/ merger covenant
      2. Capital Lease Obligations allowed under (1) or (2) above
    4. Mergers, Consolidations, Sales of Assets and Acquisitions, except:
      1. AIG and Subsidiaries may purchase and sell assets in the ordinary course of business
      2. AIG and Subsidiaries are prohibited from any other Asset Sales unless:

                                                                   i.      consideration at least 90% of which is cash,

                                                                 ii.      such consideration is at least equal to the fair market value of the assets being sold

                                                                iii.      the Net Cash Proceeds of such Asset Sale are applied to pay down the Fed's loans

    1. Other Indebtedness and Agreements.
      1. any waiver, supplement, modification, amendment, termination or release of any indenture, instrument or agreement pursuant to which any Indebtedness of AIG or any of the Subsidiaries is outstanding if the effect of such waiver, supplement, modification, amendment, termination or release would materially increase the obligations of the obligor or confer additional material rights on the holder of such Indebtedness in a manner adverse to AIG, any of the Subsidiaries or the Fed

     

    Essentially, the government loan precludes AGFC, as a restricted subsidiary, from incurring additional indebtedness, creating liens (priming our bonds), engaging in sale-leasebacks, and selling substantial amounts of assets (subject to certain conditions).  These covenants go a long way to protecting our bonds for so long as the government loan is outstanding.

    There are further protections in the government loan document.  Under the Events of Default section, if AIG or any Subsidiaries were to fail to pay principal or interest on "Material Indebtedness" when due or payable it will result in an Event of Default under the Fed's Loan.  Furthermore, if any subsidiary were to file bankruptcy, either voluntary or involuntary, this would constitute an Event of Default under the Fed's Loan.  Under any Event of Default, the loan from the Fed would become immediately due and payable.

    We should point out that there is language in the federal loan that states that the government could ask AGFC to guarantee the government loan so long as in so doing the government did not trigger the equitable and ratable clause.  We do not see a way for the government to get around the clause.  We therefore believe there is little risk of the government priming our loan by taking a secured position in the assets.

    In fact, if anything the government loan appears to be a potential source of liquidity for AGFC.  In September, the Company borrowed $420 million from AIG Funding, Inc., an AIG-guaranteed funding affiliate, in the form of a demand note.  AIG Funding appears to be an entity that has access to the government credit facility funds, and therefore is an intermediary through which AGFC can access government funding.  The Company noted in its September 10Q that if necessary it would look to tap into the government funding facility, so this is consistent with the loan we saw in September.  It is also consistent with the government's incentive to keep AIG's subsidiaries out of bankruptcy.  In fact, one could argue that the demand note was not really necessary (given that AGFC finished the quarter with over $2 billion in cash on the balance sheet), and that the loan was more a symbolic gesture by the government and AIG that AGFC would be supported.

     

    Support Agreement / AGFC Credit Facility

    Another powerful document protecting AGFC lenders comes from the Support Agreement that was put in place concurrent with the renewal of the Company's 364-day credit facility on July 10, 2008.  Essentially, the banks that are party to the facility required that AIG sign the Support Agreement to protect the banks' investment.  The credit facility, which is unsecured, is pari passu with our bonds so we get the benefit of the arrangement the banks negotiated to provide for their own protection.  As part of the Support Agreement, AIG agreed to have AGFC maintain an Adjusted Tangible Leverage Ratio of no greater than 8-to-1 at the end of each calendar quarter.  The Support Agreement thereby effectively limits the amount of debt at AGFC, particularly if portfolio write-downs or discounted asset sales erode the Company's net worth.  At current debt levels, the Support Agreement requires that the Company maintain approximately $2.9 billion of tangible net worth.  With $3.2 billion of tangible equity as of the most recent quarter, this covenant also serves to put a limitation on how much the Company can dividend up to the parent, as dividends would reduce tangible book value.  As further protection, the Support Agreement states that AGFC must maintain minimum Consolidated Net Worth of at least $2.2 billion.  This last piece is not meaningful for now given the $2.9 billion requirement discussed above, but at some point it could come into play. 

    AIG's commitment under the Support Agreement will terminate automatically upon the earlier of (a) the termination of the 364-day credit facility, (b) payment in full of the loans outstanding under the 364-day credit facility, (c) AIG providing an irrevocable and unconditional guarantee of the outstanding amounts under the 364-day credit facility, or (d) consent from the majority of the 364-day credit facility banks.  Because the Support Agreement provides the first layer of defense for bondholders, we need to understand the risk that any of these four events might come to pass. 

    The termination of the facility considered in part (a) above is irrelevant because AGFC already drew down the facility in its entirety.  There is therefore no way to terminate it without paying off the debt, which is actually part (b).  As we will show, part (b) above is the only viable option, and as such we will address it after discussing parts (c) and (d).  Regarding (c) above, which contemplates AIG guaranteeing the credit facility, our read of AIG's loan with the federal government indicates that AIG is limited in its ability to guarantee any debt (including the credit facility) of AIG's subsidiaries.  This language can be found in Section 6.09 of the loan document.  Part (d), consent of the majority of the banks, is even more unlikely as there is no incentive for the banks to provide such consent in light of the current stress of the parent.    

    This leaves (b) above, paying off the balance of the credit facility, as the only viable option.  If AGFC were to pay down the credit facility, it could conceivably work itself out of the restrictive covenants.  We think this is unlikely prior to maturity, however, because the credit facility provides two-year paper (it is extendable for an additional year at AGFC's option) and it would be irrational for the Company to repay two-year paper when it has nearer-term maturities coming due.  We think it is unlikely that the Company would put itself at risk for nearer-term maturities by removing its safety net in the two-year credit facility just to escape the covenants.  Furthermore, the optionality AGFC would create by relieving itself from the restrictive covenants in the credit facility would not be that great.  We therefore believe we will benefit from the credit facility's protection until its maturity in 2010.

     

    AGFC Bond Indenture

    The bond indenture is fairly light on covenants, but it does afford bondholders some protection beyond the government loan and the credit facility discussed above.  The indenture (and the credit facility for that matter) has an "equal and ratable" clause.  The clause specifically states that if the Company takes on any secured debt then the notes we would be purchasing would receive a lien that is "equal and ratable" to any new secured borrowings.  What this means is that any prospective secured bondholder would have to be comfortable with a security claim that is diluted by the existing unsecured debt.  This would probably be unpalatable to most any lender given the amount of senior unsecured debt ($25.4 billion unadjusted, or $23.4 billion net of the $2.0 billion of excess cash).  With that being said, there are a few loopholes in the clause that create some risk exposure provided that the protections afforded by the government loan on the limitation of additional indebtedness and liens are, for whatever reason, not in effect. 

    One loophole is that the Company has a carve-out that allows it to secure up to 10% of its consolidated net worth.  This would work out to $282 million, which is a relatively small amount.  A bigger loophole risk is the potential ability of the Company to structure a repo transaction as a "sale" that would allow it to finance its mortgage receivables.  It is an important distinction that it be deemed a sale because as we have shown, AGFC is effectively prohibited from creating secured debt by the equitable and ratable clause.  The grey area is caused by language in the indenture (Section 1007(b)(12)) that could be interpreted to give AGFC the ability to pursue repos, provided that the transaction is limited only to the assets being sold (in other words, it must be non-recourse). 

    We think that the Company's entering into a repo transaction would be challenged by bondholders, but that might not prevent the Company from attempting such a transaction.  If successful, AGFC could pay off loans in front of ours (or underwrite new loans, which would probably be value neutral to us at worst).  In this case we would be worse off because there would now be new "debt" in front of ours in a secured position (this would not be as big a problem if the repo arrangement matured after our bonds mature).  Given the multitude of liquidity options available to the Company, we think the risk of repos is very low.

    Another related risk that we need to consider is the potential for the Company to sell assets outright.  The indenture does not specifically preclude asset sales.  There is therefore a risk that management might sell assets, potentially at a discount to fair value, in order to extend the life of the Company by paying off maturities as they came due.  While this is a risk, we believe it is manageable because even if the Company were to sell off all of its assets at 70 cents on the dollar, the bonds would still return 82 cents.  Even under this fairly absurd scenario, a recovery of 82 cents would be a perfectly acceptable investment result given our low purchase price.  Furthermore, large discounts on large sales of assets would significantly reduce management's tangible net worth and risk violating covenants in the Support Agreement for so long as the credit facility is outstanding.  Again, given the multitude of liquidity options available to the AGFC, we think the risk of heavily discounted asset sales is very low.

    Finally, in the event that AGFC can not meet its maturities or otherwise would like to renegotiate the terms of the indenture, a 100% vote is required to pass changes to any of the important money terms.

     

    Unlawful Dividends

    As asset stripping is the primary risk we are concerned about, the main protection bondholders have is a restriction on the ability of the Company to pay dividends up to the parent.  Although there is no specific limitation on the ability to dividend up to the parent in the bond indenture, there is strong protection against "unlawful dividends," which is a dividend that would leave AGFC insolvent.  This effectively puts a limit on the value of assets that could be distributed out of AGFC.  Solvency is a simple test based upon the assets being worth more than the liabilities - if they are not the company is considered technically insolvent.  There is nothing in the document that would prevent AGFC from up-streaming all of its net worth (which was $2.8 billion as of September 30, 2008).  However, if the Company were deemed to be insolvent as a result of the dividend, we believe the directors of the Company would be held personally liable.  As such, we do not believe that a lawyer would let the Board dividend out more than half of AGFC's current net worth, nor do we think any Board member in his right mind would even consider putting himself personally at risk with such a transfer.  This puts an effective cap on the amount of a dividend to AIG at about $1.5 billion, if at all.  We would not be happy to see that happen but it certainly does not come close to impairing value on the bonds. A $1.5 billion dividend would first eat into equity value and if it did end up impairing the bonds, it would represent a 6.4% hit ($1.5 / $23.4 billion).  Again, such a dividend would not be possible for so long as the credit facility is outstanding.

     

    Conclusion

    We repeat, does this sound interesting?

     

    -          Strongly cash flowing entity

    -          Senior unsecured debt trading at 50-60 cents

    -          IRRs held to maturity of 35-45%

    -          Assets are whole loan mortgages with 4% delinquencies

    -          To be impaired dollar one at 50 cents would need 100% defaults and 57% severity on mortgage (66% severity on average home price)

    -          Likely government support

    -          Assets for sale

    -          Liquid bonds - over $20 billion

    -          No secured debt ahead of you

     

    We think these bonds are a no-brainer bet. 

     

     

    Disclaimer:

     

    We are neither a registered investment advisor nor a broker/dealer. 

     

    Readers are advised that this report is issued solely for information purposes and should not to be construed as an offer to sell or the solicitation of an offer to buy any security. The opinions and analyses included herein are based from sources believed to be reliable and written in good faith, but no representation or warranty, expressed or implied is made as to their accuracy, completeness or correctness.

     

    YOU SHOULD VERIFY ALL CLAIMS AND DO YOUR OWN RESEARCH BEFORE INVESTING IN ANY SECURITIES MENTIONED IN THIS REPORT.

     

    We accept no liability whatsoever for any direct or consequential loss arising from any use of information in this report.

     

    We may on occasion hold positions in the securities mentioned; however, these positions may change at any time.  We currently own AGFC bonds.

     

     

    Catalyst

    - Sale of the company.

    - Liquidity picture crystallizes.

    - Government support.

    Messages


    Subjectcredit facility
    Entry01/22/2009 06:41 PM
    Memberbriarwood988

    On your post you indicate that b/c the notes are pari passu with the facility, noteholders get to enjoy the same protections as long as the facility remains outstanding. Can you walk me through how that works exactly? If there is a default on the facility, one would think the company would renegotiate better terms with the facility (possibly to the detriment of the noteholders), and cure the issue. For example, they could lower the net worth covenants in return for paying off a certain amount of principal, granting a security interest, etc. Thanks in advance Max.


    SubjectRE: credit facility
    Entry01/23/2009 11:15 AM
    Membermadmax989

    Completely fair point.  If it wasn't emphasized enough in the document, we benefit from the terms of the facility while it is in place, but clearly we can't rely on it as bulletproof protection.  Our equal and ratable clause prevents the credit facility from taking a secured position though, which is our main concern.  Also, the legal argument is important, but fundamentally this investment's strength resides in the numbers and the low likelihood that we could be impaired based on the quality of the underlying loans and the extreme levels of defaults required to dent our bonds.


    SubjectRE: region
    Entry01/23/2009 11:36 AM
    Membermadmax989

    This is what the company discloses in its 10K:

    CA 14%

    FL 6%

    OH 5%

    IL 5%

    VA, NC, PA, CO 4% each

    Other 54%


    SubjectCouple of other issues and questions
    Entry01/29/2009 12:15 PM
    Memberneo628

    1.) Given that their weighted average life is so short, that would seem to imply that the current portfolio has almost entirely been originated within the last 3 years (i.e. the bubble period).  That would seem to open up the possibility for much higer defaults

    2.) Looking at the bonds, there are verious maturities, if you think the company is going to not have a liquidity issue, why not buy the 2009 or 2010 bonds.  On an IRR basis, the 2009 bonds seem pretty high

    3.) The bond market is discounting these bonds - i guess because of both liquidity and solvency concerns.  What do you think the bond market consensus is are to why current prices are the right ones?

    4.) What are the next milestones that you look for for the full year and Q4 earnings release and beyond?


    SubjectRE: Couple of other issues and questions
    Entry02/04/2009 12:08 PM
    Membermadmax989

    1. You can get a very good sense for the years in which the mortgage loans were originated by looking at the company presentation.  There is a chart that shows the vintages.  Be careful to separate the mortgages from the other loans as the other loan categories have much shorter maturities.

    2. That is up to you.  It's a trade-off between IRRs and absolute dollars and just depends on what you are looking for.  

    3. This is not something we have spent a lot of time thinking about, but I would tell you two words probably carry the day: "AIG" and "sub-prime".  Most investors who aren't taking the time to understand what they own will not want to own anything associated with these two nasty words.

    4. We don't have any specific expectations.  Of course, we will be most focused on the liquidity picture and loan delinquencies.


    SubjectRE: update?
    Entry03/20/2009 07:14 PM
    Memberdoggy835

    I'd like to hear an update, too. I think BK is pretty likely, but recovery should provide upside if you buy cheaply.

    There seem to be two series of notes, "retail" (CUSIPs 02639xxxx) and regular MTNs (CUSIPs 02635xxxx). Any difference between them?

      Back to top