Description

I will not spend a lot of time on the business as there are three prior VIC write-ups to reference and to note for how the company has changed over the years. AGO is a mono-line insurer for financial guaranty. Translation: They insure bonds against default. They are also the last mono-line insurer of financial guaranty, depending on what happens to MBIA's attempt to create a new subsidiary to insure public finance.

Financial guaranty is the ultimate long-tail insurance. While structured finance deals pay premiums quarterly or yearly, public finance pays the entire premium up front and it is non-refundable. Premiums are then earned over the life of the bond issue. 75% of their book is on municipal bonds and 25% on structured finance bonds, compared to about 55% public finance and 45% structured finance in 2005. This dramatic shift occurred because most of their new business over the past three years has been municipal and the fact that structured finance runs off faster. Both areas have been receiving a lot of bad press that has investors shooting first and asking questions later with regards to AGO's legacy business.

CEO Dominic Frederico has overseen AGO from its separation from ACE in 2004 to present. He is no shrinking violet, which has proved valuable during the past few years. When S&P announced their proposed changes this past January, Dominic came out swinging with an investor call that tore into the proposed changes. He also has skin in the game, owning $8 million in shares with 1.3 million in options. He was an active buyer with his own money at $3, $11 and $14. Investors also have Wilbur Ross, who owns 8.7% of AGO, on the board. Ross bought 10.6 million shares in 2008 at $23.47, bought more during the equity raise and has not sold a share.

So let's look at the three issues and then I will get into more detail about the upside.

1. Structured Finance/RMBS

The great irony for AGO is that they were AA-rated for many years and lost a lot of RMBS structured finance business to then AAA-rated Ambac and MBIA. Meanwhile, Ambac has filed for bankruptcy and MBIA has had to restructure and put the structured finance business into run-off. AGO ended up winning by losing. However, the reserves relative to the net par outstanding are not large, which makes investors unduly nervous..

There is a huge issue about the underwriting quality of residential mortgages. It is being played out via the representations and warranties ("R&W") clause that allows holders of mortgages to put back bad mortgages to the originators. The result is that MBI and AGO are battling BAC, WFC and JPM, which you now see popping up in earnings calls and the like.

An amazing statistic from AGO's 4^th quarter2010  conference call:

"As of year-end 2010, we had reviewed over 53,000 individual loan files, aggregating to $8.5 billion of loans, and identified breaches of rep and warranties on approximately 48,000 of those loans equal to $7.8 billion. To date, we have now obtained repurchase commitments from six counter-parties across 23 transactions, a significant improvement."

 Meanwhile, JPM lost a suit involving use of sampling of mortgages:

http://www.huffingtonpost.com/2011/03/28/jpmorgan-loses-court-ruli_n_841484.html

 If you understand the situation, the following is pretty funny:

http://www.propertycasualty360.com/2011/03/25/banks-say-mbia-undervalued-losses-by-as-much-as-10

The banks that took out insurance on their securitizations for the benefit of the bondholders are trying to push MBIA into insolvency. [Separately, from what I have heard, the Article 78 suit brought against MBIA by the banks is unlikely to succeed because they have to show that the NY Insurance Department did not follow procedure, which is a high hurdle.] Since the financial guarantors pay the bondholders directly, the banks do not receive any of the loss proceeds unless they are also bondholders. The banks are not spending millions to help the bondholders - they want these mortgages buried so they don't boomerang back to them. The problem for the banks is that the financial guarantors are making full use of the put-back feature under the R&W - and the banks will do anything to try and stop the put-backs.

AGO has certain rights within the securitizations trusts, allowing them to make important changes. They learned very quickly that some mortgage servicers do a poor job with preventing and handling delinquencies, especially those servicers who are part of the selling bank and thus conflicted. By exercising mortgage servicing rights, AGO can improve loss outcomes through more pro-active servicing carriers and by removing bad loans. AGO estimates that they can recover $1.4 billion through this process. Due to a quirk in FAS 167, once AGO exercises their rights, they have to consolidate the trusts onto their balance sheet because they are deemed to have contol.

I believe that all of the banks will ultimately settle with Ambac, MBIA and AGO, even as MBIA and AGO recently turned up the heat by accusing the banks of fraud before a hearing at the NY state assembly. Some banks have already settled, where the insurance policies were commuted - essentially cancelling the policy. The last thing that the banks want is a lawsuit revealing what actually went on, particularly those banks (BAC, JPM, WFC) who bought their problems. While the insurer has to refund the premiums, all losses associated with the policy are erased from their books. It is a good trade for AGO.

A lot of the problems have been with 2^nd liens, a.k.a. HELOCs. The good news is that the HELOC problems appear to have peaked. In the 2010 10-K, on p.96, it shows a credit of $805.7 million for HELOC's for Present Value of Next Expected Loss and LAE to be Paid. A decent chunk of that relates to R&W and with one bank. You can guess the bank by looking on p. 35 of the 4^th Quarter 2010 Financial Supplement for the Top 50 US Structured Finance Exposures:

http://www.assuredguaranty.com/Documents/AGLFS%20December%2031,%202010.pdf

The settlement with Bank of America will be remove a significant overhang.

Due to the nature of repayments and amortization, the structured finance portfolio will likely shrink in half over the next three years. From 12/31/2009 to 12/31/2010, US RMBS net par outstanding ran down from $29.2 billion to $25.1 billion, or a 14% decrease. As an aside, CMBS exposure stood at $7.1 billion of net par exposure as of 12/31/2010, which is much smaller than RMBS.

2. S&P Changes

Standard & Poors unveiled proposed new requirements for financial guaranty insurers in January 2011 and requested comments through 3/25/2011. AGO responded early and often with some blistering comments:

http://www.assuredguaranty.com/Documents/Second_Letter_to_SP_with_attachments_03252011.pdf

Transcript and presentation of 2/1/2011 investor call on S&P:

http://www.assuredguaranty.com/Documents/SP%20Call%2002022011%20Transcript%20vF.pdf

http://www.assuredguaranty.com/Documents/SP%20Presentation%20for%20Call%2002-01-2011v2.pdf

AGO makes numerous arguments against the S&P proposal. The most cogent is that the S&P uses the same 40 year-old Hempel study of municipal bonds during the 1930's depression for their proposal as was used initially back in the 1970's to set up the current scheme. There are two main concerns with the proposal: One involving the exclusion of unearned premium reserves (UPR) and lower leverage tests.

a. UPR - This is the biggest concern because it could potentially require AGO to raise $1.9 billion of capital to maintain its AA-rating. If UPR is included (which their primary regulator, the NY State Insurance Department, does), AGO would already qualify for AAA. That is a big difference, but not that it matters. AGO needs the S&P more than the S&P needs AGO. If you are into conspiracy theories, one could say that the S&P is looking to kill the financial guaranty business. A more likely approach is that the S&P needs to re-establish its credibility and is taking a hard-line approach.

The ratings business is very lucrative. A review of McGraw-Hill (MHP) shows that while the S&P division accounts for only 25% of revenue, it provides over half the operating profits. Operating margins are in the high 40's for the S&P, so this is a business that will be hard for them to walk away from.

I think there is only a 25% chance of the UPR exclusion sticking. The main two regulators of financial guaranty insurance in the U.S., the state insurance departments of New York and Maryland, both include UPR in their claims paying ability for financial guaranty. My understanding is that statutory accounting does not use UPR and recognizes the premiums immediately and thus available to pay claims. The wild card is what the guys in the JC Penney suits decide but the fact that their UPR proposal goes against insurance accounting shows how flawed it is.

b. Lower Leverage Test - The second concern is lower leverage tests. While no official test currently exists, many believe it is 150 to 1, meaning for every $150 of par value of bonds insured, there needs to be $1 of capital backing it. The proposal calls for 75 to 1 leverage for AAA in public finance and 20 to 1 for AAA in structured, effectively doubling the capital requirements at the minimum. Fighting this requirement will not be easy with the Ambac bankruptcy poster child hanging around.

One clear benefit from the overall S&P proposal would be transparency and objectivity in the rating levels. There is a general sense of how the rating agencies have determined levels in the past but it has been more "double-secret probation" in nature. To be credible, the S&P has to establish that they are objective in how they rates things and are not "AAA" sellers.

Most people have seen or heard the doom-and-gloom predictions for municipal bonds in 2011. Besides the information presented by AGO showing debt service levels for municipalities are actually much lower than during the Depression, Bloomberg had this interesting tidbit in an article about the Vallejo bankruptcy:

"Since 1937, 619 local government bodies, mostly small utility or sewer districts, have filed for bankruptcy, according to Spiotto. In contrast, there were more than 11,000 Chapter 11 filings, used by companies to reorganize debt, in 2009 alone.

Local-government bankruptcies will "be minimal and isolated to mismanaged or weak credits," BlackRock Inc., the world's largest money manager, told clients on Dec. 7. States such as Pennsylvania and Rhode Island have become increasingly active in helping prevent them, it said."

http://www.bloomberg.com/news/2010-12-14/vallejo-s-california-bankruptcy-failure-scares-cities-into-cost-cutting.html

AGO did a presentation last summer and Frederico made several insightful comments. He noted that outside of general obligation bonds, many munis are tied to specific streams of revenues. [Check out the Denver Convention Center Hotel idea recently submitted by nha855 as an example.] Defaults tend to be more about unwillingness to pay as opposed to inability to pay. Governments own two unique assets - the ability to tax and the ability to apply tax liens. Plus, governments cannot be liquidated and only about half the states even allow municipalities to even file for bankruptcy. That is why muni bond issuers say that they have few real losses. AGO may make interest payments to bondholders for a while, they rarely have to pay principal, and then they recoup what they paid from the municipalities. Sort of like pay-day loans for governments.

Most problems tend to be cities and counties, where things can get out of control. Noted examples in the news are Harrisburg, PA and Jefferson County, AL. In reading the news about both situations, bankruptcy is not really an option and their respective state governments are now involved in the process. While both have received considerable news coverage, they are largely isolated instances stemming from corruption and mismanagement.

Nouriel Roubini has indicated that there might be between $30 billion to $100 billion of defaults over the next five years but with almost 100% recoveries:

http://www.financialsense.com/node/4679

---------------------------------------------------

Debt - AGO has between $22-32 million of debt maturing over the next three years. $189 million of debt will be due in 2014, but then no meaningful debt will be due until 2034.

Buffett - I figure this will come up, so might as well address it. Berkshire Hathaway famously entered the financial guaranty market in 2007 to insure muni bonds and then pulled out in 2009 because of less than firm pricing. Is that an indication that the financial guaranty business is bad? I do not think so because Buffett was looking for opportunistic pricing in what is a small side business for Berkshire. Berkshire had a very small presence that limited meaningful deal flow. Meanwhile, WEB is willing to accept lower returns in bigger businesses like Burlington Northern and Lubrizol.

Earnings - With a $10 billion investment portfolio, AGO generates almost $2/sh in pretax investment income per year. As the bond book seasons, losses will drop. Analysts' estimates are $3.11 to $3.85/sh for 2011, and $2.91 to $3.56/sh for 2012. A lot depends on the outcome of the above issues and it is difficult to extrapolate much from the past few years.

A rough gauge would be to consider $2+/sh in EBIT for investment income and add an equal amount of underwriting gain. Taking out taxes from $4+/sh in EBIT gets into the range of the analysts.

AGO should benefit from several trends. Last year, new premiums were down despite a rebound in municipal bond deals that totaled $429 billion. The President's Build America Bonds program, a.k.a. BAB's, provided interest subsidies or tax credits for qualifying municipal bonds, making the bonds attractive to investors without the need for insuring the bonds. BAB's ended in 2010 and it is unlikely to gain support in the Republican controlled House, so AGO should be able to wrap more bonds. Also, there is going to be a point where the US government decides to get out of insuring mortgage loans and there will be more non-agency MBS issued than the $63 billion issued in 2010 now that the credit rating liability contained in the Dodd-Frank Act has been nullified.

Due to the nature of financial guaranty accounting, even if AGO went into run-off, it would continue to earn premiums over time. The duration of the insurance is about 14 years and that may trend longer because the deals written over the past three years are less likely to be refinanced. The company provides visibility into premium and losses recognition over the next 20 years in the 2010 10-K, starting on p.189.

To at least get some back of the envelope numbers, let's start with a few basics about the market. There are $3.2 trillion in tax-exempt bonds outstanding of which $1.0 trillion are insured (Source: http://www.sifma.org/research/item.aspx?id=23188 ). Long-term, around 1/3 of tax-exempts are insured. On a yearly basis, insurance penetration peaked at 57% in 2005 and clearly shows that impact of under-pricing the risk. Going forward, there will continue to be demand for bond insurance. These smaller issues rely on an insurance rating more to attract buyers, while making interest rates reasonable for issuers. US municipal bond issuance has averaged $330 billion over the past 15 years (Source: http://www.sifma.org/research/statistics.aspx ). If AGO captured 15% of that, it would represent $50 billion in bonds wrapped. The upfront premium is, very roughly, 1% of the total issue. Thus, AGO could write $500 million of new premium per year. While that won't translate into significant earned premium in the first year, a series of similar years will accumulate.

Financial guaranty is like the oil business. The assets have a limited life expectancy and new assets have to be continuously acquired to replace the existing ones. The key difference is that AGO gets all the cash up front and most of it is placed in UPR and in the investment portfolio. The balance sheet represents a better picture of the economic value of AGO than the income statement.

Valuation - Book value was $20.65 (82%) and adjusted book value was $48.98 (35%) as of 12/31/2010. AGO calculates adjusted book value to include the present value of new premiums, referred to as PVP. By comparison, MBIA trades at 69% of BV and 26% of ABV. I suspect that MBI will trade up when the Article 78 suit is dismissed, and it is obvious that AGO is financially stronger than MBI.

Book value is understated due to the FSA deal. Since AGO paid under 40% of equity for the business, there was $1.7 billion of negative goodwill generated. Normally, this would be run through the income statement as a gain all at once but, instead, they had to increase UPR (thus, it is in ABV) and amortize it over time.

BV and ABV are understated by $3.30/sh because of VIE consolidation. The VIE's represent $607 million of net liability on the balance sheet by virtue of AGO exercising their contractual rights but there is no actual recourse. It is an accounting quirk that FASB has yet to address.

Competition - Before the financial crisis, there were seven financial guaranty insurers. Now there is AGO and MBI. You will probably note that most of the competition now is not with other financial guarantors but with other means of credit enhancement. On a deal basis, AGO insured 14% of all tax-exempt deals, equating to 8.4% on a par basis, implying that they handled many small deals. That does not include BAB's because they are not tax-exempt. With BAB's not being renewed, AGO will likely gain market share in 2011.

Upside - If you have read this far, now is the time to talk dirty. On the surface, there is not one main catalyst but a series of them: 

• S&P Ratings Proposal - Whatever is decided, AGO and the market will have clarity for once on rating requirements.
• Settlements with the banks - obviously based on today's news
• Monopoly/Duopoly - Pricing and market share will be more rational.
• Structured Finance Portfolio - As this shrinks, it frees up capital because public finance has lower capital requirements. It also frees up cash, as they unwind RMBS credit derivatives that have required posting of margin. The BofA settlement includes reinsurance that will free up more capital.
• Buybacks - AGO is authorized to buy back up to 2 million shares. This is on hold until their capital requirements are known. When looking at buybacks from an adjusted book value basis, this would be hugely accretive, even if they went into run-off.

Outcomes

1. Base Case - AGO goes into run-off because the S&P goes through with the proposed UPR exclusion and lower leverage tests.

2. Likely Case - The S&P recants the UPR proposal but keeps the lower leverage test. AGO will probably need to raise a small level of capital. According to AGO, they should be rated AAA, if this is the case; however, the S&P currently has them rated AA+. The S&P has enough egg already on its face, so they will engineer a way to justify last fall's downgrade of AGO. Given the

3. Best Case - Minimal impact from the S&P proposal, they get AAA rating back and it is off to the races.

Risks:

- The S&P proposal requires a capital raise

- Fewer public bond deals

Catalyst