Many pundits and analysts have panned investing in financials, especially banks, given the regulatory issues brought on by the Financial Crisis, namely the Dodd-Frank law which neutered most of the profit drivers that banks enjoyed prior to its passage. The popular theme being purported by investors is that financial companies have become utilities with a fixed and stated rate of return that minimizes risk but also minimizes returns on capital.
I'll post a passage from the recent Whitebox Tactical Opportunities Fund fourth quarter 2013 commentary stated:
We also don't have an unstable financial system. If anything it is excessively stable in our opinion. It is hard to think of any objective to which the U.S. government has become more devoted than the immortality of money center banks. True, as Dodd-Frank takes hold it becomes ever more evident that the banks have been neutered, that we can expect less than ever from them in support of entrepreneurial capitalism. We are in an era of do-nothing banks, banks as public utilities, banks as government pets. Among the many, many things the big banks will not do in the future, however, is also this: they will not go broke.
This does not mean that important American public companies lack access to capital. Comfortingly, important companies need less help from the banks than they have perhaps ever. With their customary sources of capital first in disarray and now in retreat, these companies have increasingly become self-financing by accessing bond markets directly and more conservative in their use of cash by dramatically cutting back spending. There are downsides to this: restrained business capex by U.S. firms has been probably the single biggest reason the recovery has been mediocre in our opinion. The upside is that many American companies have lots of cash and balance sheets are in great shape, making U.S. firms more resilient in a downturn.
In other words, the high capital requirements brought on by the regulators would decrease the leverage and bring ROEs way down. But the counter to that is Jaime Dimon, CEO of JP Morgan Chase & Co., stating that this analysis is true, if nothing changes. All else being equal, if you raise requirements for tier 1 capital, then ROEs will go down. But his viewpoint is that banks will not sit idly by and accept the lower rates of returns. They will re-price loans to reflect those higher capital ratios, increase fees, exit or enter ancillary businesses, and adjust their rates of return accordingly. They won't simply accept that it is a new paradigm of lower ROEs and sit on their hands.
What am I getting at? Well, I agree with Dimon and believe that most financials are going through a period of reorganization and refocus. One such financial stock is Ambac Financial (AMBC), who is undergoing a similar transition in order to come out from its bankruptcy with a stronger, healthier business model. This is a highly complex business, with many moving levers of possible upside benefit to shareholders. Additionally, given the complexity, uncertainty surrounding a lot of its claims, and the angst towards financials in general, the situation provides a possible strong risk-to-reward payoff.
Prior to its bankruptcy in 2010, the company underwrote insurance for a wide variety of debt instruments, including RMBS, municipal bonds, and student loan ABS. The RMBS credits that it insured began to default in large quantities starting in mid-2007, leading to massive losses. Over the subsequent three years, the company entered into many agreements, raised capital, and settled claims, but in June 2010, formally entered bankruptcy to restructure its debt burden and settle other claims.
The bankruptcy process was lengthy, but it finally emerged in May of last year, issuing 45 million shares of common stock and 5 million warrants. The stock trades under the ticker symbol (AMBC, formerly ABK), while the warrants can be purchased under the symbol (AMBCW). Its main line of business is described below from its 10-k:
Ambac's principal business objective is to reorganize its capital structure and financial obligations through the bankruptcy process and to increase the residual value of its financial guarantee business. Ambac's financial guarantee business was executed through its primary operating subsidiary, Ambac Assurance. Ambac is also exploring entering into new businesses, apart from Ambac Assurance and Everspan Financial Guarantee Corp., as it prepares to emerge from bankruptcy. As part of its efforts to increase the residual value of its financial guarantee business, Ambac Assurance has pursued certain loss mitigation strategies, including seeking recovery of paid claims, commencing litigation to recover losses or mitigate future losses, entering into commutations of policies at discounts to their expected losses and purchasing Ambac-insured securities (collectively "de-risking").
The legal process and entities are complex and lengthy, but you can read about them in the latest 10-Q (HERE). Essentially, it is still a monoline insurer but has several other pieces that have to do with its reorganization following bankruptcy. The new company began trading on May 8th after shedding its debt and settling its litigation with the IRS the month before.
In general, it is probably more helpful to think of the stock as a collection of disparate assets attached to a now small monoline insurance business. It is those disparate assets that add significant value to the share price.
I describe these disparate assets below:
Roll-off of RMBS Reserves
The thesis here is that the housing recovery has driven the prices of its underlying RMBSs up and lessened the need for reserves against losses. Below is Ambac's book of business by vintage year, type, and gross par outstanding. The total adds to $16.882 billion.
And against that is the loss reserves by type:
(Source: Q3 2013 10-Q)
Clearly, the company has substantial reserve levels against those boom-year vintages in RMBS to the tune of roughly one-third. In other words, one-third of the remaining RMBSs could be written off based on reserve levels. The question is, how much will the actual losses be compared to what it's prepared. This is where the mispricing lies. The prices of these Ambac-insured RMBSs, according to Bloomberg, have increased substantially over the last several years. (Ambac, in its effort to increase transparency, allows the download of a spreadsheet with all its RMBS securities HERE). Meanwhile, the level of reserves has only trickled off, dropping $154.3 million in the third quarter after an $18.7 million benefit in the second. As stated in the third quarter press release: "Third quarter 2013 results were driven by lower estimated losses in first and second lien RMBS and student loans, partially offset by higher estimated losses in the municipal finance portfolio." We'll get to the muni portfolio in a moment.
The prices of the RMBSs average to an approximate estimate of 16.5% off par. In other words, the market is assuming about half of the market risk that Ambac is, which would mean a 16.5% reduction in reserves or $2.77 billion in actual losses, leaving $2.63 billion to be released out of the reserves in the coming years. That equates to 2.5x the current market cap of the company. We will use that as our best-case scenario below in our sum-of-the-parts valuation.
But even if the actual losses turn out to be 23% of losses, it would mean that the company would still have to release approximately $1.6 billion, or 160% of the current market cap. Being conservative, let's assume the latter as the low end in our model. On the flip side, the market may still be a bit conservative in its pricing, and losses may turn out even lower than they are expecting. At 13% of losses, the losses turn out to be $2.19 billion or just 40.5% of reserves.
It will be difficult to ascertain just how much it is going to expel out of reserves and into equity, but it is very safe to say that over the coming quarters/years, it will be releasing between $1.5 and $3 billion of reserves on these insured RMBSs. This is a substantial sum considering the market cap of the company and the fact that it has zero debt.
Representations and Warranties
Another source of value is the recoveries of losses resulting from breach of contracts of representations and warranties (R&W) on securitized mortgages insured by Ambac's pre-bankruptcy predecessor. Fellow monoline insurers Assured Guaranty and MBIA won significant court rulings while Ambac was in its bankruptcy proceedings. I won't go into all the legal rulings, but these lawsuits essentially ruled that the counterparties misrepresented the quality of the mortgages they securitized, which led the monolines to insure them based on false data. The monolines then suffered massive losses during the housing bust.
As such, Ambac then filed similar R&W lawsuits to recover losses from its counterparties, namely: Bank of America/Countrywide, JPMorgan, First Franklin, Nomura, and Credit Suisse. The precedents set by the prior cases likely mean the lawsuits will be settled rather than going through the lengthy court proceedings. It also means that Ambac is likely to have leverage in the settlement talks for more favorable terms.
The total benefit is likely to be north of $2.4 billion that it previously wrote-off, and that will be credited to shareholders. Of that total, $1 billion is within the Bank of America/Countrywide complaint alone. If the company were able to recover 50% of the losses, it would be $1.25 billion in awards, or 100% of the current market cap.
However, the three monoline firms (MBIA, Assured Guaranty, and Ambac) report these forecasted recoveries as a line item described as "Subrogation Recoveries". As I stated above, Ambac's subrogation recoveries are noted in the third quarter release as:
According to the data provided by the other two monolines, the actual recoveries have been trending above 100% of the expected subrogation recovery. For instance, MBIA has recovered over $1.7 billion in recoveries but had only expected to recover $1.5 billion, while Assured Guaranty recovered $107 million in its lawsuit against Flagstar when it had only expected to recover $43 million. Against Bank of America, Assured recovered $1.1 billion versus an original expectation of $1.04 billion. In other words, either the monolines themselves are assuming conservative recovery rates and getting that, or slightly above, or the recovery rates have been exceptionally high.
Now this rate of recovery has already been recorded, and is thus "baked into" the current stock price. But the icing on the cake is that the $2.5 billion in total outstanding judgments that the company is expecting represents just under half of the total policies written. However, no lawsuits have been issued, nor any recoveries baked into its estimates. Meaning the company could, in effect, engage in new lawsuits with the counterparties of the remaining approximate $2.6 billion, creating further upside catalysts for the shares.
I think we would be conservative in assuming another $750 million in additional litigation recoveries, with a best-case scenario close to $1.25 billion.
Net Operating Losses (NOLs)
The company currently has on the books over $5.1 billion of net operating losses (NOLs), which potentially could be valued at $35.70 per share [$5,100,000,000 * 35% tax rate = $1,785,000 / 50m shares = $35.70 per share]. Granted, there is a time value of money involved here, as not all of these losses can be realized immediately. However, the price per share allows us to put the value of the NOLs in perspective.
The NOLs are split between its two operating segments, with roughly $4 billion in its main operation Ambac Assurance Corporation (AAC) and its holding company, Ambac Financial Group (AMBC) with the remaining approximately $1.1 billion.
How can this value be realized? Well, the most apparent way would be for the firm to start earning taxable income again on its traditional operations. The value can be realized substantially faster if the company receives approval to restart its insurance business. However, on the conference call, the company seemed open to acquiring another AAC-like business in order to facilitate the realization of the NOLs' value to shareholders.
So how should we account for the value of these assets? Looking at the old Ambac entity (which traded under ABK) within Bloomberg, the effective tax rate for the company from 2000 through 2006 was between 23.3% and 27.6%. I averaged them all and got a mean effective tax rate of 25.3%. During the time period, its average cash taxes were approximately $208 million, which gives us an NOL utilization of $822 million going forward.
Now there are a couple of different ways we can value these NOLs: the merger method, which is the preferred methodology, or the traditional NPV method. It is hard to give a precise figure given we do not know what a good assumption is for pre-tax income. One scenario I ran through would be to use pre-tax income from 2000-2006 as a baseline going forward. The average during that time period was $810 million per year. I admit the simplification of this process, but since we have no idea when/if the insurance business will be reignited and what it would be producing in pre-tax income, we cannot really give an accurate picture of potential value to the NOLs.
Below is the traditional methodology using simplistic assumptions.
(Source: original analysis using company inputs)
Again, it is difficult to value these in a business such as Ambac given we cannot accurately forecast pre-tax operating income over the next several years. But the above gives us a general idea of the value of these NOLs. For valuation purposes, there will definitely be some delay to the actual realization of value unless it goes out and purchases a similar insurance or other financial business. To do so, it would have to raise capital, but it would then have an operating business with immediate pre-tax income with which to offset these NOLs. The $884 million would be an absolute best-case scenario, but I think $300 million would be a decent base-case estimate to use in our model.
Since the IPO back in May of 2013, the company has been a bit obtuse in its dissemination of information. It is assumed that this has weighed on investor sentiment. For some reason, the company did not give a second quarter conference call. That has begun to change, as it had its first conference call following the third quarter earnings release. It also issued a PowerPoint presentation on its Puerto Rican debt exposure, something that has also been weighing down its share price.
Management has been difficult to reach, as I still cannot find information detailing the top executives' pay packages. I have to assume given the lack of insurance operation that it is mostly in stock awards, meaning we could see share count rise from here. In any event, the next 10-Q will be released on or about March 3, and I hope to see that information then. With each release, it seems to be becoming more transparent and open, but it has been a slow process.
Puerto Rican debt became the center of the municipal bond world not long after the Detroit bankruptcy rattled the muni market. The tipping point came when Barron's highlighted the US Territory's issues on the cover of its magazine last August. The debt burdens of the island commonwealth are well-known now, with between $50 and $70 billion in total debt plus another $30 billion of unfunded liabilities placing it third among the fifty states in total debt behind only California and New York. However, the island has just 3.67 million people compared to the 38 million in California and 19.5 million in New York, making Puerto Rico many times worse on a gross per capita basis.
When the market reacted to Puerto Rican debt exposures, investors took down Ambac's stock price from approximately $24 to $15 over the course of 2 months. The nadir hit when the company finally issued its Puerto Rican slide deck on October 10. The stock price rallied substantially over the next two days, by roughly 14%, and thereafter getting back to $24 by the fourth of December.
The reason for the rally was that investors saw that only 10% of the total $2.5 billion in Puerto Rican insured debt were Commonwealth General Obligation (GO) bonds. Like in Detroit, there is a large difference between owning Detroit GOs and Detroit Water and other revenue bonds. 90% of Ambac's Puerto Rican debt are revenue bonds, with the largest exposures to highway and transportation authority bonds, hotel occupancy taxes, rum taxes, and sales tax revenue bonds.
(Source: Ambac presentation materials)
Even within the GO bond piece, 76% of it are GO guaranty bonds from the Public Buildings Authority. These are bonds issued to finance the construction of schools, health facilities, correctional facilities, and offices, for government use. The bonds are paid from lease payments to which the good faith and credit of the commonwealth are pledged. While that is all well and good, they are subject to non-impairment covenants of the commonwealth and supported by guaranty. Now, I take that with the grain of salt its worth, but you can make the argument that these GOs are slightly safer than plain-vanilla Commonwealth GOs.
So what are the possible losses from its Puerto Rican debt exposure? Well, I think a conservative case would be that all $250-million worth of GO bonds are wiped out, while the revenue bonds are paid in full. From here, you can add another $100 million to get to a worst-case scenario, in my opinion, while best-case would see no haircuts on the current debt as economic reforms take hold.
Exposure to Detroit is much more limited than Puerto Rico, with $170 million in total outstanding municipal bonds. Of that total, the company has $78 million in unlimited tax GOs and $92 million in limited tax GOs. Management addressed the Detroit issue on the recent conference call:
The geo's are very small portion of the city's liabilities. We are actively engaged as appropriate and what is now a very complex situation. To various means, including litigation and leviation, we're seeking to maximize economic outcome in Detroit. In fact, whereas just last Friday, we filed a lawsuit against the city of Detroit relating to the city's failure to segregate certain pledged property taxes to be used for making debt service payments on Ambac insured bonds.
Given the situation and the emergency managers' rhetoric regarding GO bonds that the city has issued, I would say it is safe to expect near 100% loss on these issues.
On November 8th, Ambac filed a suit against the city of Detroit asking that the city be required to segregate the property taxes pledged to pay the city's GO bonds insured by the company. Ambac's counsel stated that Michigan law is very clear on the matter, "the city is required to segregate the pledged property taxes and then only use them to pay debt service on the bonds. And bankruptcy law is equally clear on such matters: Michigan law must be followed."
Even still, I think it is a total loss of Ambac, and in the case of being conservative in our analysis, we'll assume that. Anything recovered would be immediately accretive, as investors seem to be assuming a total loss.
Student Loan Reserve Additions
While RMBS prices are moving higher due to lower default rates and higher home prices lowering the need for reserves, student loans are doing the opposite. Default rates for student loans, according to the Department of Education, have increased for the sixth year in a row. Two-year cohort default rates measures the number of borrowers whose loan repayments began in the 2011 fiscal year and defaulted (failing to make a payment for 270 days) by September 20, 2012. In that cohort, 475,000 defaulted on their loans, with the rate rising to 10% from 9.1%. The three-year default rate rose from 13.4% in 2009 to 14.7% in 2010.
There is a large dichotomy between for-profit and non-for-profit education institutions. The three-year default rate at for-profit colleges was 21.8%, although that is down from 22.7%, while nonprofit colleges have an 8.2% three-year default rate.
Student loan reserves fell by $120 million to $945 million in the most recent quarter. However, of the $5.6 billion in outstanding ABS on student loans, nearly 59% are from for-profit issuers:
(Source: Q3 2013 10-Q)
If we do some simple math from the table above, we see that it has reserves of approximately 26% against for profit-issuers and just 2.6% against nonprofit issuers. The question is will that be enough to protect against future, rising default rates? Here is how the company estimates default rates:
We develop and assign probabilities to multiple cash flow scenarios based on each transaction's unique characteristics. Probabilities assigned are based on available data related to the credit, information from contact with the issuer (if applicable), and any economic or market information that may impact the outcomes of the various scenarios being evaluated. Our base case usually projects deal performance out to maturity using expected loss assumptions and interest rates utilizing the projected forward interest rate curve at the reporting date. As appropriate, we also develop other cases that incorporate various upside and downside scenarios that may include changes to defaults, recoveries and interest rates.
And how it estimates reserves off of that:
In estimating loss reserves, we also incorporate scenarios which represent remediation strategies. Remediation scenarios may include the following; (i) a potential refinancing of the transaction by the issuer; (ii) the issuer's ability to redeem outstanding securities at a discount, thereby increasing the structure's ability to absorb future losses; and (iii) our ability to terminate the policy in whole or in part (e.g. commutation). The remediation scenarios and the related probabilities of occurrence vary by policy depending on ongoing discussions and negotiations that are underway with issuers and/or investors. In addition to commutation negotiations that are underway with various counterparties in various forms, our reserve estimates may also include scenarios which incorporate our ability to commute additional exposure with other counterparties.
In another note within the 10-Q, it states that changes to certain assumptions could make its reserves "underestimated". These changes include higher interest rates, regulatory issues, an increase in default rates, and loss severities on the collateral due to economic factors. On student loan debts, it has an estimate of expected loss at the end of the third quarter (a "high-stress" scenario) that the reasonably possible increase in loss reserves could be approximately $659 million.
In my analysis, I think it would be conservative and prudent to assume at least $300 million in additional reserves against losses, as default rates and interest rates rise over the coming several years. Worst case would be $659 million, with a best case of zero.
Now we get to what it means for the intrinsic value of the stock. The next question is how do we value an asset like this? Certainly a discount cash flow model will not work, since the company does not have a traditional, ongoing, recurring revenue stream. Typically, large financial conglomerates are valued based on book value of equity (think Berkshire Hathaway or Citigroup). We'll start with the current adjusted book value of equity straight from its 10-Q. What is adjusted book value? Here you go:
And here is its current adjusted book value calculation:
(Source: Q3 2013 10-Q)
So we start from a deficit of $498.4 million and go from there. Let's sum up the pieces we analyzed above and see where it comes out.
The sum-of-the-parts analysis is detailed below:
You can see the asymmetric payoff on the adjusted book value per shares. The worst-case scenario, with 22% downside, is definitely a worst-worst case which I place a low probability of occurrence. The base case is, in my opinion, highly likely, and the best-case scenario would be if all the stars align.
The next question is the time frame for these catalysts to play out in order to reap the benefits. Some of the catalysts will take longer to materialize than others. The nearest upside benefit will likely come from litigation settlements, namely with Bank of America, who appears to want to put its Countrywide issues behind it as quickly as possible. I expect Ambac to settle these sometime in 2014, probably in the second or third quarters. When MBIA settled with Bank of America on May 6th, the stock popped 55.5%! The RMBS roll-off is likely to play out slowly, with each quarter the company releasing a bit more in reserves as it feels more comfortable with the housing situation and the underlying exposures in its risk models.
The NOLs and Puerto Rican debt situation will take longer to unfold and will largely offset each other. Student loan loss reserve additions are likely to ramp up in the back half of the year, if at all. And lastly, Detroit could take years to play out as the lawsuits and restructuring plans go through the court system.
In general, a lot of the upside should be coming sometime this year or early next, but several of the overhangs will linger for several years to come. But the risk-reward ratio is very favorable, offsetting the overhang risks. And all of this ignores the potential of the insurance business re-starting probably post-2014, which would also be a significant value boost to the shares.
- RMBS loss reserves being released into equity/book value
- NOLs on the books providing a large asset
- R&W litigation claims likely to provide significant upside potential
- Muni portfolio safer than investors are giving credit for
- Potential of the restart of the insurance business
Ambac is definitely a black-box special situation play that isn't getting a lot of press or sell-side coverage. For those that have read "You Can Be A Stock Market Genius" (and if you haven't, you should!), this is the type of situation that Greenblatt pines for. There are several potential sources of optionality for investors, any one of which will add significantly to the share price. The lack of coverage, along with investors shunning many segments of the financial sector are the individual investors' gain. I see substantial upside over the next two years as these catalysts play out and the company gets its traditional insurance business up and running.