Ambac Financial Group (AMBC) – Litigation – Potential for 100%+

Current Price: $13.8

Target Price: TBD

Upside: Potential for 100%+

Expiration Date: Fall 2022

Over the last couple of months we have received a fair amount of inbound questions on the Ambac Financial Group litigation case. This is an old, quite well-known, and, admittedly, a very intriguing situation that has tempted and frustrated investors for over a decade now. Every AMBC write-up over the last decade nearly always started along the lines of ‘now is the best time to jump in, because…’

Today, once again, the finale seems to be close. Admittedly, for us this case is in the ‘too hard pile’ due to various litigation and accounting intricacies, that smarter investors than us were unable to assess/predict accurately so far. Nevertheless, given inbound requests, we have looked at the situation and are publishing the key aspects of the AMBC investment thesis below. Keep in mind, that litigation cases are inherently very risky and unpredictable (refer to RENN as a recent example).

For more detailed information, please refer to the article and comments on VIC as well as Andrew Walker’s post and updates on YAVB (paywalled, but really worth subscribing to).

Ambac is one of the financial guarantors/bond insurers, which in the early 2000s amassed exposure to billions of toxic real-estate-backed mortgages (e.g. NINJA mortgages) and then got wrecked by the housing market explosion/GFC before filing for bankruptcy in 2010. Since then, many bond insurers, including AMBC, have sued banks on the basis of contract breaches. All of these suits have so far ended in substantial settlements in favor of the claimants (see more in the “Historical suits” section below). The only major litigation remaining is AMBC against Bank of America’s subsidiary Countrywide. This litigation against BoA is comprised of two suits seeking $2.2bn (+$600m in related court case) in damages and doubled that figure in interest rate accruals over the last 14 years (vs AMBC market cap of $670m and a book value of $880m). The larger suit, after dragging on for more than a decade due to countless appeals and delays, seems to finally have a definitive trial date set at the 7th of September’22. Given that all of the previous similar insurer/bank litigations ended up with a settlement (none of them failed or reached court), it seems that a settlement can also be expected here before September. A successful settlement is likely to have a material impact on AMBC share price.

The upside here will depend on several factors:

1. The likelihood of litigation settlement/award – seems very likely given the precedents of almost identical situations from the GFC involving banks and bond insurers.
2. The total damages amount – AMBC expected damages from BoA amount to $2.8bn.
3. The accrual rate, which will be used to calculate interest on the awarded amount over the last 14 years since the damages occurred. Expected between simple (not compounding) 5% – 9% annually. In 5% scenario, the total damages, including accrued interest would be $5.3bn and in the 9% interest scenario – $6.3bn.
4. In case of a settlement, the discount is applied to the expected award amount. Trying to guess this won’t add much value here but the main point is that even a steep discount still results in a very substantial settlement for AMBC.
5. Use of proceeds – at the moment there is a lack of clarity on how the award/settlement proceeds will be used after these land in AMBC’s pocket and what portion of these would result in shareholder distributions as opposed to only book value gains.

Even assuming the lower 5% accrual rate and 40% settlement discount, the expected settlement award would stand at $3.2bn, resulting in c. $2bn BV gain for AMBC (as around $1.2bn estimated win has already been recorded on the balance sheet). The company will pay no taxes on the settlement award due to $3.8bn NOLs.

Will elaborate on these points in more detail below.

The total damages amount – in case of a settlement with BOA, it is likely that the other smaller suit (with $600m damages) would be included as well, with total claimed damages amounting to $2.8bn (in line with what is indicated on BoA financial reports). Investors seem to more or less agree on this figure but in case this goes to court there is a risk that the final award (before interest accruals) will be below the $2.8bn for both suits. The total litigation asset value on the AMBC book stood at $1.7bn as of Q3’21 and included several other smaller litigations besides BoA (Nomura with around $230m in damages and First Franklin with hundreds of millions in damages). These non-BoA cases are at earlier stages and noone is yet talking about the trial dates. We can assume out of the total $1.7bn litigation assets on AMBC book, $1.2bn are reserved for the BoA litigation. This number itself is just an accounting entry (discounted estimate of the litigation value made by the lawyers and includes various risks like the case not making it to court, trial loss, lower than expected damages award, etc.) and likely does not represent the true expected litigation awards from BoA case.

Interest accrual rate – the claims are also accruing interest throughout the whole period from occurrence to prejudgement. At the moment it’s not clear what rate will be used to calculate actual interest accrual and this will also need to be decided by the judge or in the settlement. The details on which rate sould prevail are quite complicated and there does not seem to be a unanimous expectation among investors. The AMBV/BoA litigation takes place in NY, where, the statutory rate was simple 9% (not compounded). Initially, it seemed that the statutory rate is the one that will be used in this suit, however, later on, as AMBC lost the fraud claims to Bank of America (which would have allowed AMBC to pursue punitive damages as well), the contractual rate appears to be more likely to be used. While the base expectation is for the contractual rate o be at 8%, court documentation suggests that it might end up being materially lower at 5% (see comments #49 – #54 on VIC). Nonetheless, because this litigation has taken so much time, both 5% and 9% accrual rate scenarios add very substantial upside. The contract breaches occurred in 2008 or earlier – 14 years and counting till today. This turns $2.8bn in initial damages from BoA suit to $5.3bn (in case of 5% rate) and $6.3bn (in case of 9% rate).

Use of proceeds – the award/settlement proceeds would land at the AMBC subsidiaries and not at the hold-co . So the whole argument, that AMBC holdco is non-recourse to subsidiary level operations, kind of loses its validity when thinking about the shareholder value creation from the litigation proceeds. Most likely the funds received will be used to at least partially reduce leverage and make further acquisitions as management is trying to transform the business. As we understand, management has no clear plans of returning cash to shareholders. Thus in the base case, the majority of the litigation award/settlement proceeds will simply accrue to AMBC book value rather than actually land in the pockets of shareholders. Historically, similar businesses traded at 0.65x BV, suggesting that out of the potential $2bn book value gain impact on AMBC market cap would be $1.2bn, which is still a +200% upside the current prices. That might be overly optimistic and at least a part of litigation win in excess of the litigation asset balance sheet value is probably already incorporated in today’s market price.

Aside from these points, there are several other caveats and uncertainties involved, which obviously also explain the visual “cheapness” of AMBC relative to the potential settlement award:

• The trial might get delayed again. This has happened several times already due to countless appeals and process dragging by BoA. However, the last two delays (Feb’20 and Feb’21) happened due to COVID. In fact, the last adjournment was initiated by AMBC itself as it pushed for an in-person trial instead of a Zoom call. However, we’re not exactly sure if BoA won’t be able to come up with something again to adjourn the trial.
• There might be no settlement and the whole thing will go to court instead. Given the historical results of such litigations, it seems that AMBC should win. However, the fact that BoA has been dragging this one case for so long while at the ame time settling with multiple AMBC peers before (see more below) gives a bit of an uneasy feeling. If everyone before has settled and the chance of winning is that small, why would BoA drag this, especially when the damages are accruing billions in interest?
• Aside from the contract-breach claims, AMBC tried to go for fraud claims on BoA as well, which would have allowed AMBC to pursue punitive damages (potentially >100% of contract-breach claim damages). However, these fraud claims and an appeal afterward were dismissed. Moreover, by losing that fraud claims case, AMBC also lost the jury decision option, meaning that this will be a bench court and the final decision will be made by the judge. Jury court was expected to be more advantageous for AMBC due to the emotional appeal of the GFC crisis, bank frauds, etc.
• We haven’t found much on the discussion of the attorney fees in case of a large award/settlement. It’s likely that ongoing fees have already been covered by AMBC, whereas the final win fee is expected to be insignificant relative to the total settlement amount. VIC author estimates $50m.
• Not sure whether/how negative this is, but the judge for this legal case has recently changed. Apparently, the previous judge went to private sector practice last year. The new judge has stalled the process for many months so far for unknown reasons (bureaucracy) before setting the final trial date to the 7th of Sep’21. The delays might also be covid related.
• Downside protection is difficult to assess. The operating business is difficult to value (black box of monoline insurer). AMBC has exited bankruptcy several years ago and is now trying to transform the business/move into other insurance types (specialty, etc.). Over the years, the company has generally been profitable. However, it still has about $29bn of net par exposure to the legacy portfolio. $6bn+ of that was accounted as adversely classified (very high risk) – a substantial amount relative to AMBC book value. Management has done a good job of gradually reducing the total net par exposure (used to be $60bn in 2017), but obviously, any kind of unexpected negative development could make a huge impact on the equity value. The positive thing is that the operating business of the company is held on at a subsidiary level and the debt is non-recourse (as mentioned above, ‘non-resource’ is kind of fades when thinking about litigation award).  Meanwhile, the parent company holds around $6/share of cash and net assets, so in the worst-case scenario, the share price should stay above that level at the very least – unless management makes more acquisitions/investments and spends that cash in the meantime.

Historical litigations

• AGO vs BoA in 2011 – $1.6bn settlement. AGO marked a $334m book value gain. Shares went up 24% after the settlement.
• Syncora vs BoA in 2012 – $375m settlement.
• AGO vs JPM in 2013.
• MBI vs BoA in 2013 – $1.7bn settlement. $200m book value gain. Shares went up 45% after the announcement.
• AGO vs UBS in 2014 0 $358m settlement.
• FGIC vs BoA in 2014 – $3.6bn settlement.
• Syncora vs JPM in 2014 – $400m settlement.
• AGO vs Credit Suisse in 2014
• AMBC vs  JPM in 2016 – $995m settlement. $272m BV gain.
• Syncora vs Greenpoint in 2018 – $335m settlement. $285m BV gain. Shares jumped 40% after the settlement.