GAAP book value seriously understates the value of bond insurance companies Ambac (ABK) and MBIA (MBI). Assets are understated because the present value of future installments and unearned premiums is not included: also, liabilities are overstated due to mark to market losses on insurance protection written in CDS form. The amounts involved are substantial: and Ambac's recent settlement of the AA Bespoke transaction provides an early indication of the implications for investment decision-making.
For years, ABK and MBI have published NonGAAP “adjusted book values” in their quarterly supplements, which are available on their websites. They reconcile these figures to GAAP, and the following table displays numbers from 3/31/2008:
click to enlarge images
When presenting their 1Q 2008 results, MBI CEO Jay Brown proposed a new NonGAAP metric, which he calls “analytical adjusted book value.” This adds back mark to market losses on insurance protection written in CDS form to the extent they exceed impairments as estimated by the company. Starting with adjusted book value and adding back cumulative mark to market losses less cumulative impairments, he develops an analytical adjusted book value of 43.63.
Here is Moody's take on this line of thinking:
In public disclosures, the guarantors have indicated they believe expected actual credit losses on the insured CDS portfolios well be materially lower than implied by estimated market values ascribed. This view is broadly consistent with the analysis performed by Moody's in our modeling of the guarantors' portfolios, as disclosed in our research.
To extend the above table, the MBIA figure are from their presentation, the ABK figures are my calculations as noted :
Ambac's settlement with the sole counterparty to its AA Bespoke transaction provides the first hard evidence on the extent to which mark to market losses overstate actual losses. ABK had 1 billion of mark to market on the transaction, vs. a 789 million impairment. They settled by paying 850 million, so 62 million of the 211 million difference between mark to market and impairment was realized. That is roughly 30%, so 70% of the mark to market was in excess of actual losses. It is hazardous to extrapolate from one settlement, but certainly this is an encouraging sign that mark to market losses on insurance provided in CDS form may overstate reality for ABK and MBI.
Merrill Lynch, when they announced their settlement with SCA and the sale of CDOs at 22 cents on the dollar, mentioned that they are also negotiating with MBIA in an effort to secure a settlement with them. If and when such a settlement occurs, it will provide another look at the accuracy of mark to market losses as compared to impairments recognized by bond insurers. The SCA settlement was for 500 million of insurance proceeds on 3.6 billion net par outstanding. Because SCA was on the brink of insolvency and Eric Dinallo, NY Superintendent of Insurance was involved, there is probably a substantial discount since Merrill really had no alternative but to take what they could get before the window of opportunity closed.
Another implication of ABK's settlement arises from the rating agency stress case scenarios which are used to evaluate bond insurers' capital cushions. The 850 million was less that the agencies' stress cases, perhaps by a substantial amount. The Net Par outstanding on the AA Bespoke case was 1.4 billion. Noting this was a CDO squared containing mezzanine sub-prime RMBS collateral, very possibly a stress scenario would consider it a total loss. The size of the possible increase in capital cushion is very important to ABK's ratings and its ability to re-enter the municipal bond insurance market through a triple A rate subsidiary. Obviously it simplifies Dinallo's thinking in that he no longer has to weigh the huge but remotely possible worst case losses on the AA Bespoke transaction.
Market response to ABK's settlement announcement was dramatic: the stock was up 50% for the day, and MBI was up 30%. Options activity was florid, with calls in heavy demand. A commentator quoted by the WSJ attributed it to short-sellers buying protection, “There's a realization that the stocks are not going to zero.”
Both companies are scheduled to report second quarter results this week.
I will be watching the results closely – my focus will be on the components of adjusted book value and/or analytical adjusted book value – especially any information I can garner from the conference calls on the extent to which mark to market losses may revert over the next 6-9 months. Another area of interest would be the success of remediation efforts. Both companies have suggested that they may be able to assert various rights and remedies to reduce losses. Finally, if any more settlements such as ABK made on the AA Bespoke transaction occur, the effect on share prices could be dramatic.
I think both companies have the potential to trade at a share price somewhere between the adjusted book value and the analytical adjusted book value. Possible catalysts would include:
- Buybacks. ABK's board recently announced a 50 million repurchase authorization and MBI had 374 million remaining on its authorization as of 3/31/08. Since their ratings were downgraded, both companies have capital available to do the buybacks. The capital is no longer required to protect the triple A ratings. In point of fact, the ratings reduction was partly due to limitations on financial flexibility caused by low share prices: if buybacks were sufficient to increase the share price, it would add to their financial flexibility and bring them closer to the profile for a triple A rating.
- Stabilized earnings. Both companies continue to earn premium from long term business that is already on the books. My guess is that they will need to recognize somewhere between 700 million and 1 billion each of additional losses over the next few quarters. However; that is much less than the accumulated mark to market losses to date. Eventually, the mark to market losses will reverse and they will report substantial profits.
- Settlements or remediation. When Merrill Lynch announced its settlement with SCA, the company mentioned that they were also negotiating with MBIA. Because solvency is not an issue with MBIA, logically the settlement would be approximately the same as MBIA's estimated loss as reflected in their financials. Otherwise, MBIA would simply pay principal and interest when due – their obligations cannot be accelerated without their agreement. Accordingly, any settlement would be less than the mark to market and would improve reported earnings on a GAAP basis. Also, the settlement would be less than rating agency stress case assumptions, and would improve the size of the capital cushion.
- Short covering. These stocks are not going to go to zero. At recent prices, short sellers have little to gain, and much to lose – remember, adjusted book value is approximately four times as high as the present share prices. If you add back some portion of the mark to market, the undervaluation is even more extreme.
- Successful establishment of triple A subsidiaries to resume writing municipal business. Ambac has applied to Dinallo for permission to use 850 million of excess capital for this purpose. MBIA is also working with Dinallo on this solution. Moody's reportedly is willing to consider the triple A rating under these circumstances.
Disclosure: Author is long shares of MBI, long shares and calls of ABK