No, no, nobody asked for my opinion; rather, I am referring to the Financial Accounting Standards Board and their lovely pronouncements of the last several years. In short, they have shown the world which tools not to use in the valuation process while making it more painstaking to extract crucial figures.
The crucial figures are those that indicate how much cash security holders will get and when. Yes, the discounted cash flow, "when do I get the skrilla?", model is the mother of all valuation. Actually building precise models is mostly an exercise in futility when the subject is something as volatile as a bankrupt financial guaranty holding company with a subsidiary in rehabilitation. A multitude of ultra-sensitive and constantly changing assumptions overwhelms conventional precision. Even a single variable can confound a model. For example, assigning a single discounted value to the possibility of permanent impairment by actions of Ambac's rehabilitator is impractical.
A more intuitive approach would be assigning probabilities to a few outcomes representing the real diaspora of possible future events. Said again for the statistician, take a sample representing the population. That is easier said than done when the outcome turns on the decisions of a few individuals.
Continuing with an example for Ambac Hold Co's Sr. Unsecured bondholders, one would then need to apply the same process for other major unstable variables - in this case repurchase litigation, bankruptcy, and loss developments. Better to dump the whole thing into black-scholes and go for a coffee. The inherent difficulty of discounting such components tends to elicit a desire to pack our bags, buy Microsoft, and extrapolate the Office Suite sales off into the sunset. But THAT desire to walk away is precisely what ought to draw at least an initial interest.
Catastrophic past losses and the possibility for total loss should also incite interest along with natural and prudent alarm. Having established interest still leaves us a long way from an investment decision. Lets take a step back before we move forward. Lets take capital structure out of the equation.
Without the firm's own securities, the main components of an economic balance sheet are investments and actuarial accounts. Specifically that is investment portfolio, net expected recovery assets, reinsurance, and premiums receivable on the left and net expected case loss reserves on the right. All of the numbers we will use will continue to be Ambac's and for these they are:
Investments+Subrogation Recoverable+Reinsurance Recoverable-Loss Reserves=Net Major Assets
With this number we will keep in the back of our minds that we must: 1) adjust our valuation for risks in the insured portfolio; 2) adjust our valuation down for the present value of costs; and 3) adjust our present value up to account for the present value of installment premiums. Due to the economy of time I spend on the couch doing this, we do not bother with the latter two numbers as they are overshadowed by the forces of loss and recovery developments. As for the first item, we note that as a Sr. Unsecured bond holder downside is limited to zero and our view is that recovery potential skews insured portfolio risk to the upside.
Now lets approach capital structure. Here, we all conservatism to reign and so value all liabilities at par though market value may be as little as 15% of that. The predicament of Ram Holdings has proved this method relevant by certain large preferred shareholders being unwilling to tender at deep discounts. Back to the perspective of a Sr. Unsecured Ambac creditor, liabilities valued in this way make for a negative $2,850M ($2.05B AAC surplus notes + $800M AAC Pref. Shares) adjustment bringing our earlier $3,773.9M number to $923.9M. That is approximately 3/4 of Sr. Unsecured bonds.
This process has taken us past three of the unstable variables we identified. We bypassed repurchase litigation and other loss reserve developments by taking the balance sheet estimates at face value. We dealt with bankruptcy by taking a value of the assets below the reorganizing entity. We have given conservative valuations of subsidiary liabilities. This process has not had any effect on the actual uncertainty of the underlying value. What we have done is found a number to start working with.
Another way - which we will not go through here - to arrive at a preliminary number would be to start with the rehabilitator's base case projections and adjust them for developments since 6/30/10. Using either method, one starts by relying on the assumptions of potentially biased constituents.
Whichever method is chosen, one garners a starting point from which she can apply adjustments. For instance, an increase in repurchase recoveries of $1.5B after bankruptcy would make Sr. Unsecureds 200% of par equivalent. If repurchase litigation hits the skids, any AFGC security is worthless. What are the chance of the rehabilitator permanently impairing AAC equity value?
The issues of repurchases and rehabilitator impairment are the two biggest factors we wrestle with. And while we have developed views on these points, it's sunny outside so we will leave that for another time. So lets get to an ending thought before the cane hooks us off the stage.
As we mentioned earlier Ambac is a far cry from the blue chips. In approaching such a complex situation, one cannot simply model out revenues and expenses into the horizon. Rather, three basic steps should be taken: 1) identify the major risks, 2) get a starting number (or several) that deals with those risks, and 3) apply your views of those risks to the valuation and see what happens. Total loss is a viable outcome but not much more - if at all - likely than a parabolic upswing in price.
We invest in such a way that makes a zero OK. The culture and/or mandate of many professional investment shops precludes taking such risks. Most individuals simply won't approach or don't know how to approach such risks. We postulate that is why distressed investing can be so rewarding.