Sunday, December 16, 2007 6 comments ++[ CLICK TO COMMENT ]++

Ambac Margin of Safety Analysis

This post will capture the final piece of analysis I'll do before purchasing Ambac (still depends on how events unfold over the next few weeks). I initially evaluated Ambac a few months ago so read that to get details on Ambac's historical profitability and what it is capable of. The initial analysis, which occurred after the 1st sell-off in the stock but before the massive 2nd sell-off, didn't really go into valuation or risk. It quickly became obvious that valuation isn't the problem with Ambac (it's undervalued by almost any measure); instead, the real issue is risk.

Margin of Safety

Before I say anything, I should note that Ambac has a chance of going to zero! As Mark Sellers (don't know who he is but just saw the interview at The Motley Fool), says in this interview, a company like MBIA can go bankrupt so he doesn't invest in it. So if you can't afford a total loss, none of the bond insurers are for you. If you want a beaten down financial that won't result in a total loss, something like C, BAC, MER, etc should be investigated. I know Warren Buffett's rule #1 is never to lose money but I'm willing to face the possibility of a total loss (note that even Benjamin Graham has implied that adverse events can bankrupt anyone).

shares outstanding: 101.55 million

market price: $22.81 ($2.32 billion)
book value per share: $55.64 ($5.65 billion)
adjusted book value per share: $88.07 ($8.94 billion)

2006 earnings (CASE I): $876 million
Pre-2004 5yr median earnings (CASE II): $433 million


  1. Most numbers from Yahoo Finance as of Dec 16 2007; balance sheet data from end of Q3

  2. Adjusted book value is "stockholders’ equity (book value) and adding or subtracting the after-tax value of: the net unearned premium reserve; deferred acquisition costs; the present value of estimated net future installment premiums (discounted at 5.6% and 5.4% at September 30, 2007 and December 31, 2006, respectively); and the unrealized gain or loss on investment agreement liabilities." (Ambac 3rd quarter 2007 quarterly operating supplement) This is basically what the company is truly worth if it doesn't go bankrupt in an optimistic scenario.

  3. For earnings, I'm picking a superbullish case (2006 earnings) and a reasonable case (pre-2004 5yr median). The 2006 number is likely peak earnings for the time being so I would not rely on that. I expect income from structured products to drop off significantly over the next few years (US CDO market is toast). Wit low income from structured products, Ambac's earnings would be similar to the $433 million it earned in 2001 and 2002.

The way I'm going to look at Ambac is treat the current crisis it is facing with RMBS and CDO losses as a one-time issue. From an investment point of view I think it is fair to treat it as a one-time issue. In reality, however, the losses will be spread out over time, with likely large losses in 2007 and early 2008 with losses declining in 2008 and 2009.

First I'll ignore the one-time loss and see what the company would be worth if it lost a big chunk of its structured products business. Pretend that this was 2000, before the big run-up in CDOs. If Ambac can handle the losses it will incur over the next few years, this is the value of Ambac.

After that, I'll look at how much loss is being priced in by the market.

Earnings Valuation

A very rough guide (ignoring company specifics, industry valuations, interest rates, etc) to picking an earnings multiple is to use the following Benjamin Graham formula for growth stocks:

value = current_(normal)_earnings x [8.5 + 2 x expected_annual_growth_rate]

For a zero growth company, the P/E ratio should be 8.5. Let's assume Ambac has zero growth. If we use Case I (2006 earnings), then Ambac's P/E is 2.64; with case II, it is 5.34. Or another way of looking at it, if the P/E ratio should be 8.5, then Case I yields a market cap of $7.4 billion and Case II yields market cap of $3.4 billion. You can clearly see that Ambac is cheap.

Case II implies around 50% appreciation potential from current prices. This means that, ignoring the current one-time loss issue, if the company earned what it did 7 years ago then it is 50% undervalued. The appreciation potential is higher if you use more optimistic scenarios. For instance, if you think earnings won't drop off to 2000 levels then the potential is even higher. Furthermore, if you think earnings can grow (instead of being flat) then the valuation would be even higher. But I like to be pessimistic so my view is that the potential is around 60% if earnings were back to the 2000 levels before the recent CDO boom.

Book Value Valuation

The stock is currently trading 60% below book value and 75% below adjusted book value (non-GAAP measure). I'm just a newbie but I think book value and adjusted book value are good indicators of the true value of Ambac. Unlike other types of companies, there isn't inventory, plant & equipment, intangibles like brands or patents, or other assets that can overstate the book value. With an insurance company, what you see is sort of the value of the firm. What is unknown is the potential for losses on the insurance that was written. Ignoring the RMBS and CDO of MBS insurance (loss on these are clearly not accounted for and I'll look at that below), I would imagine that management took proper loss reserves for the rest of the underwritten insurance. So looking at the market price should tell us what losses are expected by the market.

To sum up my thinking... I am assuming that the stock should trade at exactly equal to book value or adjusted book value (historically it has always traded above book but we can't be sure that the market will give it the same multiple in the future (highly unlikely)).

What the stock price says is that the market is pricing in $3.33 billion in losses from a book value point of view; and $6.62 billion in losses from an adjusted book value point of view.

Let's assume a reasonably pessimistic bear case generally has Ambac losing $4 billion. Note that a superbear case can be made for bankruptcy, but everything I read likely implies that is unlikely. Things will have to be extremely severe and/or Ambac's underwriting has to be truly incompetent.

Based on the numbers above, I believe the market is discounting something close to the worst case of $4 billion. If you go with the book value number and a $4 billion loss (meaning short by $600 million), then the stock can drop even more (-30%). But if you go with the adjusted book value then the stock can go up (+113%).

Of course, if the loss is less than $4 billion then the numbers will look even better. Or if the losses are spread out over time (meaning present value of losses is slightly lower) then the upside will be slightly better.

What all this means to me is that Ambac's stock is pricing in a loss approximating the bearish estimates (but not the superbearish bankruptcy scenario). Another way of looking at it is that the market is pricing in a lot of shareholder dilution so the price is unlikely to drop with further dilution (this is what happened with MBIA, where the stock actually jumped even though they announced a $1 billion dilutive deal).

Do note that we don't know what the actual loss will be. If it is, say, $7 billion as some superbears suggest, then even the current adjusted book value is not pricing in any of that.

To Sum Up...

If you think the losses will be around $4 billion or less, the stock looks to be pricing in the losses... If you think that the company can survive the current crisis then the company should be worth at least $3.4 billion under reasonably pessimistic scenarios (eg. long term income of $400 million with zero growth rate).

Is Ambac Buffett's American Express in 1963?

I have a habit--possibly a bad one--of trying to fit present opportunities into something that successful investors like Warren Buffett encountered. I also have a habit of looking at Buffett's AMEX investment in 1963 but I think it offers a good template for contrarian investing. Ambac looks like American Express in 1963, when Buffett bought after the infamous Salad Oil scandal. I can be wrong but there seem to be some similarities.

In both cases, there were massive one-time losses (in the AMEX case, it wiped out shareholder equity to zero) and some thought bankruptcy was a possibility. AMEX's bankruptcy threat wasn't as severe as Ambac. AMEX only dropped around 40% whereas Ambac is down 75% from peak.

Buffett also looked at whether AMEX's future earnings were impaired. He came to the conclusion, after observing AMEX customers at a restaurant, that AMEX still retained its brand and customers didn't flee the company. Ambac's impact on customers is hard to decipher but some articles I have read seem to imply that the market is still using bond insurance (not necessarily from Ambac but that was before the rating agency reviews). One thing to note though is that earnings from structured products (big chunk of Ambac's earnings) will likely be lower in the future (but even that is uncertain because the rising spreads on risky assets mean that Ambac will earn higher premiums).

Buffett was also comforted by management honesty and, although I can't tell from the distance, Ambac's management seems to be clean. They did try to combat misleading statements in the press, as well as buying some stock with their own money.

I hope this turns into something like AMEX back then :)

Once the Tribune (TRB) takeover closes (likely December 20th) then I'll redeploy that capital into Ambac. My investment in Ambac will be very large (25%+ of portfolio) but my portfolio is small. I'm switching to focus investing and would ideally have 4 holdings of 25% or 5 @ 20% each. Ambac will likely determine a big chunk of my portfolio performance over the next 3 years. If the original investment thesis holds and if Ambac dilutes shareholders early next year, I'll likely add (I will look favourably upon any exchange-traded convertible bonds or warrants). I may always change my mind but that's my plan for the time being.

Tags: ,

6 Response to Ambac Margin of Safety Analysis

December 17, 2007 at 10:28 AM


Great post and analysis.
ABK did mention that they were lining up other capital measures (but did not specify). Maybe they are finally getting ahead of the curve, after languishing in denial for too many weeks.

What kind of entry point are you looking, or hoping for, if you were to take a position? 25% is a large piece. Do you leg in, or jump in full bore?

December 17, 2007 at 11:04 AM

From an article referring to the FBR upgrade today --

In a note titled Capital Demands Appear Manageable, the brokerage said while it did not rule out some form of equity raise in future, it did not expect the size of any potential capital raise to dilute book value for Ambac.

"Moody's affirmation indicates that expected losses and capital charges should be under $2 billion, which is ABK's excess capital above its minimum triple-A requirement by Moody's," analyst Steve Stelmach wrote in a note to clients.

Stelmach said his new rating was based on losses coming within his targeted $1 billion to $3 billion range, but added that this range was not a given, considering the uncertainty in the credit markets.

He also said his rating assumed that Ambac would attract sufficient capital to fill the void left by the losses.

Stelmach has a price target of $35 on the stock.

December 17, 2007 at 2:10 PM

What I think we need to do in buying these companies is to have a longer view of things. For example, in the longer term, I just don't think that CDO will no longer be part of the business for monolines. Companies like CFC are generating large number of bonds to be invested by investers. Somebody will want to insure them. No doubt everyone would be a lot more careful about risk evaluation and so on.
Further, sometimes when one stares at something up close, they tend to get whipsawed by recent events that may or may not come to be. AccruedInterest, for example, is one of the most knowledgeable guy I have run into in this business. If you look at his blog about the monolines, and for ABK in particular, his position has changed several times in the span of a couple of weeks. Naturally, underlying conditions don't change that much.

Same as we don't look just in front of us when we drive, we should focus on longer term and at the same time ensure that there is a good chance the company will get there from here.


December 17, 2007 at 5:14 PM


My portfolio is very small compared to most of you so to minimize commissions, I buy all at once. So 25% to me is probably like 5% for most of you.



I think the CDO market probably won't recover anywhere near what it is right now. It is possible that the CDO structure is just flawed and has no useful purpose (eg. if investors can't tell what's in the CDO, I don't know why one would invest). If anything, CDO-squared is pretty much gone IMO.

Having said that, there will be other structured products that will offer opportunities for Ambac. I'm just not so sure about pooling together debt and assuming it becomes less risky.

December 18, 2007 at 12:39 PM


Still, companies like CFC and other banks that originate loans will not hold on to all their portfolio and will sell them in some form to investors. Pooling together of debts does not make it less risky, but does turn it into blocks that could be bought by the average debt buyers who buy them in large blocks.

Perhapse in the future, the flight by night loan originators will not be able to sell their loans and only big companies like major banks who can stand by their product will be able to sell them. I still see a day when mortgage debt will be invested by debt investors. The market will find a way.


December 18, 2007 at 4:25 PM

NEADERTHAL: "Pooling together of debts does not make it less risky, but does turn it into blocks that could be bought by the average debt buyers who buy them in large blocks."

The problem I see is that, doing that causes lack of transparency. If someone bought a straight bond or some sort of a debt obligation, they sort of know what it is. But when you pool things, no one really knows what the mixture in the pool is.

Post a Comment