Ambac is Effectively Insolvent & Will See More than $8 Billion of Losses with Just a $2.26 Billion Market Cap

I came to this conclusion after a detailed analysis of Ambac's portfolio (at least what Ambac has made public, which was sufficient) covering exposure in the Structured Finance, Sub-prime RMBS and the Consumer Finance business. Ambac's management was forthcoming enough to publish a portion of their insured portfolio which allowed me to review each structure.

I am short Ambac and MBIA (for whom I have also released research), so be aware of my position as I present this opinion. I profit not necessarily from whether ABK can continue as an ongoing concern (which is in doubt and wouldn't hurt my shorts to say the least), nor from an infusion of capital (whether it be debt or equity, either of which would be a poor investment from my perspective) but from the significant decline in value of the existing shares in which I have taken a bearish position. To determine my short position, I calculated relative nominal book valuations, actual economic book valuations and produced standard financial forecasts. Of interest is the loss tail analysis wherein I have estimated the present value of the future losses.

As it stands now, ABK's equity value will be totally wiped out with a 175 basis point move in their insured's underlying – which seems like a very, very likely possibility. My Ambac analysis is much more granular than that of MBIA's (see A Super Scary Halloween Tale of 104 Basis Points Pt I & II, by Reggie Middleton ) where I discussed the predicament of the ratings agencies, the monolines, and in particular, MBIA.

The referenced predicament is exacerbated by the fact that some, such as Ambac, are truly insolvent, thus a mere waiving of the "Magic Ratings Wand" will not pay the claims when they come due. More to the point, the monolines have grown too big for their capital base, specifically their equity base. They are insuring much more than they can handle in the case of an outlier event. I don't consider the burst real estate bubble and the consequent mortgage debacle much of an outlier, for anyone could have seen it coming if they simply opened their eyes.

Now, if a big monoline fails or is even downgraded, a large part of the credit market goes with it.  This level of catastrophe may be too much for the powers that be. This portends a bailout of some sort or fashion, or maybe the players will just be forced to take their market medicine. Only the future will tell.

Six Degrees of Separation: Guess who Ambac insures!

Bank of America issued a report on the monoline insurers on July 30^th, 2007 that states that ABK's RMBS exposure to troubled companies is limited to only 4 cos. with vintages primarily in the early years excluding two relatively well performing underwritings. Despite this, they failed to include in this caveat the consumer finance insureds:

• Countrywide, which probably has one of the worst performing portfolios in the industry;
• GMAC, who has also suffered significant losses that GM has been forced to cover, hence hampering a clean sale of the company;
• Indymac, another company that is saddled with mortgage related losses that is on the insured's list (Indymac and Countrywide have had their shares more than halved in the last few months. I was short these companies. CFC may go bankrupt);
• Lehman brothers has some losses to contend with as well, but I don't know to what extent since I don't follow it - I do know that they are the 2nd largest MBS house on the street, next to Bear Stearns;
• Greenpoint Mortgage Funding is defunct, wound down due to losses;
• Then we also have Citimortgage (SIV king whose own mortgage portfolio is a mess);
• Accredited Mortgage Loan (bankrupt or close to it);
• Wachovia (just reported a billion plus writedown on mortgage assets);
• Countrywide Revolving Equity Trust/Alt-A trust (need I say more about undocumented 2nd lien loans from this lender);
• Option One Mortgage Trust (nearly defunct due to mortgage losse);
• BofA, mulit-billion dollar mortgage asset writedown;
• and Newcastle – who I believe is either out of business or close to it. I stopped following it some time ago.

These are the companies and exposure that I am familiar with, at first glance in the consumer finance portion of Ambac's portfolio, without any research. Just imagine if I took a real hard look at the insureds.

Now, using some common damn sense, would you think that the company that is insuring these guys' mortgage and finance products with 90x leverage may be having some problems that they may not be coming forward with. I have over 100 pages of proprietary analysis and calculations costing me weeks of analyst hours, that tell me Ambac may be out of business soon - but I really didn't need to do all of that math and research if I just glanced at the bullet list above. I used the loss statistics from the BofA report as a baseline for the losses in my models on Ambac. I know they are too conservative (and to be fair to BofA, they were contrived before this mess got worse), but that should only lend credibility to my findings. Click here to download ambac loss tail.pdf.

The ACTUAL quality of the ABK's insureds is truly suspect in my opinion and the underwriting quality of their insureds needs to be investigated further. Unfortunately, I have very limited resources. I literally told my team that "this is worth digging in and spending time on, for there are many who are now trying to go long on this stock due to its price and nominal book valuation. If they are wrong, it can be a very profitable opportunity". Well, that's what we did. By investigating the losses on similar books written by the originators for the vintage in question, one can guess the performance of the books underwritten by AMBAC. The policy terms must be examined to see where the breakpoints are for losses, of course. Exposure to Countrywide alone is a cause for suspicion, IMO. As stated earlier, the default estimates in the B of A analysis are assuredly too conservative, but are used for the sake of prudence over alarmism (with some mandatory tweaks to edge them towards reality). Why do I say they are conservative??? Take a look at the REO rates and land value forecasts in my blog, and then look at the target prices for the insurers in question on the first page of B of A's analysis, right before you query the prices of these stocks today. For those that don't have access to the report, I will reveal just this one tiny part:

Bank of America Top Picks (June 2007)

Ticker	Rating	Price	Target	Price as of 11/29/07	Profit on the BofA Call	% Profit
SCA	B	$23.60	$37.00	$6.69	($16.91)	(71.65%)
MBI	B	$60.33	$85.00	$30.04	($30.29)	(50.21%)
Least Favorites
NONE

You really can't get rich listening to these guys. Hopefully, you can see where the use of their default data is a conservative approach (even a bit rosy), albeit tweaked ever so slightly for the sake of reality. As you may have ascertained, I do not put a lot of faith in sell side research. I have even less faith in the big three rating agencies research (although Fitch is trying to be taken seriously). Thus, even if they deem ABK and MBIA not in need of more capital, that is near meaningless in my book. These are the same companies that rated the insured portfolios AAA a year or two ago that are now taking up to 20%+ losses.

We also have to contend with the moral hazard/bailout issue. If you read my earlier missive on MBIA, I detailed the rating agencies' dilemma.

The calculations in this analysis are only estimated losses in 4 insured categories (of many, they are enough to generate significant losses). I am expecting higher losses in Public Finance as well due to the loss of property tax revenues (lower tax base) and income tax revenues led by housing value declines and loss of corporate revenue and jobs, respectively. Many municipalities created huge budgets during bubble times (like everyone else) and failed to prepare for the bubble to burst. Now unfunded services run rampant. The shortfall will have to be covered somewhere, and default on debt service is not out of the question.

In the base case scenario created, we expect the company to report losses to the tune of $8 billion+ in its Structured Finance, Subprime RMBS and the Consumer Finance portfolio. This loss will wipe out the company's remaining equity and it will need to raise an additional $2 billion in order to function as an ongoing concern. Moreover, we think the company will need to reinsure a higher percentage of its portfolio in order to transfer risk and free up capital.

"The Truth! The Truth! You can't handle the Truth!"

I calculate that Ambac will need to raise an additional $2 billion in order to continue as a going concern. In order to maintain AAA status they will need $5.4 to $7 billion, according to how I perceived the comments of its CEO in the last conference call (they say they are an average of $1.4 billion above what is needed to maintain a AAA status from the three main rating agencies – without my little economic reality marking here). In the base case scenario below, Ambac will need to bolster its reserves by $6.8 billion. A fellow blogger that I follow, Mike Shedlock, commented that Citibank has recently sold approximately 5% of itself to a foreign investor to raise $7.5 billion dollars. Citibank is much more diversified, with a much larger capital base, than Ambac. Let's be realistic here – no let's not - Let's be highly optimistic with pretty rose colored glasses, and say that ABK can fetch a significant premium to Citibank's valuation. ABK's current market valuation is $2.26 billion. Where in the world will they get this kind of capital from and who will be the risk cowboy to give it to them??? These guys are in a real solvency dilemma, and it is a shame that the ratings agencies and the sell side guys have yet to admit it. I guess it takes entrepreneurial investors and bloggers such as me to ferret out the truth, and the truth is hard to find in detail. You remember what Jack Nicholson said in "A Few Good Men"? "The Truth! The Truth! You can't handle the Truth!" I had two analysts and I work on MBIA and ABK for weeks, when I should have been able to just buy a report… Yet, everyone expept Ackman from Pershing Square was unrealistically optimistic. There are some big losses ahead of us folks. If the real estate bust was the impetus for the current debacle, we have a long trip ahead of us because the real estate bust has just started!

My full analysis is bulky, but well documented, and will be posted as a .pdf if I get enough requests. Most should be satisfied with this lengthy summary.

Here we have done a loss tail analysis of the forecasted losses of the Structured Finance, Direct RMBS and Consumer Finance portfolio, expecting the losses of the vintage year 2005 to be paid over the next 5 years in 2006-2010. We have calcluated the loss ratio of the company which is deteriorating from 2007 onwards (denoted by Paid losses/Written premium ratio).
Base Case Analysis				Calendar year payout
Year	Gross written premium	Expense ratio	Total Expected losses	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012
2003	1,144	172	972	22
2004	1,048	157	891		61
2005	1,096	164	932			200
2006	997	150	847				504
2007	1,006	151	855					1,260
2008	766	115	651						1,928
2009	690	103	586							1,880
2010	635	95	539								1,741
2011	597	89	507									1,437
2012	561	84	477										671
		Calendar year paid losses		22	61	200	504	1,260	1,928	1,880	1,741	1,437	671
		Cumulative losses		22	83	283	787	2,047	3,975	5,855	7,596	9,033	9,704
		Report year written premium		1,144	1,048	1,096	997	1,006	766	690	635	597	561
		Paid Losses/Writtem Premium ratio		2%	6%	18%	51%	125%	252%	273%	274%	241%	120%
		Outstanding loss reserves		950	1,780	2,512	2,856	2,451	1,174	(120)	(1,321)	(2,251)	(2,446)

Alternatively, we have calculated the provisioning for losses that Ambac will need to make every year on the basis of the anticipated losses that the company will have to pay in coming years. In doing so we have assumed that the 85% of the premium written from 2007 onwards (excluding 15% as underwrting expesnse) will be transferred to the loss expense reserve every year. The loss reserve uptill 2007 is taken from comapny's balance sheet. The losses have been calculated on the basis of various default probabilities assummed in Strucutred Finance, Direct Subprime RMBS and Consumer Finance portfolios. We have assumed a duration of 5 years to spread the losses on various vintages over the coming years. We anticipate the company will have to create a provisoin of $ 6.8 billion under the base case scenario.
Base Case Analysis				Calendar year payout
Year	Gross written premium	Loss and loss expense reserve		2003	2004	2005	2006	2007	2008	2009	2010	2011	2012
2003	1,144	189		22	-	-	-	-	-	-	-	-	-
2004	1,048	254		-	61	-	-	-	-	-	-	-	-
2005	1,096	304		-	-	200	-	-	-	-	-	-	-
2006	997	220		-	-	-	504	-	-	-	-	-	-
2007	1,006	279		-	-	-	-	1,260	-	-	-	-	-
2008	766	930		-	-	-	-	-	1,928	-	-	-	-
2009	690	1,517		-	-	-	-	-	-	1,880	-	-	-
2010	635	2,056		-	-	-	-	-	-	-	1,741	-	-
2011	597	2,563		-	-	-	-	-	-	-	-	1,437	-
2012	561	3,040		-	-	-	-	-	-	-	-	-	671
		Calendar year paid losses		22	61	200	504	1,260	1,928	1,880	1,741	1,437	671
		Cumulative losses		22	83	283	787	2,047	3,975	5,855	7,596	9,033	9,704
		Provision for losses			4	(150)	(588)	(1,202)	(1,276)	(1,294)	(1,202)	(930)	(195)
		Total	(6,832)

In our base case analysis of the CDO and the Subprime RMBS portfolio, we have assigned default probabilities based on collateral; wherein we have assumed an average default probability on its subprime collateral of 6% and on its ABS CDO mezzanine a default probability of 25%.

Average default probabilities (by Collateral)
Subprime RMBS	6%
Other RMBS	6%
ABS CDO High Grade	6%
ABS CDO Mezzanine	25%
CDO Other	10%
Other ABS	10%

In our base case analysis of the consumer finance business, we have assigned default probabilities largely based on ratings. We assigned a average default probability of 2% on its AAA rating portfolio and 11% average default probability on its BIG (Below Investment Grade) portfolio.

Average default probabilities (by Rating)
AAA	2%
AA	5%
A	6%
BBB	8%
BIG	11%

Valuation

In the case of Ambac, and most of my analyses, I draw a distinction between accounting (or nominal) book value and actual economic book value – the stuff I get paid for as an investor. Below you will see comparable valuation based upon nominal book value which actually has ABK underpriced. You will also see the forensically scrubbed economic book value, which in the most optimistic scenario (which I can tell you now, just ain't gonna happen) has Ambac valued at about $9 per share. You don't want to know what the base case and pessimistic scenario portend.

Ambac Financial Corp
Relative Valuation
Nominal Book Value			FY2007
All Figures in Millions of Dollars, unless othrerwise stated			Mean Multiple	High Multiple	Low Multiple
BVPS			53.67	53.67	53.67
Equity Value Per Share			$22.5	$34.4	$12.7
Current Stock Price			$21.8	$21.8	$21.8
(Discount)/Premium to Fair Market Value			(3.11%)	(36.60%)	70.93%
Book value as marked to market (Optimistic Scenario)
			FY2007
All Figures in Millions of Dollars, unless othrerwise stated			Mean Multiple	High Multiple	Low Multiple
BVPS			21.4	21.4	21.4
Equity Value Per Share			$8.97	$13.71	$5.09
Current Stock Price			$21.8	$21.8	$21.8
(Discount)/Premium to FMV			142.89%	58.93%	328.49%
Book value as marked to market (Base Case Scenario)
			FY2007
All Figures in Millions of Dollars, unless othrerwise stated			Mean Multiple	High Multiple	Low Multiple
BVPS			-14.0	-14.0	-14.0
Equity Value Per Share			($5.87)	($8.98)	($3.33)
Current Stock Price			$21.8	$21.8	$21.8
(Discount)/Premium to FMV			(470.93%)	(342.71%)	(754.38%)
Peers
Name	Ticker	P/B '07	Price	BVPS '07
MBIA Financial	MBI	0.38	22.3	58.5
Assured Guaranty		AGO	0.64	20.17
The PMI Group	PMI	0.25	10.45	42.43
Primus Guaranty		PRS	0.59	5.91
Security Capital Assurance Ltd		SCA	0.24	5.32

Price to Book Value

Average		0.42
High		0.64
Low		0.24

The Effects of Adverse Spread Movement

An Increase in spread of 175 Bps would erode the entire equity

Residential Mortgage Back Security and CDO Exposure

Here you see Ambac has significant exposure to some of the worst vintage years, and as detailed above has some of the worst possible clients one would want ensure. These ingredients mix to become a very toxic cocktail, indeed.

	AMBAC
Total subprime exposure with in insured portfolio
Total MBS portfolio	53.9
RMBS subprime exposure	8.8
% of total RMBS portfolio	16.3%
Sub prime porfolio by vintage
vintage 1998-2001	1.2	13.6%
vintage 2002	1.2	13.6%
vintage 2003	2.4	27.3%
vintage 2004	0.8	9.1%
vintage 2005	1.6	18.2%
vintage 2006	1	11.4%
vintage 2007	0.6	6.8%
Direct Subprime RMBS	8.8	100.0%
36.4% of the subprime portfolio belongs to vintage years of 2006-2007 when credit writing standards has been on its low.
Total CDO portfolio (in US$bn)
High yield	24.3	34.0%
Investment grade	8.6	12.0%
ABS > 25% MBS	29.2	40.8%
ABS < 25% MBS	3	4.2%
Other	2.80	3.9%
Market value CDOs	3.60	5.0%
	71.5	100.0%
Breakdown of CDO of ABS's subprime collateral by rating	2Q 07	3Q 07
AAA	3.8%	7.4%
AA	39.7%	39.0%
A	47.2%	36.9%
BBB	8.6%	8.7%
Below investment grade	0.7%	8.0%

<>

<td valign="bottom" style="BORDER-RIGHT: m

Sensitivity Analysis - Default probabilities - Base case
	Vintage	Sub-prime RMBS	Other RMBS	ABS CDO High grade	ABS CDO Mezzanine	CDO other	Other ABS
	1998	2%									Average defualt probabilities (by Collateral)
	1999	2%									Subprime RMBS	6%
	2000	2%									Other RMBS	6%
	2001	2%									ABS CDO High Grade	6%
	2002	5%									ABS CDO Mezzanine	25%
	2003	5%									CDO Other	10%
	2004	8%	5%	5%	15%	10%	10%				Other ABS	10%
	2005	8%	5%	5%	15%	10%	10%
	2006	15%	8%	8%	35%	10%	10%
	2007	15%	8%	8%	35%	10%	10%
Sensitivity Analysis - Deafulat probabilities - Worst case
	Vintage	Sub-prime RMBS	Other RMBS	ABS CDO High grade	ABS CDO Mezzanine	CDO other	Other ABS
	1998	5%
	1999	5%
	2000	5%
	2001	5%
	2002	10%
	2003	10%
	2004	20%	10%	10%	30%	15%	15%
	2005	20%	10%	10%	30%	15%	15%
	2006	30%	15%	15%	70%	15%	15%
	2007	30%	15%	15%	70%	15%	15%

Comments

[Rao]

Remember what happenned to Jack Nicholas in the movie. Blurting dialogues just does not cut it.

I agree with your analysis and that AMBAC is fraught with danger. You made any analysis of what is going to probably happen - a worst case scenario for the longs.

On the other hand, have you done any scenario study - a worst case scenario for the shorts, if it turns out that the rating agencies come out with favorable ratings, and in addition, Ambac comes out with more news that they have found XXX that is ready to infuse more than the needed cash? (As we see more and more capital infusion is coming into the market when it is needed - CITI got it, Etrade got it etc.)Even though the euphoria may be short lived while nobody will know for a couple of months at least how much truth is behind the rosy picture presented by the agencies and AMBAC, given that the shorts are huge in number/volume, the Fed cutting interest rates is more than 50%, and market events may help Ambac in the short run, how badly will the shorts get burned ??? Will this go beyond $35, towards $40, with all the help from the shorts of course. A couple of Hedge funds may go down under, given that they have betted against MBI too.

Life can be a bitch, even when you are right. I have a feeling, a suspicion that the shorts will get burned on this one.

[Andy Gaudet]

Your summary is a lot of mental masturbation with no factual analysis or credible results. You could have at least provided historical claims payout data by Ambac e.g. losses to date. Don't include any of your gloom and doom forecasts on mark to market which have no meaning anyway in yerms of Ambacs portfolio strngth and earnings capability. And yes, you are cut out of the same mold as Ackman.

[Thomas Davis]

I found it interesting that you mentioned a drop in Public Finance revenues due to a drop in housing prices in various municipalities. I thought these revenues were based on a home's assessed value, which is stable, and not the home's market value. It seems there would be no lossed in the area of Public Finance, just a leveling off.

Also, I thought that Ambac would be required to make interest payments over the life of the bonds that it insures, not over the 5 year periods that you mentioned.

Your analysis seems to indicate that Ambac would be responsible immediately for the value of the bond plus any interest payments. But the homes have underlying value and homes are still be sold, albeit at a slower rate that a year ago.

There will be some pain while the real estate market sobers up, but I think that Ambac will be around for a few more years.

BTW, I am long on the stock with a few put options to cover my downside over the next few months.

[Reggie]

This seems to be a controversial topic. Let's address the questions:

@ Thomas Davis
The assessed values are tied to how the municipality in question wishes to assess values. I live in NYC, where the assessed value was hiked three consecutive times until it reached the level of the market value in my tax statement. Many of these levels (marked to market, basically) either no longer stand as accurate or soon will not. There is also the matter of corporate tax and personal income in the wake of a recession. Taking NYC for example, the real estate and finance industies are just not going to add to the coffers like they use to. As a result, Mayor Bloomberg, a finance guy, made sharp cuts across the board. The same has happened in California and many other municipalities. Whether the cuts are sufficient remains to be seen.

Financial guarantee insurance is a short tail insurance policy with a relatively short payout, as opposed to mediacal malpractice liability which is a long tail casualty line. Ambac does have relatively short payout periods, although it is most likely dependant on exactly what is being insured. The 5 year duration is plausible as a generalization. I addition, it is not the bonds that will cause Ambac it's biggest headache. It is the structured products, which are taking significantly losses - some of which have manifested in the latest quarter.

As for the home having value, depending on the complexity of instrument (ex. structured products) there is a significant disconnect between what was the underying asset and the derivative vehicle. Even if there wasn't, the values that some of these homes where borrowed against are easily more then 100% of the debt outstanding, particularly in the west and southeast as time moves forward and the real asset market reverts to the mean.

[Reggie]

@Rao
I actually like that Nicholson quote :-)

I don't do analyses based on what benefits longs and shorts. I just analyse prospects and let the facts speak for themselves. I would just as easily be long ABK if I thought it was worth it. Alas, I don't believe that to be the case.

I did run three different scenarios though, and they are in the blog post - Optimistic, Base Case, and Pessimistic. Whether the rating agencies give ABK their blessings or not does little to alleviate their solvency issue in the present. They still have a book of business that has signiicant potential losses lined up. My scenarios indicate that ABK will need between $2billion and $16 billion dollars of capital to continue business. They currently have a market cap of $2.6 billion. Using the optimistic scenario, what do you think happens to the share price with the 50% dilution effect of an injection of 100% of the value of the shares outstanding. From a straightline calculation, roughly halving the share price - and that is the best case scenario. This if from a fundamental persective. I am sure they will get a trading pop with a big investment. Hey Countrywide got a one day pop from their investment via BofA.

[Reggie]

@ Andy Gaudet

Hmm! Doom and Gloom forecasts and mental masturbation. I think you have been the 1st rude poster on my blog, ever. Oh well, so be it. ABK posts is first significant loss ever, we are in the midst of the worst housing recession in recent history, financial institions are taking record losses on their real asset related financial holdings, credit and liquity is freezing up globally as solvency is in question in both the household and corporate balance sheets, home building companies are reporting billions of dollars (each) of value loss in their inventories while operating with deep negative margins as we have near record low interest rates that are not slowing our decent into recession. Now, exactly where in there is my gloom and doom forecast???

I think the blog post is quit detailed - a least for a blog post (including a loss tail analysis), but I'll tell you what - I will post additional data on the proposed defaults of Ambac's portfolio via pdf (it won't fit in this narrow blog in html). I am not doing it because of your "mental masturbation" comments, but because I want the more credible patrons of the site to have a much info as possible to come to their own conclusions. So, does the Bloom blog populace desire more info???

[Thomas Davis]

Thanks for the time you took to reply.

I'll hold my position, but watch it carefully.

What do you think about the puts to protect the downside? Wise choice?

[john]

Time for shorts to eat crow.... Early next week will be fun to watch. To 40 by end of the week. Thanks to the shorts who will cover like crazy. We will see what analysis you can provide thereafter.

[George Spritzer]

Your analysis is interesting but seems static and weighted toward the negative and does not put enough weight on potential (totally unexpected) bullish events that could happen.

Have you noticed the recent upturn in the ABX numbers? ABX-HE-AAA 07-2 is now 73.56, up more than 10% from the recent low of 66.41. Suppose it hits 90 by year end?

[Reggie]

@ Thomas
I don't give invesment advice via this blog, and you would be best served not to accept it either. It has always been my experience that those who don't hedge always regret it more than those that do (but you didn't hear that from me:-)

@ George
If things turn up significatly, then the optimistic scenario would take precedence. That is why I gave three scenarios: good, likely, and bad. I truly don't think the analysis is weighted toward the negative, only the findings.

BTW, and index that shoots up by year end does not make things fine and dandy. We are in a hell of a bind here in the real asset and credit markets, and a sharp uptick in the ABX will not fix it. Read my FrankenFinance post. The problem was never subprime, it was lax underwriting standards. ABK's subprime exposure is no where near their worst problem. It was the structured products which were poorly underwritten and imprudently purchased.

[Arnie]

I did not see where you factored into your analysis the collateral subordination in the RMBS pools that ABK insures. ABK typically insures AAA tranches that have approximately 20% subordination. That would mean that there will be no payout on policies until the default rate exceeds 20%. At that time ABK would make timely payments, not the whole obligation, while seeking recovery on the underlying assets. Given the subordination I would think that recovery would be very high. What am I missing?

[David Hug]

You may be correct regarding your ananlysis of ambac, Reggie, but the chart shows the early stage of a rally. I'd cover the shorts until the rally fizzles.

[Reggie]

@ Arnie, good to see you're paying attention, I think. Starting at page 4 of the pdf, subordination is addressed for about 30 different ABS CDOs, each having an individual subordination, of coursse. But all hovering somewher around 20%. I see you have done your homework. This Ambac, analysis represent an awful lot of work ad research, I hope you get something worthwhile out of it.

Due to the amount of work that went into it, it is impractical to post every single calculation, assumption, and justificiation, but pdf has a lot of it.

[keith]

Mr. Davies comment needs a bit of clarification on Public Finance revenues (property taxes). Revenues are impacted in 2 ways. First, strapped borrowers stop paying their property taxes as many states cannot force sale for 5 years post initial delinquency. Second, as happened in the early 90's borrowers appealed their property tax assessed values and had them reduced.

[Mark]

Before taking a short position on Ambac (which I have not yet done but plan to), I decided to begin to look at things optimistically over the week-end. I reasoned that Ambac's exposure to subprime mortgages in 06 and 07 was not that high, and the prime and midprime stuff for these years will probably perform worse than other years, but may do OK. In addition, despite the market price drops on Ambac's ABS CDOs, who really knows what will happen to these?

That was until I decided to look at one randomly selected auto securitization for 2006 -- Triad. Here is what I gleaned from the prospectus.

Total collateral $1.174B
Ambac insured layers: $1.092B (i.e., 7% subordination)

New car collateral: $277M, average FICO 565
Used car collateral: $898M, average FICO 567

The weighted average APR for the total pool is 17.31%. Losses for 2002 and 2003 pools are around 13%. 2004 and 2005 look a little better.

It is somewhat surprising that losses have not begun to chew away at the securities that Ambac insures on the 2006 deals, but I am willing to bet that this will start very soon.