Ambac Financial Group Inc (AMBC) 2008 Q2 法說會逐字稿

Greetings and welcome to the Ambac Financial Group Incorporated second-quarter fiscal year 2008 earnings conference call.

At this time, all participants are in a listen-only mode. A brief question-and-answer session will follow the formal presentation. (Operator Instructions). As a reminder, this conference is being recorded.

It is now my pleasure to introduce your host, Sean Leonard, Chief Financial Officer for Ambac Financial. Thank you, Mr. Leonard, you may begin.

Thank you. Good morning, everyone. Welcome to Ambac's the second-quarter conference call. I'm Sean Leonard, Chief Financial Officer of Ambac. With me today are Michael Callen, Chairman and CEO, and David Wallis, Chief Risk Officer.

Our earnings press release, quarterly operating supplement, and a slide presentation that follows along with this discussion are available on our Web site. I recommend that you review the slide presentation as we speak today. Also, note that this call is being broadcast on the Internet at www.Ambac.com.

During this conference call, we may make statements that would be regarded as forward-looking statements. These statements may relate to, among other things, management's current expectations of future performance, future results and cash flows, and market outlook. You are cautioned not to place undue reliance on these forward-looking statements, which reflect our current analysis of existing trends and information as of the date of this presentation, and there is an inherent risk that actual results, performance or achievements could differ materially from any future results, performance or achievements expressed or implied by such forward-looking statements. These differences could arise from a number of factors. Information concerning factors that could actually cause results to differ materially from the information we will give you is available on our press release and in our most recent Form 10-K and subsequently filed Form 8-Ks. You should review these materials for a complete discussion of these factors and other risks. Copies of these documents may be obtained from the SEC Web site.

I would now like to turn it over to Mike Callen, who will comment on the current market environment and Ambac's strategic initiatives. Mike?

Thanks, Sean. Thank you, ladies and gentlemen, for joining us to discuss our second quarter. After a few opening comments, I will return the floor to Sean to analyze the quarter's financial results, and then David Wallis, our Chief Risk Officer, will talk about our -- the portfolio.

From my point of view, the high points of the past three months have been several. First, we have made progress in managing the risk in the portfolio by negotiating commutations of our CDO of ABS exposures.

Clearly, the bid offer spreads in the marketplace are narrowing so that meaningful discussions are being held and progress is evident. The transaction announced recently is the so-called AA Bespoke contract, which released a substantial amount of rating-agency capital. A number of market analysts published widely distributed estimates that included this transaction as 100% impaired. In the event of course our payment of $850 million to a sophisticated counterparty was about 60% of that amount.

Every contract is quite different, as you know, with respect to structure, asset quality and timing of cash flows. I would caution against using simplistic ratios of settlements compared to par amounts for purposes of determining so-called good from bad deals.

Our policy continues to be consummating these agreements after careful scrutiny of economic terms that benefit Ambac's owners and policyholders. We do not conclude agreements that simply reduce the par value of the CDO portfolio. Nevertheless, we are encouraged by the significant progress that continues to be evident as credit market participants work to resolve their exposures.

A second major development this quarter was the result on the direct mortgage portfolio. Primarily by means of arduous, meticulous analysis with the assistance of outside professionals, we have managed to identify clear, egregious violations of reps and warranties that will enable us to recover approximately $260 million. This estimate is quite conservative in that it was subject to only breaches that had been substantiated on eight problematic transactions; that's eight problematic transactions. So therefore it covers a small portion of the total portfolio. In other words, as our investigations continue, we anticipate greater recoveries.

In addition, the lost projections on several portfolios have improved measurably, permitting positive reserve adjustments of $234 million this quarter. Those who follow Ambac closely may recall that we estimated ultimate losses as high as 80% on two of our transactions last quarter. Many were incredulous at losses that high, but we have historically employed a disciplined and consistent methodology on these matters and let the chips fall. That same approach this quarter produced a much happier outcome.

The third high point calling for comment is the progress on bringing our AAA entrant, Connie Lee, to reality. Detailed plans for Connie Lee has been pretty much completed. This vehicle will bring a highly transparent and pristine balance sheet together with a team of experienced and respected underwriters to a public finance marketplace that demands the product. We can talk about that demand later.

Our primary regulators in Wisconsin have been extremely helpful in this effort, though we have to say that we have not received formal approval yet and our dialogue with the rating agencies has commenced in earnest. The fact that this company will be totally independent of Ambac and that by its bylaws will restrict its business to the area of public finance will certainly buttress the case for a AAA stable rating, at least we think so. The steps that need to be traveled are not entirely within our control, of course, but we still think in terms of commencing business as early as the fourth quarter of this year.

Fourth, I direct your attention to the deleveraging that has been achieved since the beginning of the year. Our total outstanding insurance exposure has declined from $524 billion to $487 billion as of the end of the second quarter. Our capital position has strengthened in step. Our capital under the Moody's model now exceed the requirements for a AAA. While we do not have the updated S&P analysis, we do believe that we are come comfortably at the AAA margin of safety. We expect that trend to continue for a while as we complete amendments and commutations of our CDO contracts and generate internal capital at a rate which has been exceeding our estimates at the beginning of the year.

Lastly, we have had to further impair our CDO portfolio this quarter by an amount slightly in excess of $1 billion, 90% of which comprises three transactions. Underlying assets have continued to be downgraded, if not by as rapid as a rate as the first quarter, certainly at a rate that tipped several of our deals to levels below investment grade. There are signs that the pig through the python theory has more and more credibility, and we may talk about that later if you wish, but the results of that phenomenon for portfolio evaluation purposes cannot be anticipated in this quarter's results.

I have mentioned several times already that we are progressing and ameliorating these problems and we'll inform everyone as soon as reportable results are available.

Thank you for your attention and let me ask Sean now to walk through our financials.

Sure. Thanks, Mike.

I will now provide a brief overview of the financial results for the quarter. Net income was $823.1 million, or $2.80 per diluted share, up 68% on a per-diluted basis from the second-quarter 2007 results. The increase was primarily driven by the net, non-cash, mark to market gains in our CDO of ABS portfolio, record accelerated earned premiums resulting from refundings and negative loss provisioning related to transaction and performance of certain direct RMBS transactions exceeding our previous expectations and estimated recoveries from remediation efforts on the direct RMBS portfolio. Partially offsetting these positive results were there-than-temporary market value declines in our financial services investment portfolio.

To assist users of our financial statements and analysis of our reported earnings, we also report earnings on an operating and core-earnings basis. Both of these earnings measures are considered non-GAAP measures.

Operating earnings excludes the net income loss impact of net gains and losses from sales of investment securities and mark to market gains and losses on credit, total return and non-trading derivative contracts that are not impaired. Core earnings exclude the net income impact of accelerated premiums from re-fundings.

Operating earnings per share in the second quarter were a loss of $1.53 per share, and core amounts to a loss of $1.83 per share. Remember that, in calculating operating earnings and core earnings, we exclude the impact of unrealized gains and losses from our CDO portfolio but do not exclude the impact of estimated credit impairment within this portfolio.

In the second quarter, Ambac recorded credit impairment of $1.06 billion, pretax, or $2.45 on a per-share basis. This amount is reflected in our operating and core earnings results.

Normal earned premiums amounted to $166.3 million in the quarter, declining $11.7 million, primarily due to the December 2007 Assured Guaranty Reinsurance contract. If not for the seed to Assured, normal earned premium would have been down only about $4 million, given that we have written very little business in 2008, our steady premium earnings as a result of the value provided by our embedded book of business.

Accelerated premiums increased significantly, up almost 300% to a record amount of $159.2 million. Refunding activity was heavy and auction-rate and variable-rate debt notes due to the illiquidity experienced in those markets since really this year. We believe a large percentage of those transactions of work their way through the refunding refinancing process, but we expect to see continued elevated levels of accelerated premiums throughout 2008. This refunding activity enhances our capital position and delevers our balance sheet.

Investment income increased $14.1 million, or 12%, to $127.3 million on increased volume driven by net positive cash flows from operations and the net proceeds from the March capital raise.

Now, I will discuss our mark to market and credit impairment activity for the quarter. During the second quarter, Ambac recorded net mark to market gains related to its credit derivative portfolio amounting to $961.6 million. The net mark to market gain was benefited greatly during the quarter by an adjustment related to Ambac's own credit spreads, which widened significantly during the second quarter, especially in June after the rating agencies downgraded Ambac Assurance, our operating company, to the AA level.

Subsequent to June 30, our credit spreads have declined significantly. In fact, we estimate that, if we had used Ambac Assurance's credit spread as of July 31 instead of June 30, our net mark to market would have been approximately negative $1.3 billion, instead of the positive $962 million in the quarter.

Offsetting the favorable credit spread adjustment at June 30 were unfavorable price declines on our CDO of ABS obligations, coupled with several downgrades to below investment grade for certain of these transactions. But as our recent AA Bespoke commutation illustrates, mark to market on a transaction at a given point in time is not necessarily indicative of ultimate loss.

For the second quarter of 2008, estimated credit impairments on CDOs amounted to approximately $1.06 billion. The impairment was driven by notable credit deterioration of inner-CDO collateral across several CDO of ABS transactions during the quarter.

At the end of last quarter, we had on our books three recent vintage CDO Squared transactions with a notional value of $2.4 billion. Today, one transaction has been commuted and the other two have full impairments charged against them, so those deals are completely behind us from an impairment perspective. David will provide additional details later in this call.

I would like to now discuss our loss provisioning on our direct RMBS insurance portfolio. Loss reserves on our direct mortgage exposures declined from last quarter. We adjusted our RMBS net loss reserves down by approximately $339 million during the quarter. The downward adjustment is primarily a result of our remediation efforts in this area, as we had discussed during our last earnings call.

Estimated remediation recoveries for select second-lien RMBS transactions recorded during the quarter amounted to approximately $260 million. David will provide more information on those efforts, but it is important to note that, to date, we have only considered the worst examples of the verified breaches of reps and warranties through detailed file reviews to credit-worthy counterparties. We assume that we will collect on these breaches in three years' time and have discounted the amount accordingly.

In addition, the lower reserve is also the result of overall net improvement in the performance of our second-lien and Alt-A exposures, which contributed to the remainder of the quarter's reserve reduction. There were no other sizable adjustments across the remainder of the portfolio.

Total net loss reserves at June 30 amounted to $1.1 billion. This compares to approximately $1.5 billion at March 31, 2008. Case reserves of $520.2 million at June 30 are up $169.6 million from March 31. During the quarter, actual net claims paid amounted to approximately $67 million, much of most of which related to HELOC RMBS transactions.

Active credit reserves of $555.5 million at June 30 are down $575.8 million from the end of last quarter, primarily due to the aforementioned expected recoveries from remediation efforts, transfers from ACR to case reserves on certain transactions that have defaulted, and improving performance on other transactions. Our below-investment-grade exposures increased by almost $13 billion during the quarter to $29.7 billion or approximately 6% of our total portfolio. $10.8 billion of the increase was in our CDO of asset-backed security portfolio, and $1 billion in our direct second-lien RMBS portfolio.

Now, I would like to discuss liquidity at the holding company and rating agency capital position. Total holding company cash amounts to approximately $169 million at June 30. That amount of cash is equivalent of approximately 1.5 years of debt service for the holding company. We dividended approximately $54 million at the holding company in July, and plan to dividend a like amount in October, which would grow this cash position to an expected $210 million by year end. That is approximately 1.8 times the holding company's annual debt service needs. Share repurchases in the remainder of 2008 would reduce the expected cash position.

In early July, we announced an authorization to buy back $50 million of our common stock. We have not utilized any of the authorization to date, as the Company was in a self-imposed blackout period starting July 4, in advance of our second-quarter earnings release. Additionally, at the time of our capital raise, certain underwriters of our common stock offering purchased shares. Certain repurchase restrictions applied to our ability to execute the share repurchase under the US Securities law. As of today, one underwriter still holds shares purchased in the offering. We can begin executing share repurchases [on the earlier] of the sale of the underwriters shares or an agreement with the underwriter to terminate the offering.

In the quarter, we continued to release capital as exposures amortized down and refinanced. We currently exceed Moody's AA3 level of capital by an estimated $3 billion and believe, considering our AA Bespoke commutation reported last week, that we met Moody's AAA stress target as of the end of July. We also believe that our S&P capital model results were comfortably in excess of both the AA flat and AAA levels of capital.

From an earnings perspective, our deferred earnings, representing future earnings on premiums already collected and the future value of installment premiums, will carry us through periods of low CEP. These deferred earnings amounted -- amounting to $5.9 billion will be recognized as earned premium and realized gains from credit derivatives in the future over the life of the related exposures. $2.3 billion represents cash already collected and invested in our conservative investment portfolio, while $3.6 billion is cash that is contractually due to be paid to us in installment premiums over the lives of the transactions.

As a result of additional CDO of ABS credit impairments recorded in the second quarter, Ambac has breached its net worth covenant leading to its $400 million credit facility. We had amended the facility earlier in the year to exclude mark to market on credit derivatives and total return swaps, but actual credit impairment charges are included in the net worth covenant test. We have been and will continue to discuss cost-effective potential solutions with the banks in the near future, but please keep in mind that this credit facility has never been drawn upon.

Now, let me spend a little time focusing on our Financial Services segment, which comprises our investment agreement and derivative products businesses. In the first six months of the year, we terminated, restructured or otherwise mitigated approximately $1.5 billion of notional amount related to our cost of fund swaps. This reduces our cost of fund swap notional balance from $2.2 billion outstanding at December 31, 2007 to approximately $700 million. We expect to execute another $300 million of terminations in the third quarter, reducing the notional size to approximately $400 million.

The severe dislocation in the structured credit markets also continues to negatively impact our Financial Services investment portfolio. We reported unrealized market losses of $150 million, comprised of credit impairment losses of $99 million related to other-than-temporarily impaired securities and $51 million for mark to market adjustments on a portfolio of securities that we maintain for liquidity purposes. A detailed analysis of the affected securities shows that the actual expected credit losses are a fraction of the market value losses that had been booked.

It is also worthy to note that our investment agreement balances have declined by $1 billion in the first six months of 2008, and we expect them to end the year at approximately $6 billion. That would be down another $700 million from June 30, 2008.

That concludes my prepared remarks on the financial results. David Wallis will now talk through some detailed elements of the portfolio. David?

Thanks, Sean. Good morning, everyone. As you would expect, the focus of my comments will be RMBS-related in the two areas where we have taken this exposure, the direct RMBS book and within CDOs of ABS.

On the direct RMBS side, we have had three more months of seasoning [monitor] trends and performance, and as our transactions mature, we consider that future outcomes become moderately more certain. In addition, given the progress of our active RMBS remediation efforts which I will summarize later, this is the first quarter where we have felt able to recognize the benefits of these efforts.

On the CDO of ABS side, we continue to be disappointed with the pace of deterioration, especially with regards to CDO collateral within our transactions. However, as we have now progressively written down this constituent element of our CDO transactions, we anticipate less pain to come with respect to this important historic driver.

I'm now going to review the portfolio in a manner akin to last quarter's earnings call. This chart shows aggregate outstanding reserves and impairments for both RMBS-related product constituents, further breaking out the segments of each.

The left-hand side shows RMBS reserves with HELOC reserves now representing 52% of the total $896 million RMBS reserve and closed end second representing 21%. Mid-prime or Alt-A exposures comprise approximately 20%, while subprime exposure, as we have noted previously, benefiting from very limited and almost exclusively fixed rate 2006-2007 originations, continues to be a source of only relatively moderate concern.

Next, on the right-hand side, we show our CDO impairments. Having satisfactorily commuted our largest CDO Squared transaction last week, we were essentially fully reserved for our remaining two large, below-investment-grade CDO Squared transactions.

Some of the high-grade deals in the remainder of the book were a source of continuing disappointment in the quarter as the adverse CDO structural features of ingredient or constituent CDOs, combined with general weakness in the RMBS sector, have moved nine transactions to below investment grade and drew impairments of approximately $1 billion.

As mentioned, the constituent CDOs within the CDO of ABS transactions continued to experience significant downgrade during the quarter. Briefly, as with the inner CDOs within our residual CDO Squared transactions, rating downgrades of the RMBS collateral supporting these constituent CDOs can of itself trigger adverse cash flow consequences from the perspective of the mother, high-grade CDO. Examples of this structural weakness include picking, where a security no longer pays cash but instead accrues, or write down, perhaps as a result of liquidation.

Let me move to a more detailed review of, firstly, our direct RMBS book, starting with second-lien transactions. The two performance charts show delinquency and default rate data since inception. The thick black lines show, on an unweighted basis, the average of the performance for the transaction calls listed. Outside of normal volatility, a reasonable take-away is that delinquencies appear to be flattening out whilst default rates, with a lag as one would expect, may have begun to decline.

With respect to second-liens, it is in this broad context that we will discuss reserving activity. Before getting into the detail, note that we have applied the same roll rate methodologies in the last quarter and that, despite the possible stabilization or even improvement in some collateral performance, this methodology still imputes high growth, cumulative losses for a limited number of select, adversely performing transactions. For example, with respect to the Bear Stearns and First Franklin transactions, our model currently postulates cumulative losses of 70% and 65%, versus 82% and 79%, respectively, in the first quarter.

Let's now move into the second-lien reserves themselves. HELOC reserves increased modestly. The net $36 million increase was largely driven by one lagging, underperforming deal for which a new reserve was established. Overall, the aggregate $464 million of HELOC reserve arises from seven transactions representing 27% of our total HELOC par. I will discuss the profile of these transactions in a subsequent slide.

The aggregate closed end second reserve arises from eight transactions representing 51% of total closed end second net part. Closed-end second reserves now stand at $187 million after a $448 million reversal in the quarter. The closed-end second reserve benefited from two factors, firstly, improved performance per the previous slide, and secondly, acknowledgment of the progress we are making in our remediation efforts. Disaggregating these two elements, performance trends contributed $234 million of the reversal, and estimated future remediation recoveries contributed $214 million.

Let's now talk a little bit about these remediation activities. As mentioned, we are undertaking aggressive remediation activity on the direct RMBS book, and have now reflected initial estimates of recoveries in our reserve position. The overall recovery estimate is $263 million on the present value basis across eight second-lien transactions. The recovery estimate is based upon professional scrutiny of around 1500 loans, many of which contain more than one substantive breach of the relative transaction representation or warranty conditions. Our various advisers found these breaches in a review of 1800 loans, an approximately 85% hit rate. These results were obtained from a limited initial sampling of mortgage loans chosen on a variety of bases, some random and some adverse. In total, there remain more than 83,000 loans that have not yet been reviewed in the eight subject transactions.

Beyond this group of eight second-lien transactions, we are expanding our investigation to include an additional six second-lien transactions and nine mid-prime transactions. As always, we will continue to exercise all of the contractual rights and revenues available to us, and our estimated recoveries will be revised and supplemented as appropriate.

The next two slides review the performance segmentation that we have previously pointed out in our second-lien portfolio -- first HELOCs, then closed end seconds. The HELOC chart segments our $11 billion HELOC portfolio into three categories. Notably, all of our reserves relate to the middle column -- seven transactions, all non-bank originated in 2005 to 2007. The performance of transactions in the other two columns, pre-2005 transactions, and (inaudible) originated by banks making direct mortgage loans to their client base within the context of a likely broader relationship are continuing to perform satisfactorily.

Onto closed-end second transactions -- here, all of our reserves relate to six transactions, all piggyback second liens. Piggybacks are seconds supporting a first clean where the LTV of the combined loans is close to or in excess of 100% at inception. Per the HELOC portfolio, there is a marked performance differential between the two overall CES subsegments.

As noted at the outset, we do believe that the [affluxion] of time and the consequent relative stability of performance, be it adverse or otherwise, progressively assists in diminishing the possible range of future loss outcomes.

Let me move on briefly to the mid-prime and subprime portfolios. Overall, there has been little net change in the aggregate performance of our mid-prime portfolio over the quarter. Here, eight transactions account for the 14% at par in this category that we rate below investment grade.

As discussed a quarter ago, we continue to monitor high foreclosure and real estate-owned buckets in some 2006 and 2007 transactions. We anticipate that cumulative losses could rise quickly, as these buckets are liquidated. Whilst the timing and magnitude of these likely events is difficult to predict and lost severities have generally been modest in this product, transaction credit enhancement is relatively low. Anticipating some increase in severities, we project cumulative losses up to, exceptionally, 30% in a number of transactions where we have focused the bulk of our mid-prime reserves.

In the subprime portfolio, we have one transaction that comprises 88% of our total reserve. This is the 2007 transaction with current net par balance of $535 million, and represents the only subprime transactions that we insured in 2007.

In closing on the direct book, we can summarize as follows. Our reserves are concentrated in the 2006 and 2007 vintages and largely within select product subsegment -- subsegments of these vintages. Our reserving process is not generic, but instead transaction-[cause] specific. We have a strong and ongoing focus on remediation activity.

This slide provides a summary snapshot of our 2007-originated RMBS portfolio which we hope is useful in the context of the segmentation comments I have just made.

Let's now turn to the CDO portfolios. I will begin by providing some industry context (inaudible) ratings migration in the quarter. This data is sourced from a recent UBS study.

In the aggregate, approximately 47% of the [absolute] assets underlying ABS CDOs have been downgraded as of the date of this study. As a result, approximately 78% of the mezzanine CDOs of ABS and 50% of high-grade ABS CDOs have hit an event of default. With these dramatic statistics, let me move onto our CDO book.

The quarter's activity centers around two developments, the successful commutation of our largest CDO Squared transaction shown in the top row and the continued underperformance in the 2006 and 2007 high-grade CDOs of ABS, shown in the bottom row. For completeness, the middle row includes the remaining two approximately $500 million CDO Squared transactions, which are effectively fully reserved, as noted previously. As the slide illustrates, it is the combination of the high-grade ABS CDOs and the commitment transaction that have performed disappointingly over the quarter.

I will comment briefly on the commitment transaction and review the high-growth portfolio in a later slide.

Per our Web site disclosure, the CDO securities within the commitment transaction are predominantly high-grade CDOs that were AAA or AA at issuance. More than 80% of these securities were issued before 2006, including almost 50% in and prior to 2004. However, despite its face value of markedly better vintage distribution of these assets, there have been substantial rating downgrades over the quarter. As of June 30, more than 50% of the CDOs were rated below investment grade by a one or both of Moody's and S&P, compared with less than 20% as of March 31. Whilst an existing $910 million deductible must be exhausted before Ambac could become obligated to issue a policy, the above ratings performance has resulted in an impairment against this transaction over the quarter.

Now, let's discuss the commutation that was briefly mentioned earlier. As reported, we paid $850 million to eliminate the $1.4 billion AA Bespoke CDO Squared transaction. 98% of the underlying collateral of this transaction was rated below investment grade. As a reminder, all underlying collateral was AA at inception. Idiosyncratically and directly resulting from the unusual structure -- and this is a key point in thinking about the economics of the transaction from our perspective -- Ambac would have had to pay sizable claims in the immediate future.

We view the commutation achieved as an attractive outcome. The exit price is less than the likely hypothetical second-quarter accumulated impairment would have been. The exit price is less than the pre-existing $1 billion adverse mark. Finally, the commutation is capital-accretive, given that the exit price was lower than rating agency stress case losses. We remain very active with regard to other commutation and/or restructuring discussions concerning select elements of our CDO of ABS portfolio.

The next slide segments our portfolio of later vintage, i.e. 2006 and 2007 transactions, into two groups. The middle group shows those with modest exposure to inner CDOs out against the right-hand column, which consists of transactions where the CDO bucket is larger than the original overall transaction subordination.

Current high-grade CDO of ABS performance continues to be dictated in large part by the inner CDO buckets. For deals in the right-hand column, 70% of the inner CDOs, on average, carry a below-investment grade rating. Furthermore, more than half the overall underlying collateral, inclusive of the inner CDOs, also carries a below-investment-grade rating. Consequently, all but one transaction in the above subsegment has an Ambac below-investment-grade rating, or has hit an event of default trigger. As one would therefore expect, the bulk of our impairment charges relate to transactions in this category.

Whilst we have also impaired transactions in the middle column, we have done so to a lesser overall extent. One distinguishing feature is the relatively lower proportion of below-investment-grade collateral of these transactions. Consequently, a lower proportion of these transactions have hit an event of default, and the impairment is relatively muted.

Very briefly on methodology, Ambac's impairments and internal ratings are increasingly based on cash flow runs, post making haircut assumptions with regard to CDO buckets. In relation to the cash flow runs, it is noticeable that some of the transactions which are less exposed to inner CDOs appear to be more exposed to a relatively aggressive portfolio of ultimately underperforming RMBS assets. The CDO haircut assumptions are in keeping with those we have historically developed for our CDO Squared transactions, including the one that we have now commuted.

Overall, we've assigned below-investment-grade ratings to an additional eight high-grade ABS CDOs in the quarter. Additionally, it is becoming clear that the older vintage transactions are now tending to perform better than more recent vintages. This segmentation is also evident in rating agency ratings, where super senior transactions are predominantly in the A- to AAA range.

Please note that we have chosen to expand on disclosure in relation to our high-grade ABS CDOs by supplementing our usual webpage with a one-page summary of each of these transactions. The summaries show select performance characteristics of the RMBS assets, as well as our analysis of what percentage of the CDO bucket we have written down in our ongoing evaluation.

Finally, let me briefly discuss the timing of our claims payments. We expect the majority of HELOC and closed end second claims to occur over the next five to six years. Currently, claims are being paid on seven HELOC transactions, whilst we have recently paid our first claim on a new vintage closed end second transaction. Year-to-date, RMBS claims paid are approximately $103 million, and we project approximately $180 million over the remainder of 2008.

For reference, these numbers compare to our first-half investment earnings of $247 million. We have paid approximately $1.6 million in interest claims on our two large remaining below-investment-grade CDO Squared transactions, and we believe, conservatively, assume that principal will fall due in around five years' time.

In relation to our high-grade ABS CDOs, we do not expect to incur interest claims until 2013, and principal is not payable until the earlier of the legal final or the date on which knowing a collateral exists. The resultant delayed aspect to claims highlights a fundamental strength of our business model.

Obviously, all of the above numbers exclude any commutation-type payments that we may or may not choose to incur.

That concludes a brief review of the RMBS-related portfolio. I will now hand it back to Sean.

Thanks, David. That concludes our prepared remarks. We would now like to open the call up to your questions.

Thank you, gentlemen. (Operator Instructions). Andrew Wessel, JP Morgan.

I guess a couple of questions -- one on the use, the implementation of FAS 157. You know, how do you put your kind of -- help us think about using your ability or the market's perception of your ability to pay your debt against the obligation you have under a CES contract to pay in the event of default. I'm kind of having problems figuring out how that kind of leads to a $5.2 billion write-up when in actuality it would have been a $4.2 billion write-down.

Sure, Andrew. What the theory is in the accounting rules and it would apply for a known set of cash flows versus what we are dealing here with, is a potential contingent set of cash flows that we potentially are obligated to pay over time, but the theory is the same, in essence that one should include a market-based discount rate. So one should not just discount at, say, a risk-free rate which is kind of the rate that we use. We actually use a portfolio rate of 4.5% to discount potential claim payments in our insurance portfolio. But the accounting theory is one should utilize a market-driven rate. Now, when one considers what a market-driven rate would be, we have chosen to look at market indicators relating to the credit spreads of Ambac Assurance, which is effectively the company that would have to make the pavement. Those discount rates, if you will, rise, the present value or the fair value, if you will, of those obligations decline. So that is consistent with how the market would look at our debt obligations at trade-out in the market and so forth, and I think it's consistently applied I think amongst other market participants.

Okay. Then I guess, looking at it from the perspective of the statement has always been made "economics aren't affected by temporary impairment, rating agency models aren't affected by temporary impairment." Why take the step now to change GAAP results when the focus has always been on operating results, and the focus for the rating agencies has been on the same?

Well, Andrew, we are required to report GAAP results and we are required to follow FAS 157. That goes hand-in-hand with reporting in compliance with US Generally Accepted Accounting Principles. So we don't have a choice on the GAAP.

We've always provided operating and core earnings for these very reasons. It provides an additional analytical measure for folks to look at for these reasons, that the volatility that's produced in fair valuation adjustments may not reflect the intrinsic value, or our view of the ultimate impairment that underlines the transaction. So we have not done anything different. I don't know if you are implying that we did not do -- adopt FAS 157 this quarter. We adopted it at the beginning of the year, which was required to be adopted, so there's nothing, from our perspective, unusual about that.

You know, what did happen and part of my comments and the reason for my comments is our credit spreads did spike in June, largely due to some actions by the rating agencies in the middle of the month. The credit spreads reflected that. They have subsequently come in and further have come in recently based upon the announcement of our AA Bespoke transaction. So I was trying to give folks a sense, in my comments, for the effect of that. We do expect that gain to reverse in July.

[Alan] here -- let me just make one comment. We are trying our best to be as transparent as we can be. As you know, we have been publishing our mark to market on a monthly basis, plus whatever information is available to us, so that you have the opportunity to put this in the context for analytical purposes that you need to put it.

No, no, that's definitely been helpful. Then my last question and I will jump back in the queue -- just the reserve release on the direct RMBS transaction. I mean, the remediation effort, I think it's fantastic and shareholders should be very happy that you are undertaking this effort, which is of course ridiculous in the past. I mean, I can't even imagine the labor involved, but I think, from a perspective of that three-year recovery period, is that reflective of expected legal activity, expected lawsuits? I would think, if it was cut and dry enough, that would be a much shorter period of expected recovery.

I sort of agree, but for the moment, they've got the money, and you know, clearly, they are not keen to open wallet. So it is a protracted process; it isn't easy to gather the data. There is a degree of, candidly, obfuscation or intransigence in terms of handing over that data. Then you've got to analyze it. It isn't easy. Sometimes there are fields missing which inhibit the sort or the sampling that you what to do to ascertain what the whole looks like, and so on and so forth. Then obviously, you need to locate the files. Each file, by the way, takes an hour, an hour and a half to go through. Where does the guy live? Is it consistent with his Social Security number? Where does he work? Does he really work there? Phone them up; what is his income? And so on and so forth. It's a very, very long-winded process. You have to build the case over a period of time.

We are in no rush to do that, but I can assure you we will be very diligent and very steady and unremitting in our efforts, and the three-year period is our best guess at an average life type basis as to what it will take to bring home the gains that we expected.

There are significant [haircuts]; I mean, we don't think we are being overly aggressive. Clearly, we are not booking gains in any transaction in excess of the gross cumulative loss that our modeling of the mortgage securities involved project. We can't do that. What that means, in some cases, is that already, in one particular case, we've got breaches; I think it's in excess of three times of the reserves that we've posted. So obviously what we would do is bring down that reserve to 0 on a net basis, and in a sense we would have the ability -- I mean we won't be stepping forward with this needlessly -- to settle for $0.30 on the dollar or something in respect o these known, substantive and documented breaches.

So hopefully that's helpful and rounds out the comments a little.

Okay, thank you.

Steve Stelmach, FBR.

Just real quick on the FAS 157, again I apologize for circling back to it, but, Sean, did you mention that the $961 million benefit in the quarter would have been a negative $1.3 billion mark to market lost as of 07-31? Did I catch that number right?

Yes. What I had said is, if we had used similar, you know, the July 31 spreads and kind of pushed that back to the June 30 balances and the state of affairs at June 30, just using that market level discount rate, yes, it would have reversed the gain and produced a loss of that amount.

Okay, so I'm just thinking about the swing of what, like, $2.3 billion? That would wipe out stated equity. Is that the right way to think about that? I know obviously the operating numbers are the more appropriate measure and I am in agreement with that. But just from a GAAP perspective, would that have resulted in a 0 equity value?

It would have had a reducing effect on equity, and then you would need to tax-effect the amounts.

Oh, so take the $2.3 billion, tax-affect it, and then sort of net it against equity?

That's right.

Okay. Then Mike, on Connie Lee, how much capital sits at Connie Lee today? How much do you expect to sit or see Connie Lee with in the fourth quarter?

Right, we have $150 million in the vehicle at the moment. We will take down another $850 million, start with about $1 billion when we get the AAA.

The portfolio today, the insurance claims are about $650 million of very high quality assets that have been there, very mature, with no loss reserves -- healthcare and universities, that type of thing. So we will be starting with a very clean slate.

Okay. Then will it sit below the holding company, or will it sit below the current operation?

It will sit below the insurance company, the operating company.

So dividends still have to go through op co. up to hold co., from Connie Lee to op co. to hold co.?

But remember, we have a prohibition built into the plan of any dividends for at least three years.

I got it, okay. Then on the legislative front, there's been obviously a push for sort of corporate equipment ratings on the Muni side, and the rating agencies themselves have sort of began to migrate that way. Could you just give us some color on sort of the business outlook there?

Bob Shoback is here. He probably knows as much about public finance as anybody. Let me ask him to comment on that.

Thanks. You know, with regard to the corporate equivalent ratings, there's been pushes in that direction for a number of years. One or more of the rating agencies have implemented that to some degree in the past. At that point in time, we didn't really see a significant decrease in the utilization of insurance as a result of those efforts on their part.

Okay, great. Thank you.

Geoffrey Dunn, [Valling and] Partners.

Sean, first of all, can you give us the net capital benefit from the commutation of the AA deal?

Yes, sure. It's different between the two agencies, based upon the levels, or at least the process that they undertake. In the case of Moody's results, Moody's does not tax affect; one thing they don't do is tax affect any stress losses or any theoretical losses that they have in their models, so they have just gross numbers, and they compare that against our claims payment resources.

Moody's had a very low view of that transaction, and particularly combined with the nature in which they calculate their theoretical losses utilizing a 30% margin, so they calculate it at a 1.3 times level. With that in mind, coupled with the tax impacts of that transaction, the estimate of the capital relief just on that transaction is about $800 million.

Relating to Standard & Poor's, their view, they include an element for taxes and theoretical losses in their calculations, so you don't have that differential. They don't have a similar construct even though they do require a margin of safety of 1.25 times. But effectively, we think that is probably between $50 million and $100 million.

Okay. Then probably for Mike, with the structure of Connie Lee under AC, how are you making it such that that is walled off and the market perceives that as clean $1 billion of capital?

We will have a separate board, separate bylaws. The legal structure will be such, and the regulators, remember, in Wisconsin and for that matter in New York, will be involved in this. We have asked our legal counsel and they tell us, our external counsel, in building this plan, that we want this subsidiary to be totally independent, which they claim to have found unusual but were happy to accommodate us. And then the documents that the agencies and the regulators are reviewing, I think it makes it very clear there will be at least seven directors, four of whom will have nothing to do with Ambac. Certain things will require a supermajority vote. So there are all sorts of restrictions being placed on the vehicle, including agreements on multiyear termination of dividends, or no dividends to be upstream and so forth.

I don't think, in terms of the separateness, as we call it, that we have expanded on, we've had any doubts from the constituencies we have to satisfy.

Okay. Then the last question for David, understanding all of your comments, it seems like maybe you are starting to get your hands around the RMBS but each quarter, for the last couple, we've seen significant impairments on the CDOs. How can you get ahead of this issue? How can we get to a point where the impairments are declining significantly, rather than being surprised by an incremental $1 billion each quarter? Is it purely getting ahead of the ratings of the internal collateral, or is there something else where we can gain confidence that maybe the worst is behind us and now we are facing an improving result from here on out?

That's a good question. I think it's a combination of things, as I tried to sort of get through what has really been damning is the first change in the credits of these deals has been the degradation in the CDO buckets. So a high-grade deal has X% RMBS but sometimes it's significant CDO buckets and sometimes buckets in excess of the subordination of the deal. So when that really degrades, that's a kind of immediate first-order effect.

You know, the [salutory] fact there is that we've progressively been writing off, and the details of this are on our Web site in respect of each individual transaction -- the portions, in some cases, very large portions, of the CDOs. So we think that -- we hope and believe that we are through the worst there.

The other thing, the other component to the credit of these deals is obviously the RMBS collateral. As I intimated in the prepared remarks, what we really are reliant upon now in terms of looking at these things is cash flow modeling. I think we've convinced ourselves, and probably others share this view, that some of the CDO models that we've used historically in addition to cash flows are somewhat flawed. So what that's about is obviously running through all of the (inaudible) of all the RMBS, in all the different vintages, using different assumptions, taking account of the fact that they are at different points in their loss curves, you know, the cash flows through all of the deals. So what that depends upon, in terms of outcome, is the assumptions that you use, and have you got the process right?

We believe we've got the process right. We have hired an external party and sort of done an apples-to-apples type comparison in relation to the results using similar assumptions and so on and so forth, so we feel good about that. What that leaves you with, in terms of variables, is the assumptions that you use. We believe our assumptions are reasonable. We do believe, as a general statement, Mike's comment, right, the pig and the python earlier, that loss curves are very front-ended, and that therefore there will be an amelioration in the rate of decline in terms of cumulative loss and so on and so forth.

So, we think we are applying reasonable cash flow assumptions. We think that we have progressively degraded the first order problem as it were, which is the CDO buckets. Therefore, God willing, in terms of the environment and the assumptions we are using, we are hopeful that we won't have this kind of volatility going forward.

Geoff, let me reiterate an observation I made in the beginning. It's a very good question, by the way. The ability to have meaningful conversations with counterparties now versus, let's say, the end of the year or at the end of the first quarter, has improved significantly. I want to be broad in my comments here. That is quite encouraging. As we have announced the Bespoke transaction, frankly, at the end of the year, probably at the end of the first quarter, the consummation of that transaction would not have been possible. But counterparties' positions, counting positions, perceptions change and as I said, if anybody asks me to express one comment I made in my opening remarks, it's that clearly the bid offer spread on these possibilities is closing down, which is a sign of more liquidity in the market, and we have about -- well, I probably shouldn't characterize, but a number of efforts underway which have been underway for some time. So I think that's going to have impact, too.

Great, all very helpful. Thanks.

Darin Arita, Deutsche Bank.

A question on Connie Lee -- in terms of getting this transaction done, what are the key hurdles?

The major hurdle is the AAA stable. Our conversations with the rating agencies have probably reached their second step with, I'm going to guess, three or four more steps to go, in great, rigorous detail. They are putting us through the ringer, but I quite honestly believe that we are up to meeting that test.

We have developed a plan with appendices to it that we needed to rent several mules to get over to them. They look at this as a project. They are working it very hard. They have a lot of very good questions. They want to know what plan Bs are in the event that our initial plan, in terms of risk profile, are not good.

I cannot speak for the rating agencies, I wouldn't dare, but I can say that every criteria that we know of, including the restriction to public finance, which is going to be in the bylaws, has been incorporated and the test and the questions that they are putting us through in the models are only going to be helpful.

I am encouraged in both of the major rating agencies, in the sense that our discussions have been very serious, they have not had any negative attitudes on display, just quite in-depth questions. But clearly, the hurdle is the AAA stable.

Some think that the issue is going to be market acceptance. There is nothing I have seen, and we have put that to the test as much as we can do by putting on a road show and then leaving the room and getting feedback through our investment bankers over with. There's nothing I've seen that says that's going to be a problem.

Perhaps Bob Shoback could comment on that himself if you need some elaboration.

Sure. I think we've gone out to a lot of the institutional investors and our client base and have received indications from all parties that the need has never been greater at this point in time. The capacity that has left the municipal markets is the vast majority of what had been there, 75% or more of past capital and past market share. The need is still great. So, there's a huge imbalance between supply and demand, and the investors, the issuers, want this product and need it.

In terms of the ratings, would you be willing to have this rated AA or split-rated?

We are not even thinking about that at this point! (LAUGHTER) I don't think it's a bridge we will come to. I'm quite encouraged.

I should stop there, but as my mouth has never been disciplined, I will tell you, if it ever came to that, would we go forward and prove that this is a AAA vehicle that can do business effectively? Absolutely. There's just been enormous effort put into this, and when we put our underwriters on display, as we have, I think it just strengthens the case all the more.

Great, that's helpful. Then in terms of the second-lien RMBS, the analysis on breaches of reps and warranties, of the 1800 loans that were analyzed, David, you mentioned that some were randomly selected. How many of these were randomly selected?

I haven't got the numbers in front of me. What we did, for the purposes of trying to figure out which deals, candidly, to go after first, is initially, in some cases, look at 60-day delinquencies, especially where those buckets were large. On any reasonable hit rate, we would get meaningful results, we believe, in relation to our reserves. So it's a mix. I haven't got the precise specifics. I can come back to you on that, but we do think, I guess, if there's a take-away in all this stuff, is that there's 85,000 loans we haven't looked at. I will say that the results show that many, many thousands of loans I think will fall within the crosshairs of rep and warranty breaches.

Yes, that would be helpful, David, if we could have that, because the 85% hit rate on the 1800 loans, I thought that was very high. But I was curious. What was the hit rate on the randomly selected loans?

Oh, much less, I think it was kind of in the 30s, but obviously then you're looking at the entire transaction, not just a bucket. But I will come back to you.

Great, thanks very much.

Arun Kumar, JP Morgan.

Good morning. Just a couple of quick questions on your statutory numbers -- if you could quickly walk us through how you arrived at a statutory gain after you took the $1 billion credit-impairment charge. What percentage of the credit impairment charge actually walked through your statutory numbers?

Secondly, I know a lot have been asked about Connie Lee but just quick one on that end -- given that you are on stage two of what you anticipate to be three to four stages on the ratings process, when do you expect that to conclude, given that you've stated that, by October 1, you'd like to get into the business?

Nothing that I have encountered in this process would cause me to change the October 1 date in terms of doing business. If that changes for some reason that I can't foresee, we will try to communicate that fact. I said in my remarks "the fourth quarter". That's kind of a (inaudible) prepared statement. But we still have -- October 1 is still what we have in our minds around here for planning purposes.

Sure, it is Sean. Dealing with a statutory numbers, as you are probably aware, statutory accounting requires the recording of loss reserves for defaulted items, so case reserves. We have recorded case reserves in the second quarter, and that's reflected in the second-quarter numbers of approximately $170 million.

What also is going on there is the premium revenue recognition pattern under statutory accounting is a much slower revenue recognition pattern. So when we have refundings, particularly heavy in the public finance area, what happens is the accelerated premiums are much higher under statutory than under our GAAP results, so there's an impact there as well.

What also you have is -- you are right. We are recording impairments to our statutory results. There's some unique aspects of some of the contracts that would require, would only require recognition upon default. So we haven't recognized the entirety of the $1 billion of impairment in the quarter, but a substantial part of that.

The last thing that's going on is there's some tax benefits for the impairments that are running through the quarter. When you kind of combined all of those together, that's how you come up with the income results.

You do see, on Page 31, which is perhaps what you're looking at in the slide deck, some of the details of the accounts. You can see that the actual surplus from the end of the first quarter to the end of the second quarter has declined slightly, but obviously that includes a substantial amount of impairments already in that number with a very sizable policyholder surplus amount.

What you don't see in these numbers is we also are required to establish a contingency reserve, and that number is [sightful] as well. At the end of the quarter, the contingency reserve is $3.26 billion, which is kind of a loss reserve in a sense that reduces policyholder surplus. So you have a substantial amount of reserves built into the policyholder surplus number.

Fair enough. Sean, just a quick question on the holding company cash -- you said you are going to be at about $210 million at the end of the year. Would I be right in assuming that, if you do some of the share buyback in the second half of the year, the $50 million that you've discussed, that $210 million would then be down by that amount or whatever extent you do the buyback?

Yes, absolutely. Page 32 in the slide deck provides a very in-depth kind of reconciliation, if you will, on how we arrived at our numbers. You can see all of the elements in their and we provide the debt service coverage and the like on that particular slide.

Great, thank you. That's very helpful.

Scott Frost, HSBC.

Just a couple of things -- could you -- I had some housekeeping stuff. The notional amount of FAS 133 liabilities outstanding that are included in the $487 billion of total notional, I might have overlooked that. Could you also remind us of the cumulative amount of true credit impairments included in the derivative liabilities, if I overlooked that there?

What I'm trying to make sure of is you are saying that October 1 is when you expect to be open for business with Connie Lee. You expect rating agencies to have completed reviews and if you can't get AAA, you expect a comment then or earlier. Have I got that right?

Yes. If we should run up against the wall with the rating agencies, which I do not anticipate, but if that should happen, I think that we would certainly tell you about it. We'll tell you what our plans are then.

Okay. The other question I had is, is it true -- I mean, the way I'm looking at the Citigroup commutation, they took less in consideration than what they would have been owed under the original terms of the contract, which raises the question of why they did that. You said, if I'm not mistaken, that you would have been subject to considerable loss payments, absent the commutation. In other words, it looks like they took less than they were entitled to because they were afraid that you might not be able to honor the original terms of the contract. Would that be a fair description of what happened? If it isn't a fair description of what happened, how would you describe it?

I will let David describe it because he does it very concisely (inaudible).

That's pressure.

I think the point you're making on a counterparty's opinion of the particular enhancer, in a sense, regrettably, (inaudible) that doesn't really apply to us. I think that perhaps to other, less fortunate participants in the industry, you could make that point.

I think the issues here are much more complicated. Obviously, we believe we will be around; I think others absolutely believe that too. I think it gets complicated in relation to how different entities have accounted for particular positions, what their view of the underlying cash flows is, particularly in terms of timing.

There's another variable. Quite often I think, when transactions were intercepted, perhaps a counterparty took about a hedge on the counterparty risk to ourselves. Clearly, if they did that, that, then despite the recent dramatic reduction in our CDS spread, they're on a very, very large gain. So the combination of a check, i.e. cash -- obviously with 0 discounting applicable to that -- and the ability to close out a [CDS] position at a very large profit can be an enticing one. I think there are all sorts of reasons why you have to look at the particular situations in terms of what this means for different counterparties.

Let me rephrase it a little bit here. Absent the commutation, would you have been able to pay these obligations as agreed?

Absolutely, 100% yes.

Okay, thank you.

Yes, in fact, it was quite close to our first-quarter impairment number, and we disclosed that in our press release.

Relating to your question on CDO impairments, I would direct you that the number, the cumulative number, as of June 30 is $3.1 billion. Clearly, we paid $850 million, so that number will go down. We are paying down some of that estimated and credit impairment liability, if you will, so that's the number. Page 18 of our operating supplement provides those details.

As it relates to par, we also have a very in depth slide presentation on our Web site that provides the numbers for overall CDOs and exposure. That number is slightly above $60 billion.

Yes, I had $63 billion a quarter -- in the first quarter. You said that is in the supplement. The CDO are -- that's everything that is related that's FAS 133? That's right? They are the same number; is that correct?

Yes, it is. We also footnote the commitment that happens to be for peculiar reasons; that also is included in our FAS 133/FAS 157 calculations.

Yes. Again, you're saying that of the 7 -- am I right in looking at this? The $7.4 billion of derivative liabilities you've got on the balance sheet at the end of Q2, and that includes the $3.1 billion of true impairments, less the $850 million you paid? That's what I'm trying to get that, that number, what's real in terms of the losses you're going to have to pay, and what do you view as something that's going to revert over time back to gains? Do you see what I mean?

Yes, that's where the reconciliation on Page 18 of our supplement comes in.

I got you, okay. Yes, I don't -- all right, thanks.

(multiple speakers) I told you and I just point you to that. You can look at it at your leisure, but effectively the differential, if you will -- because the derivative liability on the balance sheet includes some interest-rate swaps and some other things, but if you want to just isolate credit default swaps, the kind of unrealized mark, if you will, that's on the balance sheet, that is not impaired, is about $3.7 billion.

Great. Okay, thank you.

Bob Ferguson, Barclays.

Yes, regarding the remediation efforts, considering that you are expecting approximately a three-year recovery time for those efforts, and they will obviously meet significant legal challenges from the originators, do you really feel it's wise at this point to actually reduce your reserve today? Why not put that reduction of the reserve off a little further until you actually see some actual results from putbacks?

No, we absolutely feel it's wise; we wouldn't do it otherwise, clearly. I don't really see it as a whole lot different from recoveries in other sorts of transactions. We've been very diligent; our auditors have been very diligent. Candidly, we have some historic experience of these matters which I'm not permitted to go into, and we looked at that as helpful guidelines in relation to what we should do here.

As I mentioned, we've I think undertaken a pretty small sampling, in relation to the total anticipated potential gain here. We are covered many times over in terms of the reserves that we've taken. So no, all in, we feel very good about it. We absolutely acknowledge that people aren't running towards One Street State Plaza with open wallets. That's why we've discounted the projected proceeds by three years. As I say, there's great coverage there. So no, we feel very comfortable about it. We are very realistic about how tough it's going to be, but we are very good at this and we will pursue it intensely.

[Nick Maguire], [Pershing Square].

Can you discuss the $850 million payment that you made on the commutation payment? How does that cash payment compare to the actual cash payments and the timing of the cash payments that you would have expected to make on the contract otherwise?

I will take a go at that. Obviously (inaudible) Sean.

This is a very unusual transaction. Let me briefly explain why. The structure of the deal is such that, once the existing, or the pre-existing when the deal was around, 30% subordination was eaten through, then for each successive inner CDO event, credit event, then we would have been on the hook, at that point, for the principal payment of the relevant charge. It was about $40 million a go, I believe, in terms of each tranche.

As I indicated in my prepared remarks, circa 98% of the underlying inner CDOs were noninvestment grade. As we know, in the marketplace, liquidity is returning, and what that means is there are buyers and sellers and obviously liquidations. Obviously, that would have been and was extremely detrimental to this transaction.

We go through, as we do in our remaining CDO Squares and the high-grade deals, all of the constituent CDOs, looking at the terms of each CDO in relation to the difficulty or ease or relative closeness to liquidation-type events. When we did that, we felt very confident that we would be paying several hundred million out over the remainder of this year. So, we think that was an unusual thing in relation to this particular deal.

Clearly, as we've discussed previously, our normal transactions are timely payment of P&I and therefore in effect much more beneficial in terms of our business model, as I alluded to.

Right, but if you were going to pay $700 million or several hundred million over this year --

Several, not seven.

Right, sorry, several hundred million over this year, that's just this year. What were you expecting to pay over the life of the contract?

Well, we think that, when you've got 98% of the underlying CDOs as noninvestment grade, I think one can expect to pay pretty damn close to $1.4 billion over the course of time. The liquidation element of this transaction, meaning that it's not a timely [IMP]-type deal, means that you haven't got the benefit of effectively spinning the process out to 10, 20, 30 years, which obviously makes it a much more advantageous transaction for us to settle, and that's what we did.

What would the timing of the -- if you would have paid nearly close to $1.4 billion on this contract, what would the timing of that have been?

Well, as I said, over the remainder of this year, we have spent several hundred million. I think the number was sort of $280 million, $300 million, with an even bigger sum, I believe, over the next year.

So, would it be accurate to say, then, to the extent there were analyst estimates assuming a 100% loss on this, that those estimates were actually fairly accurate?

No, they are not accurate because I believe the present value is kind of important. What we've done here is to pay I think a very mutually advantageous sum of money to us and our counterparty, given their position, to make a deal that makes sense.

I think many people -- I don't know what your views were -- would have anticipated a complete $1.4 billion write-off today, tomorrow. I think what we've done has invalidated that proposition.

[Jason Spindel], Aurelius Capital.

This is actually [Eleanor Chan] from Aurelius Capital. My question is that, going back to the remediation efforts on the RMBS transaction, it seems like you guys are expecting a very high probability of success in terms of your remediation efforts, so I'm just curious. Like what stage of negotiation are you at with your various originators? Also, does an originator have to know of the fraud in order for them to buy back the loan? Meaning, do they -- like what is the standard in terms of buying back loans from pools because of a breach of reps and warranty? (multiple speakers)

Sure. The kind of technical standard is that they are supposed to -- generally, which is I think a reasonable statement to make -- is that they are supposed to buy back or replace loans on the earlier of 90 days, or when they knew or should have known about the respective breached loans in terms of reps and warranties.

Going back to the first part of your question, where are we in the process, particularly in relation to the different counterparties, what we've done is we are in the process and have submitted very detailed breeched notices, loan by loan, pointing out that we have a bus driver earning $460,000, or somebody is apparently working in a McDonald's in an owner-occupied mortgage 200 or 300 miles away from the mortgage property. So you really have to get down and dirty and provide very firm evidence and then await a response. I think, right now, we are in the kind of holding period, for the most part, in respect of awaiting a response. We think the response will be inadequate, and obviously, you move through the legal process from there.

Okay. Then also, regarding the -- I think you mentioned that the expectation right now is that you would settle for $0.30 on $1. (multiple speakers).

No. To be clear, I didn't make that statement. What I said was that, in certain cases where the breaches already on the work we've done cover our reserves by three times, then in fact we could afford in a sense to settle for $0.30 on $1 and prove out the reserving position adequately. Obviously, we are not necessarily going in there saying "Well, we want to settle at $0.30 on $1." That isn't what the deal documents say.

Okay. So what net par or what reserve number does this $260 million correspond to?

Well, it responds -- because you can't make a gain, you know, clearly you can't take remediation gains in excess of the reserve that you've posted, so the answer is the identical number.

Right, oh, okay, so I see. So you are expecting a gain exactly in the amount of your reserve?

Precisely, in the cases where the coverage is one or higher. But clearly if we have a reserve of $100 million -- and thus far we've only substantiated breaches of $20 million -- then the reserve would be $80 million. If we have a reserve of 100 and substantiated breaches of 300, then the projected gain in that sense would be 100, because you can't take a gain. Obviously, the trust would gain, the subordinate bondholders would gain, but we can't be and neither should we be over-remunerated as against our policy commitment.

Just one last comment -- it's related to the transactions. To be perfectly clear, the remediation elements are related to the transactions that we have had reserves posted against. So when you look at the first-quarter reserves, the 260 relates to those transactions.

Now, there could be elements where we could choose to look at remediation for other aspects of the portfolio to obviously look to provide additional cushion against loss, but for clear reasons, we focused on the most problematic transactions in order to get that process going and recoup the amounts as quickly as possible.

Then is it fair to say that, with respect to that $260 million, you are not assuming any, like, discounts, settlement discount, or you're not assuming anything for, you know, if the originator has gone bankrupt or is in receivership, that the recovery from the originator might be less than par?

No. I mean, clearly, the first thing you do when you are contemplating action of any sorts in these matters is (inaudible) you look through the reps and warranties to make sure that they are substantive and, B, look at the entity who is sitting behind them. So for a bankrupt entity, don't waste your time. So we are not looking at that with any hope or expectation, and none of the numbers that we've talked about this morning are, in effect, from anything other than what we firmly believe to be creditworthy counterparties.

Okay. I have a question on the Financial Services segment. Has the insurance company taken any reserves on any potential losses that the Financial Services segment might take as a result of ratings downgrade, which triggers termination and lateralization requirement? Because it seems like the Financial Services (inaudible) portfolio is $1.5 billion or so under water, so I'm just curious if the insurance company which ensures all of the GIC contracts, whether the insurance company has taken any reserves for that part of the portfolio.

Yes, I can answer that question. We've looked at it from a number of different ways. One would have to look at -- we have to employ several different accounting regimes here.

Let's start with statutory. From a statutory perspective, one would need to recognize a reserve, if you will, for any case reserves for defaulted items. We clearly don't have that in this particular case, being the obligation of the insurance company, Ambac Assurance, to its affiliate, the GIC businesses. So that would be one element of the calculation.

If we were to have to, say, provide unsecured lending, one would have to look at that asset and whether or not that asset that has been provided the loan, if you will, would be non-admitted. So one would look to the resources of the Company at that time, in this situation, and make a conclusion there. Most likely, the conclusion would be that would be a non-admitted asset. So that's kind of the statutory aspect of it.

From a GAAP consolidated aspect, clearly we don't have this issue to deal with. That full mark to market has been recognized either through the income statement as other-than-temporary impairment or through our equity section because the entire portfolio is valued at fair value. So that's all recognized on a consolidated basis, that decline.

We also filed a subsidiary GAAP statement in our SEC filings of the insurance company consolidated. One would look at that and our view of that has been that we would look to a book value basis. If there were to be further declines in our credit rating, we might have to revisit that, but at this point in time, the book value -- we provided a chart, I slide on Page 45 of our presentation to address that. The book value of the business, based upon detailed reviews of potential for tranche losses of the at-risk securities, mostly the originally AAA, Alt-A RMBS that we have in our Financial Services portfolio, yield very little tranche losses. They have suffered significant market value declines, but the detailed analysis we've done, including using some third-party experts and analysis to also take a second look at that, would support our view. So, we have not recognized anything at that level either.

Okay. So it sounds like it's only at the point where you really anticipate a payment of a loss when you would actually recognize a reserve. Is that correct?

Well, as I stated, the payment of a loss is obvious but there might be a need, if we were to be downgraded, to -- and we have collateralization requirements, particularly at the single-A and the A3 level beneath that, we might be required to make some type of loans from the insurance company to provide the collateral. There might be a consequence of that particular point in time, as I described.

[Alastair Lumdsen].

It seems, on the high-grade CDO portfolio, you've taken around about a 5% impairment. I recall, from earlier quarterly calls, that you've always made a very strong point about how you only have to pay principal out on those deals at final maturity. Given that it's so far away from the market value of high-grade CDOs, can we assume that those are very unlikely to be commuted?

No, and let me just correct you. You know, the principal payment is the earlier of legal final or when there's no collateral, so meaning 0 collateral in the underlying deal. So, that's the kind of basic rule for virtually all of the transactions that we have.

I think the question of commutation, and I'm not going to go into this in too much detail, we remain, you know, very active in the area but I think, if you think about it from a counterparty's perspective, a lot depends on where they've booked particular transactions. It's pretty clear that the markets are increasingly rewarding certainty with respect to any of these assets, be those assets effectively contingently on our books or via the policies or CDS contracts that we've issued on others' books. So I think there are all sorts of incentives in terms of equity, discount rates and the like, to settle. In addition, to the extent that different counterparties entered into counterparty hedges to protect their counterparty exposure against us in respect of the dealer conception, there are substantial gains. So I think there are all sorts of things going on here. I think the deal we settled was kind of more obvious in some ways because of the cash flow timing profile that we've talked about previously.

I think it's time that we finished; we've gone over the time. The questions have been very good.

I'd like to reiterate one last point I made earlier just as a closure here and continuation of David's comment. There will be no commutations unless they are economically justifiable for our shareholders and policyholders. Rest assured, we get many calls that start out by saying "You are desperate and we are here to help." Then there is a varying degree of time that they come back and realize we are not maybe so desperate, and then serious discussion ensues, and we either come to something that we think makes sense, mutuals sense for ourselves and the counterparty, or not.

So I just wanted to close by saying we are not desperate to do deals just for the sake of doing deals. They have to be good deals. We expect the market is intelligent and would be able to make that judgment.

If there are any further questions, we apologize for not getting to them, but I urge you to call us directly. We will be available the rest of the week, particularly Pete Poillon and Vandana Sharma, as well as Sean, David and myself will be very happy to take your questions and appreciate your time and your interest. I wish to thank you very much on behalf of the Ambac management and the team here.

This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.