Radian Group Inc (RDN) 2007 Q2 法說會逐字稿

Welcome to Radian's second-quarter 2007 earnings conference call. At this time all participants are in a listen-only mode. Later we will conduct a question-and-answer session. (OPERATOR INSTRUCTIONS) As a reminder this conference is being recorded.

I would now like to turn the conference over to our host, Ms. Mona Zeehandelaar, Senior Vice President of Investor Relations and Corporate communications. Please go ahead.

Good morning and thank you for joining us today. With me on the call are S.A. Ibrahim, Chief Executive Officer of Radian; Bob Quint, Chief Financial Officer; Mark Casale, President of our Domestic Mortgage business; and Steve Cooke, President of our Financial Guaranty business.

As I do every quarter, let me remind you that today's conference call will contain statements that are forward-looking. As you know these statements are based on current expectations that are subject to risks and uncertainties and Radian's actual results may differ materially from those expressed or projected in these forward-looking statements.

Factors that could cause Radian's actual results to be materially different than those in the forward-looking statements are described in the Safe Harbor statement that is included with our webcast and in the risk factors detailed in Item 1a of part 1 of our annual report on Form 10-K for the 2006 fiscal year and in the joint proxy statement prospectus for our 2007 annual meeting.

Today S.A. will begin with some remarks and he will be followed by Mark who will comment on the NIMs product. Bob will then discuss NIM reserving and will review his normal key financial metrics for the quarter. And then we will take your questions. In the interest of time today, I ask that all participants please limit themselves to one question and a follow-up if necessary. Then please return to the queue. This way we can give as many people as possible the opportunity to ask questions.

For those logged onto the webcast at www.radian.biz, the slides are provided as background and should complement our remarks today. Now it is my pleasure to introduce you to S.A.

Thank you, Mona. And thank you all for joining us today. Let me begin by saying that our second-quarter results clearly illustrate the credit challenges in today's mortgage market. But I believe they also reflect positive trends for the business. We have three key objectives for today's call.

First, to review the quarter's financial and credit results including the NIM book and comment on the external environment. Second, to emphasize Radian's financial strength, stability and diversification and importantly our ability to withstand the pain of the current credit and housing cycle. And third, to highlight the positive trends for our mortgage insurance business.

Net income in the first quarter was $21.1 million and diluted net income per share was $0.26. In spite of an extremely challenging environment, our book value per share increased 10% from $46.95 a year ago to $51.53 at the end of the second quarter and is stable from year end. Adjusted book value rose 8% year-over-year from $65.61 to $71.14. As a reminder, adjusted book value is a non-GAAP financial measure and a reconciliation to book value is included with the webcast slides available at www.radian.biz.

Our balance sheet remains solid with a highly rated investment portfolio of more than $6 billion and total loss reserves of more than $900 million. While our mortgage insurance business results were disappointing and I will address this in a moment, our financial guaranty business continues to demonstrate the benefit of our diversification by contributing net income of $22 million this quarter.

The Radian share of Sherman's second quarter reported pretax profit of $71.6 million was $26.5 million. Our share of C-BASS as reported pretax profit of approximately $50 million was $23 million. Bruce Williams and John Draghi from C-BASS joined MGIC's earnings call last week and provided valuable insight into the difficult market conditions they are facing. Bob will provide an update on C-BASS in his remarks.

For our mortgage insurance business, credit results were disappointing this quarter. There is no denying that today's environment is more difficult than anticipated, perhaps the most challenging environment we have seen in a long time. While our paid losses of $112 million were within our range of expectations, incurred losses of $180 million were higher than expected. The rising interest rates and home price declines were the primary drivers of the increase in default rates across the board with primary areas of deteriorating performance in California, Florida and several midwest states.

The ARM portion of our portfolio had the most impact on our default rate this quarter as resets forced many homeowners to default without the ability to refinance their mortgage. In California and Florida in particular, where loan amounts are higher and the Alt-A product more prominent, we saw an increase in the default rate and average default size. In the recent past, these states have shown unusually low credit losses and defaults.

This quarter we experienced $61 million of pretax losses on our NIM book of business driven by the weakness in the subprime market which Mark and Bob will discuss in more detail.

While the quarter was disappointing, since its inception NIMs have been a very profitable product for Radian with approximately $240 million of cumulative premiums booked to date and it's a product we plan to continue to write prudently dependent upon market conditions. This morning Mark Casale will elaborate on NIMs followed by Bob's comments on our reserving process and the quarter's financial results.

Our Alt-A business experienced an increase in its default rate from 5.7% at March 31 to 6% at June 30 and is flat with year-end 2006. It is important to note that our high FICO Alt-A business which represents over 80% of our Alt-A risk in-force had a default rate of 4.7% in the second quarter which compares favorably to our prime book of 3.2%. We continue to be disciplined in our approach to this business choosing our partners carefully and maintaining prudent underwriting standards.

ARMs continue to be a declining portion of our portfolio. ARMs with the reset terms less than five years, decreased from 23% of primary risk in-force in the beginning of 2006 to 16% this quarter. While we continue to see negative reports on subprime and other mortgage related performance as a headline, we are encouraged by continued strong employment statistics and reports about the overall economy.

As an insurer of credit risk, these are difficult market conditions for bottom-line growth. But history demonstrates that this is fundamentally a cycle; one of the many that the mortgage business has seen and one that's going to be with us for a while. It is important to remember that we are in the business of taking credit risk and we prepared for it by diversifying our risk and protecting our book with captive reinsurance and Smart Home, which comprised nearly 50% of our primary risk in-force as of June 30, 2007. And we believe we have priced appropriately for the risks we took.

Those of us who have been in the mortgage industry over the past several years have seen a market where competing products threatened the viability of mortgage insurance and the long-term franchise value of our Company. While we are feeling the pain in today's environment, this pain has created a more desirable environment for our products which should translate into increased franchise value.

Heightened awareness and concern over risks in the mortgage market are generating a renewed demand for mortgage insurance. This comes at a time when spreads are widening, lenders are tightening their underwriting standards, the subprime market is shrinking and greater value is being placed on MI over alternative products. This led to an increase in [slow] new insurance written of 60% over last year reaching $10.6 billion in the quarter. Persistency also improved to 71% versus 63% for the same period last year.

Turning to our pending merger with MGIC, most significant regulatory and stockholder approvals have been completed with only the approval of two state insurance departments remaining. As MGIC mentioned in their earnings call last week, we expect that the approvals will be complete and that the deal will close late in the third quarter or early in the fourth.

Mark will now provide additional details on NIMs. Following his and Bob's comments, we will take your questions and I will close with a few remarks. Mark?

Thank you, S.A., and good morning everyone. Before I discuss our existing NIMs exposure, let me take a step back and explain the product. NIMs are cash bonds. They come from subprime mortgage securitizations and bondholders are entitled to receive certain cash flows that the mortgages in a trust provide generally from the excess interest and prepayment penalties. When we wrap a NIM bond, our policy covers the principal and interest shortfalls on a AA NIM note. The underlying bonds that we insure are typically rated BBB or BB based on the amount of subordination and other factors.

These transactions are insured based on the collateral quality of the underlying portfolio as well as the merits of the structure that effectively provides adequate cash flow from the underlying transaction in order to offset the risk of prepayments and expected losses. Our NIM risk is generally concentrated within the first few year's performance of a deal and can be significantly impacted by both the amount and timing of losses during this period.

In addition, our NIMs can be impacted by the timing and amount of prepayment penalties placed to the trust and the over collateralization which is contingent upon both delinquency and loss levels. We have been injuring the NIM bond since 2000 and have booked approximately $240 million of premium to date. As of June 30, we have $795 million of risk in-force pertaining to our NIM portfolio across 42 deals.

As noted in our press release, during the quarter we recognized a $61 million mark-to-market write-down. The majority of this write-down is being driven by the poor credit performance on the underlying securitizations on three deals and it is our expectation that we will have to pay claims on these particular deals over the next few years.

These underlying deals are experiencing losses a lot sooner and in excess of our expectations. This has the effect of decreasing the amount of excess spread that is available to pay down the NIM bondholder which may result in a claim or potentially the NIM bond extending. The majority of the increase losses to date are being driven by the amount and poor performance of the second lien collateral that was included as part of the underlying collateral which was securitized.

It is well-documented in the market that delinquency and loss trends have been very poor for the 2006 subprime vintage. These trends have been further aggravated by poor second lien performance whereby the second lien losses are being recognized a lot sooner within the securitization structure versus the first liens. With first liens, it takes a certain amount of time in which collateral has to be foreclosed upon and ultimately liquidated. During this time period, mortgage servicers normally continue to advance cash flow to the trust which is then used to pay down the NIM bond.

The rest of our portfolio is performing within our expected range. We continue to perform surveillance on our NIM portfolio and will update our valuation in the third quarter including loss and prepayment assumptions to reflect as best as possible our expectations concerning the performance on the underlying deals at that time.

In addition during the first quarter of 2007, our loss in prepayment assumptions were updated to reflect the state of the industry and our views on future credit performance. In 2006, our expected loss was approximately 4% to 5% versus current expectations of 6% to 8% including a more front-loaded loss curve. These are also on deals with both better underlying collateral and increased enhancement levels which increased from but the 1% to 4% range in 2006 to the 3% to 7% range in 2007. In general the business written in 2007 can withstand 1.5 to 2 times our increased loss expectations.

And now I will turn it over to Bob.

Thanks Mark. By virtue of the some of the features in the NIMs, they are accounted for as derivatives and therefore premiums are broken out separately on our P&L and all other P&L activity on the product is run through the change in fair value of derivatives line. That is why we have repeatedly told everyone that although most historic activity on the fair value of derivatives line has been spread and price related and not credit related. If there is ever a real credit loss, we will always break that out and highlight it separately.

We have done NIMs since 2000 and there hasn't been any credit losses until the fourth quarter of 2006 when we had a $6 million credit loss which we spoke about and highlighted. This quarter the $61 million represents a real credit loss that we expect to pay claims on over the next several years. We go through the following process to set these reserves.

We've established a model that uses market observed data to set a spread or price related mark on the NIM and which is run each quarter. That mark has been and remains minimal. We also observe the credit performance of every NIM bond. If the performance is within our modeled range or better, we will most likely not be paying any claims. If the performance is worse than expected, we rerun all of the bond cash flows, incorporating the actual performance and we project what amounts, if any, we will ultimately pay out. That is how we arrived at the $61 million on the three bonds.

As of June 30, all other NIMs are either performing as expected or better or when we run the cash flows, we do not project that we will being paying claims. We continually update our analysis for actual performance and believe that we are always capturing our best estimate of what will happen in the future at any point in time.

And now as always, I will be providing some detail about our second-quarter 2007 financial results. MI premiums are up from last quarter due to the insurance in-force growth resulting from higher persistency and a higher penetration for the MI product in the mortgage market especially in a slow channel of business. We expect to continue to grow our insurance in-force over the next few quarters. Persistency was 71.1% for the year and 72.3% for the annualized quarter. The breakdown by channel was 65.8% for structure, and 73.3% for flow for the year ended June 30.

The reserve for losses on the MI side was up significantly from first-quarter levels. This reflects an increase in the number of delinquencies; larger loan balances on delinquent loans including more delinquent California and Florida loans, Alt-A loans and recent vintage loans; a continued aging of our delinquencies; and an increase in our projected [role] rates.

As for the second lien transaction that we have been tracking, our net reserve on that transaction was $29.5 million compared to $28.2 million at March 31. To date we have paid $36.4 million of net claims on this transaction including $15.4 million during the second quarter. With regard to total MI paid claims, we have paid $208 million for the six months and we expect that each of the next two quarters of claims paid will be in the $130 million to $135 million range.

We believe that our balance sheet loss reserve at $746 million for the MI business is extremely strong by any measure including reserve per dollar risk in-force, reserve per loan in default and number of quarters of claims coverage. MI operating expenses have continued to decline. Please note that our merger expense line is shown separately.

Financial Guaranty had another solid quarter. Much like the prior two quarters, Conseco and Trade Credit paid claims represented substantially all of Financial Guaranty claims paid in the second quarter. Similarly to the first quarter, the negative incurred loss line in Financial Guaranty came primarily from a reversal of trade credit reserves on exposures that had expired.

With regard to exposure to subprime mortgages in our Financial Guaranty book, we have a total exposure of $561 million to subprime RMBS deals consisting of 175 million of direct business and $386 million of reinsurance. This represents one-half of 1% of our total net par outstanding in Financial Guaranty as of June 30. The direct business is all from 2004 and prior.

Approximately $111 million of the $386 million of reinsurance is from the 2006 vintage and the underlying ratings of such business is 43% AAA, 11% A and 46% BBB. Only a tiny fraction of such reinsurance business is on our watchlist as of June 30.

Sherman had a typically strong second quarter and we expect similar strong results throughout the year. For C-BASS, there was a return to profitability as spreads briefly stabilized in the second quarter. Last week C-BASS closed on its acquisition of Fieldstone which used about $90 million of their cash.

To support C-BASS's liquidity position, Radian and MGIC each provided C-BASS with a $50 million credit line, $100 in total, which was fully drawn over the last week. The current subprime mortgage environment with continuing margin calls has drained cash resources and challenged liquidity for all market participants including C-BASS which is currently pursuing a number of options to help strengthen its liquidity position.

In the second quarter, a widening of corporate credit spreads was the primary cause of a negative mark-to-market result in our financial guaranty corporate CBOs of approximately $32 million which is contained in the fair value of derivative line, none of which we consider credit related. We also booked a $10 million mark-to-market loss in our book of MI credit default swaps which is also in the fair value of derivative line, none of which we consider credit related. The remaining $61 million relates to the NIMs which we went through in detail.

With a respect to stock repurchase, we repurchased over 300,000 shares during the quarter and we do not plan on any further repurchases prior to the merger closing.

Finally, please remember that preservation of book value is a key factor in the way we run our business and has led to transactions like Smart Home. Despite the negative credit cycle that we are in, we are confident in the stability of our book value, with a strong balance sheet and conservative loss reserve. We are confident that our book value can be preserved throughout the cycle and that the foundation for future book value growth is currently being created with our deeper penetration into the mortgage market and our ability to attain stronger pricings in a market that is currently demanding our product.

We would now like to turn the call over for questions.

(OPERATOR INSTRUCTIONS) Rob Ryan, Merrill Lynch.

Good morning. With CDOs as such a hot topic, I was wondering if you could go over the nature of your portfolio, the heavy emphasis on corporate bonds as opposed to CDOs of ABS?

We would be glad to, Rob, and Steve Cooke is going to take that question.

All right. The overwhelming majority of our portfolio consists of corporate CDOs as opposed to CDOs of ABS such that when you look at the exposure vis-a-vis the subprime that we are talking about, what it constitutes, that $560.7 million which is our total subprime prime exposure in the RMBS sector, that is our total exposure overall is $110 billion. The exposure within this [CDS] line is predominantly corporate, overwhelmingly corporate. I think we have three CDOs of ABS deals on a direct basis which are primarily RMBS but not subprime RMBS.

Okay. But spread widening still led to a mark-to-market unrealized loss --

That's right. That was $32.7 million for the FG line, but again, I'm not tied to as Bob said, any direct credit issues in the portfolio. It is solely tied to the widening of spreads.

Could you on that issue even though it is not directly credit related, could you let us know about your tolerances for how much net income volatility you are willing to accept and how you mitigate that potential volatility through attachment points and other issues about the magnitude and nature of the credit derivative exposures you guarantee?

Rob, it's Bob. We, clearly we take the risk that we are going to have, marks in any given quarter based on spread changes. I think a couple of things mitigate that. One is we play very, very high in the stricture so our risk is extremely remote. Typically it is several times the AAA attachment so there tends to be less spread volatility in that area.

The other thing is we do set limits on the amounts of exposures and the amounts to each obligor etc. But clearly by being in the business of wrapping corporate CDOs, we are taking some risk and although we do believe it is credit risk, you saw this quarter that some corporate spread widening caused the mark this quarter.

Great, thank you.

Geoff Dunn, KBW.

Thanks. Mark, I was hoping you could go back over the NIMs. A lot of the sentiment out there is that a lot of the subprime prime '06 vintage is going to be wiped out. And these are securities that are actually collateralizing. As I understand it, the excess spread of a subprime RMBS deal. How do you get comfortable with the underwriting of that? Why shouldn't the market be afraid of a much bigger notional exposure than you marked down this quarter? And what specifically the structure that keeps you comfortable looking at that going forward?

Those are good questions, Geoff. And I think -- I just want to remind everyone but it's not just -- when we get paid back on a NIM bond, we get paid through not only excess interest, but prepayment penalties are pledged to the trust. We also get paid based on the release of over collateralization. So there is really three different ways that we can get paid, Geoff. So when we look it, how do we get comfortable? A majority of the risk that we wrote that is on the books now was in the last half of 2006 and 2007.

How we got -- one of the things we did to get comfortable first was to pick our partners. 97% of the risk in-force that was written in the last three quarters has been with our large national accounts. We made a choice to kind of diversify away from the monoline subprime guys that caused a lot of the 2006 subprime vintage and surprise surprise, that is where the problems are in the NIM book. So we effectively are not working with those guys anymore. So that is one; picking your partners is paramount and we made that adjustment in the third quarter of 2006. And I think we said that either on the call before or the one prior to debt.

And second, it is really looking at the ultimate default rates and the duration of the loans. And as I mentioned at the tail end of my comments, we took off our loss assumptions pretty significantly in the first quarter of 2007 to reflect our changes in what we saw in the 2006 vintage.

So I think those are questions and then from the performance of the bonds to date, as Bob mentioned, all are within our expectations. And these things pay down pretty quickly so we will know even within the next three to six months -- to see that risk continue to pay down.

And if these bonds fail their step-down test and you don't get a release of OC, have you run that scenario and what the potential severity issue could be (multiple speakers)?

Of course and we (multiple speakers).

-- strong enough to sustain their performance past that?

Yes, we run it across many scenarios. We run it to make sure that we are fine even if the OC does not step down. And for instance in one of the deals we did in the first quarter, the delinquency trigger for OC not to step down or to fail is close to 40%.

Okay.

They are pretty high.

Okay. S.A., just to follow up on the merger. Obviously the stocks are trading in new territories than we've ever seen. And you are starting to get speculation once again that maybe the merger could be at risk not necessarily even because of what is going on fundamentally in the companies, but because of how far the stocks are falling. I'm not sure how you can address this, but is there in your mind any risk of this merger not going through due to the movement in your stock price?

All I can say there, Geoff, is that we've been working very closely with the MGIC team. We've now been operating on not only making sure neither of us get surprise but also working on the integration planning and like we expect, like Curt said in his call, we expect that our merger will close in -- by early fourth quarter.

Okay, thanks.

David Hochstim, Bear Stearns.

A couple of things. I wonder, Mark, could you just repeat again what you said about the coverage ratios? And then I guess wonder if you could relate the premiums associated with the deals that you took the $61 million credit adjustment on?

I'm not sure I understand the first question, David. I will answer the second one. I don't have the numbers for the deals with me right here but we expect on the whole portfolio to earn approximately $60 million in 2007 on the book.

What I was referring to in the first question was you made some reference to coverage ratios and having adjusted for higher collateral levels and I wonder if you could just repeat what you said?

What we said was just in 2006 we expected, we had about a 4% to 5% cumulative loss expectation and over collateralization levels and the deals back then were probably in the 1% to 4% range. In 2007 we took that up -- we took our loss expectation up to the 6% to 8% range and we also front-loaded some of the loss assumptions to earlier in the structure which obviously stressed it more because remember we get paid within the first 24 to 36 months. And the OC levels went -- they went up in the market from -- because of the rating agencies really to the 3% to 7% range.

So from a cash flow, [MTB] of cash flow standpoint, we feel like we're in a much better position than we were for similar deals 12 months ago.

Because basically the collateral is higher (multiple speakers)?

And quite frankly that is why we've written more business in the first -- in the fourth again, to reiterate strategically where this goes, we said we are going to target the top lenders in the country and we're going to work with them across all of their needs obviously making money for rating in the process. But we felt like when spreads started to widen in the fourth quarter and the first quarter, the opportunities to both price and structure properly came to us and we believe we took advantage of them.

Okay. And then I wonder if Bob or someone could talk about the kinds of coverage you are getting from captives? And remind us about the coverage you I guess could talk about Smart Home? And sort of is the protection you get from Smart Home and [deep cede] excess or loss captives concentrated in higher risk business or is it kind of distributed like the portfolio?

Dave, it's Mark. I can comment just on Smart Home in particular. Our insurance in-force at June 30 for subprime was $11.6 billion; approximately $4.5 billion is covered through Smart Home. And just to give folks some direction on when that would attach, right now if frequencies of loss hit in the probably the low teen range, say 12% to 14%, Smart Home begins to attach. So under again, a catastrophic scenario where frequencies go to 10% or 15% and cumulative losses are in the 10% to 12% range, subprime becomes a pretty significant contributor for our reinsurance.

And captives?

Yes, Dave, I think approximately 50% of her primary risk in-force is covered either by Smart Home or captive. The captives generally are attaching at lower levels than that. So if -- but that is mostly prime business although that includes some other business. If those -- if loss levels in our prime business increase, they have to increase more for the majority of the captives to attach. But if they do we've got much of our risk covered by captive.

Okay, so if we start to see even substantially higher losses, then the captives would kick in, so there is some cap?

Absolutely.

On that portion of the business?

There is definitely protection. The captive on much of the prime business that we've written.

And of the 50% that isn't covered by Smart Home or captive, is there some characteristic about -- something you could say about that business what kind of bonds are those?

On the subprime side, I think the majority of it is either pre-2005 so it is earlier vintages that are performing fine, David.

Okay.

And a lot of the other stuff that is not included in the captive would be our modified pool with deductibles and things like that that we've written a lot of and we include within our primary risk in-force.

Okay, thanks a lot.

And just again, this is just adding one thing around the modified pool which is predominately around our Alt-A. Our coverage ratio on our Alt-A book, our insurance in-force is $35 billion at the end of June. Our coverage ratio is approximately 14% which is significantly lower than our prime book which is 24% to 25% range and also our subprime book which is in that same range.

Steve Stelmach, FBR.

Thanks. Good morning. Just on the $386 million of subprime CDOs, I believe you were disclosing -- on the '06 vintage, a 43% at the AA, 11% A, 46% BBB. Is that -- that's the attachment point and you insure everything above that or is that discrete exposures?

Those are the attachment points.

So you do insure up to the AAA level or where does your -- where do you generally detach?

That business that we described the $386 million is reinsurance so it would be typically it would be a primary financial guarantor that would be directly insuring those. And would either be wrapping a BBB to a AAA --

So you're all the way up the capital structure?

Right.

Okay.

Just as a point of clarification, I mean we have ARMs, CDOs of ABS are all direct deals. We have no assumed CDOs of ABS with underlying domestic RMBS in them. So our discussion of that is just the direct deals that we have on our book -- where at least the attachment points tend to be 3 to 4 times the AAA attachment point.

Okay. And the second question, on your expectation for peak losses '05, '06 vintage, and I know you gave a little bit of color but could you rehash where you think peak losses or peak delinquencies come in over the next few years?

On all of our business typically we said that the prime peak losses are I think years three to five and on nonprime they are two to four. So I think those pretty much hold true. Now we've been seeing some earlier loss development especially in the subprime latest vintages. But those are typically what we've seen on a historic basis.

Okay. Great, thank you.

Mike Grasher, Piper Jaffray.

Bob, just to follow up on that comment. Would you expect your peak to shorten given the acceleration of the '06, '05 books?

It is hard to tell. Still pretty early Mike, but we've seen a more significant development early. Does that mean they are peeking earlier or the curve will look the same and it will just be higher? It is hard to tell right now.

Okay. And then I don't know if Mark or who wants to address this but just understanding the change in demand for MI and I guess the question is how are you approaching the market, your customers, to avoid sort of the pitfalls of the '06 underwriting standards? Or to make sure that this influx of volume that you are seeing isn't being pushed upon you and therefore resulting in lesser underwriting standards?

Mike, it's Mark. Those are very valid questions. I think we feel very comfortable on the amount of business and the quality of business that we've written in 2007. There was clearly a change in the market kind of with the February meltdown of subprime where folks really continued -- very quickly tightened their guidelines. A lot of those guidelines didn't tighten until -- they didn't come into effect really until April/May fundings so we are seeing it there.

And it's a surprise, our volume has actually increased in the second quarter. A lot of that is because of people just really, really -- I don't want to say fleeing -- where they embraced risk without fail in the last three years, now they are avoiding risk at all cost. So we are getting MI on loans that we wouldn't have gotten MI on two years ago because the folks were self insuring.

So taking a step back and looking at it from an actuarial standpoint, and we feel like the increase in volume and the amount of business that we are getting and the pricing that we're getting on it, we've tightened our guidelines across the board and continue to tighten them to try to make sure certain folks are not adversely selecting. So we're either limiting the amount of production they can do, cutting off certain programs that are raising pricing. But all in all, we feel very comfortable with the book of business that we are writing.

Thank you for the color on that.

Matthew Roswell, Stifel Nicolaus.

Yes, good morning. The question is actually on C-BASS. Could you go through the liquidity situation at C-BASS again? I understand you all providing or you and MGIC together providing the credit line to essentially pay for Fieldstone. What sort of lines do they have remaining in terms of unused capacity?

We are really, because they are in a -- the sale process that we're in right now it's not really appropriate to discuss the specific liquidity situation. But I think we just should reiterate that the whole market is going through a tough challenge with regard to liquidity and that includes C-BASS.

Okay. And I guess sort of as a follow-up to that. Do you absolutely have to sell down the stake in C-BASS to close the transaction or is there some way of taking the ownership down?

There is no requirement in our agreement to sell down the interest in C-BASS prior to closing. When we made the determination that we needed to sell down our interest, it was based largely on our achieving certain ratings that will (inaudible) rating agency comfort.

We are in an environment which is pretty unusual. We'd have to deal with the reality of the environment in the best interest of our shareholders and with the right kind of discussions with rating agencies and other parties.

Okay, thank you very much.

Andrew Brill with Goldman Sachs.

Thanks, just a couple of questions here. Just hoping if we could circle back up to C-BASS; just wanted to understand this a little bit better. Are they currently in the process of liquidating some of these assets at depressed values? Because if I remember correctly, they did talk about on the call last week that the mark to model effectively assumes that they hold the bonds to maturity. And to the extent they are forced to liquidate that could, I guess, dramatically shift the market values in their portfolio.

I don't know where you got that information. I'm not -- I don't think that was ever spoken about.

And they are not selling any bonds, Andrew.

Well, that is what I am asking, because their basic assumption is that they don't sell the bonds. And to the extent they are forced to, to cover margin cost, that obviously poses a big risk. My question to you was, are they in that position right now where they are forced to liquidate some of these positions?

No, they are not.

I guess just a second question, just related to the captive reinsurance programs. I mean, how much higher do 2005 and 2006 losses have to get relative to where they are today for some of those programs to kick in?

Well, it is still pretty early in those books. So typically, the attachment points on our captives are either 4% or 5%. So if you think about cum losses of 4% or 5%, that is where the attachment points are. Obviously, in the earlier or the most recent vintages, they are not there yet. But based on the current development, could they get there? They could get there.

Do you have a sense of based upon how they're currently seasoning, how quickly do they get there?

Andrew, it's Mark. Just to clarify, the 4-10 40 works where it's 4 claims out of 10, so it's a frequency rate. So the cum loss is really 1%. So in the 2006 vintage, it clearly could pierce some of the captives.

I guess one quick question. How are the 2007 books performing, I guess in the first six months, relative to the '06 and the '05?

It's Mark. The '07 book is performing, if not better than '06, but pretty much in line.

Thank you.

Mark Patterson with NWQ Investment Management.

Hey there, a couple of questions. First on the reserving issue with the mortgage insurance. With the paid to incurred at 62% this quarter, I mean obviously you are getting in front of what you are seeing in the delinquency trends and all the bad news that has been out there. But really over the last four or five years I mean that's been a number that has been consistently at roughly 90%. You kind of have to look back to the beginning of the decade and the recession period where you get under something even into the lower 80s or below.

And I'm just wondering if -- you talked about paid claims being a little bit higher over the next couple of quarters. It is the reserving -- is the paid to incurred ratio going to stay at this kind of a level? I mean you actually set aside $60 million, $70 million extra this quarter.

Mark, I guess we never give guidance on incurreds because that ratio will depend on one, new delinquencies that occur over the next quarter. And two, the development of the (technical difficulty) existing delinquencies. So if there is a stabilization, we can easily see that paid to incurred change dramatically like we've seen it go the other way. However, if we get new delinquencies and delinquencies keep rolling to claim at a higher rate and larger loan balances keep coming into the books, it's conceivable that the incurreds are -- remain significantly higher than the paid.

It really depends on the development of the delinquencies. But I agree, this is a very -- it's an unusual level of paid to incurred. But the reason is because we believe that it's warranted based on what we expect to pay out in the future.

Right, okay. Second question on the premiums, on the new business. I think people following the industry are understanding that there is better underwriting coming out of '07 and you are talking about getting higher pricing and I assume that is offset with some weaker expectations on home price appreciation or depreciation, if you want to call it that.

So I'm just wondering, the new insurance written growth has been accelerating very nicely and insurance in-force I'm sure will follow. But just on the premium revenue, the premiums earned, is there an expectation that we should see the new pricing start to affect that line more significantly?

Hey, Mark, it's Mark. Yes, we clearly -- and let me go back to the fourth quarter and the first quarter calls, we didn't expect premiums to grow much at all in 2007 and didn't really see an uptick until 2008. Clearly the premiums earned to date are higher than our expectations and I would expect that increase to continue over the next several quarters. It's again, with a lot of the credit issues, it has been painful, we'll admit. But I think as S.A. had mentioned, it really has provided renewed interest in the product and this renewed interest is really rolling its way into insurance in-force.

This is the first time in a while where NIW and persistency are both rising. So we are again going through a difficult period, but setting the seeds for higher premiums earned in the 2008 and 2009 timeframe. And you saw a little bit of it this quarter and you'll see it over the next coming quarters.

Great. Thanks a lot.

In fact, Mark, I mean that is a great question, this is S.A. If you look at it -- if you put it in perspective, we don't like the pain we're taking right now but that pain we are taking right now is the admission ticket to what is creating the conditions for a long-term, stronger business franchise that is more valued by our customers.

Understood. I think the pain that is being felt in your stock is -- the silver lining I'm kind of licking my chops at the merger completion and the accretion available from the buyback. But that's the downside protection. Thanks a lot.

[Oscar Wu] with Barclays.

Hi, can you could you again why the credits reserve for the NIM is included as a FAS 133 adjustment?

I'm sorry, can you repeat that please?

Can you go through the accounting for why the credit reserve for the NIM was accounted for as a FAS 133 adjustment?

Yes, the product, the NIM product, has features that under the accounting rules require it to be treated as a derivative. Now if you think about our Financial Guaranty business, much of our Financial Guaranty business is also treated as a derivative although we look at the business as financial guarantor, guaranteeing principal and interest so we don't look at it differently. But the features in the product require it to be accounted for as a derivative and the NIM as one of such product.

All right. Most of my other questions have been answered I guess just one point of clarification.

Can you speak louder please?

Is it safe to assume that your press release makes -- mentions the fact that it is I guess you are speaking on behalf of MGIC and yourselves when you say that you expect the merger to close at the end of the third quarter. Is it safe to assume that you have discussed these results with MGIC and they are on board with your statement in the press release?

MGIC on their earnings call talked about, Curt talked about the timeframe for the merger closing. While we still run our businesses as separate companies and we compete separately for businesses we also work very closely on making -- getting to the point where we get can close the merger on an integration that follows and we make sure that neither side is surprised by anything that is happening on the other side.

Thank you.

Jim Delisle, Cambridge Place.

Good morning, folks. Going back to the three AA wraps that you've written down, I mean you were wrapping the AA tranche, correct?

Excuse me, let me clarify. It's Mark. Are you talking about our CDO book in Financial Guaranty or the NIM bonds?

The NIM bonds.

The NIM bonds are not -- we wrap them to AA. The underlyings are generally BB to BBB.

Okay, you wrap all the way up to AA?

Our wrap enables it to be AA.

Okay, great. That answers my question. Thank you.

Jim Fowler, JPM Asset Management.

Good morning. Mark, a question for you looking outward on your NIM bond coverage. How will the interplay between loss mitigation, deferments, repayment plans, and the noncollection of prepayment penalties result in -- could result in pressure excess spread? How does that play through in the future in terms of the potential for future write-downs and charges?

I think that's a great question. Remember when prepayment penalties are pledged, the way it words in the trust and this is important to note, especially we talk about first liens, we mentioned it a bit before. If the servicer deems the loan to be ultimately collectible, they are required to advance cash flows to the trustee. The trustee then obviously use those cash flows to pay down the bond. So someone could ultimately default on the loan and we have already paid off the NIM bond. I think that is a very important point to remember around the structure.

Also around prepayment penalties, a servicer still is obligated to advance the prepayment penalties to the trust regardless if they waive them or not.

So if there is a --

(multiple speakers). In the modification, the modification again if it gets cash flow into the trust earlier, we think we are in better shape than before.

So if a loan is in a forbearance or a deferment or a repayment plan, there is still an obligation to collect the prepayment penalty?

Yes, if they waive it, it is the servicer's decision to do that.

How is that monitored so that that is actually collected and remitted?

It is part of our surveillance process.

Okay, great. Thank you very much.

[Satish Bhulla], Merrill Lynch.

Hi, this is Satish Bhulla from Merrill Lynch Principal Finance in London. I understand from what you have just discussed about the NIM transaction, in your cash flow modeling, can you give us a feel for how stressed your individual cash flow modeling would be for each of the deals that you have actually done some of these analysis for? For example, do your stresses take into account the existing -- the existing delinquencies that you already see in the pipeline or do you actually run much harsher stressors than that? Thank you.

It's Mark. I think it depends at the time of origination we, most of this stuff is newly originated. If there is delinquency, that is monitored into it. In general the deals can withstand 1.5 to 2 times our expected loss and that means 1.5 to 2 times earlier in the cycle also so that is important.

During our surveillance process and when we revalue the deals, we obviously take into effect impact current performance and then stress out what we believe is going to happen and run scenarios around that.

Thank you.

[Ken Schlamel], [Woodpoint Capital]

C-BASS, I was just confused. Did you say the line of credit was to help pay for the Fieldstone acquisition or for margin calls, or both?

The Fieldstone acquisition used part of that cash, so that was a stress on their liquidity as on margin calls.

Okay. Is the line of credit secured?

No.

No, it's an unsecured line of credit. And you had said something I think at the end of that comment that they were looking at other alternatives? Does that mean that they are looking for additional liquidity or are they experiencing more margin calls or what is happening there?

There are a variety of ways that they are working on to improve their liquidity.

Is there a need for additional liquidity at the moment?

We said last week their cash resources are fine at the moment. Okay. All right, thank you.

Geoff Dunn, KBW.

Thanks. Bob, I wanted to clarify something. On your paid loss guidance, I think you said 130 to $135 million per quarter in the second half?

Yes.

Can you give us an idea within that what does the second paid claims trend look like?

Approximately flat to the second quarter, Geoff.

Okay. And then the other thing I wanted to follow up on on the [bulk] production in the quarter the only other company we have seen so far this quarter actually stayed back away from the market. You are writing a fair amount of business. Can you give us a little color on what you are willing to do in this type of environment?

Again, predominately most of the business written, Geoff, was in modified pool with the agencies with attachment points. Again, we are writing very little deep MI through bulk because there is very little deep MI in the market to be written.

Okay, great. Thanks.

Okay. We have I think exhausted the questions. Let me now close with some closing comments. Thank you all for participating in this morning's call. While this mortgage cycle is challenging, I'd like to reiterate that we have the financial strength to withstand this environment and our top-line growth and solid book value illustrate the long-term success, strength and franchise value of Radian.

Thank you all for your participation and your ongoing interest.

Ladies and gentlemen, that does conclude our conference for today. Thank you for your participation and for using AT&T executive teleconference. You may now disconnect.