Radian Group Inc (RDN) 2003 Q3 法說會逐字稿

Good morning. My name is Lynn and I will be your conference facilitator. At this time, I would like to welcome everyone to the Radian Group third quarter conference call. All lines have been placed on you to prevent any background noise. After the speaker's remarks, there will be a question and answer period. (Operator Instructions). At this time, I would like to turn the conference over to Mr. Frank Filipps, Chairman and CEO. Please go ahead, sir.

Actually, this is Mona Zeehandelaar -- good morning -- Vice President of Investor Relations for Radian. With me on the call are Frank Filipps, Bob Quint, Roy Kasmar and Tino (ph) Kamarck. Before I turn the call over to Frank, let me, as always, remind you that today's conference call will contain statements that are forward-looking and as you know, these statements are based on current expectations that are subject to risk and uncertainties. Radian's actual results may differ maturely from those expressed or indicated by forward-looking statements. Factors that can affect our performance are described in our 10-K and in our other filings with the SEC. Now I'd like to turn the call over to Frank Filipps, Chairman and CEO of Radian.

Thank you, Mona. Good morning to all of you out there and thank you for joining us. I'm pleased to be here and to report to you another record quarter for Radian. I will begin by highlighting our results and some areas of interest, and then turn it over to Bob for a more detailed look at our financial statements, and then we will of course take some questions.

For those of you who have logged onto our webcast, there are slides that are provided and those slides should be looked at as background. We will not speak to them in any order or use them directly in our reports today, but they should be used for your background and history and amplification of some of the things that we're talking about.

As we said, the third quarter, we report net record net income, despite what was a challenging environment in the mortgage insurance industry. Net income reached $114 million, up 7 percent compared to the 107 million that we reported last year, and earnings per share were up to $1.20, up 12 percent from a year ago. And for those of you look at our earnings before gains and losses, that was $1.15, up 7 percent from last year. The component mix continues to demonstrate the success of our diversification strategy. Mortgage insurance produced 62 percent of our net income, while financial guaranty provided 29 percent, and our mortgage and financial services group contributed 9 percent. That compares to a year ago where mortgage insurance contributed 68 percent, financial guaranty 26 and mortgage and financial services 6.

For the nine-month period, you should note that financial guaranty's contribution was about 28 percent, versus 20 percent for the same period of last year. That is important, because as we have described to you for the last couple of years now, this is a trend that we are expecting to continue. We stated a couple of years ago that our goal by 2005 was to have the mortgage insurance income component of our net income to be about 50 percent of the total. We are well on our way to achieving that. That remains our total. We are very comfortable and confident that we will get there.

For the third quarter, premium (indiscernible) was very strong, up 22 percent over last year to $286 million, while with both the mortgage insurance and the financial guaranty business showing very strong growth. Mortgage insurance was up about 20 percent, financial guaranty was up about 25 percent. There again, you can see the results of our diversification efforts with mortgage -- within the mortgage product areas, 44 percent of our written premium is coming from what we define as our traditional mortgage insurance business, while 24 percent is coming from the nontraditional mortgage insurance. That includes our second liens, net interest margin securities, etc. And there again, that is very much part of our strategic plan. We see the products in these markets diversifying, we see them changing, we see the opportunities here to change. Lots of people have asked us about the threats that are (ph) opposed to traditional mortgage insurance by the -- what is known as an A-10-10 product (ph). We see our participation in the second mortgage lien product area as our answer to that threat. And as Bob will explain in a little bit more detail, the premium from that product has been growing quite substantially.

Thirty-two percent of our total written premium was produced by our financial guaranty groups -- public finance, structured finance and trade credit -- and that all grew about 33 percent in direct business and 24 percent in our reinsurance businesses, third quarter over third quarter. Remember that most of the financial guaranty written premium is earned over a long life, typically somewhere between 5 and 20 years, sometimes as much as 30 years. That unearned premium reserve is now $581 million, up about $100 million from a year ago. That is embedded into the future growth prospects of our company.

In the mortgage insurance, net new insurance written was up substantially year-over-year. It was down a bit in the sequential quarter, as was our insurance in force. A couple of things have happened in that market. Persistency remains very, very low. Persistency in the mid-40s is rather an odd phenomenon, although it has been going on for much of this year. We are expecting that -- we're hoping that persistency will start to turn up as interest rates appear to have, stabilized. And with the growth in the economy, we would expect that, if there is a bias in the interest rate marketplace, it is upward rather than downward. That should lead to persistency improving in 2004.

As we look at earned premium, we were very pleased that earned premium was up 25 percent in the quarter to $261 million, compared to 210 million last year. Again, that demonstrates the earnings power of the business and the franchise that we've built. We think that this clearly shows -- the earned premium combined with the net new written premium shows that the earnings growth and the earnings power of the business and the franchise is strong and will remain strong.

When we look at our reserves, we also think about strength and we've continue to strengthen the balance sheet by building our reserves in total, as well as within each of our business segments. We have had a consistent methodology, in terms of reserving in both the mortgage guaranty and the financial guaranty businesses. We will continue to maintain our consistent policy as it regards reserve methodology. And as such, you should expect to see a strong balance sheet and strong reserve provisions as we go forward.

Another area we feel poised for the strength of our balance sheet and the business franchise is our book value and our adjusted book value. Remember that adjusted book value is not a GAAP number; it is our calculation, and it is fully described in the slides and on our website. Adjusted book value continues to increase, and at 9/30, was 4825 (ph), up sequentially. And here again, a real showing of strength.

In our mortgage insurance business, as we had said, we expected that claims would be going up a little bit this quarter; they have. We would expect that, as we look out over the fourth quarter of '03 and into the first half of '04, we would expect that our mortgage insurance claims will not go down and will likely tick up in each of the next couple of quarters, although we're not expecting anything dramatic. As we look at the way our delinquencies have increased and as we look at the mix of loans in our portfolio, we would expect that claims paid over the next couple of quarters will not reverse and go down, but will be within expected limits and increase slightly.

A number of people have asked us about our strategy around captives and the use of captives and captive reinsurance in general. When we look at our captive business, we look, at a number of things. We look at the total relationship, we look at the cost involved in doing business with that customer and we look at the return on equity for the relationship. And I can tell you that, those of you know us, return on equity is probably the paramount measure of our success when we talk about it amongst ourselves, and we're not going to be doing anything and we have not done anything that will put the return on equity to Radian shareholders in jeopardy. We believe that our relationships with customers are strong and we will negotiate captives as we do all other business relationships to produce a satisfying return on equity to our shareholders.

As we look at the forecast for the mortgage business for 2004, current estimates for mortgage originations are in $1.5-$1.6 trillion range. That is obviously down quite substantially from the 2003 levels that are running at about $3.5 trillion. We would expect that purchase originations will be over the trillion-dollar level and the rest will come from refinancing. So we will be back to about a two-third, one-third mix as we go into 2004. That is substantially different from this year, where it's probably 70 or 75-- 70/30 or 75/25, in some cases, with the 75 being obviously the refinancings.

As we look out into growth opportunities, one of the growth opportunities for us is international. We received our UK license to write a very broad-based credit enhancement business in the UK and in the EU as a result, And, in September, our international mortgage group closed its first UK residential mortgage transaction. And for us, that was quite a milestone. The transaction was -- the underlying mortgage transaction was issued by the UK's largest and most experienced originator of nonconforming residential mortgage loans. We participated by selling a credit default swap backed by a financial guaranty issued by Radian insurance on 15 million pounds of triple-B rated MBS.

As we turn to the financial guaranty business, the public finance market is still producing opportunities at record levels; probably $400 billion new offering market, or in that range, for this year. Our book in the financial guaranty business grew about 45 percent in the third quarter and was up 42 percent year-to-date. On the direct side of our public finance business, we grew 33 percent quarter-over-quarter and we are seeing very, very strong opportunities for us to continue in the near and not so near future as we look out. There is no restraint on our participation in the market at the double-A level. In fact, we think that this differentiation is offering the market something that they are finding an interesting alternative, and we're going to continue to explore that market interest.

We received -- I think when we spoke last, we told you that we had just received our license in California. And in the third quarter, we have insured three deals in California and we have participated in the secondary market as well. We think that our trading value has been established and we are looking forward to writing business in this area. California has been much in the news lately. We're very comfortable with the credit in the state.

On the structured products side, even though credit spreads have tightened, the low default activity, the growing need for global institutions to hedge their credit risk means that there are still a number of transactions that we're interested in, and we had a very strong quarter there as well. The market has been growing at a rate of somewhere between 30 and 50 percent a year for the last five years. We would expect this market to continue to grow in the 25-30 percent level over the next couple. That could be very, very conservative as this market takes on new products, takes on new shapes and the growth for the demand continues very, very strong.

People have asked us about our financial guaranty reinsurance business. We're in this business. We like the product, we like the business, in terms of what it does for us in diversifying some of our risks, and we will continue to participate in this business on an ongoing basis as it meets our return on equity hurdles. In terms of our corporate strategy, our strategy has been and will continue to be to optimize our capital allocation. Return on shareholders' equity is the driving force which we measure ourselves against every day, and we're committed to write a portfolio products that will provide our shareholders with an adequate return for the risk that we take. Now let me turn it over to Bob.

Thank you, and good morning. As always, I'm going to provide some of the details behind some of the financials. Starting with the MI segment, which posted pretty solid year-over-year revenue growth, and that was really for two reasons, the first being the 11 percent insurance in force, growth from the third quarter of '02, end that is coupled with the significant premium growth from second mortgage insurance and other nontraditional mortgage credit enhancement products that Frank mentioned. Despite a record low persistency in the third quarter and a reduced level of structured insurance writings, and that was really due -- reduced the level of structure was really due to our deliberate effort to control our portfolio mix. Despite that, we were able to take advantage of opportunities in high-quality seconds and other nontraditional products. Premiums earned from this business for the third quarter of '03 were $28.1 million; that compares to $22.1 million total last quarter and $7 million a year ago. So this will account for much of the increase in earned premiums. And for those of you who calculate average on a premium rate, this is also accounting for much of the increase in the average earned premium rate. And to reiterate -- on seconds, we're only insuring high credit quality prime second (indiscernible).

We do think persistency has bottomed and will slowly increase beginning next quarter. I don't expect to see too much, but it should start to turn, and then it will happen more meaningfully as the year 2004 unfolds. We have spoken about a number in the mid 70s for persistency by the end of 2004, and that -- we still think that holds true. Regarding captive reinsurance, we wrote captive new insurance written of 6.5 billion; that's about 40 percent of new insurance written. That is on the high side for Radian, and it is pretty much due to the relatively small amount of structured business -- structured doesn't have captive attached to it. We didn't write very much structure this quarter, and therefore, the higher percentage. Captive premium for the quarter was $18.6 million. That represents about 9.8 percent of MI premiums, 10 percent year-to-date. That is pretty close to where it has been. It's pretty close to where we expect it to be. So we don't expect that 10 percent to change materially over the next few quarters.

Based on all of the housing and economic data we see, we are expecting a continuation of normal loss development delinquency rates in the current ranges and claims continuing to rise slowly as the business ages. Consistent with what we said last quarter, we are still expecting claims in the fourth quarter to be in the mid 70s; that could be on the higher side of the mid 70s, just maybe a tick higher than we thought when we protected in the second quarter. We have talked about normalized loss ratios in the MI business in the low- to mid-40s, and that is what we continue to expect to see. And regarding the MI business -- but this also holds true for the financial guaranty business -- the returns that we generate on the business are generated over the life of the business, the life of the insurance on the mortgage side. Normalized, that's 4 to 5 years; on the financial guaranty side, that return could be established over a much longer period. Returns in individual years throughout that life are going to be higher or lower than that lifetime average. So remember that as loss development curves develop, you might have years in which the returns are lower, but overall, we are expecting a long-term ROE over the life of the transaction.

Talking about loss reserves. On the MI side, we increased our loss reserve position by adding about $10 million to reserves. That supports the increased delinquencies that arose this quarter. However, reserve per delinquency and reserve per dollar of risk in force (ph) have stayed in a very consistent range for several years. We remain comfortable with that range and we continue to expect to reserve consistent and in a way that we have always done it. The movement of loss reserves will typically track the incremental number of delinquencies. Obviously, there were other factors as well, but that is certainly an indicator of where reserves will be going.

Expenses on the MI side were down a little bit in the second quarter, and I think we can expect some minor reductions over the next few quarters as the volume of new business is reduced. Contract underwriting costs for the quarter were $15.1 million; that was down from 15.4 in the second quarter. It's only down a little bit, and that is where over the next several quarters, we can see some decent reductions and expenses.

Onto financial guaranty. Frank spoke about the strong premiums. Clearly, premiums written were very, very strong, highlighted on the direct side by both on the direct side and on the reinsurance. The reinsurance results were down a little bit from prior quarters due in part to the termination of some of our current treaty (ph) business. This is consistent with the emphasis of more profitable direct business and we're only offering reinsurance when the terms of the relationship are acceptable to us, and we have made that clear for a long time, and that is obviously a key to the financial guaranty business.

Earned premiums continue to be solid. It was down from last quarter. If you heard us last quarter, we did say that there were a couple of unusual things that happened that caused the higher-than-anticipated increase last quarter. There was a nonrecurring items that was $10 million that was booked in the second quarter as well. Premiums earned from refunding were higher in the second quarter, down in the third quarter. Refundings is a number that is very, very difficult to project, and that is the reason for the earned premium decline from the second quarter. Generally speaking, nothing has changed. We have booked a great amounts of premiums written and premium earned growth should be slow, steady, the way it has been almost every quarter since the acquisition.

We continue to build reserve on the financial guaranty side to support the growth. Claims paid you will see this quarter were unusually high, at $12 million. We did break that out because trade credit is typically -- the majority of the claims we pay on the financial guaranty side, that was not any different this quarter. 8.5 million of the 12 did relate to trade credit. There were a few large individual payments on the trade credit side that were previously reserved for and happened to be paid in the third quarter. And there was also a $2 million very old claim from Van Am (ph), which is runoff and is running off exactly according to plan. However, there are some loss reserves that had been booked previously that will be -- have to paid out (ph) and there were $2 million this quarter that we paid out.

You will notice on the financial guaranty supplemental information, the present value of the future installments went down a little bit this quarter. Two reasons for that. There was a deal that was previously booked as an installment deal that was changed to a single premium upfront. Therefore, what we really did there is we traded future installments for current unearned premium, so it is still future earned premium, but it is characterized as unearned premium as opposed to future installments. The other thing we did -- and take note of this, because this is something that could happen and is likely to happen in the future -- we took advantage of corporate spread tightening in the credit default swap market to trade out of a previously booked CDO, and we locked in the gain on that transaction. In doing that, we gave up some future premium. From time to time, that could be a strategy that we employ either to lock in a gain by accepting a cash settlement on a deal, or to make a cash payment to capital loss on an underperforming deal. All such settlements will be disclosed in our quarterly reporting, if we do that.

Just a bit about the services segment, another very strong quarter. We are very happy with C-BASS and Sherman's performance. We do expect similar strong results in the fourth quarter. And Rating Express had a small loss for the quarter, revenues of $5.6 million and expenses of 6.4 for the quarter.

Finally in the third quarter, most of you saw that we were able to close a $150 million soft capital security for Radian Asset. This is a couple market transaction. We view this (indiscernible) capital as being very, very efficient, helping support our financial guaranty capital needs to support the growth in that business. For the near-term, both the MI financial and the guaranty side for the business, we are confident that we can support the growth in those segments with internally generated capital and potentially additional soft capital. Okay, now we would like to turn the conference over to questions.

(Operator Instructions). A.J. Grewal, Smith Barney.

Hi. Could you give us some color around -- the year-over-year growth and premiums earned in the furnished financial guarantee business seems to be slowing. Could you give us some color on that? And also, what is driving the derivative gains? In which business -- where -- are these hedges, or in which business do you usually -- has this been driven by? Thank you.

The earned premium growth on financial guaranty has been so substantial, it is not going to continue at the level that it had been. So I think that is just natural, growing the business and maturing the business. I think the earned premium growth continues to be very, very strong year-over-year and the written premium is very strong as well. So obviously, that's going to slow over time. It's not going to remain at the level that it was at. The gains are going to be caused -- a lot of it was corporate spread tightening, so that is going to cause gains on the CDO book. The other gains could also be caused by the equity markets. Because as you recall, the other part of that line item is our convertible securities, which has a debt component and an equity component, and the equity component is really what produces the change in gain loss on that line. So that is going to be the other component.

Could you just briefly go over the credit picture? As I recall, you were saying you have (ph) billed reserve, and you expect a certain amount of loss -- I didn't catch all the numbers you threw out regarding your reserving and losses expected.

The only number we threw out, and it is consistent with what we said in the second quarter, was that claims in the fourth quarter would be in the mid 70s, and we revised that just slightly upward and said it would be on the higher side of the mid 70s for the fourth quarter. And then the reserve building of $10 million, those are the only numbers.

10 million?

Yes.

Okay, thank you.

Geoff Dunn, Keefe Bruyette & Woods.

Good morning. A couple of questions. First, can you talk a bit about your guidance on Radian insurance company? You've had some big premium out of the seconds -- what should we be thinking about over the next 12-18 months for that kind of level?

It is a hard one, Geoff, because as you know, a lot of this business is done on a structured basis, deal-by-deal, as opposed to flowing on a normalized level. I think we can tell you that we expect it to continue to grow. I cannot say it's going to grow at the levels it has been growing -- in fact, it won't grow at those levels. But we do see a lot of opportunities in seconds where we're writing a regular amount of business. And so I think, growth, we can expect growth, we can expect a reasonable level of growth, but it is not going to grow at the level that it has.

And then on the loss side, you've given some guidance for fourth quarter. As you look out onto '04 and you're looking at delinquencies, do you think the delta of sequential increases could slow, based on what you're seeing now?

Well, it slowed -- from second to third quarter, it slowed. I don't know that it is going to slow further. And, I think as Frank said, we expect sort of consistent increases. We don't expect dramatic increases, but I don't think it is going to slow to a very, very small level of increase.

If you go back and look at the delinquencies, and hence, defaults and claims, this part of the business is really very dependent upon the change in employment and unemployment. And until we see employment growth and a backup in the employment rates, it is unlikely that we will see any change in our delinquency default in claims pattern line. And so as I have said for a long time, that is the leading indicator. When you see the change in employment and unemployment turning positive; six months after that, you will likely see that the change in our delinquencies and then defaults and then claims improve after that. That is the sequence.

Last question. Could you give a little detail -- I think you had about 4.5 million of financial guaranty losses in the quarter -- what were those related to?

Claims (ph), Geoff?

Yes.

Well, the two was the Van Am (ph).

So the (multiple speakers) 4.5 was Van Am?

And the other were -- there were a couple of deals that we're (ph) making small payments on that are just sort of normal public finance deals.

Remember -- Van Am is a company that wrote reclamation policies, and so they expect, and in fact, fully bake in claims payments as the reclamation process goes on. So as Bob said before, these claims were fully expected, anticipated, projected at the time that Radian acquired enhanced, and they're just now developing according with that expectation.

The reserve on Van Am was about $6 million or so. So over the next few years, we're going to have payments offsetting those reserves.

Okay, thank you.

Paul Miller, Friedman Billings Ramsey.

Thank you very much. My question is on financial guaranty. I'm just trying to understand basically on a year-to-year or quarter-over-quarter basis, how these numbers move. If you back out your derivatives and security gains in financial guaranty, you made about $24 million on that, which is down not only quarter-to-quarter, but also year-over-year. And I know you said quarter-over-quarter that the second quarter had a lot of onetime gains. But can you just discuss -- was it the same thing in the quarter ago period? Is it just because it was that a lot of these re-fis, you recognizing that income upfront?

I'm sorry, Paul -- can you repeat that?

On the financial guaranty, on the business segment, when you back out security gains and derivative gains, you guys only made $24 million, which was down not only from the second quarter, but it was also down -- this is the net profit - from the year-ago period. And I know that you mentioned on the call that there was some onetime gains in the second quarter '03. Was at the same situation from the year ago period? Can you guys hear me?

Yes. No, that was not the same. You have to look at the entire business segment. You have premiums earned growth, which has been very substantial. A lot of loss provision of financial guaranty is adding to reserve. So, yes, that hits the P&L, but a lot of it is adding to reserves to support future payments, growth of the business. Operating expenses obviously have gone up in that segment as well, not only because we have grown the infrastructure and the ability to run more business, but because we have corporate allocations now that are coming from Radian Group and being allocated to the financial guaranty segment. So I think all of that has to be taken into account. But I think it is also suffice to say that the income from financial guaranty overall for the year has grown pretty substantially.

When you write a policy, do you take the reserves upfront on that policy, or do you take the reserves over time for unexpected losses?

We put up non-specific reserves as we earn premium. And then -- so it is not like an IB&R (ph) goes up like P&C, but we do put up a non-specific reserve that is general and there to support the business, and it does not relate to any one deal. Then we'll move that to case reserve if there are problems with a particular deal.

So when you're writing -- if your premiums written is growing, your reserves (indiscernible) growing in hand, that could put pressure on the bottom-line earnings?

It will impact the bottom-line earnings. We can't put pressure, but it will certainly impact it. As we grow the business, write more premiums, we put up more loss reserves.

In the trade credits, (indiscernible) said you took losses I believe $8.5 million. Is that stabilizing and is the pricing still there for that business for you guys to (indiscernible) that business?

If you look at the year-to-date loss ratio in that business, and it's in the supplemental information -- we give you to premiums and the losses -- it is about 50 percent. That is the loss ratio for that business. We have said that -- it is really -- it is more like a P&C combined ratio that's sitting in financial guaranty. But if we can write trade credit at 50 percent loss ratio, that is good business and a good return on allocated capital.

My last question is -- what type of premium yields are you getting off of these second lien structured programs? Is there a range that you guys can give us?

Obviously, it depends on the structure and it depends on a lot of things, but the average is in the 1.5 percent range. And there could be deals that are much higher or much lower than that within that range. The average is not necessarily the same as every deal -- any deal can be much higher or much lower.

Thank you very much, Bob.

Alex Orloff (ph), Bank of America.

Good morning. A couple questions on financial guaranty again. If I look at the loss and expense ratios as related to premiums in Q3, it seems that the polished (ph) acquisition costs ratio actually increased by about 4 percentage points, compared to Q2. Is there any reason for that?

I think you have to look at that on a really a year-to-date basis. From quarter to quarter, there could be items that don't necessarily reflect the future. We have been building infrastructure in this area. It's obvious to us that we need risk management expertise, we need product expertise. And in order to be in this business, the deal structures are very complicated, we need to have the proper infrastructure built, and we have done that. That is a part of the increase in expenses.

The other part is international business in the UK. That is a business that we have created an infrastructure for and have the expenses embedded in the business. However, we have not booked the income yet, so that is certainly something that will catch up on the income side.

I'm just kind I think slightly surprised that you book it through a polished (ph) acquisition cost, rather than through other operating expenses.

If it is underwriting and production people, then they are booked through policy acquisition costs.

Right. And as you mentioned, since the revenue has not really come in for the UK contracts, you would not be able to (indiscernible) defer acquisition costs and amortize it. Just last questions. You mentioned your -- a target loss ratio in the trade credit business, about 50 percent. Just remind me -- trade credit business -- is it primary insurance or reinsurance?

It's reinsurance.

Thank you.

Jonathan Gray, Sanford Bernstein & Co..

Thank you. Could you give us some idea of the balance between contract underwriting revenue and contract underwriting expense? In other words, as the refi boom winds down, can you give us some reassurance that the net effect on your pretax margins within the MI segment will be positive as I think most expect?

Yes, Jonathan. The short answer is that there's definitely a positive trend there. Contract underwriters are down 25-30 percent, and I would expect that to continue to go down as the volume and the market changes, so there will be a positive impact by that.

Can you give us some idea of the balance between the revenue and expense associated with contract underwriting, in dollar terms?

We gave the dollars -- 15.1 million for contract underwriting expenses. Revenues for contract underwriting for the third quarter were in the -- priced $7 to $8 million range.

Thank you.

Robert Hodginson (ph), Goldman, Sachs & Co.

Just a quick follow-up and clarification on the financial guarantee reserving. Three questions, Bob and Frank. First, you do formulaic general provisioning in your -- in terms of the income statement provisioning for losses. Second, do you maintain general and specific case reserves on the financial guaranty business, transferring specific general reserves to case reserves as specific losses become known? And third, do you target along with the rating agencies your double-A business, a specific combined ratio that would normalize out, once the growth and some of these growing pains get normalized into the ongoing revenue pattern?

Bob, let me take a couple of those. We have a quite extensive and regular watch committee that looks at each insured transaction in the financial guarantee business in our surveillance group, and it progresses from a watch list to a case reserve when a case reserve is necessary. So the answer is, yes, we do establish case reserves on a very active case-by-case managed basis. And that's something that literally goes on a monthly basis. So it is a very active and ongoing process.

With regard to combined ratios and relationship to the rating agencies, we do not have specific targets there that we have established. We have working guidelines, obviously, but no hard targets that have been established. And third, yes, there is a formulaic general reserve, as Bob was talking about, in terms of a function of written premium that is accrued every quarter.

Is it an accrued by line or (multiple speakers)

I'm sorry -- earned premium, not written premium.

But by earned premium -- would we be able to detect that from looking at the earned premium and then working back, or is there a different formula for each line of business?

It goes by line of business and by product.

The one thing that we tried to do is break out trade credit, because that distorts the numbers some. So I think that will be helpful to you in performing your analysis.

Do you consider the trade credit business, for instance, to be a fourth (ph) loss business that you're just going to accept a given level of losses, and they're just going to be low severity losses, and that is built into the expectations?

Let me back up. If you were looking to formulaic number, I can give you -- I think the best I can do would be to give you a range, and to tell you that the range is likely to be 10-15 basis points earned on percent -- on earned premium. That is what it works out to be formulaic.

But the high end, it might be 15 basis points for trade credits, and 10 basis points for -- not 10 basis points, but 10 percent of earned premium?

Yes. Trade credit would be substantially higher than that. Trade credit probably has a loss ratio built into the product of somewhere around 50 percent and sometimes higher than that, depending upon the product. Now what we have done in that product line is to pull back our attachment points in the reinsurance business, and a year ago, started writing considerably more excess of loss coverage than quota share participation. So hopefully, that will change the loss development in that product as well. And that was something that we decided to do some time ago, right after the acquisition, and we have proceeded along those lines. Do you have a mix on that that you could add?

I don’t have those numbers in front of me, but last year and so far this year, in terms of both the treaty terms and the performance and expected performance of the trade credit primaries, have been excellent years in the industry. So we're happy to participate on a (indiscernible) a proportional kind of basis. But if that business were to head into a cyclical turn of much higher loss ratios, what we would try to do in managing our relationships with our key clients there is to move, as Frank as indicated, more away from the proportional participation and more towards XOL (ph) kinds of lines, which is a loss profile much more similar to financial guaranty. So (indiscernible) lower top line and higher profitability is the trade off.

I know it's simplistic, but for now, it's 10-15 percent on non-credit, not trade credit business and maybe 50 percent on trade credit, heading down as you restructure the risk severity.

I think that is a good set of assumptions, yes.

Thanks.

One other thing I'd want to throw in, in terms of the targets about the expense ratios, which is that, as Bob was talking about earlier, as the earned premium ramps up as the business matures, and as you begin to see earned premium as a greater percentage of written premium, which is typical of the more mature triple-A primaries. And at the same time, we are still growing the topline so that we're getting a better leverage off of our in-place cost structure. I would expect Frank to expect from me that those expense ratios in financial guaranty will begin to go down.

Bradley Ball, Prudential.

Thanks. Just a follow-up to an earlier question. Did you in fact say that your delinquency trends monthly in the quarter were improving, and that you would expect that improvement to continue?

No, we said just the opposite, that delinquency trends had gone up, and they may go up in the fourth quarter as well.

And so September notices were higher than July notices?

It is difficult for us to give month-by-month reports. We get this information from our mortgage servicers. And intra-quarter, there is just a lot of noise that gets developed as late reporting, there's lack of reporting from some servicers on a timely basis. So intra-quarter, it's just not a consistent number and we often find that things just don't get reported accurately to us. And therefore, the information that we're able to give out is not consistent. So intra-quarter is not good indicators in this industry.

Separately in the MI business, contract underwriting expenses are expected to come down. Are you also expecting other opportunities to cut costs, and where would those opportunities lie?

That would be clearly the bulk of expenses or reductions in expenses. There may be some sales cuts and some other operational efficiencies, but the bulk would be in contract underwriting.

Thanks. Finally on the financial guaranty business, do you have any exposure to the S&P downgrade of Pittsburgh, or at least indirectly, what do you think the implications of that are for your financial guaranty business?

I will give an (indiscernible) and maybe Tino (ph) can comment. We have about $85 million of GL exposure. It is all reinsurance, none is direct, and that is par (ph) outstanding. And remember that, even though we don't expect anything to happen here, if it were to happen, that $85 million exposure (indiscernible) over a very, very long period of time, and we are just responsible for P&I over that time period.

Certainly don't expect at this point on the information that is available to be paying claims on any of our exposure to Pittsburgh. The economy there, the demographics are robust. The problems are, as is usually the case in public finance of fiscal management, not fiscal fundamental (ph). So looking ahead, A, we don't expect to pay a claim on the information we have now; and B, expecting that as is frequently the case in such situations in public finance, that there will be some refinancings, restructurings of debt going forward with stronger security features. There may well be good additional opportunities for both reinsurance and direct business for us there (ph).

Great, thanks.

Ed Groshans, Moore's & Cabot (ph).

It looks like the contract (indiscernible), you said 15 in expenses and 7 to 8 million in revenues -- is that correct?

Correct.

With captive relationships, I'm wondering -- what are the other items that you have in there that you look at and the full relationship to generate the appropriate ROEs? I thought the contract underwriting was one of them, and that doesn't seem like that would enhance ROEs, but take away.

There are a few others of contract underwriting, and clearly one of them -- and the uses of contract underwriting does not enhance ROE. There is sales expense, relative to how much effort we need to expense in order to drive volume, and there are other ancillary expenses around that. Just generally, also what kinds of mix of product we're getting from them and the mix of ROEs associated with that.

Certainly, losses. Loss expense is the highest percentage expense in the business. So quality of business and performance of business is very meaningful in ROE by customer.

All right, thank you.

(Operator Instructions). Mekiko Coakley, Endeavor Capital.

Hi. How much was that CDO both gain that you took this quarter?

The unrealized gains on derivatives was 6 million. Is that what you wanted or the deal that we accepted in payments?

The payment one.

That was about $4 million.

And that goes to unrealized gain, or goes to premiums?

We book mark-to-markets through the gains and losses, so that gain had already been booked through mark-to-market. Here, we locked in the cash, so the gain disappears, but we have the cash.

But there is no operating income (indiscernible)?

Correct -- it has already been booked.

The gain was actually recognized in the second quarter. We took the cash early in the third quarter.

Did you say that you expect paid claims to go down in the second half of next year?

No, we didn't say that.

In the second half of next year?

Yes.

No, we did not give any forecast. What we said was that, for the first half of next year at this time, we are expecting a slight -- a couple of ticks up in claims paid for the first half of next year.

So, claims will continue to go up until second quarter of next year, and from there, we don't know where it is going?

We have not given any statements on that.

Great, thank you very much.

A.J. Grewal.

I was under the understanding that you don't (ph) offer contract underwriting to those that have captive relationships. Can you clarify that for me?

We do offer contract underwriting to any number of customers. Some of them have captive, some of them do not.

Thank you.

At this time, there are no further questions. Are there any closing remarks?

Nothing official, other than to thank everyone for their participation, and we look forward to talking with you next quarter.

This concludes the Radian Group third quarter conference call. You may now disconnect.