Radian Group Inc. (RDN) Q2 2013 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by, and welcome to Radian's Second Quarter 2013 Earnings Call. [Operator Instructions] As a reminder, this conference will be recorded. I'd now like to turn the call over to our host, Senior Vice President of Investor Relations and Corporate Communications, Ms. Emily Riley. Please go ahead.

Thank you, and welcome to Radian's second quarter 2013 conference call. Our press release, which contains Radian's financial results for the quarter, was issued earlier today and is posted to the Investors section of our website at www.radian.biz. During today's call, you will hear from S.A. Ibrahim, Radian's Chief Executive Officer; and Bob Quint, Chief Financial Officer. Also on hand for the Q&A portion of the call are Teresa Bryce Bazemore, President of Radian Guaranty; David Beidler, President of Radian Asset Assurance; and Derek Brummer, Executive Vice President and Chief Risk Officer of Radian Group. Before we begin, I would like to remind you that comments made during this call will include forward-looking statements. These statements are based on current expectations, estimates, projections and assumptions that are subject to risks and uncertainties which may cause actual results to differ materially. For a discussion of these risks, please review the cautionary statements regarding forward-looking statements included in our earnings release and the Risk Factors included in our 2012 Form 10-K, as well as subsequent quarterly and other reports and registration statements filed with the SEC. These are also available on our website. Now I would like to turn the call over to S.A.

Thank you, Emily. Thank you all for joining us and for your interest in our company. On today's call, I will provide highlights from the second quarter, including the attainment of several important milestones. And I will also share my thoughts on the topics and trends important for our industry. Following my remarks, Bob will cover the details of our financial position. Earlier today, we reported a net loss for the second quarter 2013 of $33 million or $0.19 per diluted share. This includes net losses on investments of $130 million and the impact of fair value and other financial instrument gains of $88 million. Most importantly, our mortgage insurance business achieved profitability absent the impact of fair value gains and losses for the quarter and 6 months. Needless to say, at Radian, we've been working hard to achieve this significant milestone and we are extremely encouraged by the improved performance of our mortgage insurance business. I'm also pleased to point out several additional milestones achieved by Radian. We continue to write a leading share of the new mortgage insurance business in an extremely attractive market, perhaps the best business we have ever written. We recorded our second highest volume of primary flow business in Radian's history in June and for the second quarter. We have successfully improved our portfolio, and in the second quarter, as expected, the profitable high-quality books of mortgage insurance business written since 2008 represented more than half of our total portfolio. You can see on Slide 17 from our webcast presentation that for the 6 months ended June 30, 2013, the primary earned premiums less incurred losses from our 2009 and later MI vintages far exceeded the comparable negative sum from the 2008 and prior vintages. In the second quarter, the MI incurred loss ratio dropped to less than 70%, which represents another significant improvement. Finally, since 2008, we have been successfully reducing the exposure in our financial guaranty business and have reduced it by 76%, including many of the riskier segments of the portfolio. Radian Asset has only $7.9 million of exposure to general obligation bonds for the highly publicized municipal bankruptcy in the city of Detroit, for which we hold a loss reserve of nearly half that exposure. Radian maintains strong holding company liquidity of approximately $816 million, and Radian Guaranty's risk-to-capital ratio is a competitive 19.7:1 at June 30, 2013. We continue to expect to maintain a risk-to-capital ratio for Radian Guaranty of 20:1 or below for the foreseeable future, while also preserving a strong level of holding company liquidity. Bob will cover the upcoming GSE eligibility requirements, which we expect to be issued by the end of the year. New mortgage insurance written continues to consist of high-quality loans. We wrote a total of $105 billion of new MI business from 2009 through June of this year, which is expected to produce substantial profits and attractive returns. And our primary insurance in force now stands at $151 billion compared with $130 billion a year ago, an increase of 16%. It is the new vintage mix, size and performance of this growing insurance in force that is expected to drive our future earnings. Slide 18 in our webcast presentation breaks out the recent performance by vintage. And I would draw your attention to our profitable post-2008 vintages, as well as our legacy book of business written in 2005 and prior, both of which contributed to our profitability in the quarter. As I mentioned earlier, a high-quality business written after 2008 represents 53% of our primary risk in force and now outweighs our legacy book of business. We believe this achievement is unique to Radian among legacy MI companies and is the direct result of our market share gains over the last 2 years. You can see the details on Slide 19, where the most problematic 2006 and '07 books are now down to less than 17% of the total portfolio. The slide also shows that including HARP refis, which improved the credit profile of our legacy book, new business underwritten since 2009 now represents 64% of Radian's mortgage insurance primary risk in force. As you know, the HARP program was extended through 2015, which should allow additional deserving borrowers to take advantage of the program. It is important to note that we continue to expect a marginal level of mortgage insurance profitability for the full year 2013. As a reminder, this forecast excludes the impact of stock price changes on our long-term compensation expense, as well as net fair value gains and losses. Now, I'd like to highlight the progress made against 3 important priorities for Radian: writing new mortgage insurance business, mitigating losses on our mortgage insurance portfolio and reducing our financial guaranty exposure. First, we wrote $13.4 billion of new mortgage insurance business in the second quarter, an increase of 60% year-over-year. While it's too early to provide volume for July, we are on track to reach net new business comparable to or surpassing our outstanding June NIW. As I've mentioned before, mortgage originations over the past 2 years have been driven by high volumes of refinance activity. We are beginning to see signs of a marked refi slowdown and a shift in the market to more purchase business, which is likely to reduce the overall origination market. Thus far, interest rates do not appear to have meaningfully slowed down purchase activity, and it must be noted that the MI penetration for purchased loans is about 3x to 4x greater than for refi loans. While further increases in interest rates could reduce purchase activity, they are also likely to increase our persistency rates, which will help us grow our insurance in force. On average, every 1% increase in persistency means that approximately $1.5 billion of insurance in force remains on our books each year. Also worth noting is that purchase patterns tend to be seasonally skewed, slow in the first quarter, strongest in the middle quarters and then slowing again in the fourth quarter. We also continue to grow and diversify our customer base, which has helped us grow and maintain our share of new business. This year, we signed up 117 new lender customers, and our pipeline of prospective new customers remains strong. And the new customers bring more volume, as more than 25% of our new insurance written in the first 6 months came from customers new to Radian within the last 2 years. We also continue to gain more business from existing customers, which is a good indicator of how positive they feel about us. NIW coming from lenders we've worked with for more than 2 years grew 46% from the second quarter of last year. We're supporting our customers with a continually enhanced industry-leading sales and training team made up of a combination of seasoned as well as new talent. In the first half of 2013, our trainers reached 50% more customers than we did in all of 2012. The Radian Foundation series, our newest suite of courses, focused on basic underwriting and processing, is gaining popularity. After only 6 months, more than 250 participants have completed the course to earn their certificate. Next, focusing on our continued efforts to mitigate losses in our mortgage insurance portfolio, the total number of primary delinquent loans declined by 8% from the first quarter of this year and 21% year-over-year, as you can see on Slide 22. And the default rate on our primary book fell further in the first quarter to 9.7%. We maintain $2.7 billion in loss reserves, and our primary reserve for default increased slightly from the first quarter to $30,932, up from $28,410 a year ago. HAMP and other modification programs continue to produce steady results and borrowers show signs of commitment to their homes and mortgage even in the latest stages of default. As you can see on Slide 12, 31% of borrowers whose loans were in default made at least 1 monthly payment in the second quarter. And repeat defaults represent 78% of new defaults in this quarter. Historically, redefaulted loans have been less likely to result in a claim than first time defaults. Finally, our financial guaranty business continues to serve as an important and unique source of capital for Radian Guaranty. We have successfully reduced our exposure in that business from a peak of $115 billion in June 2008, when Radian Asset stopped writing new business, to $27 billion in the second quarter of this year. Radian Asset paid a dividend to Radian Guaranty in July of $36 million for a total of $420 million paid since 2008. We expect Radian Asset to continue to pay dividends in the future years. As of June 30, 2013, Radian Asset maintains statutory surplus of $1.2 billion and additional claims-paying resources of $396 million, including $248 million of contingency reserves. The private mortgage insurance industry continues to slowly but steadily regain share from the FHA. We estimate that penetration for our industry was 8.7% in the second quarter, up 12% from the first quarter of this year and up an impressive 30% since the second quarter of last year. The FHA's latest price increase in April, coupled with the elimination in June of FHA mortgage insurance premium cancellation, should help continue the shift of business from FHA to private mortgage insurance. In terms of housing market trends, CoreLogic recently reported May statistics citing rising home prices for 15 consecutive months and the largest year-over-year increase in the last 7 years. While rising interest rates have recently made headlines, they remain at historically low levels. According to the National Association of Realtors, existing home sales were at their highest level in May since November 2009, when buyers took advantage of the cash stimulus. Turning to the regulation and legislation affecting our industry. We continue to actively engage with legislators and other decision-makers in Washington. The qualified mortgage rule or QM, which will go into effect early next year, closely mirrors the mortgage lending environment we've successfully been operating in today. And the final Basel III rule preserves the important role for private mortgage insurance. The bipartisan Corker-Warner bill recently introduced on GSE reform also explicitly supports the increased use of private mortgage insurance for low down payment loans in the housing market. Overall, we continue to hear a resounding support on Capitol Hill for a healthier balance between public and private sectors in today's mortgage market, which is extremely positive for our industry. In closing, our improved financial performance in the first half of this year, particularly in our MI business, continues to energize our team at Radian. This continued improvement remains our top priority and will be driven by 3 important factors: First, the size and potential earnings power of the high-quality new MI business we have already written since 2008 and our growing insurance in force book. Second, the continued management of our legacy mortgage insurance book, which is now a smaller piece of our total MI portfolio than the high-quality book written after 2008. We expect the legacy book will continue to improve through HARP, modifications and home price appreciation. Third, our success in building a strong sales and operations platform to continue writing more high-quality future business, driven by our strong customer relationships that has earned us the #1 market share position. Now I'd like to turn the call over to Bob for details of our financial position.

Thanks, S.A. I'll be updating you on our P&L activity and trends for the second quarter 2013, our capital and liquidity positions as of quarter end and some updated expectations regarding 2013. The MI provision for losses was $136 million this quarter compared to $132 million last quarter and $208 million a year ago. This is the second consecutive quarter with an MI loss ratio of approximately 70%, reflecting the improving delinquency trends and stable claim submission, as well as a steady increase in earned premium from our growing book of business. We have not changed our frequency expectations on either new or aged defaults. MI earned premium in the second quarter also benefited from an approximate $9 million reduction in the premium refund accrual relating to future rescission. For the balance of 2013, we continue to expect a much improved incurred loss line from the comparable 2012 quarters, driven by a relative decline in new defaults and no material changes to net low rate expectation. Paid claims for the full year 2013 are expected to be approximately $1.4 billion. The reduction in average claim paid this quarter is due primarily to lower average loan balances and coverage percentages on claims paid in the quarter. Curtailments have remained elevated, which has also positively impacted several. We've been successful in reducing our primary pending claims inventory from 17,625 at year-end 2012 to 15,018 as of June 30, 2013. And we anticipate that pending claim inventory will continue to decline. Single premium business written since the beginning of 2009 has continued to perform well in 2013 with loss ratios similar to those for monthly premium business. While our pricing is in line with the general industry pricing, we have chosen to sell the product to our customers as a competitive product to FHA. We expect to see our new business mix shift towards more monthly premium business due to the higher percentage of purchase volume in the mortgage origination market and the decline or elimination of borrower paid singles from the implementation of the Dodd-Frank qualified mortgage rule early next year. Borrower paid singles are currently about 28% of our singles volume and 9% of our overall production. Beginning April 1, we reduced the quota share percentage for new business ceded to our external reinsurance partner from 20% to 5%. That helped reduce the amount of ceded premiums written significantly in the second quarter. In addition, we will also have the option to recapture a portion of the ceded quota share risk at year-end 2014 and 2015. Those decisions will be based on the performance of the business and our capital position at that time. The unrealized losses in our investment portfolio this quarter were caused by rising interest rates, which reduced the market value of our fixed income portfolio. These losses are temporary, as we expect to hold the investments, are GAAP only and are not recorded for statutory purposes and, therefore, do not impact our risk-to-capital ratio. Most of our investment portfolio is classified as trading, which requires us to run unrealized gains and losses on those securities through the P&L. As a reminder, each quarter, we disclose the sensitivity of our investment portfolio to interest rate shift in our 10-Q, which provides a reasonable estimate for potential changes in portfolio value. Fair value gains for the quarter were caused primarily by an upgrade of MBIA during the quarter. This had the effect of reversing much of last quarter's fair value loss caused by MBIA's downgrade on a large second to pay financial guaranty exposure. Slide 9 depicts our current balance sheet fair value positions, along with the expected net credit losses or recoveries on fair valued exposures. Based on our projections, we expect to add approximately $337 million or just under $2 to pretax book value over time as the exposures mature or are otherwise eliminated. That number is derived by taking the net balance sheet liability of $275 million and adding the present value of expected credit loss recoveries of $62 million. In the first quarter of this year, operating expenses were impacted significantly by $38 million of long-term incentive compensation expenses. The comparable expense in the second quarter of 2013 is approximately $19 million, which includes the impact of a smaller increase in our stock price during the quarter. Of the $19 million, approximately $7 million was a direct result of Radian's stock price increase of just under $1 during the quarter. Approximately $4 million represented a second quarter-specific expense associated with our regular, newly issued, performance-based long-term compensation awards in the second quarter of 2013. The balance of the expense this quarter or approximately $8 million is the base long-term incentive compensation expense for the remainder of 2013, to which any stock price or other impact will be added. Policy acquisition costs were down this quarter primarily due to higher persistency of recent vintages, which slows down our amortization schedule of costs that were recently added to the policy acquisition asset. Our valuation allowance against our deferred tax asset is currently a little over $1 billion or just under $6 per share. We continue to believe that our full DTA will be realized in the future as we become profitable with a full recovery still expected to occur some time in 2015. Our risk-to-capital ratio ended the quarter at 19.7:1. Our current plan is to manage our risk-to-capital below 20:1, and we have substantial holding company resources with which to do so. We anticipate making a contribution to Radian Guaranty in the third quarter to support the expected risk in force growth and to remain below 20:1. While the details of the new GSE eligibility capital requirements for MIs are still unknown, we expect them to be issued by year end. We have heard about possible requirements in the 18:1 range and potential additional capital requirements for legacy exposure and/or subsidiary capital haircuts. We fully expect to have the ability to comply with any such requirements within the implementation timeframe. Our holding company liquidity stands at approximately $816 million. We have $55 million of remaining par maturing in June of 2015, with the balance of our outstanding debt due in 2017 or later. In terms of future impact to our GAAP book value, in addition to ongoing profits or losses, the primary items to consider are the valuation allowance against our DTA and the fair value liability that I referred to earlier, which are expected to add to book value, as well as the equity component of our convertible debt, which is approximately $182 million as of June 30 and will be accreted into debt over time, thus reducing our pretax book value by a little over $1. I now would like to turn the call back over to the operator for questions.

[Operator Instructions] Our first question is going to come from Bose George with KBW. Bose T. George - Keefe, Bruyette, & Woods, Inc., Research Division: Actually, the first question was just on your average premium. We calculated a slight increase. So I was just curious if that was a blip, or is the premiums trending up based on any sort of mix shift?

Bose, as I mentioned, we had this $9 million recognition of earned premium that was a rescission refund accrual reduction. So that's going to impact, if you just take the earned premium and do it that way, it's going to impact the number a little bit. So if you pull that out, I think you'll be pretty close to our constant average premium rate. Bose T. George - Keefe, Bruyette, & Woods, Inc., Research Division: Okay, great. And you gave some details on your single premium production, but do you have the percentage of single premium of your insurance in force?

I believe, last we looked, it was about 20%. So it might be up slightly from that. Bose T. George - Keefe, Bruyette, & Woods, Inc., Research Division: Okay, great. And then in terms of the cure rates. What cure rates do you assume on new defaults? And how does that vary, just based on the level of redefaulting that these borrowers are doing?

We do tell you what our roll rate, our frequency expectations are. I believe that's Slide 11. And that's going to tell you by bucket of what our -- if you look, for example, at 3 payments or fewer with a gross roll rate of 25%, we're expecting 75% not to be claims. I think that's what you're asking. And then it's different, obviously, for the different default buckets. Bose T. George - Keefe, Bruyette, & Woods, Inc., Research Division: Okay, great. And then actually one last thing. I just wanted to follow up on your comment that you made about the subsidiary capital haircuts. Is it too early to tell what that could look like? Just any color on that.

It is too early to tell. We've only heard that, that's a possibility. We've not heard anything further. And we'll have to see once the eligibility requirements come out what that looks like, if it is even [indiscernible].

And our next question is going to come from Doug Harter with Crédit Suisse. Douglas Harter - Crédit Suisse AG, Research Division: Just another question on the single premium. Just wondering if you saw any impact from the higher rates yet, or should we expect that more to come sort of in the back half of the year?

This is Teresa. I don't think we've seen any impact at this point. Obviously, we do see a higher percentage of singles being used with the refi business. And so with the refis sort of tailing off as we would -- as we're already seeing, we would expect to see some reduction there just on that basis. Douglas Harter - Crédit Suisse AG, Research Division: And then I guess just on the persistency, has that slowed down or, I guess, persistency increased or the prepayments slowed yet? Or again, is that sort of more later?

Yes. I mean, I think you're going to see that happen. We've seen it on the more recent vintages for sure. I don't think you'll see our overall persistency for the year, which we report, change very much from last quarter. But I think you will see that increase over the next several quarters. Douglas Harter - Crédit Suisse AG, Research Division: Great. And then on the financial guarantor, you said you expect to be able to continue to get dividends out of there. At what point -- as the risk continues to fall there, might the pace of dividends be able to accelerate as -- to be able to pull that capital out of the financial guarantor?

Well, to this point, we've only paid ordinary dividends. Our goal has been to reduce the exposure, which we've done a good job of. So in terms of ordinary dividends, we would expect steady levels of ordinary dividends. But as we reduce the exposure, we're putting the financial guaranty company in a very good position. Douglas Harter - Crédit Suisse AG, Research Division: And is there a risk-to-capital level there that you would target before sort of looking to go after non-ordinary dividends?

I don't think there's any formula. But certainly, what we've done and what we continue to do at the company puts us in a much better position.

And our next question will come from Jack Micenko with SIG.

Looking at the denial reinstatement page in the slide deck, pretty big falloff in 2Q and clearly a positive. I know there's some sort of statute of limitations there, 12 months or so. Are we -- is that a number that looks right going forward? Are we sort of out of the window where those denials can be reinstated? Is that 13% the right number to think about on a go-forward?

I think the percentage that you're referring to is the most recent quarter. So that's not going to be indicative of what it's going to be. We still generally use 60% as our initial denial reinstatement rate. And then it comes down over time. So that first quarter, that 13%, that's going to go up, very likely. So we haven't changed our expectations. The 60% is our general expectation.

Okay, great. And then with your -- sort of like your late stage and your pending claim assumptions, they've kind of walked up since the beginning of 2012. So it looks -- I'm assuming you're getting more conservative on the view of those later stages and it's, I guess, to be expected, but do these numbers go materially higher from here? I guess there are maybe 600 basis points on the pending over the last 1.5 years or so? Or are we pretty close to sort of where we peak on that?

Yes. I think you've seen the changes more relating to our expectation that there'll be a reduction in denials and rescissions over time. I don't think the growth rates have changed in quite some time. But over time, we do expect denials and rescissions to come down, so the net will likely go up.

Okay. And then just one last one. FG provisions on a year-to-year basis, despite what's been a nice ongoing decline in risk commutation, that sort of thing, is that specifically tied to any municipal event or anything? Or what can you say about that increase in the FG provision?

Yes, it was pretty much Detroit for the quarter. Last quarter, as I recall, it was almost nothing. This quarter it was still very small.

And that got you to the 50% reserve on Detroit, is that the way to think about it?

Yes. Yes. Nearly 50%, yes.

And our next question is going to come from Sean Dargan with Macquarie.

I'm just wondering what it would take for you to get to change your assumption around the gross projected default-to-claim rate? What would you need to see to maybe bring that down in the more recent delinquencies?

Sean, I think it -- we use historical experience. So certainly, an improvement of cure rates would go a long way. I think that's the -- would be the biggest driver. Or in some other way, we've got to be convinced that the defaults won't go to claim. We give a lot of evidence that people don't make payments. And the aging of the defaults there, we give a series of slides that show -- or demonstrate why these aged defaults may not be paid claims. But I think evidence would be improvement in cure rate, for sure, or other evidence that they won't be paid claims.

Okay, great. And going back to the singles. You mentioned a few numbers. Did you say it was -- singles were 9% of overall production in recent quarters?

The borrower-paid singles, we specifically said, would likely decline or go away with QM's implementations. That's 9% of our overall production or a little less than 1/3 of our singles.

Okay. And so the overall single production is around 28% of the total, or has been in recent quarters?

Yes, yes. About 30%. A little less -- about 30%.

Okay. And where do you think that might shake out if interest rates continue to rise or stay where they are?

We certainly have it going down into the 20s and perhaps down to 20% over time.

Okay. So one last question. Can you remind us what the rationale for holding your investments as trading securities when your competitors and, I believe, most insurers hold them as AFS?

Yes. We classified our investments, or most of them, as trading. If you go back to 2007, 2008, we had a lot of uncertainty at that time regarding future claims and operating cash flow being negative. So there was an expectation that we would likely need to liquidate investments, and that really drove the classification to trading. And that requires us to run any unrealized gains or losses through the P&L. It ends up in the same place. It ends up in retained earnings, but -- or in the balance sheet, but we run it through the P&L because of that.

And our next question will come from Mark DeVries with Barclays.

First, I have a follow-up on your investment portfolio. Given the fact that it is categorized in trading in the income statement, at least GAAP income statement, and the volatility that creates, have you thought about repositioning that to get a little more defensive and have less duration into a potentially further rising rate environment?

Yes. There's certainly some thought into reevaluating, especially due to our current expectations regarding holding the securities that we own. So yes, we've had some thought about that.

Okay, great. Could you talk a little bit about how the distribution of new notices by vintage here has been moving? Are you seeing a clear shift towards less of the '06 and '07 and more of the newer loans?

Yes. I think over time we've been seeing gradual credit burnout. So the percentage coming from kind of the later -- earlier vintages has been decreasing pretty steadily over time.

Okay. And I would assume, as that evolves, you should see an expectation in your ultimate cure rates on those, right? I mean, I assume you're not assuming the same level of cures on an '06 as you are on a 2011 loan. Is that fair?

Well, to the extent that there are various vintages within the population that we observe, I would say yes. So if the population turns over, over time, and it's all recent vintages and the cure rates are higher, that's going to be reflected.

Okay, got it. I think -- Bob, I think you mentioned that there's a plan to downstream some capital into the MI subsidiary in the quarter. Did you say how much you're planning to push down?

No, we didn't say, Mark. It's really going to depend on the growth in the books. So we're -- obviously, we're growing our books significantly. We're writing a lot. So it's going to depend on the growth, that growth, the persistency and also, obviously, the results. But the plan is to manage to just below 20:1.

Okay. And next question. I'm just trying to get a sense of what I should be modeling for average paid claims. If you were to pay all of your claims today, all the loans -- delinquent loans you expect to pay out, do you have a sense, Bob, for what the average claim would be relative to what it was this quarter? Should we expect that to trend down, flat or be up?

Yes. I mean, I think it would be in the range that it's been this year. Quarter to quarter, the average claim that's tended to go up and down more related to what I said, the size, the coverage percentage. So it's more or less coincidental. But if you look at the average claim over a period, I think that's a fair estimate for the future.

Okay, got it. Just really quick questions. First, is there any impact on your average premium from the decision to reduce the amount that you're reinsuring? Does that flow through the premium line?

That will -- yes, that will reduce the ceded premium. But that's going to happen over time because we've got this existing reinsured book. So you're not going to see that right away. You're going to see it much more in premiums written than you will in earned.

Okay, got it. And then just finally, can we assume that for every dollar of stock price appreciation, that adds roughly $7 million of stock comp expense? Is that the right way to look at it?

Yes, that's about what it's been and what we expect it to be. There are other factors that could impact it, so it's not a perfect correlation, but that's why we said -- and it's been about that for the last 2 quarters.

And our last question is going to come from Geoffrey Dunn with Dowling and Partners. Geoffrey M. Dunn - Dowling & Partners Securities, LLC: Bob, I got a couple of questions. First, can you talk about provisioning on new notices? And maybe compare the levels that we're looking at today versus precrisis or normalized type of provisioning for normalized incidents.

The 25% roll rate is higher than -- if you go back in history, it's higher than history. Geoffrey M. Dunn - Dowling & Partners Securities, LLC: And where would you put history?

Probably closer to the 15% range. Geoffrey M. Dunn - Dowling & Partners Securities, LLC: Okay. And then as we look at -- using your disclosures from the slide deck and we look at the vintage loss ratio development, there seems to be a pretty big delta of performance looking at the '09 versus '010 book. And I'm not sure if there's a material underwriting difference there, or if '09 is getting hit by refinancing activity. Can you talk a little bit about that delta? And -- I mean, because obviously, we're looking at single-digit loss ratios on the '10, '11, '12 so far, although they're still pretty young. But '10 versus '09 really stands out.

Yes. I mean, I think -- we have said that '09 looks like it's going to perform as expected. So that was still a transition in terms of underwriting and things like that, but it will perform well. It will perform as expected. '10, '11, '12, hopefully '13, are looking significantly better. So I think the difference there is just how '10, '11, '12 stand out in terms of their performance as opposed '09. '09 is still going to be kind of as expected when we talk about an expectation of mid-teen, or we use -- the more recent ones are going to be better.

Okay. And then when we think about the '10 and after vintages, historically, we tended to look at your '03 and '04 as being the kind of peak provision years and then '04 and '05 as the paid. Do you think the '10 and after books are developing similarly, or are they just more elongated and maybe we'll see a complete development more in year '05 or '06?

They seem to be better at this point. So we -- yes, we still have peak loss period to get to. But based on just comparable, looking at the comparable periods, they're better.

Also, Jeff, in terms of your comments on the difference in the vintage default performance of the '09 versus later books, while, as Bob said, it's not that '09 is bad, '09 is good. The others are so much better. Also, keep in mind the size of the other books is larger, the more recent books.

And I'll hand the call back over to S.A. Ibrahim.

Thank you, operator, and thank you all for participating in our call and look forward to seeing you on the next call. Thanks.

And that does conclude our conference for today. Thanks for your participation and for using AT&T executive teleconference. You may now disconnect.