MGIC Investment Corporation (MTG) Q3 2016 Earnings Report, Transcript and Summary

Good day, ladies and gentlemen, and welcome to the MGIC Investment Corporation Third Quarter Earnings Conference Call. [Operator Instructions] As a reminder, this conference call may be recorded. I would like to turn the conference over to Mike Zimmerman, Senior Vice President, Investor Relations. You may begin.

Thanks, Nicole. Good morning, and thank you for joining us this morning and for your interest in MGIC Investment Corporation. Joining me on the call today to discuss the results for the third quarter of 2016 are CEO Pat Sinks; Executive Vice President and CFO Tim Mattke; and Executive Vice President of Risk Management Steve Mackey. I want to remind all participants that our earnings release of this morning, which may be accessed on our website, which is located at mtg.mgic.com under Newsroom, includes additional information about the company's quarterly results that we will refer to during the call and includes certain non-GAAP financial measures. We've posted on our website a presentation that contains information pertaining to our primary risk in force and new insurance written and other information we think you'll find valuable. During the course of this call, we may make comments about our expectations of the future. Actual results could differ materially from those contained in these forward-looking statements. Additional information about those factors that could cause actual results to differ materially from those discussed on the call are contained in the Form 8-K that was filed earlier this morning. If the company makes any forward-looking statements, we are not undertaking an obligation to update those statements in the future in light of subsequent developments. Further, no interested parties should rely on the fact that such guidance or forward-looking statements are current any time other than the time of this call or the issuance of the Form 8-K. At this time, I'd like to turn the call over to Pat Sinks.

Thanks, Mike, and good morning. I'm pleased to report that we had another strong quarter. Tim will cover the details of the financial results, but before he does that, I would like to spend a few minutes discussing the progress we have made on achieving our strategic objectives. As a reminder, among our strategic objectives are to prudently grow insurance in force, manage and deploy capital to optimize creation of shareholder value, and to preserve and expand the role of MGIC in private mortgage insurance and housing finance policy. First, insurance in force, the primary driver of our future revenues, increased on a year-over-year basis, ending at $180.1 billion. This growth reflects the expanding purchase mortgage market, our company's market share of approximately 17% to 18% and the hard work and dedication of my fellow coworkers to deliver stellar customer service. The 2009 and newer books now comprise 69% of our risk in force and reflecting the credit quality of the loans and current economic conditions continue to generate low level of losses. The pre-2009 books are still generating the substantial majority of our incurred losses. And while the rate of decline in new delinquent notices is slowing, we continue to experience positive trends relative to the number of new notices from these books. The increasing size and quality of our insurance in force, the runoff of the older books, solid housing market fundamentals such as household formations and home sales, and our improved capital structure positions us well to provide credit enhancement and low down payment solutions to lenders, GSEs and borrowers. Given the continued low interest rate environment, we have seen a modest increase in refinanced transactions, which accounted for 19% of our new insurance written in the quarter. The increased refinance activity has resulted in a slight decrease in our annual persistency rate. Most forecasts are calling for a continued increase in purchase mortgage activity in future periods, which is a net positive for our company and our industry. We estimate our industry's market share is approximately 3 to 4x higher for purchased loans compared to refinances, and MGIC tends to be at the higher end of that range. For the quarter, we wrote $14.2 billion of new business. For the first 9 months of 2016, we wrote $35.1 billion of new business, up approximately 6% from the first 9 months of 2015. Based on actual results to-date and the forecasted strength in overall purchase activity for the full year of 2016, we now expect to write $46 billion of new insurance for the full year compared to $43 billion in 2015. This should result in insurance in force increasing by approximately 4% to 5% in 2016. In addition to deploying capital to write new business in the quarter, we took additional actions to advance our strategic objective of managing our capital to optimize creation of shareholder value. Year-to-date, we have reduced the number of potentially dilutive shares by nearly 66 million and improved our capital profile. Tim will cover these activities in detail in a few minutes. We've also made progress toward the strategic objective of preserving and expanding the role of MGIC in private mortgage insurance and housing finance policy. In September, MGIC agreed to participate in a Freddie Mac pilot MI CRT transaction that will transfer risk to MI companies or their credit insurance affiliates. While it is good to see the GSEs continue to explore ways to reduce the government's mortgage credit risk exposure, this new structure should not be confused with deep cover mortgage insurance, which is a front-end credit risk transfer proposal that we and others have been advocating for. The amount of capital we have allocated to this pilot program and the associated premiums are immaterial to our financial results in this or future periods. Future participation in credit risk transfers will need to be evaluated based upon the terms offered and expected returns. More recently, on October 11, our trade association responded to the FHFA's request for input on credit risk transfer or CRT. We believe that an expansion of front-end CRT, including through the use of deep-cover MI, warrant serious consideration by the GSEs and the FHFA as a complement to existing back-end CRT. Without a more robust program, including one that has an entity-based capital component, such as deep-cover MI, the enterprise's CRT strategy would not function adequately during all phases of the mortgage credit cycle, particularly during times of financial stress, and therefore, would not fully realize FHFA's principal objective of protecting U.S. taxpayers. Further, we strongly believe that deep-cover MI will advance the 4 key objectives of a well-functioning housing finance system by ensuring that, one, substantial private capital loss protection is available in bad times as well as good; two, such private capital absorbs and deepens protection against first losses before the government and taxpayers; three, all sizes and types of financial institutions have equitable access to CRT; and four, CRT costs are transparent, thereby enhancing borrower access to affordable mortgage credit. Discussions with the GSEs and the FHFA are ongoing. As we get closer to 2018 when the GSEs are forecasted to be depleted of capital, it is important to remember that $1 of risk transferred to the private MIs is $1 less of exposure to the taxpayer. I would also note that there are still some market participants that are questioning the MIs as reliable counterparties. With strengthened risk-adjusted capital rules from both the GSEs and soon from the state regulators and a clear understanding amongst all parties of the risk we insure, I believe that a mortgage insurance company is a solid counterparty for our customers. I would remind everyone that as an industry, we paid more than $50 billion since the Great Recession, and MGIC alone has paid 100% of valid claims totaling in excess of $15 billion. When the capital markets left the market, it was the private mortgage insurers who were there every day underwriting loans and paying valid claims. Tim will now go through the financial details for the quarter.

Thanks, Pat. In the quarter, we earned $56.6 million of net income versus $822.8 million for the same period last year. The primary driver of the decrease is -- in the GAAP results was the fact that, in the third quarter of 2015, the reverse of valuation allowance associated with our deferred tax assets relating to future periods. Additionally, in the third quarter of 2016, we recorded a loss on debt extinguishment when we repurchased $292.4 million of the 2020 senior convertible notes. To make the year-over-year comparison of financial results more meaningful, we have begun to disclose a non-GAAP measure called net operating income. Net operating income excludes certain items that we do not consider to be part of the operating results of our primary mortgage insurance business. A full definition of these items as well as a reconciliation of net operating income to GAAP net income is included in the body of the press release. With that said, our net operating income for the quarter was $102.2 million or $0.25 per diluted share compared to $83.1 million or $0.20 per diluted share for the third quarter of 2015. The primary drivers of the improvement were lower losses incurred, lower operating expenses and modestly less interest expense. Lower incurred losses were lowered -- received 10% fewer new notices compared to the same period last year. And reflecting the current economic environment, these new notices are estimated to have a claim rate of approximately 12% versus 13% in the third quarter last year. As we have previously discussed, we view a 10% claim rate as a long-term average. The pace of improvement in the claim rates continue to be difficult to project given the unique performance of the pre-2009 loans. During the quarter, we also updated our claim rate assumption for older delinquent notices. In particular, those aged more than 12 months because the actual cure rate experience has outperformed our previous estimates. This resulted in a benefit of approximately $30 million to our primary loss reserves. Also in the quarter, there was an $8 million benefit primarily relating to IBNR. Keep in mind that because our older notices comprise approximately half of our total delinquent loans, even a small change in the cure rate assumption can result in significant reserve development. The pre-2009 legacy books, especially product delinquencies, while continuing -- will decline to -- will continue to dominate the notice activity for the foreseeable future. In the quarter, those books generated nearly 87% of the new delinquent notices received while comprising just over 31% of the risk in force. Additionally, nearly 84% of the notices received had been recorded delinquent previously. We continue to expect that the 2009 and forward books will generate a low level of delinquencies and very low lifetime loss ratios. Reflecting the declining delinquent inventory, the number of claims received in the quarter declined 19% from the same period last year. Net paid claims in the third quarter were $161 million. Primary paid claims were $147 million, down 23% from the same period last year. For the first 9 months of 2016, the effective average premium yield was 51.9 basis points, which compares to the first 9 months of 2015 effective yield of 52.9 basis points. As I discussed last quarter, there's going to be some volatility in this calculation each quarter for a variety of reasons, including the pace of prepayments on older books of business, which have higher premium rates than the business we are currently writing, the FICO LTV mix of new writings, premium refund assumptions that are influenced by expected claim rate, premium resets as older books pass their 10-year anniversary, and the level of the profit commission we earn on a reinsurance treaty, which is dependent on the level of losses incurred that are ceded. We expect that after considering the volatility I just described, the effective premium rate would trend lower in future quarters, however, the exact amount is difficult to predict. At quarter-end, cash and investments totaled $5 billion, including $329 million of cash and investments at the holding company. The investment portfolio had a mix of 69% taxable and 31% tax-exempt securities, a pretax yield of 2.5% and a duration of 4.8 years. Turning to our capital position under the GSE's private mortgage insurer eligibility requirements or PMIERs. At the end of the third quarter, MGIC's Available Assets totaled approximately $4.7 billion and its Minimum Required Assets are $4.1 billion. MGIC's statutory capital is $1.4 billion in excess of the state requirement. Reflecting the profitability of the new books of business as well as the improved performance of the legacy books, the cushion above the Minimum Required Assets was at the higher end of the 10% to 15% range we are currently targeting. We'll try to manage debt level by continually reviewing our use of reinsurance as well as continuing to seek and receive dividends from the writing company. Regarding MGIC's ability to pay quarterly dividends, the Wisconsin insurance regulator approved another $16 million dividend paid to the holding company in the quarter, which brings the year-to-date total to $48 million. We expect to receive a small -- similar sized dividend in the fourth quarter and are optimistic that these quarterly dividends will grow in the future, especially if the difference between Available Assets and Required Assets under PMIERs grows as we expect. Each dividend would be considered extraordinary versus regular and, therefore, requires OCI approval. As we discussed last quarter, we believe it is important to manage liquidity and capital position of the writing company to withstand a mild recession and to preserve the ability to continue to write new business without a remediation plan or the need to access the capital markets. It is also important to maintain a cushion for potentially higher volumes of primary business, new business opportunities and potential changes to the PMIERs. Now let me address the holding company's capital position and the capital actions we took in the quarter. The primary goals of our capital management activities are to continue the positive rating trajectory of the last several quarters to improve our capital structure and eliminate potentially dilutive shares that result in capital raises during the Great Recession. During the quarter, we issued $425 million of 5.75% senior notes. This is the first time we accessed the senior debt markets in quite a few years and marks another milestone for our company. We are very pleased with the investor reception to the offering as well as the terms we obtained. We used a portion of the net proceeds as well as 18.3 million shares of our common stock to purchase $292.4 million of the 2% 2020 senior convertible notes. This repurchase resulted in a pretax loss on debt extinguishment of $75.2 million. During the quarter, we used a portion of the net proceeds to repurchase 13.5 million shares of our common stock that were used as partial consideration in the purchase of the senior convertible notes. Through the close of business on October 17, we repurchased an additional 3.5 million shares. We'll use some of the remaining proceeds to repurchase the balance of shares that we issued in conjunction with the transactions. Any remaining proceeds from the debt issuance will be held at the holding company for general corporate purposes. This transaction, which only modestly increased our leverage ratio, eliminated 42.1 million potentially dilutive shares. As a reminder, earlier this year our writing company, MGIC, purchased $132.7 million of our holding company's 9% junior convertible debentures. These debentures are eliminated on our consolidated financial statement. We also repurchased $188.5 million of the 2017 convertible senior notes. Combined, these transactions eliminated 23.9 million potentially dilutive shares. While credit ratings are not inhibiting our ability to write new primary business, we think that long-term ratings will become more relevant. Therefore, when analyzing various options to restructure our capital profile as well as the ability to minimize potentially dilutive shares, we need to consider the resulting leverage ratio and interest expense of the holding company. Of course, we also need to consider the capitals being created at the writing company and its dividend-paying ability subject to insurance department approval. We will continue to analyze the costs and benefits of implementing various transactions to achieve our capital management goals. And when we determine that there is an opportunity to create long-term value for shareholders, we'll execute such transaction. With that, let me turn it back to Pat.

Thanks, Tim. Before moving to questions, let me give a quick update on the regulatory and political fronts. The review and updating of state capital standards by the NAIC, which the Wisconsin insurance regulator is leading, continues to move forward. At this time, we do not expect the revised state capital standards to be more restrictive than the financial requirements of the PMIERs. At present, there is a great deal of activity around prescribing an end state for the GSEs that would be delivered to the new administration in early 2017, and we continue to be actively engaged in those discussions. However, I continue to believe that the current market framework is what we will be operating in for a considerable period of time. Regarding the FHA, while we cannot say definitively that there will not be any further price reductions, based on public comments and actions to-date by FHA officials, we are not aware of any changes that are being planned at this time. We don't believe that it makes sense to change FHA pricing without first addressing the larger question of the government's role in housing. Simply put, another price reduction would likely shift business away from private capital and expose taxpayers to increased risk at a time when private capital, primarily in the form of mortgage insurance, is ready, willing and able to take this risk. However, if there is an annual premium reduction, we think it would primarily impact business below 700 credit scores, which was approximately 14% of our NIW in the quarter. How much of this business would be at risk is difficult to say as you would need to take into account lenders' concerns over the legal risks associated with FHA lending and the 97% LTV programs that a number of lenders have recently introduced and which continue to require private mortgage insurance. In closing, we continue to make great progress in achieving our strategic objectives. We had net operating income of $102.2 million, wrote $14.2 billion of high-quality business. The in-force portfolio grew. The level of new delinquency notices and delinquent inventory continued to decline. We improved our capital profile. Our market share within our industry is strong. And we maintained our traditionally low expense ratio. We also took advantage of our financial position to reduce potential dilution to shareholders through the debt repurchase and continued paying dividends out of MGIC to the holding company. I see many opportunities for MGIC in the coming years. I firmly believe that there's a greater role for us to play in providing increased access to credit for consumers and reducing GSE credit risk while generating good returns for shareholders, and we are committed to pursuing those opportunities. With that, operator, let's take questions.

[Operator Instructions] Our first question comes from the line of Mark DeVries of Barclays.

So it sounds like you still have some cash remaining, some net cash after your debt issuance and some of the other capital actions, by our estimate, somewhere in the $50 million range to buy back stock, Tim. Does that sound like it's kind of in the ballpark?

Yes. It's probably in the ballpark after we get done repurchasing the 18.3 million that we used as consideration. That's probably the ballpark of what we have left from the offering.

Okay. Great. And could you discuss appetite to do another senior debt issuance and try and take out the remaining 2020 converts?

I think it's something that, as we mentioned in the opening comments, we continue to look at. Obviously, look at the cost of that and then obviously look at sort of the cost of issuing the debt in the market. So it's something that we continue to look at every quarter and discuss with our board.

Okay. Got it. And then in terms of your 9% convert, is there much supply of that at this point? Do you have any willing sellers? Or were you able to buy most of the bonds out there that owners were willing to sell?

There's less of them out there now than, obviously, when we bought them back at the beginning of the year. But it obviously depends upon market conditions but assumes, depending upon the price, there would probably be some interested sellers.

Okay. Got it. And then finally, a question for you, Pat. I think you and others in the industry have commented that you expect to get your fair share of the share that Arch has mentioned they expect to lose as a product of the acquisition. Have you seen any signs yet of lenders already starting to shift share around?

No, not yet. We're basing that assumption on statements that Arch has made. But to this point, we haven't seen lenders take any action. So we're assuming that when the transaction close, the opportunity will be presented to us.

Our next question comes from the line of Phil Stefano.

Just a follow-up on that last question a bit about the Arch transaction. How can we qualitatively think about what's the bigger opportunity for NIW growth? Would it be the fallout from that transaction or the outlook for purchase origination increasing?

Phil, this is Mike Zimmerman. I mean, obviously, both would -- I mean, if you're talking about company specific, both obviously would contribute to that. I think the easier one to do certainly is the purchase market because those forecasts are widely available. The wildcard really is, is market share within the industry, but that's, again, company specific always the case.

Yes. And I guess -- and thinking about persistency, it feels like you've historically talked about this being in the low- to mid-80s. I mean, obviously, refis have been a drag there, and we'll probably see that maybe a little more in the fourth quarter. Does low to mid-80s still make sense in this low interest rate environment? Or does persistency come in a little bit lower than that as we think about this lower-for-longer environment?

Well, it's obviously -- this is Tim. It's obviously drifted a little bit lower, again, this quarter. I still think when we look at it long term, we still think 80% is probably a good bogey to look at, but it could probably be a couple points either side of it. Obviously, you can't predict exactly what's going to happen with interest rates, but we've had some little mini refi booms that have impacted the persistency again over the last couple quarters. So you could see a little bit of drift down on that, but I still think when you look at it long term, 80% is a good sort of place to sort of ground, I guess, your assumptions.

Okay. One quick last one. On the share repurchases, looks like if I do the math right here, there's about $1 million, $1.5 million left of repurchases to take out the common stock using the recent debt deals. Does valuation matter? Or are you agnostic and just want to get that 18.3 million taken out as soon as possible?

Well, I think you're right as far as the numbers go. And I think when we launch the transaction, the plan is obviously to take the shares out, not add any additional dilution. So our plan is to buy those shares back from the market.

Our next question comes from the line of Bose George of KBW.

Actually, just a follow-up to some of the buybacks. Of the $50 million that's remaining, did you say that the plan is to allocate that to buybacks as well? Or is that undecided at the moment?

No. What we said is that's for general corporate purposes at this point. We haven't announced any additional buybacks with that. We have to manage the debt service that we have at the holding company right now. We've got 2017 maturity coming up, so it has given us a little additional liquidity, but we have a little higher run rate on our annual interest expense, too, after the offering. So all those things sort of go into the mix.

Okay. And then can you just remind me for the -- what's the plan for the 2017 converts?

2017 converts will pay off the maturity. I think there's $145 million of those left right now.

Okay. Great. And then just switching to the GSE risk-sharing comments that you made. Just based on your discussions with the GSEs, does it look like the goal to pursue the deep end and sort of the front end loan level risk sharing is still progressing?

This is Pat. We are -- we continue to be engaged, particularly with the FHFA, both as a separate company as well as our trade association has made a lot of effort recently both with our response to the request for input as well as face-to-face meetings, so they continue to progress.

Our next question comes from the line of Jack Micenko of SIG.

This is actually Soham, on for Jack. My first question was just a follow-up to the last question. The assumptions on the deeper coverage settie [ph] put out by Milliman last year. Was that just going to be some reduction in g-fees to give MIs full credit for the coverage? I mean, how certain are we or the industry to reduce g-fees? And if there was no g-fee reduction, what sort of setup are we looking at in order to move from a pilot to something more systematic?

This is Pat. I think as the year has progressed and our thinking has evolved, our view is -- because whether or not there was a benefit to the borrower was a point of contention among some. We've taken a view that let's do the work, let's do the analysis. And if there should be a benefit to the borrower, then we can get into what happens to g-fees. But rather than have that debate, we've got analysis that just says the concept works, the capital structure works. So let's not right now worry about the borrower, and if something shakes out, we will. How do we move it to the -- your second question, how do we move it from pilot into something that's real. That's a ways off in the future. I mean, the GSEs have been trying different pilots. I would expect that if we can get it done in the pilot phase first, so it's very difficult to predict as to when it would be an actual ongoing business opportunity such that we could factor it into our financial forecast.

All right. And then, Pat, you mentioned more recently that there was more stability in LPMI. But what's the risk that aggressive pricing returns if rates do go up in December and we have something similar to this year where the long end flattened? And then could you just remind us your target mix on the business now that we have -- we're having a couple quarters of new pricing?

Well, the target mix is pretty much set by the market. I mean, right now, it's about 75-25. It could hit as high as 80, meaning, borrower paid versus lender paid. So we have seen some stability in the borrower paid segment. That 75% as all the MIs pretty much lined up on rates over the course of the first and second quarter of this year. The LPMI rates continue to be aggressive. People play there selectively. So it kind of depends on the lender and the situation as to who's being -- trying to win those bids. But we're just watching it play out. It's difficult to predict. If anything in the future is going to happen one way or the other, I'm just pleased that we have some stability in the market right now.

Okay. Great. And then just last one for Tim. Your investment yields look like they took a slight dip this quarter compared to the last 3 quarters. What's going on there? And how are you guys positioning the portfolio ahead of a potential rate hike in December?

Yes. I think the way we look at it, for the most part, it stayed fairly flat. I mean, there's a little bit of volatility, but I think we think it's, for the most part, flat. We are trying to transition a little bit in -- more into the muni portfolio now that we showed taxes on the income statement and some point in the future we'll be paying cash taxes so that you have to take that into consideration when you look at the yield. All things considered, though, from a duration, we've moved out a little bit further but don't see us going significantly up from where we are right now.

Our next question comes from the line of Doug Harter of Credit Suisse.

Where do you think you guys are in terms of expenses? As the portfolio grows, as you continue to write new business, do you need to add any expenses?

This is Tim. No, I think we're at a pretty good rate right now. I think that one of the things we always say is that this is a business that scales very well. And so if we have higher volumes of business, we think with our existing sales force underwriting team that we can handle additional volumes. Obviously, when you talk about underwriting, there could be little incremental cost. But I think we feel like the work that we've done over the last few years to, again, keep focus on expenses as we always have keeps us in a pretty good spot even if volumes go up from here.

And then can you just remind us of when your reinsurance deals are up for renegotiation or renewal as capital is building at the insurance subsidiary?

Yes. This is Tim, again. Our current reinsurance covers the NIW through the end of '16, so we'll be looking at putting reinsurance on for our '17 books and have had discussions with reinsurers and think that, that'll look pretty similar to our existing treaty right now. But we have the ability for the existing treaty to early terminate at the end of 2018 should we so choose. So that's the first contractual place where we have an ability to sort of -- to look at resizing that reinsurance that's in place right now.

Our next question comes from line of Chris Gamaitoni of Goldman Sachs.

It's Chris Gamaitoni from Autonomous. I think it's me. Can you give us -- do you have any sense of when the timing or what year you think you'd be able to get ordinary dividends out of the writing company, when you kind of lapped the period for the contingency reserve release?

Chris, it's Tim. I think that's still years out, and I say that for a couple of reasons under the current rules because the contingency reserve build, that is a direct deduction for our statutory income, which is a key component of the calculation for us. So I think especially if you think about anything meaningful, that's the way you have to think about it. Also, when we're in the [indiscernible] where we're getting extraordinary dividend, that also sort of bars you from getting -- government classifies as ordinary. So I think that is the best way to think about dividends right now is in the extraordinary methodology. And that, again, our goal is that those will continue to grow as our excess over PMIERs grows and that the regulator gets more and more comfortable with our capital adequacy and ability to stream dividends up to the holding company.

Okay. And how do you weigh reducing the usage of reinsurance versus capital return when you think about your capital planning actions?

Yes. I think it's a balance. I think we're very conscious of not getting overloaded from a reinsurance standpoint. But at the same token, we find it a very valuable tool. And I tell you that from a regulator standpoint, I think they take a lot of comfort of us having reinsurance if there were to be any stressed environment. And so I think we'll continue to look at those and see how it sort of plays out regarding dividend capacity. But the way we look at it right now, the more of an excess we can continue to grow and the lower risk to capital goes, I think the better case we have for additional dividend capacity.

Okay. And switching gears. There's been some trade publications that have mentioned that state regulators are looking more closely at LPMI competition. Do you have any kind of views or commentary on that whether that's occurring and what you think the impact might be in the future?

Yes. Chris, this is Mike. Actually, last quarter, we talked about that. There were -- California had inquired about some pricing in the LPMI space. As for our knowledge of -- at least from our company perspective, we've resolved that issue with California and are moving forward with our pricing strategy as Pat's articulated in the past.

Okay. So no real change from where it is today, you don't think?

Not for MGIC, no.

Our next question comes from the line of Eric Beardsley of Goldman Sachs.

Just on the premium rate, you've mentioned that it's a little bit tough to predict here. I think you previously said it would be down 2 to 3 basis points from, I think, fourth quarter '15 levels as we got through year-end here. I guess, just a couple questions on that. One, just in the third quarter, I just wonder if you could help us quantify how much of a benefit you had from accelerated amortization on the single premium policies? And then two, I guess, where should we ultimately see that premium rate settle out?

Yes. This is Tim. I think, for the quarter, we saw about $3 million more in accelerated singles than we would have saw in the second quarter. So if you're looking quarter-over-quarter, we probably had $3 million additional benefit in Q3 of 2016. That to-date was up probably even -- Q2 was probably even up a couple million over Q1 of this year. So you just take that in account when you look at sort of the run rate. As we said sort of in the opening comments, we still expect that it's going to trend down. And if you look sort of for the first 9 months of this year versus first 9 months of last year, we see the trend downward. It's tough to tell at what exact pace, but I don't think our view has changed significantly as far as that guidance that the average basis point, again over the long term, trends downward even though there could be volatility on the quarters.

Got it. So I guess, given the refi environment and whatever other trends are impacting it, could it be better than that initial guidance of down 2 to 3?

It can be depending upon the refi environment, and obviously, we talked about the level of losses under the reinsurance treaty as being something that impacts it. I would say that the headwinds are more going downward, but you can have, obviously, periods like this where you have some accelerated earnings on single that'll bring it up. And so that's why I think you have to look longer-term trends as opposed to quarter-to-quarter.

Got it. And just back to the share count. So you mentioned you'd done 3.5 million shares since quarter-end and maybe there's another, I think you mentioned 1.5 million or so to do, so let's say you have share count down 5 million versus 3Q run rate. And then I guess from there, if we think about the 2017 converts getting paid off in cash next year, can you just help us with how much of the diluted share count or that base at somewhere around 10 million, 11 million shares right now?

Eric, this is Mike. The 2017s is just about 10 million shares associated with those, so right.

Got it. So we're looking for the share count to come down another, call it, 15 million from a third quarter run rate, give or take?

Yes. But I would say you that -- right, the run rate, that -- the '17s, right, is maturing really -- they're going to be there in the first quarter, and they'll be there partial time of the second quarter as well. But long term, that would be correct.

Yes. And the other thing to think about is, obviously, the transaction that we did at the beginning of August, that full share count from an EPS standpoint was announced for the full quarter. So obviously, look at the shares that are remaining at the end of the quarter. And I think if you look to adjust the additional shares that we have repurchased since the end of Q3 and along with the '17s, that's the right way to start thinking about, I guess, if you're looking out to 2017.

Okay. Great. And then just lastly, I think you'd been talking originally about a 32% tax rate. Is that still good to think about going forward? Or is it more close to a 34% adjusted rate we saw this quarter?

So I think 32% is still the right way to sort of think about where we're trending to.

Our next question comes from the line of Mihir Bhatia of Bank of America.

My questions have been answered.

Our next question comes the line of Geoffrey Dunn of Dowling & Partners.

Tim, I think you indicated that you're expecting reinsurance on the '17 to look similar to what you already have. As you think about your capital management and risk management planning, have you guys looked at or considered anything like these [indiscernible] deals that you just ceded [ph] and looking more to the fixed income execution rather than the reinsurance market?

Yes. Geoff, it's something that we've looked at and we've paid close attention to. If you recall, we did some capital markets reinsurance deals back in the 2000s, which we call Home Re, so we're familiar with sort of the general structure of it. But obviously, you've paid close attention to what's developed over the last couple years. I would say that some of those transactions look appealing to a certain extent. I think you have to weigh what the PMIERs credit will be for it, and it's uncertain to us what credit you get not just for, say, day 1 but say 1 or 2 years out from that. And that's one of the reasons why we like the quota share transactions that we've been doing with more traditional reinsurers, but it's something that we'll continue to explore and look to see if it makes sense for us to execute on.

And our next question comes from the line of Chris Gamaitoni of Autonomous.

Do you have a sense of when -- what the pace of decline of the 2006 and 2007 [indiscernible] digital be? I'm really speaking about kind of the natural amortization to 78% and when you might see a cliff in the -- of those being in force, acknowledging some have been HARP-ed, so maybe that's pushed out?

, Yes. Chris, this is Mike. I mean, it's not just from a mathematical basis. Both had high percentage of 97s and 100 LTVs, and on a strict amortization basis, you're looking at 11, 12 years to get to the date where you hit a 78 LTV. So 11, 12 years from those vintage years. Then there's the -- when you start factoring in hope avail you [ph] of the other considerations there, whether there's a presence of a second lien and other considerations of the law from a cancellation perspective. But from a day perspective, related -- 11 to 12 years, so '18 and '19. But keep in mind a lot of those books also have been HARP-ed, so that resets the clock.

About 50-ish percent rough have been HARP-ed, so I guess we could see maybe a decent set down in, call it, '18, '19 and then kind of that half of the tail still being there from the HARP portfolio.

A lot have been -- are currently delinquent and have been historically delinquent, too, so that will weigh into that.

Yes. So this is Steve Mackey. I would not think of a kind of cliff effect at all or a significant step down as we go from '17 to '18 and '19. I kind of view the wind-down of that vintages -- of those vintages as more of a slow grind down over time.

Okay. I just ask as kind of Old Republic, which as a pure runoff book, has kind of pinpointed 2022 as when they think most of it will be gone with a similar type of mix.

Okay. I mean, so we have 5 years -- well, 5 years is, I guess, a pretty good long runway probably, so I mean it's hard to argue whether it's that far out, so that's not a cliff.

Okay. And on IBNR, do you have the gross dollar amount? You said an $8 million benefit. Is that the 91 minus 8 or is it a different number?

I think we went from around $66 million at the end of Q2 down the $8 million at the end of Q3s.

Our next question comes from the line of Mackenzie Aron of Zelman & Associates.

Just one follow-up on the loss reserves and kind of the assumptions that are baked in there. Can you talk a little bit about what you're seeing on the severity? I know in 2Q, you had noted that you had seen some improvement there, particularly in the judicial states and felt like the reserve adjustments related to severity had caught you up. But can you just talk about what you saw the last -- this last quarter and whether there's trends there that you might still be watching that haven't yet been reflected in the adjustments?

Mackenzie, this is Tim. I think Q3 to me was very similar to what we saw in Q2. We definitely didn't see any deterioration, so I think our expectation based upon our estimates are that we have caught up at this point. That's not to say that things can't change in the future, but there's nothing in Q3 on the severity side that led us to believe that things were getting any worse, not really getting any significantly better either, though, I would say.

Okay. Great. And then -- so some of the claim rate as we think out to next year, can you give us any order of magnitude in terms of how much the claim rate on new notices could potentially decline next year?

No. I think it's too early to say. I think that's been a tough one to predict. I think we did see some year-over-year improvement of Q3 this year versus last year, but the -- it gets tougher, I think, as we get to what we think is sort of the historical long run average of 10% to get that meaningful improvement. I think the bigger improvement we're going to get out of incurs, quite frankly, is from the lower new notices that are coming in the door is probably the best way to think about it.

And I'm showing no further questions at this time.

Okay. This is Pat. Once again, we thank you for your interest in our company, and have a great day. Bye.

Ladies and gentlemen, thank you for participating in today's conference. That does conclude today's program. You may all disconnect. Everyone, have a great day.