NMIホールディングス（NMIH）2026年第3四半期決算説明会：保有契約高が過去最高、調整後EPSは1.23ドル

主なポイント

• National MIは136億ドルの新規引受保険金額（NIW）を計上し、2026年第3四半期末時点の元受有効保険金額は過去最高の2297億ドルに達した。
• 総収入は1億8750万ドルとなり、2026年第2四半期は1億8790万ドル、2025年第3四半期は1億7870万ドルだった。
• 調整後純利益は9370万ドル、希薄化後1株当たり利益は1.23ドル、調整後自己資本利益率は13.9%となった。
• 保険金費用は前四半期の1310万ドルから2910万ドルに増加したが、これは通常の季節性、ならびに保険ポートフォリオの拡大とシーズニングを反映したものである。
• 同社は、2030年までに獲得するすべての新規案件を対象とする再保険カバーを、推定税引前資本コスト4%で確保した。
• National MIは当四半期中に2710万ドルの普通株式の自社株買いを実施した。経営陣は、当面の自社株買いペースについて最近の水準とほぼ同程度で推移すると見込んでいる。

主要財務データ

*投資ポートフォリオの純未実現損益を除く。

事業および営業実績

12カ月継続率は80.8%となり、第2四半期の81.4%から低下した。経営陣は、足元の低下は小幅にとどまっており、10月には安定化の兆しが見られたと述べたものの、今後の継続率は住宅ローン金利の動向に左右されるとした。

再保険コストおよび解約益を除くコア利回りは、前四半期の33.6ベーシスポイントに対し33.4ベーシスポイントとなった。経営陣は、蓄積された住宅資産（ホームエクイティ）や、ホームエクイティローンまたは信用枠に伴う最新の不動産鑑定評価に支えられ、借手主導による住宅ローン保険の解約が高水準を維持していると述べた。

National MIが報告した9月30日時点のデフォルト件数は8,571件となり、6月30日時点の8,020件から増加した。四半期末時点のデフォルト率は1.23%だった。治癒率は29%から26%に低下したが、経営陣はこれを信用状況の予期せぬ悪化ではなく、通常の季節的な変動であると位置づけた。

同社は新規デフォルト通知1件当たり約1万3500ドルの引当金を計上し、第2四半期の1万4000ドルと概ね同水準となった。経営陣は、直近のヴィンテージにおける保険金請求の頻度および深刻度の上昇について、それ以前のローンと比べて住宅価格の上昇環境が弱かったことを反映したものであり、同社の四半期ごとの想定から大きく乖離する変化はないと説明した。

資本および再保険

National MIは10月に新たな比例再保険および超過損害再保険契約を締結した。これらを合わせることで、2030年までに獲得するすべての新規案件に対し、推定税引前資本コスト4%でフォワードフロー型の再保険カバーが提供される。

経営陣は、同社が再保険を前提とせずグロスベースで新規契約の保険料価格を設定していると述べた。したがって、有利な再保険条件は資本効率とリスク管理を支えるものの、フロントエンドの価格決定に直接影響を与えることはない。

四半期末時点で、National MIが保有するPMIERsに基づく利用可能資産合計は38億ドルとなり、リスクベース所要資産の22億ドルを上回った。超過利用可能資産は16億ドルだった。

同社は第3四半期中に平均価格43.18ドルで62万6000株を2710万ドルで自社株買いした。2022年のプログラム開始以来、4億3500万ドルの自社株を買い戻し、平均価格30.54ドルで1430万株を消却した。自社株買いの承認残高は1億4000万ドルとなっている。

経営陣の見通し

経営陣は、長期的な住宅市場のトレンドに支えられた魅力的な新規事業機会が引き続き存在すると述べた一方、高水準の住宅ローン金利や広範なマクロ経済リスクが住宅取得能力（アフォーダビリティ）やNIW量に影響を与える可能性があることも認めた。

同社は、有効保険金額のオーガニックな成長資金の確保と株主還元のバランスを維持する方針である。経営陣は、当面の間、自社株買いは最近と概ね同等のペースで継続される見通しであると示した。

第4四半期の費用には、好業績の年に見られる同社の典型的な季節パターン通り、従業員報酬の引当金増加が含まれる可能性がある。

リスクと注視すべきポイント

• ポートフォリオの拡大とシーズニングに伴い、保険金費用は第2四半期から急増したものの、経営陣はこの変動について季節パターンと一致していると説明した。
• デフォルト件数は前四半期比で増加し、治癒率は29%から26%に低下した。
• 借手主導および期日到来による住宅ローン保険の解約は、継続率や保険料収入に影響を与える可能性がある。
• 住宅ローン金利は引き続き、住宅取得能力、借り換え需要、継続率、新規保険需要の重要な要因となっている。
• マクロ経済の不透明感が残るなか、経営陣は価格設定、リスク選別、再保険に対して引き続きプロアクティブなアプローチをとっている。

アナリスト質疑応答（Q&A）のハイライト

自社株買い：経営陣は、National MIは資本還元に注力し続けているものの、オーガニックな成長資金も確保する必要があると述べた。同社は、自社株買いを直ちに加速させるのではなく、概ね最近のペースで継続すると見込んでいる。

保険契約の解約：借手主導による解約の高止まりが数年間続いている。経営陣はこれについて、住宅所有者がセカンドモーゲージやホームエクイティラインを通じてエクイティを引き出す際に最新の不動産鑑定評価を取得し、それによってより早い段階で住宅ローン保険の解約を申請できるようになっていることが一因であると説明した。

損失実績：経営陣は、2022年から2024年のヴィンテージにおける10%台後半の損害率は、価格設定時の想定と同水準またはそれを上回る良好な結果であると述べた。最終的な結果は、既存のデフォルト案件の治癒状況や、今後のマクロ経済および住宅市場環境に左右されることになる。

業界の価格設定：経営陣は、民間住宅ローン保険市場は均衡が取れており規律が維持されていると述べ、競合他社の動向に最近目立った変化は見られないとした。

決算説明会（アーニングスコール）の全文トランスクリプト

決算説明会の完全なトランスクリプト

経営陣による説明

Operator

Good evening, and welcome to the NMI Holdings, Inc. Third Quarter 2026 Earnings Conference Call. [Operator Instructions] Please note this event is being recorded.

I would now like to hand the conference over to management. Please go ahead.

John Swenson

Thank you, operator. Good afternoon, and welcome to the 2026 Third Quarter Conference Call for National MI. I'm John Swenson, Vice President of Investor Relations and Treasury. Joining us on the call today are Brad Shuster, Executive Chairman; Adam Pollitzer, President and Chief Executive Officer; and Aurora Swithenbank, our Chief Financial Officer.

Financial results for the quarter were released after the close today. The press release may be accessed on NMI's website located at nationalmi.com under the Investors tab. During the course of this call, we may make comments about our expectations for the future. Actual results could differ materially from those contained in these forward-looking statements. Additional information about the factors that could cause actual results or trends to differ materially from those discussed on the call can be found on our website or through our filings with the SEC. If and to the extent the company makes forward-looking statements, we do not undertake any obligation to update those statements in the future in light of subsequent developments.

Further, no one should rely on the fact that the guidance of such statements is current at any time other than the time of this call. Also note that on this call, we may refer to certain non-GAAP measures. In today's press release and on our website, we've provided a reconciliation of these measures to the most comparable measures under GAAP.

Now I'll turn the call over to Brad.

Bradley Shuster

Thank you, John, and good afternoon, everyone. I'm pleased to report that in the third quarter, National MI again delivered standout operating performance, continued growth in our insured portfolio and strong financial results. Our lenders and their borrowers continue to turn to us for critical down payment support. And in the quarter, we generated $13.6 billion of NIW volume ending the period with a record $229.7 billion of high-quality, high-performing primary insurance-in-force.

In Washington, conversations remain focused on promoting broader access and affordability to the housing market. And we believe there is broad recognition of the unique and valuable role that the private mortgage insurance industry plays in this regard. At National MI, we are in the market every day with a clear mandate and purpose, offering a low-cost, high-value solution that helps borrowers bridge the down payment gap dramatically reduce the cash they need at the closing table and buy a home years earlier than they otherwise could.

In the process, we help to make homeownership more affordable and achievable for millions of Americans and communities across the country, with coverage that works to insulate the GSEs and taxpayers from risk and loss in a downturn. We recognize that there is a need now more than ever to support prospective buyers as they look to access the housing market, establish roots in their community of choice and build long-term wealth through homeownership. We're proud of the work we do and the assistance we provide. And to date, we have helped more than 2.2 million borrowers realize their homeownership goals.

With that, let me turn it over to Adam.

Adam Pollitzer

Thank you, Brad, and good afternoon, everyone. National MI continued to perform in the third quarter, delivering significant new business production, consistent growth in our insured portfolio and strong financial results. We generated $13.6 billion of NIW volume and ended the period with a record $229.7 billion of high-quality, high-performing primary insurance-in-force.

Total revenue in the third quarter was $187.5 million, and we delivered adjusted net income of $93.7 million or $1.23 per diluted share and a 13.9% adjusted return on equity. Overall, we had a strong quarter and are confident as we look ahead. The macro environment and housing market have remained resilient.

Despite elevated rates, our lender customers and their borrowers continue to rely on us in size for critical down payment support, and we see an attractive and sustained new business opportunity fueled by long-term secular trends. We have an exceptionally high-quality insured portfolio covered by a comprehensive set of risk transfer solutions and our credit performance continues to stand ahead. We're delivering consistent growth and embedded value gains in our insured book, and we continue to manage our expenses and capital position with discipline and efficiency, building a robust balance sheet that's supported by the significant earnings power of our platform.

Taken together, we see a clear opportunity for continued success. Notwithstanding these strong positives, however, macro risks do remain, and we've maintained a proactive stance with respect to our pricing, risk selection and reinsurance decisioning. It's an approach that has served us well and continues to be the prudent and appropriate course. More broadly, we've been encouraged by the continued discipline that we see across the private MI market. Overall, we had another solid quarter, delivering standout operating performance, consistent growth in our insured portfolio and strong financial results.

Looking ahead, we're well positioned to continue to serve our customers and their borrowers, invest in our employees and their success, drive growth in our high-quality insured portfolio and deliver through the cycle growth, returns and value for our shareholders.

With that, I'll turn it over to Aurora.

Aurora Swithenbank

Thank you, Adam. For the third quarter, we generated $13.6 billion of NIW and our primary insurance-in-force grew to $229.7 billion. 12-month persistency was 80.8% compared to 81.4% in the second quarter. Net premiums earned in the third quarter were $155.5 million compared to $157.5 million in the second quarter and $151.3 million in the third quarter of 2025. Net yield for the quarter was 27.2 basis points. Core yield, which excludes the cost of our reinsurance coverage and the contribution from cancellation earnings was 33.4 basis points compared to 33.6 basis points in the second quarter.

Investment income was $32.2 million in the third quarter compared to $30.3 million in the second quarter and $26.8 million in the third quarter of 2025. Total revenue was $187.5 million in the third quarter compared to $187.9 million in the second quarter and $178.7 million in the third quarter of 2025. Underwriting and operating expenses were $30.8 million in the third quarter compared to $30.5 million in the second quarter. Our expense ratio was 19.8% in the quarter.

We had 8,571 defaults at September 30 compared to 8,020 at June 30, and our default rate was 1.23% at quarter end. Claims expense in the third quarter was $29.1 million compared to $13.1 million in the second quarter and $18.6 million in the third quarter of 2025. Claims experience in the quarter reflects normal seasonal activity and the continued growth and seasoning of our portfolio.

Adjusted net income was $93.7 million compared to $106 million in the second quarter and $95.7 million in the third quarter of 2025. Adjusted diluted earnings per share was $1.23, compared to $1.38 in the second quarter and $1.21 in the third quarter of 2025. Shareholders' equity as of September 30 was $2.7 billion. Book value per share, excluding the impact of net unrealized gains and losses in the investment portfolio was $38.13, up 3% compared to the second quarter and 14% compared to the third quarter of last year.

In the third quarter, we repurchased $27.1 million of common stock, retiring 626,000 shares at an average price of $43.18. Since starting our buyback program in 2022, we've repurchased a total of $435 million of common stock, retiring 14.3 million shares at an average price of $30.54. We have $140 million of repurchase capacity remaining under our existing program.

In October, we entered into a series of new quota share and excess of loss reinsurance treaties, which together further extend our comprehensive credit risk management program and provide us with forward flow coverage for all new business produced through 2030 at an estimated 4% pretax cost of capital. Reinsurance has long been a core pillar of our risk management strategy, working to mitigate the potential impact of credit volatility in our insured portfolio and has consistently provided us with a deep, secure and efficient source of PMIERs growth capital. We have significant experience, strong secondary market relationships and a track record of leading with innovation across the risk transfer spectrum.

The deals we have just secured are among the best we've ever achieved in terms of their cost capacity, duration and structure. And serve to highlight the quality of our insured portfolio and the differentiation we've achieved through our comprehensive credit risk management framework. At quarter end, we reported $3.8 billion of total available assets under PMIERs and $2.2 billion of risk-based required assets. Excess available assets were $1.6 billion.

Overall, we achieved strong financial results during the quarter, delivering consistent growth in our high-quality insured portfolio, continued expense efficiency and significant bottom line profitability, returns and book value growth.

With that, let me turn it back to Adam.

Adam Pollitzer

Thank you, Aurora. We had a strong quarter, once again delivering significant new business production, consistent growth in our high-quality insured portfolio and strong financial results. We have a valuable customer franchise, a talented team driving us forward every day an exceptionally high-quality book covered by a comprehensive set of risk transfer solutions and a robust balance sheet supported by the significant earnings power of our platform.

Taken together, we are well positioned to continue to serve our customers and their borrowers, invest in our employees and their success, drive growth in our high-quality insured portfolio and deliver through the cycle growth, returns and value for our shareholders.

Before turning to Q&A, I also want to take a moment to comment on recent housing policy developments in Washington. On October 22, the President signed a broad executive order focused on lowering housing costs for American families. We share his goals and are proud to already be delivering a low-cost, high-value solution that helps open the door to homeownership opportunities for millions of deserving borrowers in communities across the country.

We're looking forward to working with Director Pulte, FHFA staff and the GSEs to continue advancing our shared objective of making homeownership more affordable and accessible in ways that protect the American taxpayer and uphold the safety and soundness of the mortgage market. Thank you for joining us today.

I'll now ask the operator to come back on so we can take your questions.

Operator

[Operator Instructions] Your first question comes from Doug Harter with BTIG.

質疑応答

Douglas Harter

You guys have been pretty remarkably consistent on the pacing of your share repurchase. With kind of higher rates and a slowdown in the market and the reinsurance deals you've done, would you consider kind of accelerating the pace? Or what would it take for you to accelerate the pace of share repurchase?

Adam Pollitzer

Yes. Look, good question. I think I'd start and say we're really delighted with the consistency and success that we've achieved with our repurchase program thus far. We've retired $435 million of stock, which represents 14.3 million shares or 17% of our total outstanding at the time that we initiated the program in February of '22. We are always thinking about our balance sheet and the right balance, if you will, between our funding needs and the sizing and pacing of our shareholder distributions. And so it's something we constantly think about.

I've also mentioned in the past that capital distribution for us is both a near-term focus and also it's a longer-term opportunity. But you talked about the NIW opportunity. If you look at it in the 4.5 years since we launched our repurchase program, we've grown our insurance-in-force by 51% compared to 14% for the rest of the industry. So we certainly consider how market conditions are going to impact new business opportunities, where the broader macro is. But we think we've got a lot of opportunity to continue delivering growth, organic growth in our in-force portfolio. And so we're going to continue to be balanced in how we execute.

And so for now, I think a good assumption is that we're going to -- as we look forward, we'll continue to pace roughly in line with where we've been.

Operator

Your next question comes from Rick Shane with JPMorgan.

Richard Shane

First, just sort of a housekeeping question. It looked like the persistency on the 2020 vintage fell fairly sharply. What's going on there? Just want to make sure I'm not missing something.

Adam Pollitzer

Yes. So Rick, I think it's actually more than a housekeeping item. So it's a valuable one to talk about because I think it dovetails with the executive order that I referenced. And what I mean by that is that our policies already have a range of embedded cancellation features in them that drive turnover in our portfolio, right? Private MI policies automatically cancel when the principal balance of a covered loan amortizes to a 78% LTV.

Borrowers today have the ability to separately initiate a cancellation based on the appreciation in value of their home. And so we see turnover in our book in all periods and in all interest rate environments, whether they're increasing or whether they're decreasing. And so that's what we're seeing come through in our 2022 book. It's those embedded cancellation features that are already coming through. We're at a point where that HOPA cancellation, if you will, that 78% scheduled LTV starts to come into focus a little bit more for the books that we wrote six-plus years ago.

Richard Shane

Got it. Okay. And it's interesting because, obviously, we've tracked this for a long time. And my sense is what you're describing is specific to contractual amortization rather than HPA-driven. And we just haven't ever seen that emerge in a single vintage as significantly as it did in that 2020 vintage in the third quarter. So is there something specific going on there? Or is there a behavioral change amongst borrowers? Were they becoming more proactive about actually taking advantage of the underlying HPA?

Adam Pollitzer

No. I think, actually, so what we've seen on our end, right, with the internal data is, obviously, we always see the scheduled contractual, we'll call it a HOPA, cancellation, it's embedded in the Homeowners Protection Act from the late 90s. So we always see that. We know at the time that we originate a policy, what is that outer date at which it's going to cancel if it hasn't otherwise turned over because the note has been repaid or there's been a borrower initiated cancellation.

But actually for the last several years, we've seen much higher utilization on the borrower-initiated cancellation as well based on value. And I think that what's underpinning that is there's a lot more equity in the homes, borrowers are aware of that, and so it's been coming through, but also in an environment where the low note rate on your existing mortgage is keeping existing homeowners in their current homes, they're not moving with the same frequency. There's not the same opportunity to refinance.

If they need to access the equity in that home, what they're doing is either taking out a second or they're taking out a home equity line. And as part of that process, because the home is collateral, you're getting an updated appraisal, and that updated appraisal then serves as a benchmark to go and initiate a cancellation. So that whole dynamic, there's nothing unique that changed in Q3. That dynamic has been a rich one that we've been seeing come through for several years now.

Richard Shane

Got it. That's really interesting. So the home equity explanation. Well, that's not easy for me to say at this hour. Really probably explains a great deal of it because I'm assuming your 2019 vintage loans, those borrowers may have actually moved along the way, but that 2020 vintage, those borrowers are so locked in because of those rates that they're starting to tap equity. That makes sense.

Operator

Your next question comes from Rowland Mayor with RBC.

Rowland Mayor

I wanted to quickly start on the prior year, I guess the reserve releases a really sharp slowdown in the reserve development. It looks like cure activity was fairly stable. Was there some assumption change there? Or could you walk through what drove the reserve slowdown?

Aurora Swithenbank

So in terms of the way that we present our reserve tables, we present prior year reserve releases. And so as you go on in the year -- in Q1, 100% of our releases relate to prior years. Once you get to the third quarter like we are right now, most of our reserve releases relate to loans that went into default status in Q1 and Q2. So those are current year releases and those get embedded in the current year reserve number. So that is a net number, which reflects the current releases.

I think you hit on the correct issue, which is the cure rate. We're not seeing anything unusual in that. So our cure rate was 26% this quarter versus 29% last quarter. That small deterioration is something that we would typically see in the third quarter just given the seasonal patterns of default behavior.

Rowland Mayor

And then I did want to ask on the quota share. I believe you said it was through 2030 at a 4% pretax cost of capital. How does that flow through your pricing engine? And how quickly does that impact your pricing decisions?

Adam Pollitzer

Yes. So we price on a gross line basis. And so it's great when we can achieve long-term favorable outcomes in the reinsurance market. It's enormously valuable as we think about capital planning, as we think about risk management, for our ability to deliver consistent returns. But when we price, we're pricing on a gross line basis, assuming we don't have reinsurance in place, the reinsurance is additive to what we can then achieve.

And so the outcome, especially when they are the best deals that we've ever achieved, as Aurora said, in terms of their price, their structure, the duration, the terms. But that's not going to necessarily flow through to pricing decisions. Those will be benefits that we realize over time on the back end.

Operator

Your next question comes from Bose George with KBW.

Bose George

Actually, I just wanted to follow up on the credit. So on the cure ratio, I guess I was seeing it as like 70%. Is that right? And that does seem lower than prior quarters. Was there anything to call out there?

Adam Pollitzer

No. So I think what Aurora mentioned is that our cure rate. So this is simply cures as a percentage of our beginning period NOD balance in Q3, it was 26%. And in Q2, it was 29%. That sort of roughly three-point movement is what we see every year between the second quarter and the third quarter because of seasonal patterns, right?

We've talked for a while about how there's a seasonality to credit development in the portfolio. Borrowers, as you get into the second half of the year, you start to see a little bit of a deterioration in performance and in the first half of the year, you get stronger outcomes because of tax refunds, some bonus income that they receive.

And then as you progress into the second half, Q3 is really defined not by any outflows, but it's the lack of the inflow, the tax refund and bonus income for households. And then you see a little bit of additional movement in Q4 because of spending for the holidays. And so that's sort of the typical cadence. There's nothing really though in the numbers that's unexpected for us.

Bose George

Okay. Great. And then just actually one more on that. The provision for new delinquencies this quarter, was that pretty similar to last quarter?

Aurora Swithenbank

Yes. As for new delinquencies, we put up $13,500 per new NOD, which is broadly consistent with the $14,000 from last quarter.

Bose George

Okay. Great. And then actually one on persistency. Just given the move up in rates and I guess there is a lag in terms of how it shows up, is the persistency going to sort of increase until the next quarter? So I guess some rates are kind of moving back down again. But just leaving that aside, just the move we saw in the last few months?

Aurora Swithenbank

Yes. I think that you've described it correctly, which is persistency is driven by rates. We have had a spike up in rates. And although you don't see that in the persistency number since what we published is a 12-month number. And so that encompasses a much broader period of time than the recent run-up in rates. And so anything going forward would be dictated by the forward trajectory of rates.

Adam Pollitzer

Yes, but I'll give you, for example, October, I'd say we don't know that it's going to go up necessarily and by how much, but we're certainly seeing -- I mean, the decline that we've been seeing has been a modest amount. So it's been sort of a shallow slope of decline, but we've seen that stabilize in the most recent data that we're looking at.

Operator

Your next question comes from Mihir Bhatia with Bank of America.

Mihir Bhatia

I wanted to go back to Rick's question a little bit on just the borrower-initiated cancellation, if you will. A couple of your peers were able to size it for us, or just how much is borrower-initiated versus HPA amortization-driven. I was wondering if you would be willing to do the same?

Adam Pollitzer

Yes, Mihir. We don't share that information and have it readily available to give to you now. We'll do some work on the back end. What I'd say is we have a broad sense, but the servicer is ultimately a function of servicer reporting isn't the cleanest. So we know when obviously a policy is canceled because the note has been repaid. We also know when a policy has been canceled, but the underlying note remains outstanding.

But in terms of a specific split between, in that scenario, policy is canceled, but note remains outstanding, exactly what flows from a borrower initiated cancellation versus a HOPA cancellation, the reporting is a little bit uneven, I'll call it. And so we're not prepared to share that. But we absolutely do see a healthy amount of borrower initiated cancellations and we have.

And one of the ways that we can sanity check that is, while the servicer reporting maybe a little bit uneven, we know the date at which a loan is otherwise going to hit a HOPA cancellation. So we bump up the cancellation that we have and the information we get versus what was otherwise scheduled and we could then see, oh, we've got more cancellations, far more cancellations coming through where the note rate is staying outstanding but that are happening far earlier than the HOPA cancellation date. But we haven't sized it externally.

Mihir Bhatia

Got it. Maybe just turning to pricing for a second. Think of some of your books, the 2022, 2023 vintages, I think that inception-to-date loss ratio is now in around 17%, 18% and defaults are also rising as those policies season. What's the lifetime loss ratio that you price for? Where does that curve end up and peak and kind of start plateauing if you will?

Adam Pollitzer

Yes. So what we price to as a lifetime claim rate, not necessarily a lifetime loss ratio. But what we're seeing on the post-COVID vintages, the '22, '23, '24, as you've noted, they're in and around that 17% loss ratio, those have largely stabilized, right? So what we tend to see is that by the time you are now for, say, the 2022 book, even if we have incremental default experience coming through, the book also is heavily seasoned. And so we're getting a significant amount of that lifetime premium flow through.

I don't want -- a lot of where does it ultimately land is going to be a function of what happens with the existing default population, how does it cure out? And where does the macro and housing market environment go from here? But that sort of high teens is, I'd say, in line to better than our priced expectations when we originated those books.

Mihir Bhatia

All right. And then just one last one for me. The reserve per new delinquent per default, if you will, I think it was like $13,500, similar kind of the last quarter $14,000, but the paid claims rose pretty sharply versus a year ago. I think even like the average claim is up to like $71,000. Just trying to understand, is there anything unusual you're seeing in terms of like when policies go into default is like severity increasing or anything to call out there?

Adam Pollitzer

No. Look, I think what we've been saying for a while, though, is that the profile of loans that are in the default population that traces to the post-COVID period, '22 and more recent, versus those that are from the COVID and pre-COVID period, they're all roughly a similar profile. We use the same rigor and discipline in managing our mix and underwriting those loans and evaluating the credit.

The key difference is simply loans that were originated more recently that are in default status didn't have the same HPA backdrop as some of the loans from the earlier period. And so that's the only reason that we're seeing differences in the number of loans that are progressing to claim payments that go into default and also the severity on those, but nothing is different with, I'd say, has shifted quarter-to-quarter that's different from our expectations.

All of that is embedded in our reserving, right? We are carrying reserves based on our expectation and for the unique attributes of each of those underlying loans in the default population.

Operator

[Operator Instructions] Your next question comes from Maxwell Fritscher with Truist.

Maxwell Fritscher

I'm calling in for Mark Hughes. Adam, you called out the discipline in the PMI industry, which is consistent with what you've said previously. But have you seen any incremental change in behavior from competitors and kind of what that implies in terms of NIW growth for the end of the year and into next?

Adam Pollitzer

No. Look, I mean I think everybody obviously manages their book and wants to do something constructive for their customers, the borrowers they support themselves, their shareholders. And so there's always and everybody -- the views on underlying markets, the macro environment, particular goes, risk cohorts, all of that is always evolving. So the market is never static. But I'd say, broadly speaking, what we observe is we're at a market that's got real balance to it, right?

I think where we are all able to fully and fairly support our customers and their borrowers. But at the same time, use the tools that we've each developed, rates and others, to deliver the outcomes that we also want for our balance sheet and for our shareholders. So nothing that's really moving. I think what will dictate and drive the NIW opportunity going forward is, right, we've got an affordability dial, if you will, and where rates are, is going to impact things.

But then also, there's just really powerful and supportive long-term secular trends that as we look out beyond Q4 and longer, have us really optimistic as to the new business environment.

Maxwell Fritscher

Great. And then switching gears to expenses. Is there anything seasonal we should be paying attention to in 4Q? I believe it's been historically seasonally higher sequentially. Should we expect a similar trajectory going into next quarter?

Aurora Swithenbank

Yes. It's fair to say that the fourth quarter does tend to have higher accruals in years that we have performed well associated with employee comps. And so that is a typical trend that we see in the fourth quarter.

Operator

That does conclude our question-and-answer session and our conference call for today. Thank you for attending today's presentation. You may now disconnect.