Enact (ACT) 2026年第三季財報電話會議：資本回報提高至6億美元

核心亮點

• Enact (ACT) 公布 2026 年第三季調整後營業利益為 1.64 億美元，稀釋後每股盈餘為 1.19 美元，高於去年同期的 1.12 美元，但低於 2026 年第二季的 1.26 美元。調整後營業股東權益報酬率為 12.4%。
• 新增投保金額為 140 億美元，與去年同期持平，因房貸利率上升壓抑轉貸活動而季減 7%。主要生效保額年增約 1% 至 2,750 億美元。
• 保單留存率維持在 81% 的高水準，為連續第 18 個季度高於 80%。約 89% 的受保貸款其房貸利率低於 7%。
• 信用表現依然穩健，儘管總違約件數季增 7% 至 26,100 件。3,700 萬美元的準備金釋出使損失率維持在 20%。
• Enact 將 2026 年預期資本回饋提高至約 6 億美元。該公司在本季透過發放股利與股票回購向股東回饋了 1.51 億美元。
• PMIERs 適足率為 158%，較規定要求高出 18 億美元。Enact 還簽訂了一項遠期成數分保協議，涵蓋約 35% 的 2028 年預期新增投保金額。

關鍵財務數據

業務與營運表現

房貸利率走高持續壓抑房市交易與房貸需求。Enact 表示，房屋供給正趨於正常化，且全美房價持續上漲，支撐了借款人房屋淨值與房貸信用表現。儘管管理階層指出人口結構帶來的需求強勁，但購屋負擔能力仍是一大阻力。

受保投資組合的風險加權平均信用評分為 747 分，風險加權平均貸款成數為 93%。疊加風險仍佔生效風險承擔額的 1.1%。管理階層表示，包括 2022 年與 2023 年在內的近期核保年份放款表現符合或優於定價預期。

新增違約件數從 2026 年第二季的 12,300 件和去年同期的 13,000 件增至 14,400 件。總違約件數升至 26,100 件，違約率季增 20 個基點至 2.8%。違約清償率維持在 50%，且 Enact 對新增違約維持 8% 的理賠率假設。

投資收益受惠於高於 5.5% 的新資金收益率。由於利率上升，未實現投資損失增加 2.48 億美元至 3.45 億美元。管理階層將未實現損失部位定調為實質上屬非經濟性損失，並指出這些投資通常持有至到期。

Enact 持續擴大其信用風險轉移 (CRT) 計畫。截至季末，其第三方 CRT 安排提供了 20 億美元的 PMIERs 資本抵免額。管理階層表示，再保險市場對房貸信用違約風險的需求依然強勁，近期交易定價位於低至中個位數資金成本區間的低端。

管理階層財測指引

Enact 將 2026 年預期資本回饋提高至約 6 億美元，理由是業務表現強勁、對底層投資組合充滿信心，以及在房貸發放市場規模縮小的背景下資本配置機會減少。最終金額與形式仍取決於業務表現、市場狀況及主管機關核准。

管理階層預期，不計重組成本的 2026 全年營業費用將落在先前提供的 2.05 億至 2.1 億美元區間的低端。

在信用表現方面，管理階層表示，正常的季節性因素可能使 2026 年第四季違約率小幅上升。在總體經濟環境穩定的假設下，隨著較新年份的放款組合逐漸成熟，違約件數在 2027 年也可能逐步增加。該公司未提供損失率指引。

風險與觀察重點

• 房貸利率走高對房市活動、房貸需求和轉貸量帶來壓力。
• 購屋負擔能力仍是潛在買家的限制因素。
• 隨著近期放款組合逐漸成熟且季節性趨勢依然不利，新增違約件數與違約率可能會上升。
• 未來準備金釋出規模可能會縮小，因為較低的滾動率假設同時減少了新增違約認列的損失以及後續準備金釋出的空間。
• 利率變動仍是其他綜合損益累計額 (AOCI) 和投資組合按市值計價波動的主要驅動因素。
• 由 FHFA 主導的房貸保險取消試點計畫細節仍不確定，包括估值方法、成熟期要求、治理機制以及誰將承擔鑑價或自動估值模型費用。

分析師問答重點

管理階層表示，政府關於住房成本的行政命令似乎著重於 FHFA 試點計畫建立後發放的貸款。Enact 尚未收到來自 FHFA 或政府贊助企業的詳細指引。該公司預期，按市值計價的貸款成數、成熟期、估值緩衝空間和納稅人風險保護將是重要的考量因素。

關於保單留存率，Enact 指出，在 2020 年與 2021 年合計約 2,000 億美元的原始新增投保金額中，目前僅剩約 30%。儘管這些龐大年份的資產逐漸流失，保單留存率仍連續 18 個季度保持在 80% 以上。目前借款人主動取消的生效保額每季約為 15 億美元，相當於約 5,000 張保單。

管理階層仍認為在當前估值水準下，股票回購是極具吸引力的資本運用方式。回購決策考量了股價相對於每股帳面價值、不含 AOCI 的每股帳面價值，以及管理階層對內在價值的評估。特別股利仍是另一種潛在的資本回饋機制。

關於信用正常化，管理階層表示，近期以購屋為主的放款組合具有較高的貸款成數和債務收入比特徵，且房價漲幅較為溫和。然而，其表現仍與定價時採納的假設一致。

法說會完整逐字稿

---

完整財報電話會議逐字稿

管理層陳述

Operator

Hello, and welcome to Enact's Third Quarter Earnings Call. Please be advised that today's conference is being recorded. [Operator Instructions]

I would now like to hand the conference over to your first speaker, Daniel Kohl, Vice President of Finance. You may begin.

Daniel Kohl

Thank you, and good morning. Welcome to our third quarter earnings call. Joining me today are Rohit Gupta, President and Chief Executive Officer; and Dean Mitchell, Chief Financial Officer and Treasurer. Rohit will provide an overview of our business performance and progress against our strategy. Dean will then discuss the details of our quarterly results before turning the call back to Rohit for closing remarks. We will then take your questions.

The earnings materials we issued after market closed yesterday contain our financial results for the quarter, along with a comprehensive set of financial and operational metrics. These are available on the Investor Relations section of our website. Today's call is being recorded and will include the use of forward-looking statements. These statements are based on current assumptions, estimates, expectations and projections as of today's date.

Additionally, they are subject to risks and uncertainties, which may cause actual results to be materially different, and we undertake no obligation to update or revise such statements as a result of new information. For a discussion of these risks and uncertainties, please review the cautionary language regarding forward-looking statements in today's press release as well as in our filings with the SEC, which will be available on our website.

Please keep in mind that earnings materials and management's prepared remarks today include certain non-GAAP measures. Reconciliations of these measures to the most relevant GAAP metrics can be found in the press release, our earnings presentation and our upcoming SEC filing on our website.

With that, I'll turn the call over to Rohit.

Rohit Gupta

Thank you, Daniel. Good morning, everyone. Before I cover Enact's results, I would like to take a moment to recognize Tom McInerney, who recently announced his intention to retire as President and CEO of Genworth and step down from the Enact Board at the end of the year. Tom has played an important role in Enact's journey from his support through our IPO in 2021 to his valued partnership as a member of our Board.

On a personal level, Tom has been a trusted mentor to me for many years, and I have benefited enormously from his counsel. I know I speak for our Board and our entire team in thanking him for his commitment to Enact and wishing him all the best in his retirement. We look forward to continuing our strong relationship with Genworth and working with Jerome Upton as he steps into his new role.

Turning to the business. I'm pleased to be with you today to discuss our performance. Our third quarter results build on a solid first half and demonstrate the resilience of Enact's business model as we approach year-end. Our strong results, favorable credit performance and strong balance sheet support an increase in our capital return guidance. We now expect 2026 capital returns to be approximately $600 million. For the third quarter, we reported adjusted operating income of $164 million or $1.19 per diluted share. Adjusted return on equity was 12%, and we generated new insurance written of $14 billion, resulting in total insurance in force of $275 billion, up 1% year-over-year.

Turning to the operating environment. Mortgage rates moved higher during the quarter, putting additional pressure on housing activity and mortgage demand. At the same time, housing supply continue to normalize and home prices continue to appreciate nationwide supporting borrower equity and mortgage credit performance. Labor market conditions remain generally healthy, although affordability continues to be a headwind for prospective homebuyers despite strong demographic demand.

Looking ahead, we believe the long-term drivers of housing demand remain intact and continue to reinforce the important role private mortgage insurance plays in responsibly expanding access to sustainable homeownership. Against this backdrop, our insurance in force portfolio continues to demonstrate resilience, supported by prudent underwriting standards, strong borrower credit characteristics and meaningful embedded home equity.

Persistency remained elevated in the quarter at 81% and continues to be supported by the rate environment with approximately 89% of loans in our portfolio, carrying mortgage rates below 7%. At quarter end, the risk-weighted average credit score of our portfolio was 747, while the risk weighted average loan-to-value ratio was 93%. Layered risk remained 1.1% of risk-in-force.

As in the past, pricing in the mortgage insurance market remains competitive, and we remain focused on maintaining prudent underwriting standards and generating attractive risk-adjusted returns for our shareholders. We continue to leverage our dynamic pricing engine, Rate360 to assess risk at a granular level, incorporating borrower characteristics loan attributes and our prospective view of housing market conditions. This capability is particularly important in an environment where economic and housing trends vary meaningfully by geography. Our focus remains selecting the right risk at the right price as market conditions evolve.

We continue to participate in the rollout of VantageScore 4.0. During the quarter, VantageScore volume remained concentrated among a few originators with broader implementation still in the early stages. We will continue to fully support our customers as adoption progresses. From a credit perspective, our portfolio is performing well with recent books performing in line with our expectations. Consistent with seasonal trends, new delinquencies were up 17% and cures were down 1% sequentially, and total delinquencies increased 7%.

Our strong cure performance was driven by favorable credit trends and effective loss mitigation efforts. This drove a reserve release of $37 million in the quarter, resulting in a loss ratio of 20%. Overall, credit performance remains strong, and we believe we are well reserved across a range of scenarios. We delivered another quarter of prudent expense management with operating expenses down year-over-year despite the inflationary environment. As a result, we now expect to end the year at the low end of our expense expectations. Dean will discuss the key drivers of our expense performance and provide more detail shortly.

Turning to capital. We continue to execute against our priorities, maintaining a strong and resilient balance sheet to support existing policyholders, investing to drive organic growth and operating efficiencies funding attractive new business opportunities and returning excess capital to shareholders. At the end of the quarter, our PMIERs sufficiency ratio was 158%, and our credit and investment portfolios were in excellent shape. Our strong capital position is further reinforced by our CRT program and the backing of our undrawn credit facility.

During the quarter, we entered into a new quota share agreement that will cover approximately 35% of expected 2028 new insurance written, further advancing our programmatic approach to CRT and our ability to secure cost-efficient capital and balance sheet protection. During the quarter, A.M. Best upgraded EMIC and EMIC NC's financial strength rating to A from A-, and Enact Holding, Inc.'s issuer credit rating to BBB from BBB-, all with stable outlooks.

We are also making progress against our strategic growth and diversification priorities. Enact Re delivered another quarter of strong performance, generating attractive risk-adjusted returns while remaining both capital and expense efficient. We are pleased with the progress of this business and continue to view it as a long-term growth and diversification opportunity. Finally, our strong performance supports continued returns to shareholders. During the quarter, we returned $151 million through share repurchases and dividends. As I mentioned earlier, we have increased our 2026 capital return expectations to approximately $600 million.

Before wrapping up, I'd like to comment on the recent presidential executive order on decreasing housing costs for the American people. We support the administration's focus on lowering the cost of homeownership and look forward to helping advance that goal. We are committed to expanding access to affordable homeownership while protecting taxpayers and supporting the safety and soundness of the housing finance system. We look forward to working constructively with the GSEs, the FHFA and our lending partners.

Overall, we are pleased with our performance this quarter and year-to-date. The strength of our portfolio disciplined capital management and continued focus on execution leave us well positioned to navigate the current environment, support sustainable growth and deliver long-term shareholder value.

With that, I will now hand the call over to Dean.

Hardin Mitchell

Thanks, Rohit, and good morning, everyone. We delivered another strong quarter of performance. Adjusted operating income was $164 million or $1.19 per diluted share compared to $1.12 per diluted share in the same period last year and $1.26 per diluted share in the second quarter of 2026. Adjusted operating return on equity was 12.4%. A detailed reconciliation of GAAP net income to adjusted operating income can be found in our earnings release.

Turning to revenue drivers. New insurance written was $14 billion in the quarter, down 7% sequentially as higher mortgage rates reduced refinance activity and flat year-over-year. Persistency was 81% in the quarter, up 1 point sequentially on higher prevailing mortgage rates that exited the quarter above 7% and down 2 points year-over-year. Persistency remains elevated relative to our long-term historical average of 78%, supported in part by 89% of our loans in our insured portfolio having rates below 7%.

Primary insurance in force was $275 billion in the quarter, up $1 billion from the second quarter of 2026 and up $3 billion or approximately 1% year-over-year. Total net premiums earned were $242 million, down $3 million sequentially and year-over-year. The sequential decrease is primarily driven by higher ceded premiums from reduced profit commissions on our quota share reinsurance transactions. This reduction was partially offset by premium growth from attractive adjacencies and growth in primary insurance in force.

Our base premium rate of 38.9 basis points was down 0.2 basis points sequentially. As a reminder, our base premium rate is impacted by several factors, including new insurance written, persistency, credit mix and premium refund estimates that fluctuate quarter-to-quarter. Our net earned premium rate was 33.4 basis points, down 0.7 basis points sequentially and primarily driven by the impact of higher ceded premiums from reduced profit commissions on our quota share reinsurance transactions. Investment income in the quarter was $74 million, up $1 million or 2% sequentially and up $5 million or 9% year-over-year.

Our new money investment yield was over 5.5% and contributed to a 10 basis point increase in the average portfolio book yield to 4.7% for the quarter. As of quarter end, unrealized losses in our investment portfolio increased by $248 million to $345 million, driven by interest rate increases. Although we generally hold these investments to maturity, we may selectively realize losses to pursue income enhancement opportunities. During the quarter, we sold certain assets that will allow us to recoup realized losses through future higher net investment income. We'll continue to evaluate similar opportunities to maximize the value of our portfolio but this does not change our view that our investment portfolio's unrealized loss position is materially noneconomic.

Turning to credit. We continue to see strong loss performance across our portfolio. New delinquencies increased sequentially and year-over-year to 14,400 in the quarter from 12,300 in the second quarter of 2026 and from 13,000 in the prior year quarter. The sequential increase was aligned with seasonal trends and the year-over-year increase was primarily driven by newer books progressing along their normal loss development curve. Our cure rate was flat sequentially at 50% and in line with seasonal trends and remains elevated. We maintained our claim rate on new delinquencies at 8%. Total delinquencies in the third quarter increased sequentially to 26,100 from 24,300 and the delinquency rate increased 20 basis points sequentially to 2.8%.

Losses in the third quarter of 2026 were $47 million and the loss ratio was 20% compared to $33 million and 14% in the second quarter of 2026, and driven primarily by higher new delinquencies and $36 million and 15% in the third quarter of 2025, driven by differences in new delinquencies and reserve releases offset partially by a lower claim rate in the current quarter. The current quarter reserve release of $37 million from favorable cure performance and loss mitigation activities compares to a reserve release of $37 million in the second quarter of 2026 and $45 million in the third quarter of 2025. Overall, we remain pleased with the credit performance of the portfolio and continue to believe our reserves appropriately reflect both current credit performance and the uncertainty in the broader macroeconomic environment.

Operating expenses in the third quarter of 2026 were $50 million, and the expense ratio was 21% compared to $52 million and 21% in the second quarter of 2026, and $53 million and 22% in the third quarter of 2025 as we continue to manage expenses prudently in the current originations market. To that end, we currently operate at our lowest head count in over a decade. As part of the second quarter 2026 reorganization, we incurred a $1 million charge that is excluded from adjusted operating income. Based on our year-to-date performance, and full year 2026 outlook, we now forecast 2026 expenses, excluding reorganization costs to be at the lower end of the $205 million to $210 million range we provided last quarter.

We continue to operate from a strong capital and liquidity position, underpinned by our robust PMIERs sufficiency and the successful execution of our diversified CRT program. Our PMIERs sufficiency was 158% or $1.8 billion above PMIERs requirements at the end of the third quarter. During the quarter, we entered into a new forward quota share reinsurance agreement which will cede approximately 35% of our 2028 new insurance written to a broad panel of highly rated reinsurers. This transaction demonstrates our commitment to disciplined risk management while providing certainty of coverage at favorable market terms. Our third-party CRT program provides $2 billion of PMIERs capital credit at the end of the quarter.

Turning now to capital allocation. During the quarter, we paid out approximately $33 million or $0.24 per share through our quarterly dividend and bought back 2.4 million shares at an average share price of $48.19, or $118 million. As Rohit mentioned earlier, we are increasing our 2026 total capital return guidance to approximately $600 million reflecting our continued strong business results and confidence in our business. As in the past, the final amount and form of capital return to shareholders will ultimately depend on business performance, market conditions and regulatory approvals.

Overall, we are pleased with our performance through the first 9 months of 2026. Our improved expense outlook and increased capital return expectations reflect disciplined execution, continued financial strength and confidence in the underlying performance of our business. As we look ahead, we remain focused on prudently managing risk, maintaining a strong balance sheet and delivering attractive returns for our shareholders.

With that, let me turn the call back to Rohit.

Rohit Gupta

Thanks, Dean. Looking ahead to the remainder of 2026 and beyond, our priorities are unchanged: support our customers, maintain a strong balance sheet and execute with discipline. The consistency of that approach, combined with the strength of our portfolio gives us confidence in our ability to continue to deliver attractive returns for shareholders over time.

Operator, we are now ready for Q&A.

Operator

[Operator Instructions] Your first question comes to the line of Doug Harter with BTIG.

分析師問答

Douglas Harter

Rohit, I know you spoke a little bit about kind of the persistency in your prepared remarks. But hoping we could just talk more -- and whether you see -- kind of given the increased focus on this by the FHFA, maybe some more media attention to it. What -- do you think you will see any change in cancellation activity due to home price appreciation?

Rohit Gupta

Doug, thank you for the question. So as I said in my prepared remarks, we support the administration's focus on lowering the cost of homeownership and looking forward to helping advance that goal. We are also committed to expanding access to affordable homeownership while protecting the taxpayers and supporting safety and soundness.

That being said, I would say, when you think about the executive order, it's essentially direct FHFA to set up a pilot focused on mortgage originations going forward after the establishment of the pilot. So as of right now, that's what we are focused on from a consumer behavior perspective, that is tough to predict. From a social media activity perspective or policy discussion perspective, those topics typically do not get reflected in consumer behavior, but too early to tell.

Douglas Harter

Great. And then just sticking on persistency, if you could just remind us kind of where we are in the 2020, 2021 vintages as far as kind of hitting the contractual termination points.

Rohit Gupta

Sure, Doug. I'm going to pass that question to Dean or Daniel.

Daniel Kohl

Doug, this is Daniel. Happy to take it. Thanks for the question. Yes. In terms of the 2020 and 2021 book years, as you know, very large books when they were written collectively about $200 billion in NIW. And actually, there's only about -- you can see this in the QFS, there's only about 30% of that remaining of those 2 books. And so we've actually seen those start rolling off our portfolio for some time now. And I think you need to take that into consideration when you think about where our persistency has been.

And as Dean mentioned, we consider anything over 78% to be what we call elevated. And this actually marks the 18th straight quarter of persistency actually being over 80%. And so despite the fact those books being very large, have high refi, thus a little bit lower running off through our portfolio and through our financials, you're still seeing elevated persistency throughout this time period.

Operator

Your next question comes from the line of Bose George with KBW.

Bose George

I just wanted to ask on buybacks. The -- I mean you guys have obviously done well in terms of performance or the stock has done well, and it's trading, looks like, $1.26 of book value. does that play into how you view capital return just for the breakout between dividends and buybacks? And also when you look at book value on buybacks, you look at that number ex AOCI or how do you kind of treat that piece?

Hardin Mitchell

Yes, Bose, it's Dean. Thanks for the question. As you made reference to, we do have several ways to return capital to shareholders, and we think about them kind of in this waterfall type mindset start with quarterly dividends. Those are obviously the most certain. When we set our quarterly dividend, we focus on making sure that those are durable across various scenarios, including meaningful stress scenarios.

I think getting to -- more to your point around share repurchases, they're more opportunistic based on market conditions. It's an ongoing assessment looking at current share price vis-a-vis book value, we look at it both on a book value with an ex AOCI. We obviously also compare that to intrinsic value, what we think the company's long-term potential value is. And just to complete the story in terms of ways to return capital, we also can return capital via special dividends. That's kind of the plug. It's the means to return capital when those other options aren't available.

Kind of getting a little bit to your point on share buybacks in light of current valuations. We still believe share repurchases are a compelling use of capital at current valuations. It's a dynamic assessment based on our day-to-day assessment and share buyback activity varies based on price fluctuations. But just speaking at a high level, we still believe that repurchasing shares at current valuation creates long-term shareholder value. And so that's been our assessment and you've seen that in our practices, including the share repurchases through October and the beginning of November.

Bose George

Okay. Great. That's helpful. And then in terms of the AOCI in the portfolio now, is there a way -- like what's the average sort of duration or the time line where that AOCI kind of rolls back into earnings, all else equal?

Hardin Mitchell

Yes. So AOCI really fluctuated by about -- well, $250 million on a pretax basis, $200 million on an after-tax basis, almost all of that is driven by change in interest rates and the mark-to-market change in our investment portfolio. So I think that's more rate driven than anything else. Obviously, our effective duration on the portfolio is 5 years. So those -- the weighted average life is probably a little bit longer than that effective duration. So it might have a 5- to 6-year life. But I think the change in interest rates between now and the next reporting period will be a major driver, probably more so than our holding those investments to maturity.

Operator

Your next question comes from the line of Rowland Mayor with RBC Capital Markets.

Rowland Mayor

I wanted to kind of go with the regulatory item from a different angle. Could you help me understand how much capital you have to hold against kind of high HPA loan books from 2022 or prior. I assume if there was some mass unlocked event, there'd be a significant amount of capital you guys are able to release?

Rohit Gupta

Rowland, thank you for the question. So as I said in my prepared remarks, we are very supportive of the executive order, which specifically focuses on a prospective implementation for loans originated after the pilot is established. So from a disclosure perspective, I do not believe that we disclose our capital allocated by book year. But if you go into our financial supplement, you can find disclosures on our risk in force, insurance in-force policies by book year.

And then in aggregate, we also disclosed home price appreciation by book year, so embedded equity kind of concepts. So I would say there are disclosures out there. But given the fact that the executive order is focused primarily on prospective loans, that's what we are focused on right now.

Rowland Mayor

And then I wanted some help understanding the pace of capital return you moved the guidance up through the year and the pace accelerated. It was kind of similar last year. What drives that? Is it credit dividend approvals or something else?

Hardin Mitchell

Yes, Rowland, thanks for the question. This is Dean. So what we look at to evaluate return of capital throughout the year is really how the business is performing. And as you've seen, the business continues to perform very, very well. But embedded in that business performance is also what are the market opportunities to deploy capital for new insurance written.

We're obviously operating in a smaller originations market. That has the effect of reallocating capital that we might otherwise deploy to production to other means, including potentially returning capital to shareholders. Beyond that, we look at the macroeconomic environment, the prevailing and prospective, how do we think that's going to impact both business performance today and in the future.

And then lastly, we think about the regulatory environment that's been pretty stable from a PMIERs perspective, from a state regulatory perspective. So really, it's been business performance that has driven the increased confidence to increase our return of capital, both last quarter to the range and now this quarter to the top of that range to $600 million.

Rowland Mayor

And then if I could just ask one more. How long is the typical cycle time from a new default to a cure? And is that cycle time changing?

Hardin Mitchell

Yes. That's a good question, Rowland. I don't have a particularly great answer. We'll go back and take a look at that offline and see if we can give you some insights into how quickly delinquencies are curing and whether that's changing? I think from -- yes, go ahead.

Rohit Gupta

I think Rowland, one way to think about it, and I don't think we have a number handy to give you in aggregate. And even if we had, I would say, it would be constructed off very different dynamics. So if you just think about first-time news, first-time news cure at a different pace than repeat delinquencies. And Dean has talked about this in the past that repeat delinquencies, which sometimes we refer to in the business as chronic delinquencies, are cash managers who go in and out of delinquencies and have higher cure rates.

And then on top of it, given the difference in embedded equity by vintage, that could heavily impact cure rates because consumers have different options in terms of getting out of a delinquency. So I think there are a lot of factors that probably create a range around it. But let us come back to you offline to maybe just give you the factors that impacted.

Hardin Mitchell

Yes. And Rowland, maybe just to get you started, if you look at our earnings presentation on Page 11, you can see intra-period cures, intra-quarter cures, those typically range between 25% and 30-ish percent you see this quarter in that range as well. And then as you progress through time, you can see those cumulative cures continue to increase. Typically, we've been seeing around at the fifth quarter mark, cumulative cures in the high 80s, maybe below 90%, 90%. You're continuing to see cure patterns continue at that similar pace.

So the more I think about your question, I think that disclosure and piecing that disclosure through time can give you a sense of the velocity of cures and whether that's changing? And I guess, as I think about that, it's -- there might be some modest moderation but staying -- like we've talked about, cure performance remains elevated to our expectations, which is driving the reserve releases that we've had through time.

Operator

Your next question comes from the line of Rick Shane with JPMorgan.

Richard Shane

Interesting to see the quota share agreement for 2028, and I'm sitting here kind of trying to wrap my mind around 2028 and that far out, but I appreciate you guys are working that far ahead. Look, we saw during the third quarter, companies that make and sell loans across different verticals experienced some disruption in pricing dislocations. Obviously, you guys aren't working that far ahead. and take an adverse market. But I am curious what quota share partners are looking at right now. And if there was any discussion about sort of the uncertainty that we experienced in the third quarter impacting terms of those agreements or anything that we should be thinking about?

Hardin Mitchell

Yes, Rick, thanks. It's Dean. It's a good question. I would describe the demand for mortgage credit default risk across the reinsurance market as very strong. It's one of the reasons why we continue to pursue 2-year forward transactions on what we believe to be attractive terms. I think you go back several years, maybe 3, 4 years ago, that type of 2-year forward demand didn't exist in the market or if it existed, it existed at prices that were less appealing.

So we continue to be able to source all of our reinsurance needs through the traditional reinsurance market, and we're doing so at attractive terms. We've talked about those terms through time being in the low to mid-single digits cost of capital. If you look at today's transactions just in general over the last 12 to 24 months, it's on the low end of that range. And I think that's indicative of ongoing strong demand in the market.

Rohit Gupta

Yes, Rick, Dean covered it very nicely. Just to add to Dean's point, from reinsurance markets and our credibility and relationship with reinsurance players. We have been a programmatic player in this market consistently since 2015, and we have proven with our risk selection pricing to our reinsurance partners that as different markets go through different conditions, we adjust our pricing and risk appetite through that.

So when you think about credit risk, different geographies going through different unemployment or economic events or home price paths, which are different from each other, we adjust our pricing pretty regularly for that. And when we back test it, we are able to show our reinsurance partners that we have an ability to navigate those markets differently based on market conditions. So we have a strong track record, and I think that gives us good credibility with our partners for them to rely on our underwriting and pricing discipline.

Hardin Mitchell

The last point I'll make on this, Rick, is we do have steps in our reinsurance contracts that kind of provide guardrail guardrails for the reinsurers, minimum FICOs, maximum LTVs and concentrations in those things, DTI, MAXs as well. So there is some boundary conditions that we collaborate on and give the reinsurers' confidence in providing immediate coverage or coverage on a 2-year forward basis.

Operator

Your last question comes to the line of Mihir Bhatia with Bank of America.

Mihir Bhatia

I wanted to start with actually credit performance. I think you've talked about -- everyone has seen credit performance has been quite exceptional for a few years. But I was wondering, 2022, 2023, some of those vintages are moving up to like 3.4%, 3.5% delinquency rates. Are those the levels that you typically would expect at this point in their life cycle, how are they performing versus -- like how is vintage performance versus what you had planned for at pricing, let's say?

And then just related to that, I'm thinking about -- starting to think about next year and as the loan season, reserve releases have started to fade a little bit. I think you're down to $37 million versus like maybe $40 million mid $40s million last year? And I understand loss ratio is an outcome, but as you're thinking about 2027 and planning for 2027, what's the right underwriting ratio we should be thinking about? Does it start drifting higher towards the mid-20s? Or is 20 percentage is the right way to think about it?

Hardin Mitchell

Yes, Mihir. Thanks for the question. This is Dean. So as it relates to book year performance, we're not really seeing anything different than what we talked about in the past. And I would -- just starting off at the highest level, I would agree with your characterization of credit. Overall credit performance remains really strong across the portfolio through Q3.

If we disaggregate and get down into the vintage level, and this really applies quite honestly, any type of borrower attribute, loan attribute and even geography. We're really not seeing anything different than what we expected. These newer books are purchase-heavy books, so they have modestly higher risk attributes. So think higher LTVs, higher DTIs, they're also aging through a much more moderate home price appreciation path. And when we think about the performance, those are things that we price for. And when we evaluate how those book years are doing relative to how we price. We're seeing performance very much in line with our pricing expectations, I think that's what we talked about in the past. That's continuing. That type of performance is continuing through third quarter.

If we think about what that means for credit performance for delinquencies, for delinquency rates on a kind of prospective basis I think as we head into the fourth quarter, much like we had in the third quarter, we have some unfavorable seasonality. So as I'm thinking about kind of delq rate for Q4, I think you could see that just simply tick up based on normal seasonality. I think beyond that, I think it's certainly possible that we continue to see an increase, let's just talk in delq rate turns for a second and we continue to see those gradually increase. into 2027 as those recent book years continue to season, up their normal loss development curve.

And again, anything I say as it relates to forward loss development is going to be predicated on macroeconomic environment. I'm giving you kind of the lens into more of a stable macroeconomic environment, if that trajectory were to change, obviously, the trajectory of delinquencies delq rate loss ratio changes as well.

Rohit Gupta

Yes, Mihir, a few things to add to Dean's point. So I completely agree that the kind of summary of that is from a book year perspective, performance of recent vintages, including '22, '23 that had higher purchase originations, so higher risk attributes from loan-to-value, debt-to-income perspective, they're performing in line with or better than pricing expectations and that also is driven by strong cure rates that Dean referred to.

I think one thing to just also think about this environment, given the persistency in some of the books, loans are staying around longer in aggregate. The books are staying longer than original expectations. And when that happens, those repeat delinquencies that go in and out of delinquency status do impact cure rate calculation. So we are very mindful of that, that when you think about delinquency rate, it might look like delinquency rate is trending up. But then we do look at cure ratios, which, to Dean's previous point, continue to hold up nicely.

And I think you asked a question about loss ratio and how to think about it. We don't provide guidance on loss ratio from a pricing perspective. What I would say, your question on reserve releases versus prior year and future years. I think part of it is when we have made a change in our assumptions on roll rate, that does reduce the opportunity set. So what that means is our losses on news will be lower because we have reduced the rate like last year by a point, and that means our losses on news on a dollar basis will be lower compared to last year if accounts were same. And on the flip side, your reserve release will be lower because there's less opportunity, everything else being held constant. So some of that is more geography that as we see more evidence of performance we take those things into consideration as we think about what our roll rate assumptions are.

Hardin Mitchell

Yes. And Mihir, I thought I heard you say $27 million of reserve release. It was actually $37 million of reserve releases.

Mihir Bhatia

$37 million, yes.

Hardin Mitchell

Perfect. And I totally agree with what Rohit said about the opportunity for future reserve releases being tied to the roll rate assumption.

Mihir Bhatia

Got it. And then just going to the housing yield, and I appreciate you don't want -- like it's forward looking on new policies, maybe just talk about some of the guardrails that you would be encouraging the FHA to consider putting in even as it pertains to newer policies and how to go about doing it and your response or your -- like, I mean I assume you're going to still want to earn the return -- the mid-teens return that you have been pricing for. So to the extent that this gets automated or they're using AVM and automatic cancellation, does that then get reflected in just price upfront? I don't know if there's anything more you can say on it, but maybe just talk -- just prospectively about what -- what is Enact's position and what are you advocating for? Or what is the industry going to the FHFA and explaining to them.

Rohit Gupta

Yes. So I would say, Mihir, just from a process and timing perspective, it's too early to tell. We have not received guidance from FHFA or GSEs at this point of time on the parameters. But I would say we've been pretty explicit on in the past that key factors as you think about prospective cancellations or new loans include mark-to-market LTV, it includes seasoning requirements many of these factors are already factored into the existing criteria.

When you think about GSE guidelines, there are requirements on how a consumer can actually cancel loans based on appraisals and then obviously, valuation methods buffers, how much risk is being transferred to taxpayers versus us putting that private capital in front of tax taxpayer risk. So those are typical considerations but we look forward to engaging with the FHFA and GSEs as and when those conversations start.

Operator

You have a question from Geoffrey Dunn with Dowling & Partners.

Geoffrey Dunn

I'm not sure these details are available yet, but I had a couple of questions on this pilot because obviously, the ability to cancel has been there for decades. And the borrower could always go out and pay for an appraisal. Who's bearing the cost of the AVM under this pilot? Are the GSEs just assuming that cost? Or are they going to put it onto the borrower? And secondly, what is the incentive of servicer to reach out to a borrower to try to cancel it.

Rohit Gupta

Yes, Geoff. Thank you for the question. So very good questions. Unfortunately, at this point of time, we don't have a lot of information beyond what you have seen in the executive order. We have seen the same thing. We have not received guidance from GSEs or FHFA on what the scope of the pilot would be, what the parameters would be and whether it's an automated valuation model or it's an FHFA price index. So governance, transparency of whatever the methodology would be is very important in setting up the criteria and its resilience through cycles would be definitely part of the discussion. So all those questions are still open, including who bears the cost of pulling these appraisals or pulling these AVMs. So that's answer to your first question.

Second question, I think servicers are very much aligned with borrower motivation. So if a consumer is looking for a cancellation, the consumer reaches out and applies for a cancellation based on an appraisal or whatever GSE conditions there are. So at this point of time, servicers will comply with the scope of the pilot and then operationalize it using their own SOPs.

Geoffrey Dunn

Do you have any data on historically the percent of borrowers that are smart enough and educated enough to realize they could cancel and reach out to get an appraisal and cancel?

Rohit Gupta

Yes. So Geoff, we can think about offline on what that data might look like, but I would just point to our appraisal-based cancellation data, which is embedded in our lapse already. So when you think about our lapse performance through different interest rate cycles, different home price appreciation cycles, that factor is already included in it. And I would say it's very parallel to what happens in refinance market. that when rates drop, even if there's a portion of consumer population that is in the money from a refinanceability perspective, not every single consumer goes and refinances.

I don't know if we have a breakdown of our lapse in our disclosures, but I'll ask Daniel.

Hardin Mitchell

We don't have a disclosure, but we do have -- what we've talked about, Geoff, in the past, having about $1.5 billion of borrower initiated cancellations per quarter. That's about 5,000 policies, if you think that our average loan size across the portfolio is about $300,000 are canceled every quarter. So that represents about 2% of our annualized lapse on the portfolio on an annualized basis. So from our perspective, there is -- these borrower-initiated cancellations are working for many borrowers every quarter under the current design.

Operator

I'm showing no further questions at this time. I would now like to turn it back to Rohit Gupta for closing remarks.

Rohit Gupta

Thank you, Kathy, and thank you, everyone. We appreciate your interest in Enact, and I would like to wrap up the call by thanking our -- all of our employees for their hard work and dedication in fulfilling our mission to help people buy and keep their homes. Thank you.

Operator

Thank you for your participation in today's conference. This does conclude the program, and you may now disconnect.