安納利資本管理 (NLY) 2026年第三季法說會：EAD 涵蓋股利但每股淨值下滑

重點摘要

• Annaly Capital Management 公布 2026 年第三季每股可分配收益 (EAD) 為 0.80 美元，季增 0.01 美元，高於其 0.75 美元的每季股利。
• 每股淨值季減 11.4% 至 17.86 美元。包含股利在內，由於利率上升、機構抵押擔保證券 (Agency MBS) 利差擴大及波動度升溫對估值造成壓力，該公司產生的經濟報酬率為 -7.6%。
• 經濟槓桿比率從 5.6 倍上升至 5.9 倍，主因股東權益減少，而避險比率則從 97% 上升至 100%。Annaly 在本季結束時擁有 85 億美元可用於融資的資產。
• 機構投資組合規模為 920 億美元，住宅信貸為 104 億美元，房貸服務權 (MSR) 為 43 億美元，分別占公司資本的 55%、23% 與 22%。
• 管理層預期 2026 年第四季 EAD 將再次超過股利，並表示目前的投資組合足以支持發放股利。其新增資本配置偏好仍依序為住宅信貸第一、機構 MBS 第二、MSR 第三。
• 截至電話會議前的星期五，管理層表示不含應計股利在內的每股淨值成長約 2%，若包含股利則接近 3%。

關鍵財務數據

較高的票息收入、較大的平均投資證券與住宅抵押貸款餘額、證券化活動以及較低的營運費用支撐了 EAD。這些效益抵銷了較高的證券化債務與附買回餘額，以及微幅上升的融資利率。

業務與營運表現

機構 MBS

第三季結束時，機構投資組合市值為 920 億美元。Annaly 名目上增加了約 20 億美元的 MBS，重點放在較高票息的特定投資組合（specified pools），主要是 5.5% 與 6% 的票息。

隨著 10 年期美債殖利率在本季上升 82 個基點，機構 MBS 利差擴大。該公司主要透過國債期貨增加避險，以限制利率風險暴露。在第四季初期，該公司在將特定投資組合的優異表現變現後，將超過 20 億美元主要為高齡貸款組合（seasoned pools）輪轉至較高票息的待交割證券 (TBA)。

管理層表示，目前的機構利差能提供 15% 至 19% 左右的新資金預期報酬率。管理層亦預期，在目前的利率水準下，較低的負凸性（negative convexity）將降低投資組合的存續期間敏感度以及避險再平衡的需求。

住宅信貸

住宅信貸投資組合大致維持在 104 億美元不變。Onslow Bay 產生了 70 億美元的鎖定量，以及創紀錄的 58 億美元單季撥款額。目前鎖定管道中的平均 FICO 信用分數為 766 分，平均貸放比率 (LTV) 為 67%。

Annaly 在第三季完成了 74 億美元的證券化交割，並於 10 月完成另外 22 億美元的定價。截至電話會議舉行時，該公司在 2026 年已跨 37 筆證券化交易完成超過 210 億美元的定價。

管理層指出住宅信貸的預期報酬率為 13% 至 16%，受到留存的證券化權益、透過非盯市（non-mark-to-market）管道融資的全額貸款以及精選的 BB 級非合格抵押貸款 (non-QM) 債券所支撐。

房貸服務權 (MSR)

MSR 投資組合增加至 43 億美元。Annaly 交割了一筆市值約 1.4 億美元的大宗資產包，而流量管道（flow-channel）規模超過 5000 萬美元，創下單季新高。

投資組合估值倍數微幅上升至 6.15 倍。提前還款速度降至 4.8 CPR，而嚴重違約率維持在 60 個基點以下。管理層表示，較低的票面利率、有限的提前還款以及託管帳戶餘額收入繼續支撐可預測的現金流。

Annaly 將 MSR 視為獨立的報酬機會，同時也是投資組合多元化的工具。較高的利率能支撐託管浮存金收益，而當利率上升時，MSR 估值可能有助於抵銷機構 MBS 的弱勢。

管理層展望

管理層預期 2026 年第四季 EAD 將再次超過 0.75 美元的股利，並表示現有投資組合與當前的新資金報酬率均足以支持該股利發放。

該公司在進入第四季時擁有充裕的流動性，並具備戰術性提高槓桿的空間，不過管理層表示目前的槓桿已能產生所需的報酬率。資本配置優先順序仍為住宅信貸，其次是機構 MBS，MSR 則緊隨其後。

管理層預期若波動度持續下降，機構 MBS 的市場環境將會受益。管理層亦引用了債券基金資金流入、政府贊助企業 (GSE)、海外投資者與銀行的支持，以及有限的房貸供給。

在非合格抵押貸款 (non-QM) 業務方面，管理層估計 2026 年發放額為 1,500 億至 1,750 億美元，並偏向區間高點，且預期該領域將保持在總房貸發放額的約 8% 至 9%。這些估計值被明確呈現為管理層的預期。

風險與關注事項

• 10 年期美債殖利率在第三季上升 82 個基點，而房貸利率達到 7.5%，推高了利率與利差的波動度。
• 通膨率仍高於美國聯準會的目標，且市場預估到 2027 年底前將再升息近三次（每次 25 個基點）。
• 較高的利率與負擔能力限制可能會使房屋成交量維持低迷，並可能對房價施加壓力。管理層表示，房貸利率持續維持在 7.375% 至 7.5% 左右，可能會導致全國房價下跌數個百分點。
• 房市狀況因地區而異。管理層特別指出包括德州與佛羅里達州在內的東南部地區表現疲軟，並指出供需月數（months of supply）已達到 10 年高點。
• 即使避險比率達到 100% 且進行積極的存續期間管理，機構 MBS 仍面臨利差擴大與按市價計價 (mark-to-market) 波動的風險。
• 季末過後住宅證券化利差微幅擴大，不過管理層仍形容投資人需求與市場深度維持韌性。

分析師問答重點

• 股利支撐能力：管理層預期第四季收益將超過股利，並對目前每股 0.75 美元的發放水準感到滿意。
• 資本配置：住宅信貸是新增資本的首要考量，其次為機構 MBS，再者為 MSR。
• 避險策略：隨著利率上升，Annaly 增加了國債期貨配置，並以高於平時的期貨比率進入第四季。管理層計劃考慮將部分期貨轉換為 Swap，以在維持避險平衡的同時提高收益。
• 信貸韌性：住宅投資組合按市價計價的 LTV 為 64%，平均借款人擁有接近 40 萬美元的房屋權益。管理層表示，鑑於借款人品質及有限的疊加風險，全國房價若適度下跌仍屬可控。
• MSR 供給：管理層預估放款機構將因房貸發放利潤率偏低、產業整併及流動性需求而繼續出售 MSR。Annaly 可能會在價格具吸引力時買進，並在銀行需求轉強時賣出。
• Non-QM 展望：管理層預期，在市場維持以購屋驅動且利率處於高位之際，non-QM 在房貸發放中的佔比將保持偏高。若房貸利率下降引發傳統房貸與政府房貸的轉貸潮，該佔比可能會顯著下降。

完整財報電話會議逐字稿

---

完整財報電話會議逐字稿

管理層陳述

Operator

Thank you for standing by, and welcome to the Annaly Capital Management Third Quarter 2026 Earnings Conference Call. [Operator Instructions]

I'd now like to turn the call over to Sean Kensil, Director of Investor Relations. You may begin.

Sean Kensil

Good morning, and welcome to the Third Quarter 2026 Earnings Call for Annaly Capital Management. Please note that this call is being recorded. As a reminder, materials for today's call are available on our website at www.annaly.com.

Today's call may include forward-looking statements, which are subject to certain risks and uncertainties that could cause actual results to differ materially and refer to certain non-GAAP measures. Please see the notices in our earnings release for important information regarding forward-looking statements and non-GAAP measures.

Participants on this morning's call include David Finkelstein, Chief Executive Officer and Co-Chief Investment Officer; Serena Wolfe, Chief Financial Officer; Mike Fania, Co-Chief Investment Officer and Head of Residential Credit; V.S. Srinivasan, Head of Agency; and Ken Adler, Head of Mortgage Servicing Rights.

And with that, I'll turn the call over to David.

David Finkelstein

Thank you, Sean. Good morning, everyone, and thanks for joining our Q3 earnings call. I'll begin with an overview of the macro environment and our performance during the quarter and discuss each of our 3 investment businesses, finishing with our outlook. Serena will then provide a more detailed review of our financials, after which we'll open the call up to Q&A.

Now starting with the macro landscape. The U.S. economy remains resilient, supported by heavy AI investment and healthy consumption. The labor market is stable, while inflation remains above the Fed's target, driven in part by higher energy prices. The bond market faced a challenging third quarter, with the 10-year treasury yield rising 82 basis points to multi-decade highs. The sell-off accelerated following the Fed's first rate hike since 2023, and markets are now pricing nearly 3 additional 25 basis point hikes through the end of next year. Higher rates also reflected concerns around government debt, heavy corporate issuance and a repricing of the Fed's neutral rate.

Mortgage rates have accordingly reached 7.5%. The level during this cycle seen only briefly in the fall of 2023, and thus housing activity is likely to stay subdued as affordability challenges persist and the existing borrower lock-in effect continues to constrain turnover. Weaker demand could soften home prices, though supply and demand remain broadly balanced, notwithstanding regional disparities, and underlying borrower credit remains fundamentally sound.

As upward pressure on rates intensified during the month of September, rate volatility rose significantly, with implied volatility measures reaching their highest levels since the administration's initial tariff announcement 1.5 years ago. And this drove mortgage spreads wider and led to meaningful underperformance in Agency MBS. Against this backdrop, our portfolio delivered a negative 7.6% economic return for the quarter. Earnings remain durable, however, with EAD of $0.80 per share, again, exceeding our dividend, while economic leverage ended Q3 up modestly at 5.9 turns.

In capital markets, we announced the redemption of our $443 million Series I floating rate preferred stock during the third quarter, and the redemption helps us maintain a very conservative capital structure. And we view balance sheet leverage as more attractive, particularly as repo financing remains readily available in terms of residential credit and MSR warehouse lines continue to improve. Relatedly, we issued $757 million in common equity in the quarter through our ATM program, with proceeds used both to fund the preferred redemption as well as to invest across our portfolios.

Now turning to our investment strategies and beginning with Agency. The sharp selloff in rates led to meaningful extension across the MBS universe, particularly in the more negatively [ convexed ] portions of the market. This created duration mismatches and contributed to selling pressure from some more highly levered investors late in the quarter, which exacerbated the volatility-driven MBS spread widening. While we are certainly not immune to mortgage underperformance, we were positioned for rate and spread volatility with conservative duration exposure and historically low leverage heading into the quarter.

Our Agency portfolio decreased modestly to $92 billion in market value, representing 55% of the firm's capital that we added approximately $2 billion notional in MBS in the quarter. Activity focused on deploying a portion of our common equity raised in a higher coupon specified pools, primarily [ 5.5s and 6s ]. This past quarter also demonstrated the value of our allocation of high-quality specified pools. On average, our pools for more than 3.5 years season and carry loan balances less than half those of new production, providing both call and extension protection. As mortgage rates increase, paves declined on call-protected pools as expected but increased on extension protected pools. And given the distribution and quality of our collateral, the average pay up on our portfolio increased modestly over the quarter.

Now early in the fourth quarter, we monetized some of this outperformance, rotating over $2 billion in primarily seasoned pools in the higher coupon TBAs, which stand to benefit more from a decline in volatility, something we have begun to see since the end of the third quarter. On the liability side, as rates moved higher, we responded by adding to our hedge position, primarily in treasury futures, to keep our overall interest rate exposure contained. And we entered the fourth quarter with slightly more rate exposure, however, as current higher yields offer attractive value. And we're also comfortable carrying somewhat more duration as the rise in rates materially reduced negative [ convexity ] in both our portfolio and the broader MBS universe. And as a result, asset durations should be less sensitive to further rate moves in the current environment, reducing the need for hedge rebalancing.

Moving to residential credit. Our portfolio ended the third quarter roughly flat quarter-over-quarter at $10.4 billion in market value, representing 23% of the firm's capital. Residential spreads traded in a tight range for the majority of Q3, with AAA non-QM spreads ending the quarter approximately 10 basis points wider, largely in sympathy with other securitized products. Our Onslow Bay correspondent channel delivered another strong quarter, with $7 billion of lock volume and $5.8 billion of fundings, our highest quarterly funding volume since the channel's inception more than 5 years ago. And importantly, growth has not come at the expense of credit discipline, with our current lock pipeline carrying an average FICO of 766 and an average LTV of 67%.

Now with respect to securitization, we settled $7.4 billion of transactions in Q3, while pricing an additional $2.2 billion in October. Annaly remains the market's largest expanded credit sponsor, and we have now [ prized ] more than $21 billion securitizations in 2026 across 37 deals. And while the new issue market has experienced marginal widening post quarter end, we're encouraged by the resiliency of the private label market, and expect 2026 to be a post-GFC record in transaction volume.

Now to discuss MSR, our portfolio increased to $4.3 billion market value, with our allocation of the sector representing 22% of the firm's capital. We took a measured approach to purchases during the quarter, settling one bulk package for roughly $140 million in market value, while our flow channel experienced just over $50 million in volume and new quarterly highs as we continue to gain market share in the co-issue sector. Both supply in the market increased quarter-over-quarter, and we expect it to stay robust throughout the balance of the year given current mortgage industry dynamics, as we've spoken about extensively over the past year. And this environment should provide ample opportunity to responsibly grow our portfolio. Year-to-date, we remain the largest purchaser of bulk MSR, reflecting the value we bring to the origination community as a strategic partner with a deep permitting capital base.

MSR valuations modestly increased, with our multiple ending the quarter at 6.15x, largely driven by higher interest rates and wider current coupon MBS spreads. Our prepayment speeds remained very low this past quarter, decreasing to 4.8 CPR, reflecting our portfolio's low average note rate, seasonal factors and higher mortgage rates. Credit performance remains strong, with serious delinquencies below 60 basis points and along with healthy income from our escrow deposits, these factors support a predictable cash flow profile that continues to be a complement to our other investment strategies and an important contributor to our EAD.

Now turning to our outlook. We're mindful of uncertainty in the market. We entered the fourth quarter prudently positioned with substantial liquidity, and we remain focused on preserving capital while dynamically allocating across our 3 businesses as opportunities emerge. Within Agency, we continue to manage [ call and ] extension risk proactively as the rate outlook evolves, while current spread levels offer compelling new money returns in the mid- to high teens. And while elevated volatility persists, the outlook for Agency MBS is constructive as spreads retraced the widest level since before the GSEs began more actively increasing their portfolios in mid-2025. At the same time, the technical picture remains favorable, with record bond fund inflows and continued support from the GSEs, overseas investors and banks, while supply is limited given seasonal and high primary mortgage rates.

In residential credit, our scaled Onslow Bay correspondent platform gives us direct access to high-quality whole loans, greater control over asset selection and supports one of the leading securitization programs across asset-backed finance. As the largest issuer of expanded credit MBS, we're uniquely positioned to participate in the continued growth in the non-Agency market, with non-QM issuance up nearly 60% year-to-date. And the investments we have made across our platform over the past several years have supported expansion, operating efficiency and disciplined credit execution.

And lastly, in MSR, our differentiated portfolio of low note rate, high-quality assets continues to generate durable cash flows, supported by our diversified network of servicing and recapture partnerships. And as demonstrated this year, we have the flexibility and capacity to grow the portfolio accretively, further enhancing our earnings and cash flow profile.

Now collectively, our 3 complementary businesses provide multiple avenues to deploy capital efficiently without the burden of a high fixed cost origination model. We remain patient, but we have ample liquidity and balance sheet capacity to put capital to work. And we believe the combination of diversified earnings, proprietary asset creation, conservative leverage and capital allocation flexibility positions Annaly well to continue delivering attractive risk-adjusted returns.

Now with that, I'll turn the call over to Serena to discuss our financials.

Serena Wolfe

Thank you, David. Today, I will briefly review the financial highlights for the quarter ended September 30, 2026. As in prior quarters, our earnings release discloses GAAP and non-GAAP earnings metrics. And my comments will focus on our non-GAAP EAD and related key performance metrics, which exclude PAA.

As David noted, the challenging fixed income market environment weighed on book value during the third quarter. Our diversified hub and finance platform nevertheless, continue to generate solid earnings, supported by higher coupon income, continued residential credit securitization activity and lower operating expenses.

As of September 30, 2026, our book value per share decreased 11.4% from the prior quarter to $17.86. Including our quarterly dividend of $0.75 per share, we generated an economic return of negative 7.6% for the quarter. Book value performance primarily reflected wider Agency MBS spreads and the impact of elevated rate volatility, including market-driven duration and convexity effects. Gains in our hedging instruments provided meaningful offset.

Earnings available for distribution increased by $0.01 to $0.80 per share, and again exceeded our quarterly dividend of $0.75 per share. The increase was driven primarily by higher coupon interest income on larger average investment securities and residential mortgage loan balances, along with lower operating expenses. These benefits more than offset higher interest expense from increased securitization debt and repo balances as well as modestly higher funding rates.

Net interest margin declined by 7 basis points to 1.69%, while net interest spread declined by 8 basis points to 1.42%. Average asset yields were relatively stable at 5.45%. In comparison, our average economic cost of funds increased 7 basis points to 4.03%, primarily reflecting a greater proportion of securitized debt in our funding mix and slightly higher repo rates.

We continue to actively manage the balance sheet through the market volatility. Economic leverage increased to 5.9x from 5.6x, primarily reflecting the decline in equity, while our hedge ratio increased to 100% from 97% as we maintained our defensive positioning. Our weighted average repo rate for the quarter increased 3 basis points to 3.87%, while weighted average repo days to maturity ended the quarter at 30 days, down 3 days from the prior quarter.

Despite interest rate volatility in the third quarter, the reefer funding market continues to function smoothly, ensuring reliable and efficient access to short-term financing for Agency MBS. Meanwhile, we continue to maintain ample financing availability for our credit businesses to support their growth, which David highlighted earlier. Total warehouse capacity across our residential credit and MSR businesses remained $8.3 billion, including $2.8 billion of committed capacity. We maintained significant available capacity, with utilization rates of approximately 57% for residential credit and 50% for MSR.

We ended the third quarter with $6.9 billion of unencumbered assets, including $4.2 billion in cash and unencumbered Agency MBS. In addition, we had approximately $1.6 billion of MSR collateral pledged and available to support future advances, subject to market advance rates. In total, we had approximately $8.5 billion of assets available for financing at quarter end, that approximately $1.1 billion from the prior quarter, primarily reflecting lower equity and reduced unencumbered Agency MBS. This represented approximately 54% of our total equity and continue to provide meaningful liquidity and financial flexibility.

Finally, our OpEx to equity ratio decreased 11 basis points to 1.29% for the quarter, while year-to-date ratio remained 1.34%. As anticipated, expenses moderated from the elevated second quarter level, reflecting lower compensation of the G&A expenses.

That concludes our remarks. We will now take questions. Thank you, operator.

Operator

[Operator Instructions] Your first question comes from the line of Bose George from KBW.

分析師問答

Bose George

Can I start with the -- just get an update on the mark-to-market book value?

David Finkelstein

Sure, Bose. As of Friday, excluding the dividend accrual, we were up about 2%, inclusive of the dividend closer to 3% on a sideways since.

Bose George

Great. And then just wanted to ask about dividend coverage, just given the expectation for current economic returns, how does that compare with the required ROE to cover the dividend?

David Finkelstein

Sure. So obviously, we outearned in Q3. We expect to outearn in Q4, and we'll see how 2027 evolves. But as it stands today, we didn't delever in Q3, notwithstanding the volatility and the current portfolio delivers the dividend. And as we laid out our projected returns, which are up quarter-over-quarter, new money gets us there as well. So we're comfortable with where the dividend is at today.

Operator

Your next question comes from the line of Crispin Love from Piper Sandler.

Crispin Love

David, on that last point, the market level returns kind of as you showed out on Slide 7, it looks like each strategy is a little bit higher than last quarter. Can you just discuss some of the dynamics there, driving returns a little bit higher? And then also just where you're most interested in adding incremental capital today in the macro environment? And I do appreciate you guys -- 3 businesses that you want to add, but like maybe if you could stack right?

David Finkelstein

Sure. So I'll start with the end of your question on capital allocation, then we'll go through each of the 3 businesses in terms of how the returns evolve heading into Q4. So in terms of capital allocation, it's not much different than last quarter, returns across the 3 businesses are incrementally higher. And as we did say last quarter, we still want to grow the residential credit business responsibly, considerate of credit that's at the margin where we'd like to grow, and agency returns are ample and certainly compelling, and so that would be second.

And then MSR, we still have room to grow. Obviously, those headline returns are a little bit lower than Agency and resi. But one point to note is that position is largely funded with capital of very little leverage. So we do have room to grow and maintain a 20% capital allocation to the sector. So the order is resi first; close second, agency; and then not far behind is MSR. So let's just talk about returns in each, and I'll start with Srini on the agency and then go to Mike and Ken.

V.S. Srinivasan

Sure, David. So basically, Agency spreads for the quarter widened about 20 basis points. Now when we say spreads widened about 20 basis points, we generally refer to current coupon spreads. And given that only is [ 5 and 6.5s ] are over par, we really can't invest in current coupons. So we are slightly below current coupon, call it, 5.5 or 6s is our preferred coupon. And for those -- compared to where current coupon was at the last quarter spends are about 10 basis points wider. So that gets us to 15 to 17 basis points with about 100 basis point increase in prospective results.

Michael Fania

Sure, Crispin, this is Mike. On the resi side, what those returns reflect the 13% to 16%. First, it will reflect the equity that we're retaining through our securitization. So call that $0.06 to $0.07 of capital on the retained OBX securities using maybe, call it, a turn of leverage. So we'll retain, call it, 10% to 12%, 1 turn of leverage gets to that $0.06 to $0.07 of deployment.

Given spreads have widened, we've seen our equity basis has been able to widen as well. And given the volatility, we're pricing in incremental premium. We can also earn mid-teens returns on just residential loans, putting on nonmark-to-market facilities. Our non-mark-to-market facilities are very accretive in the current environment and call it maybe 4 to 5 turns of leverage, and this is non-mark-to-market, you're calling it 15%, 16%.

And lastly, there is an opportunity to buy BB non-QM bonds. So these are new issue third-party shelves. Right now, they trade, call it, [ 4 25 ] per swaps. You can use 20% to 25% haircuts on those. And the financing there is in the low 100. So that would actually get you closer to that 16%, 17%. So there's a number of options at which we'll be able to get mid-teens or potentially a little bit above mid-teens.

Ken Adler

Yes. Thanks. This is Ken. On the MSR front, the change in return from the prior quarter is largely driven by all rates rising. And as capital is deployed, that capital earns a higher rate of return. Additionally, with the MSR, a bulk of the negative duration, certainly on the lower note rate collateral, is driven by the float income as we manage the borrowers' escrow accounts. So as rates rise, we, as the servicer earn a higher rate of return on those escrow balances, and that applies to prospective investments. So those 2 items are really driving the change in prospective return quarter-over-quarter.

David Finkelstein

Yes. And Crispin, just to add to that point, we entered Q3 with about 1.25 hikes priced in to the market, and we ended the quarter and sit today with about 4, including the hike that occurred in September. So just that float income increase in income alone gets you a lot more yield.

Crispin Love

Great. That makes sense. And then just one last question for me. Just on just the recent rate fall in the third quarter and then just a little bit of a change in the fourth quarter. Have [indiscernible] any state changes to your hedging strategy in kind of recent months, recent weeks, especially as rates were rising, you had a little bit of a different environment in October. So just curious on a little bit of the change in the second half of 3Q [ and ] October.

David Finkelstein

Yes. Look, as the market was selling off, we did dynamically hedge, and we predominantly used the futures market to achieve that. And we entered Q4 with a relatively higher percentage of futures than we normally operate with. So heading into this quarter, our anticipation will be to find opportunities to convert some of those futures into swaps, which gives us a little bit more yield and keeps a good balance on our hedge profile.

But generally speaking, over the last couple of weeks, we've been somewhat encouraged by the [ vol ] subsiding to some extent. You see it both in realized and an implied vols somewhat. And so we'd like to think we're out of the woods. We're still very, very vigilant when it comes to our hedge profile and making sure that we're managing the portfolio responsibly.

When we look at the environment today and the level of rates, we understand why rates are higher. Obviously, inflation is a big driver, and we are in a hiking cycle, as shallow as that may end up being heavy debt issuance. And also, the economy has proven to be able to grow at higher rates. And so the neutral rate is higher. And so we feel like the new normal is higher rates. And so the floor on rates is higher. But we do generally like the rates market. We're carrying a little bit of duration here, and we think that's appropriate given the attractiveness of rates, and we're comfortable with the positioning.

Operator

Your next question comes from the line of Doug Harter from BTIG.

Douglas Harter

You guys put up a smaller decline in economic return than your peers. Clearly, diversification helped that. But I was wondering if you could help us think how, like the agency portfolio performed this quarter as we try to compare that to peers?

David Finkelstein

Yes. I would -- look, I think -- I have a brief understanding of how others performed. And I think the agency sector was largely pretty well clustered, close to one another for the most part. I think my understanding is that folks were deliberate about managing duration exposure.

And I'll tell you something, Doug, agency -- it's not that easy to differentiate an agency portfolio, which may not sound too attractive to agency monoline peers. But at the end of the day, there's a lot of correlation amongst agency assets. You can be up in coupon, down in coupon, specs, TBAs. But when the agency market cheapens, it tends to affect the entire sector and performance, ex how you manage duration can be somewhat uniform. I think this quarter, we were probably somewhat contextual with the -- our agency portfolio with how others perform because I think everybody was pretty responsible about navigating the rates market and hedging.

Douglas Harter

I appreciate that, David. And then on the residential credit, you talked a little bit about some of the spread widening you've seen on new issue deals. Can you just talk about whether there have been any kind of other structural changes or what the depth of the investor base looks like kind of as we've gone through this market?

Michael Fania

Sure, Doug. This is Mike. I would say at the top of the capital stack, we ended Q2, call it, [ 1 30 ] over the curve. That's to treasuries. At the end of Q3, we're 5, 10 basis points wider, call it [ 1 40 ]. We just priced our latest deal, our latest transaction post quarter end, non-QM 16. That was done at [ 1 50 ]. We have seen other AAA sponsors and other shelves probably price, call it 5 to 15 basis points wider. I do think reflecting the strength of the OBX main, the programmatic nature of the platform and then also the performance as well. So I do think we're still firmly in that kind of Tier 1 category, and we're benefiting from that.

In terms of structural changes, what we have seen, it's been fairly modest, right? So the deals that have been coming, they call it maybe low 7% gross WACC, low to mid-7% gross WACC. The current coupon within non-QM right now, the 102.5 rate is probably closer to 8%. So these deals are modestly out of the money.

What the market has really demanded as a concession to be able to move the risk is that the AAA bond, which traditionally is pro rata with the AA and the single A, has turned into a full sequential. So it's a fairly modest change. The BBB and down, the M1 and down has always been a sequential pay. So you're really changing the structure of the waterfall of 2 bonds. And then what you're seeing is those AA and those single A bonds because instead of being pro rata and factoring down, they become sequential pay in the initial period before the call.

So very modest changes, but I would say that the market is still healthy. The market is very resilient. We've seen 95 billion of non-QM issuance this year. Last year, the entire year was 80 billion. 45% of all issuance we've seen in non-Agency has been within the non-QM market. So I haven't really seen any change in terms of investor depth, and it's a slight structural change to be able to absorb modestly out of the money transactions.

Operator

Your next question comes from the line of Jason Stewart from Compass Point.

Jason Stewart

We noticed a decent uptick in home equity securitization volume. Just maybe if you could talk, Mike, about the 2 sides there, how you're approaching sizing that as an asset and then how the market is pricing those liabilities today?

Michael Fania

Sure. Thanks, Jason. So on home equity, we'll classify that as both HELOCs and closed-end seconds. We've seen, in terms of total issuance, right, in the size of the entire market. We've seen about $220 billion of issuance within the non-agency market. About 15% of that has been closed-end seconds and HELOCs. It's around $33 billion. In terms of our production and what we really focus on, it's probably, call it, 5% to 6% of our production that we see through the correspondent.

We are active predominantly in full documentation, HELOCs and closed-end seconds. We've talked about that before. There's a small kind of niche market within the alternative stock. But really, the focus continues to be on non-QM and DSCR. A lot of it is scalability. A lot of it is our competitive advantages are within non-QM and DSCR. Close-end seconds and HELOCs, they trade at very large premiums. Even to be able to aggregate them through a correspondent, you need to be willing to pay [ 1 06, 1 07 ] type dollar prices. So that is something where we've been, I think a little bit more conservative than the rest of the market.

But in terms of the distribution, the securitization market, similar to non-QM, it's very healthy. It's not the same size. But I will say that it's -- we continue to see transactions, we continue to see deals. If you call AAA, non-QM right now for Tier 1 issuers, if we call that [ 1 50 to 1 55 ], I would say for Tier 1 closed-end second issuers, it's probably 5 to 10 basis points inside of non-QM. And part of that is that the market does believe it's a better convexity profile. But in terms of our strategy, we'll continue to buy those assets, but I think we're probably a little bit more conservative in terms of our scenario analysis, analytics and are willing to pay premiums. But market is open, it's healthy. And I would say, maybe, call it, 5 to 10 basis points inside the non-QM AAA market right now.

Operator

Your next question comes from the line of Marissa Lobo from UBS.

Ameeta Lobo Nelson

Just on MSRs, historically, MSRs have provided you a natural offset to the negative convexity in Agency MBS. But as we're seeing the mortgage universe become more fully extended and the convexity profile improving, how are you thinking about the role of MSRs in the portfolio today? And should we think about them more as a stand-alone return opportunity? Or do they continue to provide more diversification across rate scenarios?

Ken Adler

This is Ken. Thanks for the question. I would answer as both. The stand-alone return is certainly ample and we talked about that earlier. In terms of the hedging properties, beyond the gamma, the convexity sort of hedging, there's another form of hedging that's going on, and that's the model hedging. We use prepayment models to measure our duration, convexities and other risk measures. To the extent prepayment speeds for some other reason, come in faster or slower than the model, MSR provides a direct hedge to the portfolio. Considering if speeds come in slower than expected, that's always good for MSR. And that's in today's market with the mortgage [ university ] discount, just not good for the Agency MBS portfolio. So there's this model error that's being hedged through on the MSR.

David Finkelstein

Yes. One more point, Marissa, I'll say is in -- following up on Ken's point, is that to the extent that the mortgage market's at a reasonable discount like it is, and turnover is actually lower than natural turnover, which should be the case in the MSR portfolio is going to hedge that as well. And as I talked about a moment ago in terms of float income, it's a hedge for short rates, higher short rates increases in float income and MSR benefits from that as well. So there's still a very productive balancing component to the MSR portfolio in the current environment.

V.S. Srinivasan

I would also add that away from models and all of that, agency MBS has been trading directionally with rates. So when rates sell off, MBS spreads have been under pressure. And what we have observed over the last 3 years is when rates sell off, MSR actually trades very, very well. So from a pure mark-to-market trading perspective, it does provide cushion on book value loads.

Ameeta Lobo Nelson

Okay. Great. And then just shifting to residential credit and thinking about home price appreciation, can you talk to your home price outlook and its flat HPA manageable? And what are the assumptions that are underlying that 13% to 16% residential credit return?

Michael Fania

Sure. So thanks, Marissa. I'll address the second part of it, the 13% to 16%. Again, that's mostly the OBX platform. It's non-QM and DSCR. I will say that that's modeled to, call it, maybe a 20% lifetime speed, which is generally where non-QM has come in. In terms of [ cume ] losses, that probably represents close to 75 basis points of [ cume ] loss in terms of what our projections are. So there's significant room for upside.

If you look at the totality of our securitization program, we've issued, I think, close to $67 billion life to date. The actual amount of realized losses we have taken inside the trust is around $6 million. So it's been less than 1 basis point. So one of the areas where we've outperformed our initial expectations has been within credit modeling. We do expect to take more losses and like time to losses on a lot of non-term DSCR, maybe closer to 10 to 15 basis points as they season. But it just shows you that there is a level of conservatism in terms of those projections.

In terms of the housing market, a lot of it will depend upon what mortgage rates do and where mortgage rates ultimately settle. If we stay at [ 7 and 3/8 ] to 7.5% for an extended period of time, we do believe housing can go negative. It can go negative by a couple of points nationally. So a lot of it is dependent upon mortgage rates.

In terms of where we sit today, what we would say, the modestly bullish case for housing, still positive. September was up 8 basis points month-over-month. It's up 1% year-over-year. If you look at ZIP code level, 77% have positive year-over-year increases. There's a significant amount of homeowner equity, call it, $35 trillion to $40 trillion, depending upon what you think that the size of the market is. The average borrower is locked into a [ 4 65 ] rate. Inventory is still 15% below 2019. And new listings, which is reflecting that mortgage rate, they haven't really increased. It's only up 1% to 2%.

The caution, and this is what a lot of people are seeing now, and we do have a level of conservatism in terms of just not only our purchase assumptions but also the quality of the portfolio, as Dave mentioned, the Southeast is very weak. Texas, Florida, you see the public builders talking about competing with resale, month supply is at a 10-year high. And [ 7 and 3/8 ] mortgage payment, that mortgage payment is up 6% to 7% just over the last quarter relative to Q3. And everyone knows the affordability issues that we have, right? So if we stay at these rates, yes, we do think it could go down 2% to 3%, 3% to 4%.

Ameeta Lobo Nelson

But manageable within the -- yes. I guess within the equity cushion that you...

Michael Fania

Yes, it's very manageable. As Dave mentioned in the script, our average borrower has close to 400,000 of equity. It's a 64 mark-to-market LTV in our portfolio. The D60 plus of our portfolio right now is 160 basis points. And we are predominantly targeting upper middle-class borrowers, right? So it is not necessarily this first-time home buyer. It is not necessarily the borrower who has been challenged with their discretionary income in this inflationary environment. So I think from that perspective, we feel pretty good about the strategy.

David Finkelstein

Yes. Marissa, there's also very little layered risk in the portfolio. So we can manage and wouldn't even mind seeing HPA home prices come down a little bit to increase the affordability equation to get turnover up.

Operator

Your next question comes from the line of Rick Shane from JPM.

Richard Shane

There's been a lot of great questions this time so. Just wanted to talk a little bit -- I mean empirically, it's one thing to understand what's going on in MSR, but there are, I would describe, a lot of exogenous factors that are impacting -- potentially impacting valuation, mortgage volumes, industry capacity, potentially the impact of AI affecting speeds. I am curious when we think about, again, sort of your empirical models and then overlay all of these different structural things that are happening in mortgage right now, how is that impacting value pricing demand and where you see attractive opportunity within MSR?

Unknown Executive

Yes. Thanks for the question, and those are certainly things we're thinking about all day and actually all night. Look, I mean, the mortgage industry is in a place because of consolidation, because of technology, where people are operating much more efficiently and are operating at much lower profit margins. So because the efficiency that's priced in because there's overcapacity, because there's technology, because there's consolidation, it's tough to decouple all the reasons.

The profitability of the origination part of the business is very low. And the value of a newly created MSR today is somewhere depending on the loan type between 1 point and 1.5 points. And the profitability of the mortgage industry is substantially lower than that. And on average, when you look at the MBA numbers is in the 20 basis point range now. And since they've been measuring it on average is in the 40 basis point range.

So originators are just forced to sell MSR going forward in our view. So if loan production picks up, they have to sell MSR because they're not making as much money. Historically, when volumes picked up, margins picked up, and they picked up because of the shortage of capacity that occurred. We think because of technology that won't occur anymore. So as a result, there'll be lots of MSRs selling volumes pick up because the profitability will not increase on a per unit basis, allowing more MSR retention.

And then we think, if origination volumes don't pick up, what we're seeing is capitulation, consolidation, needs for liquidity to keep the platforms going until profitability might increase, and that's creating MSR selling as well. So we just see a continual restructuring of how the industry works. Where folks without a cost of capital like ours or better are not really going to be able to hold MSR and they're still holding a lot of legacy MSR and then go forward the new MSR, we have a tough time on seeing how they'll be able to retain that. And I think when you look at the MSR transfer data, that's public, you'll see who the sellers are, the bulk sellers are. And you'll see this thesis kind of playing out already.

David Finkelstein

Yes, Rick, I'll jump in here and just talk about the demand side and valuation. The market -- the bid for MSR is somewhat bifurcated. You have folks like us, capital participants in the market that have obviously accumulated a lot of MSR. And you also have banks, which, particularly given the likelihood of MSR not being a constraint when it comes to the Basel re-proposal, which we expect to get out the final which in the next couple of months, banks have been a strong source of demand, particularly in lower loan balance type MSR that carry a lot of escrow balances relative to the capital committed.

And so they've been a significant bid. As we talked about last quarter, we sold into the bank bid at levels that we thought were quite attractive. And then folks like us, we benefited from all the liquidity that's been generated with the activity in the MSR market and have traded quite a bit an MSR, and we still have a fair amount of appetite to add MSR and as do others like us that are portfolio companies. So we understand the market pricing. We still think it's fair. When there's a rich bid out there, we'll sell MSRs we've evidenced very recently. And I think we're -- from a portfolio standpoint, we're positioned conservatively and our pricing is -- on our portfolio is probably the most conservative in the industry.

Richard Shane

Look, I appreciate the answer. And it's an interesting thing. I think 25 years ago, we wrote something that said the procyclicality of margins would start to erode because of technology. And it feels like we're finally on the cusp of that, and I do think that there's probably no industry that's -- financial services industry that's going to be more transformed by AI than mortgage. So it's a really interesting time, and I do appreciate your insights.

Operator

Your next question comes from the line of Jason Weaver from JonesTrading.

Jason Weaver

So first, Dave, since the beginning of the year, we thought of the GSEs is showing a stabilizing bid behind the BS spreads. And I don't think we've seen any data since August. But do you have any sense as to the extent they've been involved or stepped away from purchases over the past several weeks?

David Finkelstein

Yes. Look, everything is anecdotal, subsequent to August because we don't have that data either. It did appear as though the activity was a little bit slow in the summer. But anecdotally, we have heard that they have been involved in the market. And I think -- you even got a tweet out of the FHFA director in September talking about the GSEs buying and some adjective to describe a considerable quality or quantity of MSR. So our understanding is that they probably are. They still do have a fair amount of dry powder. I think a less look at the end of August. Total capacity under the PSPA caps for loans and MBS was like $140 billion. They do still have to keep a lot of dry powder for loan buyouts and things like that. But our understanding is that they still have some ammunition left.

Jason Weaver

Got it. And second, I think you mentioned layering into some higher coupon TBAs in your prepared remarks. And I wonder if you could talk just a bit about how you see the risk reward in current coupon production, where under a sharp uptick in prepayments could occur under a flight to quality?

V.S. Srinivasan

Sure. I mean -- before we go there, I just want to define what higher coupon is. We refer to [indiscernible] [ 5.5s or 6s ] as higher coupon now, 6.5 float is quite small. They're going to start getting more issuance in 6.5 in the coming months. But 5.5s and 6s are [indiscernible] right now. And at this point, they're both with convictions. The only coupon that's changing at a [ $1 trillion ] price is 6.5. So most of the PPAs we bought were [ 5.5s and 6s ].

In terms of prepayments, we think -- TBA, particularly in production coupon, continues to deteriorate at an accelerating pace because more and more collateral is being carved out in the specified pools. And so what is left for TBA is significantly more inevitably convex. So up the coupon side, we generally don't like being in TBA for very long. We obviously -- when we want to add margin exposure, we typically do it in TBA and then over time, swap it into pools. So right now, we're not worried about our TBAs at [ 5.5s and 6s ] mainly because they are discounts, but at some point, particularly for 6s, we will swap them into specified pools.

Operator

Your next question comes from the line of Kenneth Lee from RBC Capital Markets.

Kenneth Lee

Just at a higher level, curious as to how you'd characterize your overall risk appetite just given the current rate outlook and the potential trade-off between market vol and return opportunity as you see across the various asset categories?

David Finkelstein

Yes. Thanks, Ken. Good question. So look, we're reasonably optimistic on the outlook here. Rates are higher. We do think there's somewhat of a ceiling to rate. I think when you look at the level of interest rates, it's obviously being driven by strong growth in inflation. And there's some real costs associated with these high rate levels. Obviously, the AI sector is continuing to borrow and continuing to invest. But the rest of the economy, to some extent, is being crowded out. That's households, corporates, and it's becoming more expensive for government to issue debt. So we feel like there's limits to how high rates can go, but there's also a new floor on the level of rates, given we are, as I spoke about in this new normal environment.

So that -- what that says to us or suggests to us is that vol should come down. And it has, to some extent, as we started the fourth quarter, we've reset. And our view is that things are going to stabilize or continue to stabilize, and you will see implied vols coming down because realized vols are a little bit lower. So that gives us a little more optimism in terms of our appetite for risk. And I think the position is, to a large extent, reflecting that. We let our leverage organically increase this past quarter. We didn't delever. We did delever the capital structure a little bit on the prep side. And I'll just tell you a little bit about that because that was really about capital structure and leverage optimization.

When we issued prep a year ago, we issued it at [ 8 and 7 As ] and agency spreads were 18% to 19% or thereabout. So the spread on that was about 1,000 basis points. And then you fast forward to when we made the decision to reduce our structural leverage, which was right after earnings, I believe. And at that time, we had floating rate prep outstanding and we've gone from 2 cuts priced into the market over the prior year to 4 hikes. And so the cost of that increased to around 10%.

Meanwhile, Agency spreads were relatively tight at that time and levered returns, as we spoke about on the last call, could be -- were as low as 14% to 15%. So that spread compressed quite a bit. We reduced capital structure leverage, and we felt good about it. And then you fast forward and the market cheapened again, obviously, to get closer to the higher teens returns. So here we are again and things look pretty appealing from that standpoint. And we're comfortable with our leverage. We have a little capacity to increase it tactically, but we're generating the returns we need with current leverage. And should spreads tighten, I think we show in our spread shocks for a 25 basis point tightening, it's just over 10% return from that. So we're in a position where we like to be and we're comfortable with the risk in the market, and we think all will continue to come down, and we'll be back in business like we were prior to the episode of vol in the late summer.

Operator

Your final question comes from the line of Trevor Cranston from Citizens JMP.

Trevor Cranston

Okay. Question on the non-QM market. Obviously, it's less rate sensitive in general than the conventional space, but curious, your outlook on how the increase in rates, particularly if mortgage rates hold up here, how that impacts the non-QM space and what your outlook is for origination volumes there going forward compared to where they've been in the last couple of quarters?

Michael Fania

Sure. Thanks for the question, Trevor. This is Mike. In terms of non-QM origination this year, we still think it probably is going to be and, call it, the [ 1 50 to 1 75 ] in terms of 175 billion. I would say it's probably leaning towards the higher number and the higher amount of that range. In terms of total mortgage origination, maybe we'll call it 1.9 trillion to 2 trillion. So we do think it will continue to be around, call it, 8% to 9% of the entire market.

The reason that non-QM is still an elevated percentage of originations, a lot of it is just because mortgage rates are elevated. It is a purchase market, right? And non-QM is targeting to a lot of investors, right? So over 50% of this market is investor. Investors right now are about 20% of all real estate purchases, which is a higher share. So we are getting a higher percentage of the end borrower at this point. And also, to your point, self-employed borrowers are a little bit more inelastic in terms of -- at this level of rates.

So I don't know if the share itself will grow. I think a lot of what you've seen over the last number of years has been maturation. But originators, at this point, pretty much to a T, have rolled out the product. It's still higher gross margins than the conforming market than the government market. And it's a product that originators need to have to be able to serve their brokers or serve their LOs on the retail side. So I would expect it to continue to be 8% to 9% of originations at this point.

In terms of if the market rallies, right, if the market rallies and the mortgage rate is at 6, 6.75, that percentage is going to shrink exponentially because you're going to see a lot of rate and term refinance of government and conventional. So the share is outsized a lot just because it is a purchase market. This market does target the purchase borrower, high investor share. And there's not a lot of rate and term and cash out at this point.

Operator

And that concludes our question-and-answer session. I will now turn the call back over to CEO, David Finkelstein, for closing remarks.

David Finkelstein

Thank you, and thank you, everybody, for joining us. We'll talk to you soon.

Operator

This concludes today's conference call. Thank you for your participation. You may now disconnect.