การประชุมแถลงผลประกอบการไตรมาส 3 ปี 2026 ของ Dynex Capital (DX): มูลค่าตามบัญชีลดลง ขณะที่ผู้บริหารคาดการณ์ ROE 20%+

สรุปประเด็นสำคัญ

• มูลค่าตามบัญชีต่อหุ้นสามัญลดลง 15.4% สู่ระดับ 10.91 ดอลลาร์ จาก 12.90 ดอลลาร์ ขณะที่การจ่ายเงินปันผลหุ้นสามัญที่ 0.51 ดอลลาร์ ส่งผลให้อัตราผลตอบแทนทางเศรษฐกิจรวมเป็นลบ 11.5%
• ผลขาดทุนเบ็ดเสร็จอยู่ที่ 1.49 ดอลลาร์ต่อหุ้นสามัญ เปรียบเทียบกับกำไรเบ็ดเสร็จที่ 0.80 ดอลลาร์ในไตรมาสที่สอง
• ไดเน็กซ์ระดมทุนสุทธิที่ช่วยเพิ่มมูลค่าได้จำนวน 384 ล้านดอลลาร์ ซึ่งประกอบด้วยหุ้นสามัญจำนวน 268 ล้านดอลลาร์ผ่านโครงการ ATM และ 116 ล้านดอลลาร์จากการเสนอขายหุ้นบุริมสิทธิ Series D
• ส่วนต่างอัตราดอกเบี้ยสุทธิทางเศรษฐกิจทรงตัวอยู่ที่ประมาณ 1.18% เนื่องจากอัตราผลตอบแทนจากสินทรัพย์เพิ่มขึ้นเป็น 5.00% และต้นทุนทางการเงินที่แท้จริงเพิ่มขึ้นเล็กน้อยเป็น 3.82%
• ผู้บริหารระบุว่า ส่วนต่างราคา (spreads) ของ Agency MBS ที่กว้างขึ้นและความนูน (convexity) ของสินเชื่อเพื่อที่อยู่อาศัยที่ดีขึ้น ช่วยสนับสนุน ROE ตามแบบจำลองแบบคงที่ที่ประมาณ 20% สำหรับหุ้นกู้ที่มีอัตราดอกเบี้ยหน้าตั๋วต่ำกว่า และ 25% สำหรับอัตราดอกเบี้ยหน้าตั๋วปัจจุบัน สำหรับการลงทุนใหม่ที่ป้องกันความเสี่ยงด้วยสัญญาแลกเปลี่ยนอัตราดอกเบี้ย
• สภาพคล่อง ณ สิ้นไตรมาสอยู่ที่ 1.3 พันล้านดอลลาร์ หรือมากกว่า 42% ของส่วนของผู้ถือหุ้นรวม ขณะที่เลเวอเรจอยู่ที่ 9.0 เท่าของส่วนของผู้ถือหุ้น

ข้อมูลทางการเงินที่สำคัญ

ผลการดำเนินงานทางธุรกิจและการปฏิบัติงาน

อัตราผลตอบแทนสินเชื่อเพื่อที่อยู่อาศัยที่มีอัตราดอกเบี้ยหน้าตั๋วปัจจุบันเพิ่มขึ้นเกือบ 100 basis points ในระหว่างไตรมาส อัตราผลตอบแทนพันธบัตรรัฐบาลอายุ 10 ปีเพิ่มขึ้นมากกว่า 80 basis points และสูงกว่า 5% ขณะที่ส่วนต่างราคาของสินเชื่อเพื่อที่อยู่อาศัยขยายตัวขึ้นเกือบ 20 basis points ผู้บริหารระบุว่าการขยายตัวของส่วนต่างราคาส่งผลกระทบต่อผลตอบแทนในไตรมาสนี้มากกว่า 80%

ไดเน็กซ์เพิ่มการป้องกันความเสี่ยงเมื่ออัตราดอกเบี้ยปรับตัวสูงขึ้นและไม่ได้ลดเลเวอเรจลง บริษัทระบุว่าสิ่งนี้ช่วยรักษาความสามารถในการทำกำไรของพอร์ตการลงทุน และทำให้ยังคงมีความเสี่ยงต่อแนวโน้มการแคบลงของส่วนต่างราคา ความไวต่อส่วนต่างราคาตามแบบจำลองเพิ่มขึ้นจากประมาณ 4.6% ของส่วนของผู้ถือหุ้นสามัญต่อการเปลี่ยนแปลงส่วนต่างราคา 10 basis points ณ วันที่ 30 มิถุนายน เป็นประมาณ 5.7% ณ วันที่ 30 กันยายน

ความไวต่ออัตราดอกเบี้ยตามแบบจำลองของพอร์ตการลงทุนก็ลดลงด้วย โดยแบบจำลองระบุว่าหากอัตราดอกเบี้ยเพิ่มขึ้นทันที 50 basis points จะทำให้ส่วนของผู้ถือหุ้นสามัญลดลงประมาณ 0.5% ณ สิ้นไตรมาส เทียบกับ 3.5% ณ วันที่ 30 มิถุนายน ผู้บริหารอธิบายว่าการปรับตัวดีขึ้นนี้เกิดจากอายุเฉลี่ย (duration) ของสินเชื่อเพื่อที่อยู่อาศัยที่มีเสถียรภาพมากขึ้นและความเสี่ยงจากการขยายอายุหนี้ (extension risk) ที่เหลืออยู่จำกัด

การจัดหาเงินทุนยังคงทำได้ตลอดช่วงความผันผวน โดยส่วนต่างราคาของ Mortgage repo เมื่อเทียบกับ SOFR อยู่ที่ประมาณ 12 ถึง 15 basis points และบริษัทรายงานว่าไม่มีแรงกดดันในการจัดหาเงินทุน ณ สิ้นไตรมาสอย่างมีนัยสำคัญ

การเสนอขายหุ้นบุริมสิทธิ Series D ทำให้ส่วนของผู้ถือหุ้นบุริมสิทธิเพิ่มขึ้นจากประมาณ 3% เป็น 7% ของส่วนของผู้ถือหุ้นรวม ผู้บริหารอธิบายว่าการออกหุ้นดังกล่าวช่วยเพิ่มมูลค่า และกล่าวว่าหุ้นบุริมสิทธิอาจยังคงเป็นแหล่งเงินทุนตามโอกาสต่อไป

แนวทางจากผู้บริหาร

ผู้บริหารกล่าวว่า ROE ตามแบบจำลองแบบคงที่ของการลงทุนใหม่ที่ป้องกันความเสี่ยงด้วยสัญญาแลกเปลี่ยนอัตราดอกเบี้ยอยู่ที่ประมาณ 20% สำหรับ Agency MBS ที่มีอัตราดอกเบี้ยหน้าตั๋วต่ำกว่า และ 25% สำหรับอัตราดอกเบี้ยหน้าตั๋วปัจจุบัน ในการอภิปรายกับนักวิเคราะห์ บริษัทชี้แจงว่าประมาณการเหล่านี้อิงตามเลเวอเรจประมาณ 10 เท่า เมื่อเทียบกับเลเวอเรจจริง ณ สิ้นไตรมาสที่ 9.0 เท่า

ไดเน็กซ์คาดว่าจะยังคงมุ่งเน้นไปที่อัตราส่วนค่าใช้จ่ายในการขายและบริหารปรับเป็นรายปีที่ประมาณ 2% ของส่วนของผู้ถือหุ้นรวม ผู้บริหารยังกล่าวอีกว่าขนาดธุรกิจที่ขยายใหญ่ขึ้นอาจช่วยเพิ่มเลเวอเรจทางการดำเนินงานเพิ่มเติมได้

บริษัทแสดงความมั่นใจในความสามารถในการจ่ายเงินปันผล โดยอ้างถึงการตัดสินใจไม่ลดเลเวอเรจ ซึ่งส่งผลให้รักษาความสามารถในการทำกำไรของพอร์ตการลงทุนไว้ได้ นอกจากนี้ บริษัทยังรายงานการซื้อสินทรัพย์สุทธิหลังสิ้นสุดไตรมาสอีกด้วย

ความเสี่ยงและประเด็นที่ต้องติดตาม

• อัตราผลตอบแทนระยะยาวที่สูงขึ้นและส่วนต่างราคาของ Agency MBS ที่ขยายตัวกว้างขึ้น เป็นปัจจัยกดดันให้มูลค่าตามบัญชีลดลงในไตรมาสนี้
• ผู้บริหารระบุว่า อุปสงค์เงินทุนที่เกี่ยวข้องกับ AI ผลกระทบจากราคาโภคภัณฑ์ เงินเฟ้อ ความไม่แน่นอนทางภูมิรัฐศาสตร์ และความต้องการกู้ยืมเงินที่เพิ่มขึ้นของรัฐบาล เป็นแหล่งที่มาของแรงกดดันให้ดันอัตราผลตอบแทนปรับตัวสูงขึ้น
• ความผันผวนของอัตราดอกเบี้ยแฝงที่อยู่ในระดับสูง ยังคงทำให้มีส่วนชดเชยความเสี่ยงจำนวนมากสะท้อนอยู่ในมูลค่าสินเชื่อเพื่อที่อยู่อาศัย
• ความเสี่ยงต่อส่วนต่างราคาที่เพิ่มขึ้นของไดเน็กซ์สร้างโอกาสเพิ่มขึ้นหากส่วนต่างราคาแคบลง แต่ก็ทำให้มูลค่าตามบัญชีมีความไวต่อการขยายตัวกว้างขึ้นอีกด้วย บริษัทประเมินว่าการแคบลงของส่วนต่างราคา 20 basis points อาจเพิ่มมูลค่าตามบัญชีได้ประมาณ 11.5% หากปัจจัยอื่นคงที่
• ความเต็มใจของบริษัทในการรักษาเลเวอเรจที่ 9.0 เท่า ขึ้นอยู่กับการเข้าถึงตลาดการจัดหาเงินทุนผ่านธุรกรรม Repo ที่มีสภาพคล่องสูงและมีความยืดหยุ่นอย่างต่อเนื่อง

ประเด็นสำคัญจากการถาม-ตอบกับนักวิเคราะห์

• อัปเดตมูลค่าตามบัญชี: มูลค่าตามบัญชีประเมิน ณ วันศุกร์ก่อนการประชุมทางโทรศัพท์อยู่ที่ 10.95 ดอลลาร์ต่อหุ้น สุทธิจากเงินปันผลค้างจ่ายแล้ว
• เลเวอเรจและสภาพคล่อง: ผู้บริหารระบุว่า การจัดหาเงินทุนผ่าน Repo ที่แข็งแกร่ง ยอดคงเหลือในกองทุนตลาดเงินที่มากกว่า 8 ล้านล้านดอลลาร์ และการเข้าถึงคู่สัญญาที่หลากหลาย ช่วยสนับสนุนความสบายใจต่อระดับเลเวอเรจในปัจจุบัน
• อุปสงค์ Agency MBS: บริษัทอ้างถึงการเข้าซื้อที่มีแนวโน้มเกิดขึ้นของ GSE ความสนใจจากนักลงทุนยุโรปและต่างชาติอื่น ๆ รวมถึงอุปสงค์จาก REIT สินเชื่อเพื่อที่อยู่อาศัยว่าเป็นปัจจัยสนับสนุน ผู้บริหารยังกล่าวอีกว่าอุปทานที่จำกัด ณ ระดับอัตราดอกเบี้ยปัจจุบัน ช่วยลดปริมาณอุปสงค์ส่วนเพิ่มที่จำเป็นลง
• การจัดวางพอร์ตการลงทุน: ไดเน็กซ์ทยอยเพิ่มการป้องกันความเสี่ยงอย่างต่อเนื่องเมื่ออัตราดอกเบี้ยสูงขึ้น ผู้บริหารมองว่าอัตราดอกเบี้ยหน้าตั๋วที่ 5.5% และต่ำกว่า มีความน่าสนใจเป็นพิเศษเนื่องจากมีลักษณะอายุเฉลี่ยและความนูนที่มีเสถียรภาพมากกว่า
• หุ้นบุริมสิทธิ: ผู้บริหารกล่าวว่าสัดส่วนหุ้นบุริมสิทธิในโครงสร้างเงินทุนยังมีโอกาสเพิ่มขึ้นเกิน 7% แต่การเสนอขายในอนาคตจะขึ้นอยู่กับต้นทุน การเพิ่มมูลค่า และโอกาสในการลงทุนที่มีอยู่
• ทิศทางอัตราดอกเบี้ย: บริษัทกล่าวว่ากลยุทธ์ของบริษัทได้รับการออกแบบมาเพื่อทำกำไรจากส่วนต่างราคาสินเชื่อเพื่อที่อยู่อาศัยในสถานการณ์อัตราดอกเบี้ยที่หลากหลาย แทนที่จะพึ่งพาทิศทางของอัตราดอกเบี้ยทิศทางใดทิศทางหนึ่ง

ถอดความการประชุมแถลงผลประกอบการฉบับเต็ม

บทถอดเสียงฉบับเต็มของการประชุมทางโทรศัพท์ผลประกอบการ

คำชี้แจงจากฝ่ายบริหาร

Operator

Good day, and welcome to the Dynex Capital, Inc. Third Quarter Earnings Conference Call. Today's conference is being recorded. At this time, I'd like to turn the conference over to Kaitlyn Mauritz, Head of Capital Markets and Investor Relations. Please go ahead.

Kaitlyn Mauritz

Thank you, operator, and thank you to everyone joining us today for Dynex's Third Quarter 2026 Earnings Conference Call. Joining me on today's conference call are Smriti Popenoe, Co-Chief Executive Officer and President; Byron Boston, Chairman and Co-Chief Executive Officer; Mike Sartori, Chief Financial Officer; and T.J. Connelly, Chief Investment Officer.

Before we begin, I'd like to remind you that today's discussion may include forward-looking statements. These statements are based on current expectations, forecasts and assumptions and are subject to risks, uncertainties and other factors that could cause actual results to differ materially. For additional information regarding these risks and factors, please refer to our filings with the SEC available on the Investors section of our website as well as on the SEC's website.

Dynex undertakes no obligation to update or revise any forward-looking statements. Our earnings press release was issued and filed with the SEC earlier today and is available on the Investors section of our website at dynexcapital.com as well as on the SEC's website. We may also reference our earnings presentation today, which is also available on our Investor page.

And with that, I'll turn the call over to Smriti for opening remarks.

Smriti L. Popenoe

Thank you, Kate. Good morning, everyone, and thank you for joining us today. This was a pivotal quarter in the markets, and Dynex was prepared for this moment. We had a strong liquidity position. Our experienced team maintained our position in our highest conviction trade, and we have retained the capacity to invest as spreads have moved wider.

In the third quarter, book value declined 15.4% to $10.91 per share for a total economic return of negative 11.4% (sic) [ 11.5% ]. Our portfolio operated as designed. The liquidity position absorbed the book value movement. Our financing remained deep and reliable and our hedging limited our downside as 10-year Treasury Yields rose above 5%. During the quarter, we took another step in the evolution of our capital structure with the issuance of our Series D preferred stock, our first preferred equity offering in more than 6 years. The transaction added an accretive source of permanent capital and further diversified our capital base.

The quarter's results reflect higher volatility and higher yields in the overall bond markets and wider spreads in the Agency MBS markets. These moves were driven by fundamental shifts in the macroeconomic system. The U.S. economy is in the midst of a significant positive investment cycle driven by AI and the infrastructure required to support it. During the third quarter, we experienced a simultaneous and significant risk and price rerating of several crucial factors related to AI investment.

The significant demand for capital in this sector, the price of that capital, the impact of that capital investment on the price of commodities and the follow-on impact of all of this on inflation are all higher, pushing yields up. At the same time, geopolitical uncertainty and rising government borrowing needs, both in the U.S. and globally, have reinforced upward pressure on yields. Against this backdrop, Agency MBS have actually been a sideshow, not the main act. And while flows were choppy and illiquid at times, the sector has not been the origin of market stress. In fact, the fundamentals and technicals behind Agency MBS are highly constructive. After the significant and rapid repricing, we find ourselves once again facing a highly attractive investment opportunity.

Absolute real yields are compelling. At current levels, mortgage spreads are still historically wide and provide significant compensation at a time when the value of the borrower's prepayment option is substantially lower. Refinancing incentive is extremely limited and much of the mortgage universe has already extended substantially. As a result, the duration and expected cash flows of these assets are more stable, and most importantly, the convexity profile of the mortgage market has improved dramatically. As a result, and as long-term investors in Agency MBS, we are being paid a wide spread while assuming considerably less cash flow uncertainty than is typically associated with mortgages. This combination underpins our confidence around strong forward returns.

Yes, the cost of capital has temporarily adjusted higher, but prospective returns have adjusted materially higher. This is an important moment for Dynex and our shareholders. We have been building a more scaled and resilient Dynex with the liquidity and flexibility to navigate market and operating shocks. As we grow, our investment platform becomes increasingly durable with greater capacity to deploy capital and opportunistically capture attractive returns.

I'd now like to turn the call over to Mike and T.J. to discuss our financial results and portfolio positioning in greater detail.

Michael Sartori

Thank you, Smriti. I'll now review our financial results for the third quarter. We reported book value per common share of $10.91 at September 30 compared with $12.90 at June 30. Including the $0.51 common dividend, total economic return was negative 11.5%. Comprehensive loss was $1.49 per common share compared with comprehensive income of $0.80 per share in the second quarter. We raised $384 million of accretive capital during the third quarter, net of fees, including $268 million of common equity through our ATM program while maintaining an opportunistic approach to capital raising.

We also completed our first preferred equity offering since 2020, raising $116 million of net proceeds in late September. The transaction was an important step in creating strategic flexibility in our capital structure. The offering increased preferred equity as a percentage of our capital stack from approximately 3% to 7% of our total equity base, diversifying our sources of capital as we continue to grow the business. The transaction was accretive and priced at an attractive cost of capital, particularly relative to the current yield on our common equity.

Importantly, funding markets remain deep and supportive. We maintained more than $24 billion of repo financing across over 25 counterparties. The breadth of our counterparty relationships, combined with the high quality and liquidity of our agency collateral provides substantial capacity to support additional investment opportunities. Asset yields increased from 4.9% to 5%, outpacing the modest increase in our effective financing cost from 3.79% to 3.82%.

As a result, our economic net interest spread remained stable at approximately 1.18%. Liquidity remains a key strength with $1.3 billion of cash and unencumbered securities at quarter end, representing over 42% of total equity, down approximately 9% from the prior quarter. As designed, liquidity levels have buffered the book value declines. Maintaining ample liquidity remains a core element of our risk management framework and allows us to capitalize on market opportunities as they arise. We ended the quarter with a $3.1 billion equity capital base, an increase of 24% year-to-date.

Turning to operating expenses. G&A for the third quarter was 2% of total equity on an annualized basis, consistent with the run rate we discussed last quarter. As we have grown, we have continued to generate meaningful operating leverage across the platform. G&A expenses have declined as a percentage of equity over time and was approximately 40% lower than the same period in 2025. As we execute on our business plan, we have continued to invest in talent, technology and infrastructure to support a larger, more scalable platform. We remain committed to expense discipline and focused on further enhancing operating efficiency as we scale.

With that, I'll turn it over to T.J. to discuss our portfolio and outlook in greater detail.

Terrence Connelly

Thanks, Mike. Current coupon mortgage yields moved nearly 100 basis points higher in the third quarter. The 10-year Treasury rose over 80 basis points, eclipsing 5% for the first time since 2007, and mortgage spreads widened nearly 20 basis points, much of it in the last 2 weeks of September. At quarter end, current coupon spreads to 7-year swaps were around 162 basis points, near the wide end of this year's range. We managed our portfolio actively and with discipline. As rates rose, we added to our hedges in lockstep. As a result, we limited the impact of the rate move.

Spread widening drove over 80% of the return in the quarter and our liquidity continued to do its job. It covered margin in the normal course of business, and we held our exposures and their return potential. We did not delever. As I turn to market dynamics, this quarter, I'm excited to announce the release of the Dynex Angle to accompany our earnings presentation. It gives you a deeper look into our market and macro perspectives. You can find it on our website under the Dynex Perspective heading.

I want to focus on what drove spreads wider, how orderly it was and why it leaves us with one of the most compelling risk profiles we've seen in this market. Realized rate volatility increased and implied volatilities in interest rate options markets jumped even more, causing cheapening of mortgages as expected. Floaters with interest rate caps were particularly exposed. Hedge funds were then forced to quickly exit or hedge their floater positions by selling Agency MBS. We also saw convexity selling from servicers and other mortgage hedgers as yields rose rapidly and durations extended.

Outside of the activity I just mentioned, the market functioned in an orderly fashion. Money managers continued buying Agency MBS into the wider spread, and that has continued into October. The GSEs also operated as we expected. As I said in July, they act as value-sensitive buyers, and there was anecdotal evidence of GSE buying in late September. They added just enough demand when spreads were wide to support a healthy mortgage market.

Funding markets remain very robust, and these conditions have continued in October with strong demand from money market funds intermediated well by the dealer community. Mortgage repo spreads to SOFR held near the tight of the last several years in the 12 to 15 basis point range. Moreover, for the first time in many years, quarter end came and went without any noticeable pressure. Looking ahead, we also see further structural support for financing from the expansion of central clearing and sponsored repo, potential changes to bank capital requirements and the Fed's standing liquidity facilities, all of which should contribute to the depth and resilience of funding markets over time.

Smriti described the improved mortgage convexity profile as an important part of today's opportunity. Let me put some numbers behind it and explain what it means for Dynex's portfolio today. Only about 1% of the outstanding mortgage universe has more than a 25 basis point refinance incentive and the vast majority of agency mortgages trade below par. The convexity selling that added to the push higher in yields has largely run its course, which means there's very little extension left even if rates move higher still.

You can see this in the interest rate sensitivity table at the top of Page 23 of our presentation, where the modeled impact of rate movements on book value is now much more limited. An instantaneous 50 basis point rise in rates as of September 30 is modeled to reduce common equity by only about 0.5% compared with 3.5% at June 30. That means our shareholders have the potential to capture the attractive carry I mentioned without relying on a particular path for interest rates.

Given this setup, we deliberately let our spread exposure grow. Our portfolio spread sensitivity increased from about 4.6% of common equity per 10 basis points of spread at June 30 to about 5.7% at quarter end. What this means is, if spreads tightened 20 basis points, which is roughly a retracement of the widening in the third quarter, you see the model shows approximately 11.5% of gains to book value, all else equal. Leverage ended the quarter at 9.0x shareholders' equity with the increase largely reflecting the decline in book value due to the spread widening and convexity. Our leverage reflects the opportunity set, the risk characteristics of the portfolio and the liquidity we want to maintain.

Our confidence in our current positioning is rooted in the market backdrop. The yield spread available, favorable supply and demand technicals, the convexity profile driven by more stable mortgage duration and a favorable financing environment. What matters is not simply that modeled ROEs are attractive. With more stable durations and less hedge rebalancing, the portfolio is positioned to retain more of the carry embedded in those returns. We believe we are being well compensated for the risk we are taking, which underscores our willingness to operate at the current level of leverage. We maintained our positions through the repricing, kept our financing and put new capital to work at wider spreads.

Today, we are earning attractive spreads on mortgages with more stable durations amid a funding market that remained resilient through the volatility. That is the opportunity in front of us, and we are positioned to capture it. At today's wider spreads, static ROEs on new investments hedged with interest rate swaps are around 20% in lower coupons and 25% in current coupons. And as you see in the interest rate sensitivity, a 10 bps spread tightening adds almost 6% of incremental return.

I'll now turn the call back over to Smriti for some closing remarks.

Smriti L. Popenoe

Thank you, T.J. Periodic repricing is a normal occurrence in the markets. And the mark-to-market impact of the third quarter's repricing is already reflected in our assets, our balance sheet, our book value and our share price. At this point, our focus, and we believe investors' focus should be on the 20-plus percent ROEs that the portfolio could generate in the future. We have a strong conviction in our ability to earn these forward returns. We continue to manage Dynex with a long-term shareholder perspective. That means staying disciplined through significant periods of repricing, managing risk while holding the positions where we have the greatest conviction and preserving our ability to invest as opportunities emerge.

Our focus extends beyond any single quarter. We are building Dynex to compound value across cycles with greater scale, a flexible capital structure and the capacity to invest when shareholders are well compensated for risk. As always, we remain invested alongside you in what we believe is a powerful long-term opportunity for our shareholders.

With that, I'd like to open up the call to questions.

Operator

[Operator Instructions] We'll take our first question from Doug Harter with U.S. Bancorp.

ช่วงถาม-ตอบ

Douglas Harter

T.J., I appreciate what you just said about kind of the risk of the market today and over kind of the longer term. Can you just talk about the risk positioning and the comfort with the ability to let leverage drift higher if markets continue to kind of be weak and spreads wider in the short term?

Terrence Connelly

Yes. Doug, One of the most important factors in the market that I discussed a little bit, maybe need to even further emphasize is the favorable financing environment. Over the last 2 to 3 years, we've seen money market balances grow by some 60%. $8 trillion sits in money market funds. And at the end of the day, that drives a lot of our ability to carry positions through volatility in markets. It's really notable that while realized volatility jumped in the third quarter, the big jump was in implied volatility.

There's a lot of macro risk priced into these mortgages. And we are able to continue to hold the financing through these periods of volatility. So at the end of the day, it's all about the financing market for us, and that has been an extremely healthy part of the market and gives us a lot of confidence in carrying the leverage that we're currently carrying.

Douglas Harter

Great. And then you talked about the ability to generate 20% to 20-plus percent returns. Can you just kind of walk through the math as to kind of how you're getting to that level of returns?

Terrence Connelly

Yes. That's a levered ROE that is based on our models and where we think -- where spreads are today and at the end of the third quarter, in particular. So we're getting on 5 through 6s around 20% static ROE and some of the higher coupons and 6.5s, for instance, are actually around 25% ROE based on the model. Leverage works out to be around 10% on that static ROE model.

Operator

[Operator Instructions] We'll move next to Bose George with KBW.

Bose George

Can we get an update on book value quarter to date?

Terrence Connelly

Yes, of course. As of Friday, our estimated book value net of the accrued dividend was $10.95.

Bose George

Okay. Great. And then on expenses, you noted the 2%. I mean just when I do the math, it looks like it was down a decent amount, like closer to maybe 1.5%. So is there anything unusual or seasonal in that? And then just like thinking about it going forward, I guess, 2% should be kind of the number we use.

Michael Sartori

Yes. Bose, that's right. As we indicated, G&A is down about 40% since last year at this time. We remain focused on that 2% ratio. As we continue to grow, we will see operating leverage benefits from this. At the end of the day, it's about performance and returns and what creates shareholder value. And what -- that decrease really comes from compensation expense. It's aligned with the company's performance and shareholder value.

Bose George

Okay.

Terrence Connelly

Bose, one of my colleagues mentioned I may have cut out just slightly, as I said, the book value, $10.95.

Bose George

$10.95, okay.

Operator

We go next to Marissa Lobo with UBS.

Ameeta Lobo Nelson

You characterized the current environment as a compelling forward opportunity. Could you speak to the demand side of that view, just where you see bank demand, GSE buying, foreign demand? And who do you see as the marginal buyer of Agency MBS at these spread levels?

Terrence Connelly

Importantly, as I mentioned, GSEs were very likely in the market in September. They have been a backstop bid to mortgages. And I think that's something we've been talking about for several quarters now ever since the January tweet and their active purchases in the first quarter. So that remains an important backdrop. What's really notable here is given the supply dynamics at these rate levels, you don't actually need all that much marginal demand. We have seen foreign buyers in Europe.

For instance, there was the news of the Norges Bank and the possibility that they'd be selling global sovereigns to buy agency mortgages, and they specifically mentioned U.S. agency mortgages. That's notable. I think that's a long-term flow that will underpin and really shows the quality of these assets, right? It's a yield that you can -- at these real yield levels is very compelling to the global investors. So I do expect foreign investors to be active in the -- on the demand side of the equation. Of course, the mortgage REIT community remains increasingly important to this.

Smriti L. Popenoe

Yes. Marissa, I'd also jump in and just talk about the view on forward returns being really solid here in the low to mid-20s. One of the other pieces to keep in mind is how we've been able to preserve our earnings power even when book value has declined. So one of the key decisions that we made over the quarter was the decision to not delever. And by doing so, we've kept so much of the earnings power, in fact, all of it. And that's another big reason for having the level of confidence about the compelling opportunity and the forward returns.

Ameeta Lobo Nelson

That's helpful. And then just shifting to the capital structure. You talked about the benefits of diversifying with the preferred. How should we think about the preferred mix from here? Is there a level of preferred within the capital structure that you're thinking is appropriate?

Smriti L. Popenoe

Yes. I mean we're really happy with the issuance and the timing of that particular deal. It was really massively accretive to the common shareholder. We think of ourselves as having a significant amount of capital structure flexibility here at this point. The total, I think Mike mentioned preferred in the stack is about 7%. That number has room to go higher.

As with all of our capital structure decisions, right, and capital raising decisions, it's opportunistic. It's thinking through the things like how accretive it is to shareholders, where we can put the capital to work relative to the cost of that capital. But suffice to say, I think we're really happy with, one, reentering the market, the reception that we got and then the possibility that this could be something that we could continue to use in terms of a source of capital in the future.

Operator

We'll go next to Merrill Ross with Compass Point.

Merrill Ross

If we look at Slide 24 and look at the changes to your hedge position, it would be helpful if you described when you changed the positioning and what you were thinking at the time? And are you satisfied with where you were positioned at the end of the period?

Terrence Connelly

Yes, thanks for the question. The hedges were put -- we added to the hedge book pretty much in lockstep as rates were moving. We came into the quarter with a very strategic mindset in terms of continuing to add to hedges as rates moved higher. So it was progressive over the course of the quarter.

As we broke through 5% there was a clear realization of the new profile that mortgages would offer in terms of the convexity that we discussed. So we had a very clear game plan to put in place there and continue to add the hedges. So you'll see a modest increase in some of the longer hedges, in particular, on Page 24, as you mentioned, in the -- on the swap side of the book as well as on the treasury futures side.

Merrill Ross

Yes. Okay. And then maybe, as Freddie Mac says that [indiscernible] mortgage rate is going to be 7.5% through 2027. So wondering where in that market, you would see more value. Clearly -- it seems not in TBAs, but you prefer a discounted lower coupon pools over the current coupon?

Terrence Connelly

Yes. Let me first make a comment on the notion of a static mortgage rate over time is not something we spend a lot of time thinking about. We're thinking about a range of scenarios and preparing for a significant number of outcomes, both higher and lower than that 7.5% mortgage rate that they mentioned.

So yes, our portfolio is built accordingly, right? We're built for -- to generate solid cash flows throughout a range of environments. I do think that a lot of these -- that convexity profile we discussed and the stable duration profile that's so important to earning these ROEs is most compelling in coupons, say, 5.5% and lower at this point.

Operator

We go next to Trevor Cranston with Citizens JMP.

Trevor Cranston

It looks like your overall rate exposure came down this quarter. I'm curious if you guys could just kind of share your macro thoughts on how you think about the risk of rates moving significantly higher or lower from where they are today? And generally, how you think MBS would perform if there was to be a significant move in rates either way from here?

Terrence Connelly

Yes. Trevor, I'd say overall, the most important thing to keep in mind is what's priced into swaptions markets, into the interest rate options markets that I mentioned and those implied volatilities being higher as high as they are, which is, again, notably higher than realized volatility. I think that's one of the most important constructs that will drive our ability to earn these ROEs and the returns that are available. So I do think the convexity profile in a higher rate environment for the overall mortgage market is very stable. Those durations just don't extend all that more.

Overall, you think from a very high-level macro perspective, overall housing turnover is around 4.5% per year. If you look at existing home sales, which are very depressed, and you could make a case are making a bottom formation over the last several years. So at these turnover rates that are implied in the mortgages, you've got very low turnover in the mortgages that drives the mortgage cash flows, creating that very stable cash flow. And then you have this wide spread that's there because the implied vols are so high on the macro side. It makes it very compelling for us to earn these ROEs that we're talking about.

So broadly, it's not an environment where taking significant macro exposures makes a ton of sense. We're being paid for it in the spread. So by design, very strategically, that's where our focus is on the spread and earning those ROEs without more or less agnostic to interest rate direction.

Smriti L. Popenoe

Yes. Trevor, I would say like the way the mortgage market is priced today, the level of extension on the margin for the next 25, 50, 100 basis points is very limited because we've extended 90% already, right? So any future extension from rates rising is limited. So if you're concerned about, again, rates up, those types of scenarios, what T.J. is mentioning is that it's all about the price of options in the implied vol markets and those -- that's -- the mortgage market today is already pricing significant risk premium in implied vols. So we feel like we're getting compensated for that incremental risk.

Trevor Cranston

Yes. Okay. That makes sense. And when you guys were mentioning the incremental ROEs kind of in the 20-plus percent range, I think you said you had calculated that based on the 10x leverage number, which is a little bit higher than where you sit today. So I was just wondering if you could clarify if you think the existing portfolio as it sits today provides good dividend coverage. And just kind of curious overall how you're thinking about the dividend policy in light of all the changes in the third quarter perhaps?

Smriti L. Popenoe

I think you've got to go back to the idea of earnings power in the portfolio, Trevor, not having delevered in the quarter. That just means that we've taken the mark in book value, but the earnings power of the portfolio is preserved. And over the past several quarters, we've been pretty confident about our ability to cover the dividend even before we got to this point.

So from here on, I don't think that's really changed that much because we did not significantly adjust the portfolio, definitely not on the asset side, right? So that piece of it, I think, is relatively well preserved. In terms of the marginal ROEs, we use a range of numbers. T.J., you can clarify on the low 20s, et cetera, the numbers that we quoted.

Terrence Connelly

Right. Yes, there's significant variables involved there, Trevor, but the numbers I was quoting were based on a 10 leverage number. And as you point out, we were sitting at 9 at quarter end. I will note, we have net purchased assets since quarter end as well.

Operator

We have no further questions. I'd like to turn the floor back to Smriti Popenoe for any additional or closing remarks.

Smriti L. Popenoe

Thanks, everyone, for joining us today, and we look forward to updating you on our performance in January.

Operator

This concludes today's conference. We thank you for your participation. You may disconnect at this time.