雷德伍德信托 (RWT) 2026年第三季度业绩电话会：C类公司转型与股息重置

Redwood Trust（NYSE: RWT）公布的2026年第三季度亏损扩大，主因遗留资产处置的损失抵消了其核心抵押贷款银行业务的持续盈利。该公司还宣布，计划自2027纳税年度起终止其房地产投资信托基金（REIT）地位，转型为应税C型公司运营。

核心要点

• 归属于普通股股东的GAAP净亏损扩大至4700万美元，即每股亏损0.39美元，而2026年第二季度净亏损为300万美元，即每股亏损0.03美元。
• 每股账面价值环比下降4.9%至6.56美元，反映了遗留资产损失以及0.18美元的股息分红，但股票回购带来的每股约0.25美元的增值部分抵消了上述影响。
• Redwood计划于2027年从REIT转型为应税C型公司。管理层预计初始递延所得税资产为8500万至9000万美元，将为2026年第四季度GAAP账面价值增加约每股0.73至0.77美元。
• 董事会预计自第四季度派息起，将季度普通股股息重新调整为每股0.02美元，而目前为0.18美元。按9月30日的股份总数计算，在拟进行的要约交换之前，这将使每年的普通股股息支出减少约7500万美元。
• 抵押贷款银行业务创造了2700万美元的GAAP净利润，非GAAP年化资本回报率达到26%。这也标志着该业务的回报率连续第九个季度超过20%。
• Redwood发放了66亿美元贷款，并分销了88亿美元，从而降低了库存，并将非受限现金增至2.01亿美元。

核心财务数据

业务与经营表现

Sequoia

Sequoia锁定交易量从第二季度的56亿美元降至51亿美元，但销售收益率从92个基点增至113个基点。管理层将这一改善归功于积极的贷款分销、审慎的定价以及对冲操作。

该平台分销了超过71亿美元的贷款，其中包括通过10笔证券化交易分销的58亿美元，以及通过与Castlelake设立的合资企业完成的首笔交易。整贷销售总额为7亿美元，包括与两家新银行合作伙伴的交易。

Sequoia创造了3200万美元的GAAP净利润，环比基本持平，年化资本回报率达到41%。第三季度超过30%的锁定规模来自存管机构。

Aspire

Aspire受利差走阔和利率波动的冲击更大。锁定交易量从21亿美元降至11亿美元，而毛利率与第二季度的101个基点相比基本持平。

该业务录得500万美元的GAAP净亏损，而上一季度为700万美元净利润。Aspire分销了13亿美元的贷款，包括通过两笔SPIRE证券化交易分销的9.04亿美元，以及通过整贷销售分销的3.86亿美元。

其未分销净储备规模低于4亿美元。管理层表示，新成立的Aspire合资企业在实现规模化后将拥有约80亿美元的收购能力。

CoreVest

CoreVest投放了4.06亿美元贷款，与第二季度大致持平。过桥贷款与住宅周转贷款合计投放量增长约35%，占总放款量的近四分之三，抵消了定期贷款需求的疲软。

CoreVest分销了4.27亿美元贷款，其中包括通过其与CPP Investments设立的合资企业分销的2.44亿美元。按GAAP计算，该业务大致处于盈亏平衡。剔除收购相关费用及相应税收影响后，EAD为160万美元，相当于20%的年化资本回报率。

信用表现与遗留资产

当前新增贷款的信用表现保持稳定。Sequoia和Aspire逾期60天以上的违约率均低于20个基点，而CoreVest证券化过桥贷款资产组合中的严重逾期率及止赎房产（REO）比例保持在5%以下。

遗留投资仍是主要的盈利拖累项。非应计贷款从年末的2.91亿美元降至2.05亿美元。自2025年第三季度以来，Redwood已将遗留资产分部的风险敞口缩减了近40%，仅在2026年第三季度就缩减了20%以上。

管理层正在评估对剩余遗留贷款中近70%的收购要约。2026年，每季度的持有与处置成本平均为每股0.12至0.14美元，但随着资产及相关债务得到处置，公司预计这些成本将在第四季度以及2027年第一季度开始下降。

管理层业绩指引

Redwood预计转型为C型公司将于2027纳税年度生效，且无需股东表决。公司计划利用约3.04亿美元的联邦净经营亏损结转以及超过10亿美元的州级NOL，来抵扣其运营业务未来的应税收入。

管理层预计将于2026年第四季度确认8500万至9000万美元的初始递延所得税资产。根据9月30日的已发行流通股数量计算，这将使GAAP每股账面价值增加约0.73至0.77美元，具体取决于最终的税务与会计认定。

以每季度3000万美元的应税收入为例，管理层表示，在适用限制条件下，未来数年利用NOL可使每季税负减少约每股0.05至0.06美元。

董事会还批准了1比5的反向拆股方案，计划于2027年1月4日实施。该合股将按比例减少已发行股数并增加每股金额，而总股东权益保持不变。

风险与关注重点

• 随着Redwood加快遗留资产的出售、止赎及其他处置进程，可能需要计提额外的减值准备并产生交易成本。
• 分摊至遗留资产组合的部分债务属于有追索权债务，从而限制了亏损被完全隔离的程度。
• 进入第四季度后，抵押贷款利差仍维持在近期的高位区间，而利率波动继续对贷款利润率和分销周期造成压力。
• Aspire的业绩对非合格抵押贷款（non-QM）整贷及证券化利差依然高度敏感，该利差在第三季度期间及之后均显著走阔。
• Redwood的部分递延所得税资产将计提估值准备，原因是部分州级NOL可能在被充分利用之前到期。
• Redwood采纳了一项NOL保护计划，以管理随股东结构变动可能依据美国税法第382条产生的限制。

分析师问答精选

管理层表示，降低股息所保留的资本将配置于抵押贷款银行业务的扩张、债务回购以及更广泛的企业资本结构优化。据该公司称，仅每年节省的股息支出就可在未来一年使杠杆率降低约0.5倍。

Redwood计划开展优先股置换普通股的要约交换，初始规模为5000万美元，并可根据需求扩大规模。该优先股目前预计股息率为11%。管理层表示，视最终的交换溢价而定，此次要约交换连同已完成的股票回购，有望抵消9月份发行可转债带来的潜在摊薄影响。

公司进入第四季度时，Sequoia与Aspire合计未分销储备规模约为8亿美元，管理层称其为近年来最轻的仓位之一。在抵押贷款利率和利差趋于稳定之前，Redwood打算保持谨慎，同时保留快速扩大幅度的能力。

管理层预计C型公司架构不会对公司的“发起-分销”（originate-to-distribute）策略产生重大改变。Redwood计划继续通过整贷销售、证券化及合资企业来扩大业务规模，同时控制对股东资本的需求。

业绩电话会完整纪要

---

完整财报电话会议逐字稿

管理层陈述

Operator

Greetings, and welcome to the Redwood Trust Third Quarter 2026 Financial Results Conference Call. Today's conference is being recorded.

I will now turn the call over to Anh Nguyen, Head of Investor Relations. Please go ahead, ma'am.

Anh Nguyen

Thank you for joining us today for Redwood's Third Quarter 2026 Earnings Conference Call. With me on today's call are Chris Abate, Chief Executive Officer; Dash Robinson, President; and Brooke Carillo, Chief Financial Officer.

Before we begin today, I want to remind you that certain statements made during management's presentation with respect to future financial and business performance may constitute forward-looking statements. Forward-looking statements are based on current expectations, forecasts and assumptions, which includes risks and uncertainties that can cause actual results to differ materially. We encourage you to read the company's annual report on Form 10-K and quarterly report on Form 10-Q, which provides the description of some of the factors that could have a material impact on the company's performance and could cause actual results to differ from those that may be expressed in forward-looking statements.

On this call, we may also refer to both GAAP and non-GAAP financial measures. The non-GAAP financial measures provided should not be utilized in isolation or considered as a substitute for measures of financial performance prepared in accordance with GAAP. Reconciliation between GAAP and non-GAAP financial measures are provided in our third quarter Redwood review, which is available on our website, redwoodtrust.com.

Also note that the contents of today's conference call may contain time-sensitive information that are accurate only as of today. We do not intend and undertake no obligation to update this information to reflect subsequent events or circumstances. Finally, today's call is being recorded. It will be available on our website later today.

With that, I'll turn the call over to Chris for opening remarks.

Christopher Abate

Good afternoon, everyone. Today, we announced several important developments for Redwood. I'll cover the key points before Dash and Brooke share their prepared remarks, and then we'll take your questions.

After 32 years as a real estate investment trust, Redwood plans to transition to a taxable C-corporation beginning in 2027. We believe this structure better aligns with our operating business today and will strengthen our ability to build long-term shareholder value. We expect the transition to generate an initial GAAP book value benefit of approximately $0.73 to $0.77 per share in the fourth quarter of 2026, which translates to an increase of approximately 11% over our September 30 book value per share of $6.56.

We realize that this transition has been anticipated for some time by many of our shareholders. Redwood has evolved considerably since its founding in 1994, with our transformation into an operating platform accelerating in recent years. Today, we serve as an embedded partner to hundreds of loan originators and real estate investors, including 70 of the top 100 U.S. banks. We support these institutions with deep underwriting and asset management expertise and leading securitization and whole loan distribution programs.

Our recent market share gains in mortgage banking indicate that we have only begun to monetize the full potential of these relationships. However, the earnings strength of our franchise has not translated into sustained book value growth. One reason is that we've distributed substantial capital through dividends, limiting our ability to allocate capital and compete effectively with our mortgage banking peers.

By de-REITing, beginning with the 2027 tax year, we want to share 3 near-term benefits. First, we expect to unlock significant tangible value from our existing tax assets through previously unusable federal and state NOL carryforwards. In this regard, we are unique as all REITs that come to mind withstand to lose their tax shelters by not complying with the REIT rules. our shareholders will immediately realize the opposite.

Second, we will have greater flexibility to retain earnings and allocate capital. As we disclosed earlier, we expect the Board to reset the quarterly common dividend to $0.02 per share beginning in the fourth quarter. This would bring our payout ratio in line with banks, asset managers and other C-corp mortgage businesses with comparable operating models.

We recognize that the de-REIT transition, while in the long-term interest of the company, entails a lower common dividend to fully leverage the compounding effect of earnings and associated benefits. For our shareholders who value Redwood's historically high dividend yield, particularly our retail investors, today, we announced plans for a preferred for common exchange offer that we expect to come at a significant premium. This new series of preferred stock will carry a high dividend yield aligned with income-oriented objectives. I would strongly encourage interested shareholders to fully explore it ahead of the upcoming launch.

For our common shareholders, this exchange can be thought of akin to a share buyback and will further enhance the contribution from our newly minted tax assets. Based on our projected size for this exchange, we would expect any remaining potential dilution associated with our September convertible debt issuance to be neutralized, especially when factoring in our third quarter common share buybacks.

Third and finally, the C-corp transition creates a path towards a structurally lower cost of capital. Our corporate debt costs reflect constraints associated with mortgage REITs, particularly the requirement to pay out substantially all earnings rather than allocate that capital for other uses. As a consequence, non-REIT mortgage peers enjoy broader access to the debt markets, including rated debt with covenants better suited for operating businesses.

With more than 80% of our unsecured debt callable by the end of 2028, the C-corp transition will allow us to not only pursue more attractive debt, but also opportunistically replace higher cost obligations with lower cost financing. In step with that unlock, today, we announced the upsize of our corporate secured financing facility with CPP Investments and extended its maturity to March of 2030, adding valuable funding flexibility.

As we head towards year-end, we are working very hard to resolve as much of the legacy book as possible in the fourth quarter. While Dash and Brooke will cover legacy, I want to be sure our shareholders understand that the legacy book remains very much a melting ice cube. Charges went up this quarter because liquidations went up and the outstanding balance went down.

I'd also note that our legacy book is marked dramatically lower than many who utilize CECL reserving for similar assets. This distinction is a big reason why we can pursue liquidations at this time rather than traditional loan modifications.

In closing, through all today's announcements, our mission remains the same to make quality housing, whether rented or owned, accessible to all American households. As we discussed in this quarter's shareholder letter, that mission is taking on new significance as we transition towards becoming an agentic enterprise. With our expectation that markets begin to normalize after the midterm elections, we're excited to challenge the boundaries between agency and nonagency mortgage lending in the quarters ahead.

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

Dashiell Robinson

Thank you, Chris. Our third quarter operating results reflected proactive risk management and the depth of distribution and product offerings that make our platform resilient even in times of market volatility.

We substantially increased activity with our joint venture partners, increased loan purchase and sale activity with banks and made important strides in the wind down of our legacy portfolio. The result was our ninth consecutive quarter of mortgage banking returns north of 20% despite sharply higher benchmark rates and spread widening that has persisted into the fourth quarter.

Across our platforms, we produced $6.6 billion of loans, down from $8 billion in the second quarter, while distributing $8.8 billion through securitizations, whole loan sales and joint ventures. We meaningfully reduced net inventory, freeing up liquidity for fresh working capital and accretive deployment into our corporate capital structure.

At Sequoia, lock volume was down modestly on the quarter to $5.1 billion from $5.6 billion in Q2. Gain on sale margins rose to 113 basis points, up from 92 basis points in the second quarter, resilience that reflected a proactive distribution posture, appropriate conservatism with our rate sheet and continued accretive hedging activities, particularly as 10-year treasury yields moved sharply higher in the quarter end and the mortgage basis widened.

Sequoia distributed over $7.1 billion of loans during the quarter, including $5.8 billion across 10 securitizations and the first out of the Castlelake joint venture. Full loan sales totaled $700 million, including transactions with 2 new bank partners. The majority of Sequoia's dispositions on the quarter occurred prior to the uptick in rates in September, helping to support gross margins.

Just 1 quarter after coming online, the Castlelake partnership is reinforcing how joint ventures can expand our production capacity while more efficiently deploying our shareholders' capital and increasing mortgage banking returns. Through these partnerships, Redwood earns upfront economics on loan sales, recurring management fees and a share of gain on sale profits, providing growing revenue streams without the added working capital requirements.

Amidst a more cautious operating approach, Sequoia continued to expand its product development and bring a deeper suite of solutions to our seller base. Traditional 30-year fixed rate production represented less than 70% of locks on the quarter, a nod to strategic progress with our medical professional loan offering for which we issued our second securitization during the third quarter as well as HELOCs, which we formally rolled out in May.

When rates and spreads are volatile, these products support margins, both through our unique sourcing capabilities and increased market demand for nonfixed rate product when convexity becomes more challenging to manage.

Transaction activity with our bank network also continued to trend higher with over 30% of our Q3 locks coming from depositories. While rates have rallied close to 40 basis points off their recent highs, spreads remain at or near recent wides, an opportunity for the business heading into year-end, a period of time during which many competitors have historically retrenched.

Market conditions more acutely impacted Aspire's third quarter performance. Spread widening across the capital structure and rate volatility, particularly at the short end of the yield curve, led to roughly flat gross margins, down from 101 basis points in the second quarter. Lock volume on the quarter totaled $1.1 billion of loans, down from $2.1 billion in the second quarter and the result of deliberate calibration of activity as distribution time lines extended with both originators and aggregators challenged to efficiently move inventory.

Low purchases represented roughly 88% of Aspire's volume, and our active seller network grew to over 165 sellers. Aspire distributed $1.3 billion of loans, including $904 million across 2 SPIRE securitizations and $386 million of whole loan sales.

The non-QM market remains in growth mode and Aspire's addressable market remains significant. With a net undistributed pipeline of below $400 million, the business is well positioned to capitalize as demand from insurance capital ramps back up, and we prepare for initial fundings under Aspire's newly formed joint venture, which at scale has overall purchasing power of approximately $8 billion.

CoreVest funded $406 million of loans in the third quarter, roughly flat with the second quarter. Higher rates once again weighed on the term loan pipeline, but combined bridge and RTL fundings increased roughly 35% and accounted for nearly 3/4 of volume.

CoreVest distributed $427 million of loans through joint venture transfers, securitization and whole loan sales, including $244 million through our joint venture with CPP Investments. We also launched a new pricing engine and redesigned website to support lead generation and began transitioning servicing of certain bridge loans to our new servicing partner, an effort we expect to reduce one-the-run servicing costs through time.

As production activity has increased across our platforms, credit performance has remained stable and delinquencies remain low. 60-day plus delinquencies in Sequoia and Aspire are each below 20 basis points, reflective of continued strong underwriting discipline as those businesses continue to expand.

Importantly, credit performance in recent CoreVest production vintages has also remained healthy with combined serious delinquency and REO below 5% in our securitized bridge portfolios, affirmation of the significant changes in our approach beginning in late 2022 that have delivered several consecutive years of strong credit cohorts.

Total payoffs across all bridge portfolios have totaled $475 million in unpaid principal balance since June 30, a level of turnover in line with expectations. This included $160 million of payoffs and dispositions of legacy and other seasoned bridge loans with another $70 million in UPB expected to be resolved by mid-Q4 or already under contract.

As credit from core production activities remains resilient, the legacy investment segment, which continues to shrink, remains the primary area of focus for loss mitigation and disposition. Since the third quarter of 2025, we have reduced exposure in that segment by close to 40%, prioritizing proactive measures, including note sales, refinances and REO sales. As such, modification activity in the legacy portfolio has decreased sharply in recent quarters as we prioritize expedient recovery, capital reallocation and reduction in interest expense.

As we evaluate offers on close to 70% of the remaining loans in the segment, we maintain our posture of expeditiously reducing this position. While we may encounter additional transaction costs in doing so, the dual economic benefits are powerful, reallocation of capital to our profitable mortgage banking activities and reduction in carrying costs that have averaged $0.12 to $0.14 per share per quarter thus far in 2026.

I'll now turn the call over to Brooke to discuss our financial results.

Brooke Carillo

Thank you, Dash. We reported a GAAP net loss related to common stockholders of $47 million or $0.39 per share compared to $3 million or $0.03 per share in the second quarter. Book value ended the quarter at $6.56, down 4.9% from $6.90 at June 30. The decline reflected losses associated with resolutions of legacy assets and our $0.18 per share dividend, partially offset by approximately $0.25 per share of accretion from common stock repurchases we completed during the quarter.

Our economic return on book value was negative 2.3%. Excluding the $0.39 GAAP loss from legacy investments, which included a $0.07 per share charge to position certain legacy HEI assets for liquidation, book value per share would have been up modestly from the second quarter.

Consolidated non-GAAP EAD was negative $0.02 per share compared to positive $0.15 per share in the second quarter. Core Segments generated $23 million of EAD or $0.18 per share, representing a 12.7% annualized return on equity compared to $0.25 per share last quarter.

Legacy investments reduced consolidated EAD by $0.20 per share compared to $0.11 per share in the second quarter, reflecting carry and resolution costs. Importantly, the underlying nonperforming loan population continues to shrink with nonaccrual loans declining to $205 million at September 30, down from $291 million at year-end. Throughout the year, we have not accrued on the majority of our legacy loans irrespective of payment status.

We completed no complex modifications or rate concessions during the quarter, focusing instead on sales foreclosures and other dispositions. As this portfolio continues to wind down, we expect the associated earnings drag to decline. Our goal remains to resolve as many legacy loans as possible throughout the remainder of the year.

Mortgage banking generated $27 million of GAAP net income compared to $40 million in the second quarter and a 26% annualized non-GAAP return on capital. Sequoia generated $32 million of GAAP net income, essentially unchanged from the second quarter and a 41% annualized return on capital. Margins benefited from our interest-only and servicing hedges as well as joint venture distributions offsetting the impact of lower volume.

Aspire reported a $5 million GAAP loss compared to $7 million of net income in the second quarter. Volume declined 45%, reducing operating leverage as we deliberately moderated acquisitions early in the quarter when loan spreads did not adequately compensate for emerging market risks, a decision which was reinforced by subsequent market volatility.

CoreVest was approximately breakeven on a GAAP basis. Excluding acquisition-related expenses and the related tax effects, EAD was $1.6 million, representing a 20% annualized EAD return on capital and compared to $2.8 million in the second quarter. Fundings remained relatively stable with increased bridge activity offsetting weaker term loan demand.

Redwood Investments generated a $4 million GAAP loss versus $1 million of income in Q2 as higher rates and wider spreads pressured retained investment valuations. This was partially offset by $4.9 million of positive RWT Horizons fair value changes.

Net interest income rose to $19 million from $10 million, primarily reflecting seasoned IO and servicing investments transferred from mortgage banking as our pipeline hedging benefit diminished. Segment EAV before allocated corporate costs increased to $13.6 million from $10 million.

G&A was essentially unchanged at $30.5 million. Year-to-date total expenses represented 69 basis points of production versus 88 basis points for the full year 2025, a 22% improvement.

Turning to liquidity and capital. We ended September with $201 million of unrestricted cash, up from $192 million in June. We raised $205 million through our 7% convertible offering, retired approximately $125 million of 2027 converts, repurchased $34 million of common stock, deployed $67 million into retained investments and paid $23 million in dividends.

Cash generated through loan distributions and ongoing balance sheet management allowed us to fund these capital actions while increasing our cash balance, a trend that has continued into the fourth quarter. Recourse debt declined approximately $310 million to $4.2 billion as distributions outpaced production. Recourse leverage rose to 5.2x from 5.0x, reflecting lower equity from repurchases, the net loss and dividends.

Turning to the financial implications of today's corporate actions, our planned C-corp conversion will allow us to use approximately $304 million of federal and over $1 billion of state net operating loss carryforwards from our REIT qualified businesses to offset future taxable income generated by our operating businesses where our effective tax rate at the TRS increased to approximately 29% this quarter, up from 21% in the second quarter.

We expect to recognize an initial deferred tax asset of $85 million to $90 million in the fourth quarter of 2026, increasing GAAP book value by approximately $0.73 to $0.77 per share based on September 30 shares outstanding and subject to final tax and accounting determinations. Beyond this initial benefit and an illustrative $30 million of quarterly taxable income, NOL utilization could reduce quarterly taxes by approximately $0.05 to $0.06 per share per quarter over the next several years, subject to applicable limitations.

As Chris mentioned, the Board is realigning our quarterly common dividend policy to $0.02 per share expected to take effect with the declaration of our fourth quarter dividend in December compared to the current $0.18 quarterly dividend that would reduce our annual common dividend payments by approximately $75 million at the September 30 share count and before the effects of the planned exchange offer.

Finally, the Board approved a 1-for-5 reverse split planned for January 4, 2027, to improve share marketability. It proportionately reduces shares outstanding and increases per share amounts without changing total equity. All per share figures discussed today are presplit as the split will be reflected in our fourth quarter metrics.

With that, I'll turn the call back over to the operator for questions.

Operator

[Operator Instructions] Our first question comes from the line of Doug Harter with BTIG.

分析师问答

Douglas Harter

Can you -- with the combination of the cash savings from the dividend and cash tax savings, can you talk about kind of your priorities for capital deployment, whether that be repurchasing debt at a discount in the market, continuing to invest in mortgage banking or share repurchase, kind of how you're thinking about those options?

Christopher Abate

Sure. Doug, I'll start. With respect to the common stock, we've announced an exchange offering that we think really addresses the portion of the shareholder base that prefers a high dividend going forward. I'll let the press release do the talking at this point. We expect to launch a tender next week. But I think that exchange has the potential to be very accretive to the common. We expect to take shares out of the market. And that's going to be our next step with respect to common.

I think with the new dividend policy, it allows us to be much more opportunistic with debt repurchases across the stack. We also noted that the type of debt that we're able to raise and the rates become more attractive, so we'll be focused on our entire capital structure.

And of course, mortgage banking, we're running a mortgage operation here. It needs capital to grow, pushing $6 billion, $8 billion a quarter through the pipes, retaining investments, managing the pipelines. All of those things take capital. But the earnings that we're generating at our core operating subsidiaries remain substantial. So we're very excited about having the opportunity to allocate capital in a more optimized fashion. So we'll have more to discuss in the weeks ahead. But I think for now, we've given the market a lot today to digest.

Operator

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

Crispin Love

Just following up on that last question. I think it addresses it somewhat. But just with the transition to a C-corp, just how do you expect the shareholder base to change? What have you learned from other companies taking similar steps? Just what that could look like over the next few months? I assume that the exchange offering is meant to address that, but what if that offering isn't as successful as intended?

Christopher Abate

Well, we certainly hope it's successful. I think it's well thought out, and it's a great opportunity for those shareholders who value a high dividend to step into what we currently expect to be an 11% coupon. There's a tender, a modified Dutch auction component. So the actual yield to the investor could certainly be higher. But we'll -- again, we'll have more to say about that at the launch next week.

I think for us, we've studied our mortgage banking peers. We were thoughtful in how to reset the dividend, looking at our true competitors based on the business that we're running, not based on the tax structure. And we also have historically had a very high institutional ownership percentage within the common stock.

So our retail ownership has not typically been nearly as high as we've seen in other more passive income-oriented REITs. So we're optimistic that hopefully, many or most of our long-term shareholders understand these changes or even anticipated these changes, and we're going to work hard to make sure that those investors who value a high dividend have a great option to exchange into a preferred.

Brooke Carillo

Can I add one thing too, Crispin, just I think for all the common shareholders, too, this is an atypical C-corp conversion in that we actually think it's very ROE accretive for us. It's actually really unlocking these tax assets that will provide a significant financial benefit for us.

We -- in addition to the retained capital that has a significant compounding effect to be able to reinvest that both in our businesses and potentially delever, that $75 million alone could bring our leverage multiple down by 0.5 turn over the next year and provide significant working capital for our 20% plus mortgage banking entities and the state cash taxes is significant, too. Every $20 million to $25 million is about a 20% ROE accretion for us.

So taken together, we think that this is really exciting for a prospective shareholder base that will get to enjoy in book value growth and ROE accretion from this change, too.

Crispin Love

Great. And Brooke, just on that point, when you think about -- because you've had ROE targets out in the past. So when you think about the conversion or the transition, the ROE uplift, the earnings outlook, can you just discuss kind of what do you think normalized ROEs could be?

Brooke Carillo

Yes. I think it's very similar to what we've been generating on the mortgage banking side. We have 9-plus quarters of 20% plus ROEs. Really, this is -- we will recognize this upfront benefit through book value and over time, see that through kind of the cash savings. And that, again, is really compounding from an ROE effect.

So I think we continue to have the same kind of target returns for our capital that's redeployed back into our core segments, which despite the market volatility on a consolidated margin basis, we really saw Aspire, Sequoia and CoreVest collectively over kind of a 1% margin. So really looking forward to continuing to really target those same returns in our on-the-run business.

Operator

Our next question comes from the line of Marissa Lobo with UBS .

Ameeta Lobo Nelson

I just wanted to talk about mortgage banking with 30-year rates stubbornly above 7%. How are you thinking about the strategy in banking across the different platforms and where you see the best risk-adjusted returns?

Christopher Abate

Yes. We -- Marissa, we'll tag team this one. At a high level, we've been, I'd say, extremely conservative and cautious up to this point in the fourth quarter. While we have seen benchmarks start to stabilize, spreads have continued to widen, I'd say, across the board. And we want to make sure that when we lean back into mortgage banking that we can confidently generate the margins that we've consistently generated for a number of quarters now.

So I think at a high level, we're excited to get the midterms today behind us. We have started to see things stabilize. And the really good news about the platform is just given our market penetration and our access to our seller base, we're able to lean in extremely quickly. I think we've demonstrated this through past cycles. And this one, I think, will be no exception.

So we sort of have the luxury of thinking about timing, thinking about when the right reentry point is to really step on the gas. And for us, I think we're -- we've been very conservative, I'd say, from early September through today, candidly, but we're definitely starting to see the market stabilize. We're starting to see mortgage rates stabilize and come down a bit. So the signs are there. We're looking forward to leaning in.

Dashiell Robinson

One subtext I would add there, Marissa, is just how light our position is, which is the function of 2 things, the conservatism that Chris articulated in terms of just the locks we've taken in the latter part of Q3 and to date in Q4. But also we were reasonably ahead of the curve, particularly in Sequoia about distributing a lot of the risk in the position prior to Labor Day, after which, as you know, there was a significant amount of upward pressure on rates, and there's been fairly consistent upward pressure on spreads since then through today.

So we're currently sitting at about $800 million of undistributed pipeline across Sequoia and Aspire, which is as light as we've been as we can remember. And one of the reasons that's important is when rates move a lot very quickly and spreads widen, current coupon moves very significantly as well. So we saw situations earlier or October and September where high 6%, low 7% current coupon was quickly 7.5% and 7.75%.

And for a variety of reasons, including convexity, those portfolios or those inventories become more challenging to manage. We were largely ahead of that, which is helpful. And as Chris articulated, the overall size of our position that's undistributed right now is as light as we've been in some time and does position us well when we feel like we can lean back in.

Brooke Carillo

The other thing it's done, too, is really helped us build a nice cash position, I would say, with all of the distribution activity. So we've seen $20 million to $25 million of cash inflow here that we can also use. We've made some comments around opportunistic debt repurchases for certain of our debt that's higher yielding and looks attractive to us today. So that's certainly on our radar as well.

Ameeta Lobo Nelson

Okay. Great. I appreciate that detail. And then just on legacy, as you mentioned you're evaluating offers in about 70% of the remaining loans. How should we think about potential for additional markdowns or transaction costs required to exit those positions?

Dashiell Robinson

Yes. We're obviously very focused on that. Maybe just to look at what we just have accomplished over the past 3 or 4 months. We talked about the majority of the fair value changes in legacy being linked to actual realized dispositions. That was about $0.22. That was linked to about $230 million, $240 million of resolutions realized or expected to be realized here in the coming couple of weeks. A lot of those loans are the loans we disclosed and talked about when we issued our convert a few months back as part of that sale process, and there were others.

So we will certainly evaluate everything. It will depend upon the specific assets, but we feel like that's a reasonable proxy potentially. The reality of these assets is -- oftentimes, there is a bid-ask depending on the particular situation with the sponsor. Obviously, we're evaluating everything also in the context of timing. We talked about the $0.12 to $0.14 carry cost per quarter. That is definitely to the extent we take additional fair value hits on where we have the position marked, that earn-back period and that quarterly savings is very important in terms of the interest expense and the carrying costs, et cetera.

So the big thing, just to underscore sort of the prepared remarks, is that the position continues to shrink. We reduced it by over 20% last quarter alone. We're very engaged to reduce it even more significantly than that in the fourth quarter. And our recent credit cohorts have been very, very strong. So to use Brooke and Chris' terms, this really is a melting ice cube, and we're optimistic that we will continue to move the majority of the rest of this year, the rest of the quarter and early next year.

Operator

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

Richard Shane

A couple of things. Look, tactically, the decision to pull back on volumes in hindsight, looks like a very good decision. Strategically, the conversion to the REIT -- or to de-REIT, I should say, is interesting. I'm curious about a couple of different things.

One is, and I think this all sort of ties together because we've talked about the financing and the corporate structure. But if we look at the expense structure, it's about $250 million on a run rate basis over the last 12 months. That makes it very challenging at current volumes to generate profit. So I am curious when you guys look forward and get credit for sort of that 75% deferred tax asset, what are the underlying assumptions to actually realize that value? Is there going to be operational change along with sort of the corporate and financial structure change?

Christopher Abate

Yes. Thanks. Maybe Brooke and I can team up on this one, Rick. At the highest level, we should continue to underscore that a huge amount of our costs associated with mortgage banking are variable. It's a unit issue. And if you're pushing through fewer units because in our case, we've decided to be more conservative as spreads have widened there's going to be a direct reduction in expenses associated with that throughput.

Obviously, there's fixed costs associated with the operation. I think that's where the retention of capital comes into play to continue to leverage the platform. We also -- a few other things. I'll let Brooke talk about our joint venture partners and how we're scaling on that front.

But we continue to point to an efficiency ratio based on operating expenses and revenue versus equity. This is a big reason why we've decided to move forward with the de-REIT. Our business -- when we think about our business and our operating expenses divided by revenues, we actually look better than most of the money center banks. So the operation itself is efficient, and we can dig into those numbers with investors. But the corporate structure has not been. And so we're trying to strike that balance. But certainly, we definitely want to preserve margin and margin requires, obviously, net revenues that makes sense and are consistent.

I personally think we're seeing -- we're definitively seeing signs of the mortgage market stabilizing here. We're hoping that as we exit the midterms and some of the ball that we've experienced over the last 6 or 8 weeks, it's going to be a great opportunity to lean back in. But certainly, if the market remains challenging and we don't see the volume opportunities that we expect, we're going to remain efficient. We're going to make sure that our expenses and our efficiency is in line.

Richard Shane

Brooke, before you answer, do you mind helping me with what percentage of the as you see as fixed versus variable? I think that will be really helpful as we sort of think through this.

Brooke Carillo

Yes. So about 56% of our costs are variable today, Rick. One thing that some of that noise was muted in a quarter like this where we had a pick up in predominantly legacy portfolio management costs, while G&A was flat, we did have a couple of -- we had some workout costs associated with loans that increased overall OpEx.

I would also note that while our lock volume was down this quarter, our purchase volume is actually up. This is really just because purchase volume lags by 30 to 45 days. So you will actually see our loan acquisition costs, some of those variable costs decrease in the fourth quarter on our 18% lower lock volume.

A couple of other comments there. Chris mentioned our joint ventures, we had over $1.5 billion or so of volume that we -- that ran through our JVs. Capital really has been the biggest constraint to our model, not the size of the -- not our addressable market. And so it is really both these joint ventures and the retained capital that we're freeing up through our -- both retained earnings and these cash savings that we're talking about via taxes. So this is why it's so meaningful to our ability to continue to grow volume.

As Chris noted, we have intentionally slowed over the past month. But as we look at out over a more stable market, then we really have an ability through our joint ventures and this retained capital to continue to meet the right operating leverage of our model.

I would lastly note just consolidated EAD would have been a mid-teens return if not for kind of 2 main factors. One, Aspire's net margin was really the primary driver of the decrease. That was due to a fairly extraordinary spread widening and rate volatility in the market, which is not something specific to us.

And then secondly, I would say the mark-to-market on the HEI intended resolutions actually does run through EAD. So those are just a couple of notes on the quarter as it relates to forward profitability.

Richard Shane

Got it. Very helpful. And I'm going to follow up with one thing. Can you walk through the Section 382 risk? And the reason I ask is, obviously, by suspending or reducing the dividend, there is risk that you're going to have pretty significant shareholder turnover. I know one of the strategies to mitigate that is to offer prefs to equity holders, do folks who tender and take press count for or against the 382 test?

Brooke Carillo

The exchange offer, we don't expect to meaningfully mitigate the 382 analysis because our Board has been very thoughtful together with tax advisers as to how to structure the preferred instrument so that we are very mindful of any 382 impact, which we don't expect a significant one at the time of the exchange.

The owner -- there's -- the plan that was put in place that's detailed in one of our press releases that the Board put out with respect to ownership thresholds is exactly to your point, to be mindful of this very significant tax asset that we have and this preservation plan is really very similar that's been adopted by other companies with significant net operating losses.

So a lot of time has been put into making sure that this structure is very cautious with respect to any forward implications on our 382 analysis.

Richard Shane

Got it. Okay. I'm sitting here thinking I had a lot more hair the last time I would a de-REIT note, but looking forward to dusting off that knowledge base.

Brooke Carillo

I spend A lot of time on the follow.

Operator

Our next question comes from the line of Jason Weaver with JonesTrading.

Jason Weaver

Lots of news today, obviously. Maybe to the point of the last exchange, just thinking about the preferred exchange offer. What were your assumptions around sizing that transaction? And do you have an idea, is there a hard limit on oversubscription or upsize preserve a given level of common that you want to retain?

Christopher Abate

Good question. We disclosed $50 million with the potential to upsize based on demand and subscription. We want to be thoughtful. Candidly, we are going to have all the details when this actually launches, which we expect next week. But I think what we wanted to be mindful of things like 382 issue, covenants, other things. However, we think this is an excellent option for investors -- common investors that value a high dividend. We expect it to come at a premium to the share price at the time. And I think once we have all the details ready to go, we will have more to say.

Brooke Carillo

One other point just around sizing is we anticipate at a $50 million level that this really would -- together with the repurchases that we've already commenced since our September offering of our convertible instrument that, that would really care for any potential dilution of that instrument as well, depending on the ultimate premium there.

Jason Weaver

Got it. That's helpful. And then maybe more broadly, without the REIT asset test restrictions, what, if any, changes do you anticipate for the investment or retained portfolio strategy going forward? As Brooke's spoke to and I think one of Rick's questions, would that incorporate possibly increasing the amount of retention going forward?

Dashiell Robinson

Jason, this is Dash. I can take that. Probably not. I think we have continued to lean into our originate to distribute model, either full distribution or having these joint ventures be a bigger part of the story in terms of how we grow our mortgage banking platforms.

Certainly, the ability to retain more capital does give us the ability to continue to invest and add to that portfolio. But I would expect us to continue to do more of what we've been doing, which is bring in outside capital partners either through whole loan sales or joint ventures or sort of similar structures to continue to scale in part -- in tandem with the ability to continue to grow our capital base.

Christopher Abate

Yes. We also -- while we have joint venture partners, we are a partner. We're an investment partner. And for our business to scale, and this gets back to the essence of the mortgage business that we're running, we need to retain investments in securitizations. We own MSR to run more volume, we need more working capital to fund warehouse haircuts and for hedging.

And so really to meet our business objectives, we need that form of growth capital. And if we're successful, the capital piece needs to follow. As Brooke mentioned, the real constraint to unlocking the true value of the franchise has been capital. So we're focused on the amount of capital, our cost of capital and certainly open to more joint ventures and partners.

Operator

Our next question comes from the line of Mikhail Goberman with Citizens JMP.

Mikhail Goberman

Lots to unpack here, obviously, and we've been over quite a lot of ground already. So maybe if I could ask you guys to maybe walk through the mechanics of the C-corp conversion, the timing of it, the expected time line of it, if you guys need a shareholder vote on that, stuff like that.

Christopher Abate

Yes. I'll start, Mikhail, and the team can add comments. The first thing is by way of the announcement today, we've announced the NOL plan and the Board's intention to transition to a C-corp. Once that intention has been made, those tax assets become realizable. That's why we announced, we expect an 11% increase from our September 30 book value, which is pretty substantial.

A reminder, these are NOLs that our operating businesses should be able to use more or less immediately. We've been paying for some time mid to high 20s effective tax rates on substantially all of our earnings which is very counterintuitive for a REIT. We've had the NOLs. We've not been able to use them based on the structure. So for some extended period of time, we expect to not have to pay taxes out of capital or cash, I should say. And tax obligations will be absorbed by the NOLs. So it will become very tangible to us.

I think we have an example in the review on what a average quarter could look like in the per share P&L increase from tax savings alone. So we think that the benefits will become immediate. We mentioned that it becomes effective in the 2027 tax year. So the technical requirements from a federal income tax perspective would presumably be met in the first quarter.

And I think that we are -- our next step is to kind of get through the next few weeks. We've got this exchange plan. We're quite excited about it. And I think what we're trying to do in the fourth quarter through all of the initiatives that we mentioned is really put the business in a position to be growing on a stable basis in 2027.

Hopefully, more simple, more straightforward with a business structure that is consistent with our operating model. And these calls, I think, will be -- our results will become a lot more intuitive.

Mikhail Goberman

So there's no shareholder vote that's needed?

Christopher Abate

No.

Mikhail Goberman

Got you. And if I could just make a quick little hard pivot to an operational segment. Just your thoughts on credit quality going forward. Some data has been popping up about delinquencies sort of rising in the non-QM space and a few other pockets of the mortgage industry. So maybe just talk a little bit about that and if there was any effect of that on your Aspire volumes in the third quarter.

Dashiell Robinson

Thanks, Mikhail. It's Dash. I can take that. No, in terms of the credit philosophy within Aspire, non-QM can be a broad term. I think since we launched Aspire close to 2 years ago, we've had the thesis of sticking to our knitting and being on sort of the middle lanes of the highway in terms of credit. I think we have market standard guidelines, but the judgmental underwriting that our underwriters do, how we price our rate sheet, et cetera, I think has generally guided us to, on balance, probably a higher credit quality net book than maybe the industry in total, which I think shines through with our delinquencies being at 20 basis points, which is probably in the upper quartile of the market or the best quartile of the market, I should say.

So we've taken a very conservative approach to higher LTVs to aggressive debt service coverage ratios, extremely mindful for some of the issues in valuation that the sector overall has had over the past few years. And I think that's shining through, frankly, in delinquencies and performance. Obviously, that's a day-to-day discipline. But so far, we feel like we're performing that effectively.

The volume pullback for Aspire was really largely market-driven. As we saw spreads widen and rates move higher, Non-QM in particular, did widen pretty significantly in the second half of the third quarter, and it actually has continued to widen in the fourth quarter. If you go back to the summer, maybe in August, AAAs were probably at 120, 125. Now prints are probably at 165 or so. So there's definitely significant widening in both whole loan and AAA spreads.

One thing I would reinforce, which we didn't mention on this call, but have on prior calls is that with Aspire, we move all of our risk. We sell the subordinate securities. We sell the risk retention securities to third parties. And while that execution did impact our margins in the third quarter, that risk is sold. And we're not sitting on bonds that maybe have had some negative price action thus far this quarter as spreads have continued to widen.

So I think sticking to our knitting on credit and continuing to focus on running that business capital efficiently. We have our JV. We expect to set up here shortly in the fourth quarter, and that will continue to be the story in Aspire.

Mikhail Goberman

Looking forward to the updates going forward over the next couple weeks and beyond.

Operator

Our last question comes from the line of Bose George with KBW.

Bose George

So first, just one follow-up on the legacy. Your capital there is $160 million, so it's less than $1 a share. On each position, can you walk away if the potential loss exceeds that equity allocation? And just confirming your loss is now ring-fenced at this pretty small number.

Dashiell Robinson

Well, some of the debt allocated against that position is nonrecourse and some is recourse, Bose, so that is a consideration. We have some detail on that. So that is part of the capital allocation. There is some modicum of recourse debt. But again, we're going to do everything we can to move the position expeditiously, but obviously protect value. We recognize that it is an overall challenging environment, each of these assets is bespoke and takes its own considerations. But we feel very good with the progress we made, again, close to $0.25 billion of resolution since June 30 and the opportunities we have in front of us to continue to derisk between now and the end of the year and into early Q1. We're optimistic that the bulk of this can get behind us pretty quickly.

Bose George

Okay. Great. And then actually, you noted that $0.12 to $0.14 a quarter of the carrying costs on this. Does that -- has that declined as the legacy has declined? And was that number for 3Q as well? Or was that a lower number?

Dashiell Robinson

It's been fairly consistent, Bose. And some of that is just the ongoing cost, but also, in some cases, just the cost to resolve and get to the real estate and move it. So those have been reasonably consistent. We should continue to see -- or we should start, I should say, to see those trend down in the fourth quarter and into Q1 as some of the associated secured debt is paid down and obviously, the ongoing property management and just overall protective costs come down as the position continues to shrink.

Bose George

Okay. Great. And then actually one on the deferred tax assets. The -- just given the size of your NOL, is there going to be like -- is there sort of incremental deferred tax assets that are going to have like a valuation allowance initially and can be reversed? Or is the amount you noted kind of the max on the DTA?

Brooke Carillo

No, we actually are putting -- a part of our DTA will have a valuation allowance up. We see some of our NOLs expiring before they're fully utilized. But there's a potential based on our profitability for that to be recognized over time. So we're not taking 100% of the value today.

Bose George

Okay. And is that something you could quantify now? Is that -- I mean, could that be a pretty meaningful number just given the size of your NOL, it seems like there should be more DTA eventually.

Brooke Carillo

Yes. Our VA that we're putting up is only on our -- we have $1.1 billion of NOLs at the California state level. And so we have -- we're expecting to use roughly kind of half of that state NOL today. To the extent, again, our projections change around our ability to utilize that, we would remove -- we would derecognize that VA over time.

Christopher Abate

Yes, Bose, we can follow up with the amount. But I think the big takeaway is the amount -- the big jump in book value has been heavily vetted with our tax accountants and tax attorneys, and that is what we expect to be realizable. But certainly, there's more upside if more of that allowance can be released. And some of these DTAs date back to the financial crisis. It's been just a very long time coming. So we're excited to have that unlock.

Operator

Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.