Ellington Financial Inc. (EFC) Q2 2026 Earnings Call Transcript

Good morning, ladies and gentlemen. Welcome to the Ellington Financial Second Quarter 2026 Earnings Call. Today's call is being recorded. I will now turn the call over to Mr. Alaael-Deen Shilleh, Associate General Counsel and Secretary. Please go ahead, Mr. Shilleh.

Thank you. Before we begin, I'd like to remind everyone that this conference call may include forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements are not historical in nature and involve risks and uncertainties detailed in our annual and quarterly reports filed with the SEC. Actual results may differ materially from these statements, so they should not be considered to be predictions of future events. The company undertakes no obligation to update these forward-looking statements. Joining me today are Larry Penn, Chief Executive Officer of Ellington Financial; Mark Tecotzky, Co-Chief Investment Officer; and J.R. Herlihy, Chief Financial Officer. Our second quarter earnings conference call presentation is available on our website, ellingtonfinancial.com. Today's call will track that presentation and all statements and references to figures are qualified by the important notice and end notes in the presentation. With that, I'll hand it over to Larry.

Thanks, Alaael-Deen. Good morning, everyone, and thank you for joining us today. I'll begin on Slide 3 of the presentation. Ellington Financial delivered yet another terrific quarter, continuing the momentum we have built over the past several years. Strong performance across our diversified platform once again drove strong GAAP earnings, adjusted distributable earnings well above our dividend, and also drove a further increase in book value per share. For the quarter, we generated GAAP net income of $0.43 per share, ADE of $0.60 per share, and an annualized economic return of 13.6%. These results reflected excellent securitization execution, continued outstanding results at Longbridge, solid contributions from our other loan origination partners and continued strong credit performance across our loan portfolios. Meanwhile, the financing spreads on our credit lines continue to narrow, which is providing an additional tailwind to our results. Importantly, all these drivers reinforce one another. Strong loan sourcing supports capital deployment and securitization volume. Through our securitization executions, we create attractive retained investments that help build our future earnings power. We release capital for redeployment, and we replace short-term financing with more stable non-mark-to-market funding. Moreover, our securitizations benefit greatly from increasing scale, as our larger and more frequent transactions continue to expand our investor base and have improved our execution levels over time. Meanwhile, strong loan credit performance supports the yields on our retained investments and also sustains and broadens the institutional investor demand for our securitizations. Finally, the profitability and market share growth of our originator affiliates contribute directly to our earnings, while also expanding the flow of loans available to our investment portfolio. We saw this dynamic play out repeatedly during the quarter. Ellington's proprietary residential loan portal, where we lock in loans for more than 40 unique sellers, is now generating more than $15 million of loan purchases per day at a pace of around $4 billion annually. This portal supplied a significant portion of the approximately $2 billion of loans we securitized during the quarter. And of course, we have Longbridge, which supplies their expanding pipeline of proprietary reverse mortgage loans for our investment and securitization. Foundational to all of this is Ellington's well-known and long-standing focus on proprietary research, data, and modeling capabilities. A full 20% of Ellington's employees are dedicated to research and technology, and recent advances in AI are further enhancing the output of that team. Ellington's research and analytics helps shape the loans we originate, the underwriting standards and loan programs we support, the risks we choose to retain and those we choose to offload or hedge, and the way we manage our liquidity. Some of this is clearly visible in our credit statistics, as shown on Slide 14. As you can see on that slide, inception to date cumulative realized credit losses were a mere 17 basis points on approximately $20.4 billion of residential mortgage loan fundings and just 39 basis points on more than $2.5 billion of commercial mortgage bridge loan originations. Keep in mind, these are cumulative loss amounts, with the annualized ratios being far lower. This credit performance spans multiple market cycles, including COVID, the 2022 interest rate sell-off, and the more recent commercial real estate downturn, and reflect not only the quality of our underwriting at loan origination, but also the effectiveness of our asset management and loan workout capabilities. The same discipline is evident in our securitizations. Our EFMT non-QM shelf has continued to rank among the strongest in its cohort for both low delinquencies and controlled prepayment speeds. These drivers enhance the yields on the retained tranches we invest in, while also helping reinforce the liquidity and reputation of the EFMT franchise. They also demonstrate how Ellington's competitive advantage in research and underwriting can translate into stronger credit outcomes and better investment performance. Longbridge had another standout quarter. Originations were up 38% year-over-year. Margins remain healthy, securitization executions improved, and servicing continued to add meaningfully to the bottom line. Longbridge remains one component of EFC's much broader platform, but its performance demonstrates the value that can be created when sourcing, analytics, financing, securitization, and servicing all work together. With that, please turn to Slide 5, and I'll hand the call over to J.R. to walk through our financial results in more detail. JR?

Thanks, Larry. Good morning, everyone. I'll begin on Slide 5 with our earnings summary, then review the principal drivers of the quarter, several disclosure enhancements we've made in our portfolio, and balance sheet activity. For the second quarter, EFC reported GAAP net income of $0.43 per common share on a fully marked-to-market basis and adjusted distributable earnings of $0.60 per share. On Slide 5, you can see the contribution to GAAP net income by segment, and on Slide 6, the corresponding contribution to ADE. Our quarterly results again demonstrated the strengths of our underlying businesses with continued excellent performance across the investment portfolio and another outstanding quarter from Longbridge. Looking ahead, we continue to see broad support for ADE reinforced by several factors, including attractive net interest margins, particularly on our portfolio of retained securitization tranches, robust credit performance, ample liquidity available for deployment, and of course, continued sizable earnings contributions from Longbridge. Turning to the investment portfolio. Net interest income increased significantly quarter-over-quarter, reflecting attractive asset yields and a higher average portfolio size. Earnings from unconsolidated entities also remain strong, driven by solid results in our equity stakes and loan originators and commercial mortgage bridge loans accounted for as equity method investments. Overall performance was excellent across the investment portfolio, led by our residential credit strategies, while gains on hedges more than offset net realized and unrealized losses. Credit performance across our loan businesses also remained excellent, with exceptionally low life-to-date realized credit losses across both our residential and commercial mortgage loan portfolios, consistent with the statistics that Larry highlighted. You'll notice several changes to our disclosures this quarter. These changes simplify certain parts of the presentation while adding detail where we believe it will be most useful to investors. First, we have incorporated Agency MBS into the broader investment portfolio disclosures throughout the presentation. In years past, Agency represented a substantially larger allocation of our capital, but we have since rotated much of that capital into credit strategies where we see stronger return opportunities and clearer competitive advantages. Given the smaller role Agency plays today, we believe that the revised presentation better reflects how we evaluate and allocate capital across the portfolio. Second, we have expanded our Longbridge disclosures. Starting on Slide 9, we now separately present HECM and proprietary reverse mortgage origination volumes, including the channel composition of each, providing greater visibility into the scale and growth of both product lines. We have also added submission volumes to this slide. Because loan fundings are preceded by loan submissions, we believe that submissions provide a useful leading indicator of future origination volume. As you can see on Slide 9, second quarter submissions were up substantially, sequentially, supporting a healthy pipeline entering the second half of the year. That momentum is continuing with July 2026 marking Longbridge's highest ever month for prop reverse mortgage originations and submissions. Finally, turning to Slide 10, you can see that we are now presenting separate roll-forwards for HMBS MSRs and prop reverse mortgage MSRs together with earnings generated by those. The roll-forwards separately identify overall MSR values, new production, revenue, runoff, and changes in fair value, providing greater visibility into changes in MSR value and the components of net servicing profits. We believe that this additional detail should make the Longbridge business easier for investors and analysts to understand and model. Turning to Longbridge's results, please turn back to Slide 8. Longbridge delivered another outstanding quarter across both originations and servicing. It originated approximately $590 million of loans, a 38% year-over-year increase. Prop reverse represented approximately 54% of volume and reached record levels, while HECMs represented the remaining 46%. Originations at Longbridge benefited from strong volumes, healthy margins, and gains from the 2 proprietary reverse mortgage securitizations completed during the quarter. Those transactions represented Longbridge's strongest financing execution to date for this product, as measured by overall debt spreads. Servicing also made a substantial contribution at Longbridge, reflecting both steady base servicing income and continued strong execution on sales of HECM tail pools. Consistent with Ellington's broader risk management approach, we maintain enterprise-level interest rate hedges in the Longbridge segment that are designed to offset some of the pressure that higher interest rates can put on mortgage origination volumes and margins. Despite the increase in rates during the quarter, Longbridge's origination business remained highly profitable, while the enterprise hedges also generated gains. That combination was unusually favorable in the second quarter. All else equal, we should generally expect origination profitability and interest rates to move inversely, so these hedges should help stabilize the segment's earnings across different interest rate environments. Turning next to portfolio activity, please turn to Slide 7. Our adjusted long investment portfolio increased modestly during the quarter, as growth in residential transition loans, commercial mortgage bridge loans, and retained RMBS more than offset the impact of continued securitization activity. In other words, asset sourcing kept pace with our robust securitization activity. Our shorter-duration loan portfolios continue to generate significant principal repayments, including payoffs providing internally generated capital for redeployment into new opportunities. Turning to financing, our focus remains on improving the durability, diversification, and cost of our liability structure. As shown on Slide 11, at quarter end, the weighted average borrowing rate on our recourse borrowings was 5.5%, essentially unchanged from the prior quarter, contributing to a solid overall net interest margin of 336 basis points, which was also roughly unchanged quarter-over-quarter. Approximately 29% of our recourse borrowings were long-term and non-mark-to-market, while 17% consisted of unsecured debt. In addition, the weighted average remaining term of our repo borrowings increased to 9.3 months, approximately double the level in mid-2025, reducing near-term refinancing risk and providing greater funding certainty. During the quarter, we extended and/or improved terms on several warehouse facilities, while adding a new financing relationship covering multiple residential mortgage products. Our securitization program continued replacing shorter-term mark-to-market financing with longer-term non-recourse financing. Through the first half of 2026, we securitized approximately $4 billion unpaid principal balance compared to $4.4 billion UPB during all of 2025. We continue to be encouraged by the market's reception to our unsecured debt. Our outstanding notes have recently traded at a premium, despite higher interest rates, reflecting the progress we've made strengthening our balance sheet and funding profile. We believe this positions us well to continue increasing the use of unsecured financing as well as preferred equity over time as market conditions permit. At quarter end, our recourse debt-to-equity ratio remained 1.9x to 1x, while our overall debt-to-equity ratio increased modestly to 9.2x to 1x, primarily reflecting additional non-recourse borrowings associated with recent securitizations. Turning now to our hedging portfolio on Slide 17. We continue to manage interest rate, mortgage basis, and credit risks through a diversified set of instruments designed to protect book value while preserving our ability to capitalize on attractive opportunities. As you can see on Slide 18, during the quarter we increased our credit hedges as market conditions changed and as the size and characteristics of our portfolio evolved. Turning to corporate other. Aside from recurring items, we also recognize unrealized losses in our corporate other category. As has been our long-standing practice, we carry our outstanding unsecured notes at fair value on the liability side of our balance sheet. With spreads on our debt tightening during the quarter, the increases in the prices of our outstanding debt led to the recognition of an unrealized loss. Also in this category, higher interest rates led to unrealized losses on the fixed receiver interest rate swaps we used to hedge the fixed payments on our unsecured notes and preferred equity. At quarter end, book value per share increased by $0.05 to $13.61 after $0.39 per share in dividends, and our annualized compounded economic return for the quarter was 13.6%. With that, I'll turn the call over to Mark.

Thank you, J.R. Despite rising interest rates, geopolitical uncertainty, and tremendous volatility in energy prices and equity markets, the mortgage and structured credit markets remained constructive. We had a favorable mortgage origination environment and relatively stable credit spreads, and we were able to execute our business plans consistently this quarter. Across our businesses, we continue to responsibly grow volumes, gain market share, expand our sourcing networks, and broaden our product offerings. Put simply, we bought a lot of loans, priced a lot of deals, and in so doing created a lot of attractive investments for EFC's portfolio. We also continue to support and collaborate closely with the growing portfolio of companies in which we've made equity investments. As a group, they have had phenomenal earnings this year, and their origination volumes have helped drive our securitization machine. On the commercial mortgage side, much of our loan sourcing comes through our affiliated originator, Sheridan Capital, which continues to grow its footprint and client base. We are helping institutionalize the business by expanding its capital markets capabilities and strengthening its operational infrastructure, applying many of the same principles that have served us so well with our affiliated residential mortgage originators. This is exactly the ecosystem we've been building. Our consistent demand for high-quality loans supports the growth and profitability of our origination partners. Those loans then become the raw material for our securitization platform, creating attractive retained investments for EFC's portfolio while providing institutional investors with high-quality securities. As Larry discussed earlier, those capabilities increasingly reinforce one another. Both net income and ADE again exceeded the dividends this quarter, while we continue to keep recourse borrowings low and organically created investments continue to perform well. We also continue investing in technology and automation while pushing for deeper integration across our businesses. On the residential mortgage side, with the help of the loan portal that Larry mentioned, we continue streamlining our channel connecting creditworthy borrowers seeking home financing with the vast reservoir of institutional capital looking for investment grade bonds. At Longbridge, our investments in technology, process improvements, and AI-enabled workflow look like they're paying off handsomely. For example, since January '23, the number of funded loans per operations employee has more than doubled, demonstrating how these investments are improving efficiency while supporting continued growth. This past quarter, we continued our disciplined portfolio growth while maintaining high securitization volumes. With bigger portfolios inevitably come some delinquencies. We put substantial resources into resolving residential mortgage delinquencies optimally for the company while seeking the best practical outcomes for borrowers experiencing financial difficulty. On the residential side, we are close to completing the acquisition of a loan servicer. We have redeployed substantial internal resources to help build what we believe can be a best-in-class residential special servicing platform with specialized processes for managing delinquent loans across multiple mortgage products. That acquisition should close in Q3. We believe that controlling our own special servicer will unlock significant value for us as we align incentives, share valuable data, and refine our workout expertise over time. We have a lot to build, but whether it's managing construction projects we take over from RTL borrowers or even just non-QM loans where borrowers can no longer pay their mortgage debt, we know that special servicing is going to be important to preserving value and delivering returns through market cycles. Stepping back, we are seeing an expansion of the addressable market for our business model. More and more mortgage loans are ultimately finding their way into the private label market rather than the GSEs. We expect approximately $250 billion of new issue non-agency mortgage securitizations this year. Larger new issue volumes have dramatically improved liquidity across the asset class, attracting many new investors over the past year. As liquidity continues to improve, more institutional investors enter the market, which in turn supports additional issuance and better execution. That virtuous cycle has been a meaningful tailwind for our securitization platform and for the broader private label market. We see these trends as ideally suited for integrated private sector capital platforms like Ellington Financial that can source, analyze, and securitize loans efficiently. Ellington has had a front row seat throughout this evolution, having been an early mover in securitizing non-QM, closed-end second liens, agency eligible loans, and of course proprietary reverse mortgages. As these markets continue to grow, we will continue investing in the people, technology, and infrastructure needed to support them, while continually working to improve efficiency across our platform. I'd like to finish with some thoughts on the forward MSR market, where we have one large investment that we've held since our acquisition of Arlington back in 2023. The market value of that MSR has increased significantly this year, even much more than you'd expect with the rise in interest rates we've seen. One factor at play is that for banks, the market is expecting that regulators will loosen the caps on how much Tier 1 bank capital can be in MSRs. If that happens, banks could flip from being net sellers of MSRs into being net buyers. The second factor at play is that mortgage companies with large servicing and origination arms are bidding up MSRs. Not only can those companies add mortgage servicing rights to their existing portfolio more efficiently than others, but they can also cross-sell a variety of products to what would become new servicing clients. When servicing low coupons in particular, home equity loans present obvious cross-selling opportunities. We all saw the feverish bidding war for Two Harbors that recently came to an end, and it was a large mortgage company as opposed to a pure investor that won that contest. Our forward MSR is also backed by low-coupon loans, and while we're pleased with the appreciation we've seen on that asset, we're better sellers than buyers at these levels from an investment standpoint. Now back to Larry.

Thanks, Mark. On last quarter's earnings call, I concluded with the observation that Ellington Financial was firing on all cylinders. I am happy to report that we still are, with that momentum continuing into the third quarter. I firmly believe that EFC's sustained strong performance reflects the capabilities and investments we've been building over many years, rather than the success of any single recent initiative. Ellington's investment in research, analytics, technology, and disciplined risk management dates back to the firm's founding more than 30 years ago and has been central to EFC since its formation. Over the past decade, we've steadily expanded the ways we apply those capabilities by investing in strategic originator partnerships, building a best-in-class securitization platform, expanding our proprietary sourcing capabilities, and strengthening our funding profile. As those investments have reached greater scale, their benefits have increasingly reinforced one another across the business. We've now covered our dividend for 8 consecutive quarters and counting, reflecting the increase in contribution of those long-term investments to our earnings. Looking ahead, we'll continue focusing on the things we can control, disciplined underwriting, thoughtful capital allocation, continued investment in technology and our platform, and maintaining a strong, flexible balance sheet. We also intend to be opportunistic issuers of unsecured debt and preferred equity when market conditions are favorable, further diversifying our funding sources and enhancing our financial flexibility. We are aiming for a virtuous cycle of stronger balance sheets and improved credit ratings. As we've emphasized throughout today's call, the strength of our platform is not in any single business or investment strategy. Rather, it is the way our research, relationships, technology, and capital markets capabilities reinforce one another to create an increasingly diversified and resilient earning stream for our shareholders. And finally, a word about our adjusted distributable earnings and dividend. As strong as ADE was in the first quarter, it was even stronger in the second quarter at $0.60 per share compared to our $0.39 quarterly dividend. By out-earning the dividend, not only on an ADE basis, but on a GAAP basis as well, we've been able to build book value per share, and we think that's really important. For now, we think our $0.13 monthly dividend remains appropriate. With ADE running so strong, we could see upward pressure on our dividend based on the REIT distribution requirements. For now, however, we believe that continuing to build book value per share is the best use of our excess earnings and that our current dividend remains appropriate. And with that, let's open the floor to Q&A. Operator, please go ahead.

We'll go first this morning to Trevor Cranston with Citizens JMP.

Mark mentioned the pending acquisition of a residential servicer. Can you provide any additional color around that, would it come with MSR assets or sub-servicing contracts, or any other details on what that would look like?

Mark?

Why don't you take that one, Larry?

Sure. Yes. So, well, first of all, it's a small servicer, single-digit billions of servicing rights. It does have some sub-servicing contracts, as you mentioned. It's diversified in the sense that it does service many different types of loans. And as we mentioned, we think it's going to close sometime in September. It's the type of project where we're going to try to build it as much in our image as we can. So, it's not going to bring any appreciable size of MSRs that are going to have a noticeable impact on our balance sheet, per se, or frankly even our earnings in the beginning. But as Mark said, we have big plans, especially to build out the special servicing aspects of the business. We think they already have some real expertise in that area. And as Mark also mentioned in his script, that's going to be super important to us over time to get the best possible outcomes from our delinquent loans.

Yes, I would just add one thing, Trevor, is that the motivation for this wasn't a servicing acquisition for scale. It's a recognition that over the past several years, we used to have a lot of servicing at Rushmore. Rushmore was bought by Mr. Cooper. Now Mr. Cooper is bought by Rocket. We used to have servicing on some other platforms that have been absorbed. So it's just a recognition that as our footprint in the market grows and the available third-party special servicing capabilities have been diminished, we think there's a real need for high-touch servicing and we've seen the benefit of building things organically in collaboration with an experienced management team. So that was really the motivation for it.

Got it. Okay. That makes sense. And then on Longbridge, J.R., in your commentary, you mentioned kind of the expected relationship and impact of higher rates on volumes and margins. Can you give us any sense sort of how Longbridge volume and margins are trending so far early in the third quarter with the new hiring mix?

Yes. They've been growing the volumes of prop reverse relative to HECM over the last several quarters. This quarter we broke them out separately, and you see the prop was a larger percentage than HECM. The reason I start there is we've seen that prop is relatively less sensitive to higher interest rates vis-à-vis HECM. We also have the enterprise hedge in place, which, all else equal, if higher rates impact origination volumes, should offset some of that impact. To your question about kind of forward-looking guidance, if you will, on volumes and margins, we did include submissions for the first time in our presentation on Slide 9. You can see that submissions in Q2 for loans that are prospectively closing in Q3 were $870 million in Q2 versus under $750 million in Q1. You can see an upward trend we showed over the last 6 quarters. So, I think that should give a good idea of what Q3 may look like. Of course, not all submissions lead to originations, and there's going to be some fallout in those numbers, but I think it bodes well for volumes. In terms of margins, we're not giving Q3 guidance on margins. I think a lot of the profits in prop have also come through because of securitizations, and we did 2 securitizations of prop loans in Q2. There's always going to be some noise in the profits from the Longbridge segment around securitization activity and execution. But long story short, I think the submission story is looking positive going into Q3 for Longbridge.

And let me add two things to that. First, in terms of margins in the HECM product, the real point of sale is when you securitize into HMBS and those spreads are still quite healthy and quite tight on a historical basis. So that's good. We don't see those moving significantly. On the prop side, it's really a function of securitization in terms of when we recognize a gain on those assets. Technically, those remain on balance sheet until we securitize, but when we feel we've generated a gain, that contributes to profit, and securitization spreads remain quite healthy. The other thing I wanted to mention is that when rates go up, the HECM product, the government product, has very defined rules in terms of what LTVs and principal limit factors the government will wrap. The FHA wraps those loans. The proprietary product, of course, has more flexibility. What we have found is that when rates are low, the FHA principal limit factors are often more competitive than on the proprietary side, but when rates rise, the opposite can be true. We think the government's requirements can be somewhat restrictive relative to the right economics, and so with rates higher we're able to offer proprietary loans that are attractive to customers and, in some cases, win share from the government product.

We'll go next now to Bose George with KBW.

This is Frank Labetti on for Bose. Just sticking on the Longbridge topic, can you maybe discuss an outlook, more normalized earnings run rate or contribution to ADE from Longbridge as you guys continue to gain share and scale that segment?

Sure. For the last two quarters, their contribution to ADE was $0.23 and $0.21. The average contribution in 2025 was $0.12. The portfolio is growing, origination volumes are growing, the MSR portfolios are growing, and so that recurring base servicing income is increasing. If you look at the roll-forwards included in the presentation, you can see net profits from those MSRs were about $0.06 to $0.065 per share, meaning that everything else in the segment was $0.16 to $0.17 per share for the quarter—originations, securitizations, less G&A. I mentioned earlier there's going to be noise in the segment's results because of securitizations and the execution. We did two securitizations this quarter. That securitization execution has been notably strong in the first two quarters of this year. I don't know that $0.16 or $0.17 aside from servicing is the durable run rate; it is probably a little bit high. But we don't need it to be that high to hit our mid-40s ADE run rate that we had mentioned last quarter. So if it's in the low- to mid-teens, that's plenty to carry its contribution to the overall EFC earnings stream.

Yes, I think overall, we're comfortable. If we do two securitizations in a quarter, like we did recently, we'll see a higher ADE; that definitely helped drive the $0.60. But even if we just do one, which is a modest goal at this point, we're comfortable guiding into the high 40s on ADE.

Great. That's very helpful. And then switching to the investment portfolio, you continue to see strong returns there. Given where spreads are now, where do you see the best risk-adjusted return in credit today? And then conversely, where are you maybe least comfortable adding to?

I would say that over the last year, from mid-2025 to now, credit spreads have tightened across the board—investment-grade corporates, high-yield bonds, CRT, non-QM investment-grade bonds—and we've seen the same in residential and commercial loan purchases. What has been supportive of our ADE is that when you securitize, the economics are driven by the difference between the spreads where you're buying the loans and the spreads where you're selling the primarily investment-grade bonds. That difference is what you leverage in the retained pieces, similar to how a CLO equity works. That difference has been preserved. So loans are tighter than a year ago, but the bonds we sell are also tighter, and we're not seeing a big change in expected yield on what we're retaining. That has been favorable in allowing us to grow our portfolio at similar yields despite spread tightening. As for pockets of weakness, lower FICO scores show higher delinquencies—any model will reflect that—and that difference has become a bit more elevated in the past year. We are also watching cash-out refinancing; borrowers that choose to cash out in an environment of relatively high interest rates can signal increased risk. We've kept our consumer portfolio relatively small compared to the past, which is one reason we are more comfortable. Additionally, in commercial mortgage there is increasing supply of non-performing loans. There's not going to be a lot of buyers in many of the smaller loan sizes where we tend to play—the sub-$50 million, sub-$25 million area—so we think there may be attractive opportunities, although we have not yet made a significant move there.

Just to add, in commercial mortgage we are seeing supply of stressed and non-performing assets starting to come out, and we expect there could be attractive opportunities in the smaller loan strata where we typically participate.

We'll go next now to Doug Harter with BTIG.

Can you just talk about how you're thinking—given what you just mentioned about the continued attractiveness of returns—how you would think about maintaining short duration versus potentially adding some duration to lock in those returns for longer? Has there been any change in your philosophy or how you're thinking about that?

Hey Doug, it's Mark. When we securitize, we're almost always keeping the ability to call the deals—we have call rights. That represents a longer-term investment and is a way to participate if market rates fall and spreads tighten. On RTL residential transition loans, they're short duration by design because that's the nature of the risk we want to take—properties where renovation is straightforward and shouldn't take long—so those will likely remain short duration. That matches the risk/return profile we seek. The retained securitizations and call options lengthen our optionality and give us a way to participate in tighter market spreads and lower yields in the future. Those call options can be valuable across different future scenarios.

If I could add two more things. First, reverse mortgages are long-duration assets—those are unique where we have market share and attractive returns, so there we are locking in spreads for longer periods. Non-QM is a 30-year mortgage, so that's another example of duration being part of our portfolio. But it's important to our model that we have many high cash-flowing, shorter-duration assets that return principal and provide visibility on LTVs and resolution timelines. That approach helps manage liquidity and risk. It also supports the counterparty confidence we see in our debt trading, which helps with funding and overall risk management. So you should expect a portfolio with a strong component of shorter-duration assets in many sectors, with targeted longer-duration holdings where the economics justify it—like reverse mortgages and certain non-QM assets.

That makes sense. Appreciate it. And then in your prepared remarks, you talked about the benefits of the investments in the operating companies. As you look at the benefits to the returns, how much of that comes through your stake of the ownership versus comes through the returns of the investment portfolio of the assets you retain?

Well, Mark, I'll let you sort of address the asset side. In terms of stakes, Longbridge is fully consolidated and its contribution is broken out. LendSure has had excellent earnings. Ultimately, J.R., it looks like you've got it right in terms of the actual numbers.

I first want to emphasize that the total investment amount on our balance sheet is more than $5 billion. Our equity stakes are about $100 million in total. Longbridge is consolidated and so it doesn't have goodwill. All the other stakes total approximately $97 million, with LendSure being a little over half of that. These stakes contribute to GAAP earnings because we mark-to-market the positions, which often reflect earnings at the underlying originator level. We also capture ADE contributions from larger originators that regularly distribute cash. LendSure, for example, has made distributions to owners that are multiples above our original cost basis and continues to do so on a somewhat regular basis. Quantifying it, the contribution to ADE in the $0.60 was a little under $0.05 from the originators—so a bit under 10%—and that's been fairly steady over recent quarters. The rest of the investment portfolio provided the majority: $0.23 came from Longbridge, $0.37 came from everything else, including overhead. The majority of those earnings come from the loans we buy through affiliates that we then securitize and hold residual tranches; most of that is net interest income. Many of the loans on our balance sheet are sourced by our affiliated originators—LendSure, American Heritage, Sheridan—so the vast majority of earnings contribution comes from loans we buy through these agreements. These originator stakes are punching above their weight: a modest $100 million investment out of $5+ billion is contributing meaningfully to earnings.

We'll go next now to Marissa Lobo with UBS.

On non-QM, issuance has been very robust. Can you speak to where EFC is differentiating from peers on their origination focus and how securitization execution has been trending on spread?

Marissa, it's Mark. One point is our relative performance on prepayment speeds and credit performance. We've always focused on prepayment risk because when you're a sponsor and retention investor, a significant part of your investment behaves like IO. We focus on loans with favorable S-curves—loans that won't prepay very quickly when rates drop. That can come from explicit prepayment penalties or aggregation of loan attributes. We prefer purchase-money loans with higher FICO and better-quality borrowers; these borrowers typically are buying homes with concessions on listing price, which we like. From a performance standpoint, non-QM bonds have tightened, but when we model bids we consider appropriate correlations and spreads relative to investment-grade corporates and Agency MBS. Given the market evolution, non-QM bonds may be fairly priced or slightly cheap. The mortgage 2.0 space has grown; we estimate it will be $250 billion this year, which implies substantial weekly new issue size and improved liquidity. That attracts more buyers and creates a virtuous cycle of issuance and better execution. Over time, the order book for deals has expanded with more sophisticated investors entering the space and being able to deploy meaningful capital. Post-COVID structural features in deals have also provided extension protections for bonds. Overall, we view current pricing as relatively attractive and the growing investor base confirms demand.

If I could add one thing: our portal is a differentiator. We're buying over $15 million a day. The portal applies loan-level price adjustments based on attributes—geography, jumbo size, and other characteristics—and those adjustments flow through Ellington Research. That results in different pricing outcomes and ultimately in the mix of loans we buy in the portal. We think it's working because it shows up in prepayment and credit performance of the loans.

And just on hedging, you mentioned you increased credit hedges as market conditions changed. Can you speak to how you're thinking about hedge construction more broadly under Chair Warsh's framework? And on the credit side, how you're thinking about TBA shorts and CDX sizing from here?

Those are great questions. We use credit hedges in two fundamental ways. One is tactical: as we get close to bringing a deal to market, we may buy protection on investment-grade credit indices to lock in our execution, because the historical relationship between IG indices and non-QM spreads is tight. That's a way to protect deal execution during the marketing window. The other way is macro—protecting the portfolio from an economic shock, such as substantial weakening employment or a recession. For that, we have hedges across commercial and consumer exposures, sometimes using high-yield indices or ETFs to cushion book value volatility from a sharp economic downturn. Regarding interest rate policy under Kevin Warsh versus previous approaches, the key for us is to accurately ring-fence interest rate risk. Our dividend and ADE are effectively spreads to SOFR, and we use hedging instruments to insulate the portfolio from changes in interest rate risk. We do expect the Warsh-style regime may lead to more interest rate volatility as markets react to numbers without forward guidance, but so far it's been manageable for us.

If you look at Slide 16 of the presentation, you'll see our interest rate sensitivity metrics. We've structured and managed the portfolio to avoid taking directional views on what the Fed might do. Non-agency RMBS, especially non-QM, are somewhat negatively convex, but overall we have been able to immunize the portfolio so that a modest instantaneous shock has relatively small impact on book value.

We'll go next now to Crispin Love of Piper Sandler.

This is Ben Graham in for Crispin Love. In the release and presentation, you didn't break out the agency contribution to earnings and instead included it within the broader investment portfolio segment. I'm just wondering if this is just driven by the size of agency? I might have missed this, but would you expect agency to decrease further in the coming quarters and if that decision was a function of that outlook?

Yes, thanks for the question. You nailed the main reason: its size. The agency portfolio is now sub-$200 million on an invested basis and about 1% on a capital basis. Several years ago, those numbers were $2 billion plus and 22% when agency was a much more meaningful part of the portfolio. The evolution of Ellington Financial toward originator stakes, securitizations, and owning loans on balance sheet has centered the business in credit, where we see better return opportunities and clearer competitive advantages. From a REIT test perspective, we also no longer need a large agency allocation. So over time we've rotated out of agency, built up credit exposures, and reduced other non-REIT assets. Given agency's modest size and contribution, we believe it's more appropriate to present it as part of the broader investment portfolio rather than as a separate breakout.

We'll go next now to Timothy D'Agostino at B. Riley Securities.

I appreciate the commentary on the pending acquisition. I guess thinking past that and maybe into 2027, is additional M&A and potential investments into loan originators part of the playbook? And if so, are there any areas you would look to address? Or any color about how you think about additional M&A or investments in originators?

Sure—yes. Absolutely part of the playbook. It's been a great part of our playbook for the last 12 years. We mentioned the servicer; we're also looking at another non-QM-focused origination opportunity and other residential products. We are being shown opportunities on the commercial mortgage side, and we've had a stake in Sheridan, which has been a great source of loan product and profitable. In the commercial mortgage space we expect more stressed and distressed assets to come to market, and we are watching for opportunities. So across those areas—residential non-QM origination, commercial mortgage originators, and others—we remain active. Historically, our approach has been to invest in smaller originators and help them grow, supporting them not just with capital but by facilitating warehouse lines and other operational support. That capability is attractive to smaller originators and we expect to continue building that playbook.

And then just as a quick follow-up, how do you think about funding those potential M&A or further investments?

We fund those with cash on hand. We don't explicitly borrow against them. Of course, unsecured notes and preferred equity are also a great use of capital when market conditions permit. As J.R. mentioned, our unsecured notes have traded strongly and preferred equity can be another lower-cost, diversified source of capital. Given that these operating investments often earn high returns on equity, funding them from our balance sheet or through unsecured notes can be attractive.

Thank you. And gentlemen, that was our final question for today. So we'd like to thank you all for participating in the Ellington Financial Second Quarter 2026 Earnings Conference Call. You may disconnect your line at this time and have a wonderful day. Goodbye, everyone.