Arbor Realty Trust Stock

Good morning, ladies and gentlemen, and welcome to the Second Quarter 2026 Arbor Realty Trust Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. [Operator Instructions]

I would like to now turn the call over to your speaker today, Paul Elenio, Chief Financial Officer. Please go ahead.

Okay. Thank you, Stephanie. Good morning, everyone, and welcome to the quarterly earnings call for Arbor Realty Trust. This morning, we'll discuss the results for the quarter ended June 30, 2026. With me on the call today is Ivan Kaufman, our President and Chief Executive Officer.

Before we begin, I need to inform you that statements made in this earnings call may be deemed forward-looking statements that are subject to risks and uncertainties and including information about possible or assumed future results of our business, financial condition, liquidity, results of operations, plans and objectives. These statements are based on our beliefs, assumptions and expectations of our future performance, taking into account the information currently available to us. Factors that could cause actual results to differ materially from Arbor's expectations in these forward-looking statements are detailed in our SEC reports. Listeners are cautioned not to place undue reliance on these forward-looking statements, which speak only as of today. Arbor undertakes no obligation to publicly update or revise these forward-looking statements to reflect events or circumstances after today or the occurrences of unanticipated events.

I'll now turn the call over to Arbor's President and CEO, Ivan Kaufman.

Thank you, Paul, and thanks to everyone for joining us on today's call. As you can see from this morning's press release, we had a very active quarter in the capital markets and several notable transactions that allowed us that have allowed us to increase our liquidity and and drive higher returns on our capital as we continue to navigate through this extended downturn. First, we were once again successful in unwinding one of our legacy CLOs by financing these loans through our bank lines with superior terms. In fact, we're able to reduce our pricing by almost 40 basis points and enhance our leverage by nearly 10 points, which allowed us to generate approximately $135 million of additional liquidity and increased returns on our capital. We believe it's very important to point out that we had 7 legacy CLOs with $9 billion of collateral in the height of the market, and through effective balance sheet management, we've delevered $7.8 billion of CLOs in addition to adding $2.5 billion of new vehicles for total capital markets transactions of $10 billion over the last 36 months. This leaves us with only one remaining legacy vehicle with $1.2 billion of collateral, which is currently levered at 66% that we also expect to successfully unwind in the near future.

We also closed on a $375 million convertible debt offering in early July, which we used the majority of the proceeds to pay off our September bonds earlier this week. This was an exceptional trade that allowed us to raise capital with pricing that is 400 basis points inside of spray debt and buy back a significant amount of our stock at 50% of book value. We used $114 million of proceeds to buy back stock at $5.42, which will be highly accretive to both book value per share and our future earnings per share as well as through allow us to be more aggressive in resolving our legacy loans quicker and reduce the drag on our earnings. The stock buyback portion of this trade also creates a natural hedge against the $6.10 convert price, strike price. In fact, the stock would need to trade above $9.28 a share before we would have to issue more shares that we bought back in the deal, effectively creating a convert premium of almost 100% above the current stock price.

And we just -- and just recently, we created another $185 million of liquidity from additional financing proceeds we're able to generate from one of our bank lines on existing collateral. These are extremely important accomplishments that again have enhanced our liquidity position and will allow us to work through our legacy loans very aggressively. We have also implemented several cost saving strategies given the challenging climate that will have a very meaningful impact on reducing our expense load going forward. The first of which was a reduction of headcount and certain disciplines in order to property rightsize our staff and payroll to the current environment. This was carried out last month, and we estimate the reoccurring savings after onetime severance payments to be approximately $10 million annually or $0.05 a share.

We will also continue to identify additional opportunities to reduce expenses going forward, which includes a big push to fully integrate AI across all aspects of our business, which will drive additional economies of scale through significant operational and process efficiencies.

Turning now to our production numbers for the second quarter in our different business lines. In our agency platform, we originated $1.5 billion in volume, $1.5 billion in volume, in addition to $50 million CMBS brokerage transactions for a total second quarter volume of $1.1 billion. This brings our year-to-date volume to around $1.9 billion, which is up 30% year -- over last year. The elevated rates are certainly affecting our ability to close deals quickly and pushing out the timing somewhat. However, we have a growing pipeline of larger deals, which we expect will result in stronger second half of the year and hopefully allow us to produce similar volumes as we did in 2025, although the exact time of closing is how to predict in this elevated rate environment.

In our balance sheet lending business, we originated $160 million in volume in the second quarter and just over $550 million for the first half of 2026. This business continues to be incredibly competitive, and as a result, we are being highly selective and are focusing our attention on large deals with high-quality sponsors. We guided to between $1 billion to $1.5 billion of volume for 2026, which was reflective of the current environment. The bridge lending business is an important part of our overall strategy as it generates strong level returns on our capital in the short term while continue to build up a pipeline of future agency deals. And with the significant efficiencies we continue to see in the securitization market with our -- and with our line lenders, we were able to produce strong returns on our capital despite the competitive landscape.

In our single-family rental business, we had a strong second quarter and have seen a real uptick in our pipeline now that the housing bill has been passed with the appropriate carve-outs for the build-to-rent businesses we discussed in the past. We originated $315 million of deals in the second quarter and $215 million in the month of July for a total volume year-to-date of $700 million. And again, we are starting to see a real increase in our forward pipeline, which we expect will result in a very strong second half of the year. This is a great business as it offers us returns on our capital through the construction bridge and permanent lending opportunities and generate strong level of returns in the short term, while providing significant long-term benefits by further diversifying our income streams.

We're also very active in the construction lending business and expect to be able to originate $500 million to $750 million of this product as well. On our last earnings call, we discussed the lengthy effect of the increase in interest rates is having on the timing and resolution of our nonperforming and sub-performing loan book. We believe in the current rate environment, it will take us 4 to 6 quarters from now, to resolve the vast majority of these assets, which will allow us to significantly reduce the drag on our earnings and build back our run rate of interest income for the future. Unfortunately, rates continue to remain elevated and volatile given the geopolitical landscape, which is certainly making it more challenging to resolve these loans quickly. Having said that, we feel confident that we have ring-fenced the majority of our issues and have a clear path to a resolution on these assets.

The rate increases have laid things a little bit but we are making good progress and again, expect to reduce this loan exposure consistently on a quarter-by-quarter basis. We ended up in the second quarter with approximately $525 million in delinquencies and around $545 million of REO assets for total nonperforming assets of roughly $1.07 billion, which is up nominally from last quarter's numbers as a result of things being slightly delayed due to elevated rates. We have, however, made strong progress in July, resolving $90 million of these assets this month and have another $105 million scheduled to be resolved next months that we have executed agreements on. This will bring down nonperforming loan book to approximately $875 million or a 13% reduction in the first quarter.

We also have line of sight on an additional $200 million to $300 million of delinquencies we expect to resolve in the third and fourth quarters in addition to feeling very confident in our ability to reduce our existing REO book down to approximately $300 million by the end of the year as we have been actively marketing several of these assets for sale. This progress will go a long way towards significantly reducing the drag on earnings and increase our run rate of income for the future.

As we discussed in detail on our last few calls, we continue to focus heavily on our legacy portfolio, which is down to $4.7 billion at June 30 from successfully resolving $800 million of these loans in the last quarter. $1.3 billion of the book continues to perform in accordance with their original terms and $1.1 billion are either delinquent or REO, and that we have a clear line of sight to resolving over the next several quarters. The other $2.3 billion of this book, we have been aggressively working through with the goal of restructuring and resolving $500 million of loans a quarter, which we are on pace to accomplish. This will reduce our legacy book, including our delinquencies and REO assets down to around $2.4 billion by year-end and well below $1 billion by the end of 2027.

We also continue to make progress in reducing the amount of accrued interest outstanding on certain loans in this subset by resetting the rates in today's market spreads and requiring that the borrower paid down a large portion of the outstanding accrued interest as part of the modified terms. In fact, of the roughly $600 million of legacy loans will result in Q2, on $500 million of these loans, we received approximately $15 million of back accrued interest in the second quarter, and we'll receive another $10 million in accrued interest by the end of the third quarter. This will reduce our total accrued interest by approximately $25 million and the total loans outstanding with accrued interest down to only $1.1 billion.

As Paul will discuss in more detail, we produced distributable earnings of $0.15 a share in the second quarter which was in line with our expectations and included $0.02 of onetime drag from some inefficiencies in our financing for facilities. Clearly, our earnings are being greatly affected by the significant drag from our noninterest-earning assets as well as from resetting legacy loans to today's market rates. We are taking a very aggressive stance with our borrowers and resolving our nonperforming loan book. This would continue to affect our core earnings in the short term, which is not something we are focused on. Our goals are always longer term in nature with our sight set on working through the loan book as quickly as possible, which will reduce the earnings drag from these assets and allow us to [indiscernible] build back our run rate of interest income and drive higher returns in the future. This, again, we estimate to take us 4 to 6 quarters to accomplish, and we are taking a very methodical approach through resolving $500 million of these loans a quarter and bring down the remaining legacy book to a very nominal number relative to our total loan book.

In summary, we have made tremendous progress in the capital markets with $12 billion of transactions between the unwind of our legacy CLO vehicles, the issuance of new CLOs, the unsecured and convertible debt markets we have accessed and the efficiency we have been able to generate on our warehouse lines. This has allowed us to increase our liquidity and drive higher returns on our capital. At our agency business and our diversified origination platforms are all performing well despite elevated levels. With respect to our legacy book, we have made significant progress, and we have a clear path to reducing this loan book on a quarter-by-quarter basis which will put us in a position by the end of 2027 for this to represent a very nominal portion of our total loan book and allow us to grow our earnings run rate for the future.

I will now turn the call over to Paul to take you through the financial results.

Okay. Thank you, Ivan. In the second quarter, we produced distributable earnings of $31 million or $0.15 per share, excluding realized losses of $10 million from the resolution of certain delinquent and REO assets that we had previously reserved for. On last quarter's earnings call, we guided to around $15 million to $25 million in realized losses a quarter as we look to accelerate the resolution of our nonperforming loan book. As Ivan mentioned, the elevated rate environment has pushed things out a bit, and we have seen a little longer time line to resolving certain assets, which resulted in slightly less realized losses for the second quarter than we anticipated. We are making good progress in the third quarter on resolutions. And as a result, we expect realized losses to increase and be in the range of $20 million to $30 million for the next few quarters, although the exact timing on dispositions is tough to predict and could result in fluctuations in these numbers each quarter.

Our second quarter numbers were in line with our guidance and expectations of $0.15 a share, which was reflective of roughly $0.02 a share of unusual drag from some inefficiencies related to our financing cost from a temporary overlap of interest for part of the quarter. As Ivan mentioned earlier, our aggressive approach to asset resolution is impacting our earnings in the short term, with long-term accretion expected as we continue to make more progress in this area. We have made good progress in the third quarter so far, which combined with the cost-cutting measures we have implemented and the positive effect of large buyback from our convertible debt offering will have on our distributable earnings per share makes us optimistic that we'll be able to start to experience some growth in our run rate of income in 2027 as we realize the full benefit of converting our delinquent assets into performing loans.

In the second quarter, we recorded an additional $14 million of impairment on our REO book to properly mark these assets to where we think we can effectuate a sale. We've engaged brokers to sell the bulk of these REO assets quickly and create interest-earning loans for the future. And while we expect a few additional delinquencies in REO assets as we work through the bottom of the cycle, we believe we'll be able to resolve more nonperforming loans than new ones and continue to reduce the drag on our earnings. We also booked another $22 million of specific reserves on our balance sheet loan book for total REO impairment and specific reserves of $36 million in the second quarter, which is up from a total of approximately $21 million in the first quarter.

General CECL was also elevated this quarter from a change in the outlook for real estate values, resulting in an additional $16 million in reserves in our balance sheet loan book, which is an increase of $20 million from the first quarter. And given the current environment, we expect that we could experience similar levels of specific reserves and impairments over the next few quarters as we are being extremely aggressive in accelerating the resolution of our problem loans, which will allow us to reduce the drag on our earnings and grow our run rate of income for the future.

Our book value per share came in at $10.95 at June 30 as a result of the increased reserves and impairments we booked in the second quarter as we are taking a very aggressive approach to resolving our legacy book. As Ivan noted earlier, the convertible debt offering we closed on July 6 contained a very unique buyback feature that's resulting in us using $114 million of proceeds from the offering to buy back stock and retire 21 million shares at less than 50% of book value. This is highly accretive to our book value per share, which on a pro forma basis, increases our book value per share to $11.59 from $10.95 at June 30 or a 6% increase.

In our GSE agency business, we originated $1.1 billion of volume and had $1.1 billion in loan sales in the second quarter. The margin on these loans came in at 1.33% this quarter compared to 1.86% last quarter mainly due to some larger transactions we closed in the second quarter that contained lower margins. We also recorded $12 million of mortgage servicing rights income related to $1.2 billion of committed loans in the second quarter, representing an average MSR rate of around 1.1% compared to 1.32% last quarter, again due to an increase in the average loan size and a shift in product mix in the quarter.

Our fee-based servicing portfolio grew to $36.7 billion at June 30, with a weighted average servicing fee of 35 basis points and an estimated remaining life of 6 years, and will continue to generate a predictable annuity of income going forward of around $128 million gross annually. In our balance sheet lending operation, our investment portfolio was $12.1 billion at June 30 with an all-in yield in this portfolio of 6.95% compared to 7.03% at March 31. This was mainly due to resetting rates on certain legacy loans and from the new delinquencies during the second quarter.

The average balance in our core investments was $12.08 billion this quarter compared to $12.04 billion last quarter from our second quarter growth. The average yield on these assets decreased to 7.21% from 7.50% last quarter, mainly due to significantly more back interest and default tranches collected in Q1 on loan resolutions in addition to the effect of our second quarter delinquencies.

Total debt on our core assets was approximately $10.5 billion at June 30 compared to $10.7 billion at March 31. This reduction was mainly due to the repayment of our $175 million senior notes in April. The all-in cost of debt was approximately 6.38% at 6/30 versus 6.40% at 3/31, mainly due to the unwind of CLS 17 with our bank lines in the second quarter at a reduced rate. The average balance in our debt facilities was approximately $10.5 billion for the second quarter compared to $10.4 billion in the first quarter mainly due to the enhanced leverage received on the unwind of CLO 17 with our bank lines and the full effect of CLO 21, which was issued late in March.

The average cost of funds in our debt facilities was 6.40% in the second quarter compared to 6.52% for the first quarter, excluding interest expense from levering our REO assets, the debt balance of which is separately stated on our balance sheet and therefore, not included in our total debt on core assets. This decrease is mostly due to the reduced pricing we received from the unwind of our legacy CLO vehicle and the full effect of CLO 21 issued late in the first quarter. And our overall spot net interest spreads were approximately 0.57% and 0.63% at June 30 and March 31, respectively.

That completes our prepared remarks for this morning. I'll now turn it back to the operator to take any questions you may have at this time. Stephanie?

[Operator Instructions] We'll take our first question from Chris Muller with Citizens Capital Markets.

So I know you may not be able to answer this one, but I'm going to try anyway. So you guys have been buying back a lot of stock discount to book value has persisted at pretty extreme level. So there's clearly a disconnect where you guys perceive the value in the market's perception. You guys have operated as a private company for a long time before your IPO in early 2000. So I guess the question is, if this discount remains or gets worse, is there a point where you guys would explore some strategic alternatives as several of the other mortgage REITs are doing?

Listen, our job is always to maximize shareholder value. There's a lot of paths to be able to do that. And clearly, that is more to do alternatives we consider in terms of maximizing shareholder value.

Got it. And then I guess maybe changing gears to REO a little bit. You guys talked about on the last call getting that balance down to $250 million to $300 million by year-end, including adding another $100 million or so through that period. But foreclosures in the second quarter were $121 million, and Ivan, I heard you mention $300 million by year-end now. So I guess the question is, are you guys expecting foreclosures to slow down dramatically in the back half of the year? Or are you expecting that you'll be able to sell down REO faster than you initially expected last quarter?

I think we're working on all cylinders. We are definitely looking to accelerate our sale of REO assets. And that does get impacted as is volatility with interest rates. As rates move down, there's more liquidity as rates move up, there's a little more uncertainty. So that can be bumped around a little bit. We are much more aggressive with our borrowers in terms of moving forward with them. And converting some of those loans from nonperforming into REO, and that might bump up and be a little volatile as well. And a lot of this is interest rate-driven. So we don't have control of all those variables, but our goal is to try and dispose REOs as quickly as possible. We're marking them as close to where we feel the markets and brokers are. With respect to our borrowers, if they can't come up with additional liquidity and repositioning loans, we're going to move very aggressively and move that along. And as you know, certain jurisdictions are create different problems. If you have assets in Texas or Atlanta or in areas like Phoenix, you can get a hold of those assets much more quickly. If you have assets in areas like New York or Florida, it takes a lot longer. So it all depends on all those factors. But our goals are still the same.

Yes. And Chris, it's Paul. I think I've hit on all the points that driving. It's hard to predict where this goes. Things are a little bit more delayed with higher interest rates. But just to put some finer points on the numbers, you mentioned the $120 million of new REO for the quarter. Really, that number was $80 million, which was right in the range of the [ $50 million ] to $100 million that I guided to last quarter. The other $40 million were delinquent loans that we took back strategically as REO and on the same day, flip them simultaneously. So they're not really, in our minds, true REO assets that you're holding and marketing for sale over a long period of time or putting capital into rehab. Those were just strategic opportunities that we purposely foreclosed on and immediately had to take out. So really the number was $80 million. Having said that, what we've guided to is this $545 million on our books, getting down to $300 million. And yes, we'll probably add a few here or there and sell you other ones, but the timing is just hard to predict where rates are.

Got it. And I guess, how quickly does that REO sales market react rates? Like if we get some relief on rates in the back half of the year, could we see REO sales accelerate in the back half of the year, or would that slip into '27?

I mean liquidity returns very, very quickly and the sentiment changes when rates go up, you get a negative sentiment and it gets hard to move them when rates come down, it becomes very positive, and it's very dramatic. So if we return to where rates were before the Iran issue, you've seen an enormous acceleration of the dispositions of the delinquencies in the REO in a very real manner.

We'll take our next question from Rick Shane with JPMorgan.

Look, I'd like to talk about the REO sales and a couple of things here. One, can you talk a little bit about the types of buyers that are out there. And second, can you give us a sense of what percentage of seller financing you are providing on those REO sales? Are you not providing financing? Or are you generally providing financing, or help us understand that a little bit better, please.

Yes. Let me speak about the type of buyers are acquiring these assets. Generally, what we like to do is to go to our existing borrower base who have knowledge and expertise in these markets that we have experience with. That's usually our first look. And those usually done on a consensual basis where we take an asset that's showing trouble, and we know we're going to foreclose on. We bring them in a long process. So when it gets to the actual foreclosure, we can do a simultaneous transaction and avoid a lot of friction costs. I say there is a lot of friction costs. If you have to foreclose on an asset, finance it step award into our management. That's the optimum situation and usually done with people who we have great relationships and in fact have done many transactions. We've had a lot of success. But that's the preferred profile when we have existing REO assets that we've already taken back and I guess that has to do with prior strategy of trying to take the asset improvement moving along, then we'll generally [Audio gap] go to more on those the right level. But that general strategy as of now is when we have a delinquency when we have a potential REO, we premarket that asset to people we've done. That's just we try and create a simultaneous transaction. I'll let Paul go through the numbers.

Yes. So a couple [Audio gap] When we look at these REO assets, as Ivan just laid out the preferred buyer of those assets. We are generally producing some self-financing. There are occasions, where we're just taking a cash offer. We had 1 or 2 this quarter when we took a cash offer and just walked away. But we are generally providing seller financing. And one of the reasons we're doing that is; one, we'd like to obviously put our money into a good loan if it's been recapped, and they're putting in the right amount of new equity. But two, it's a certainty of execution. This is something Ivan and I talk about all the time. sometimes in certain markets, time is not your friend on certain assets that as time marches on, things could deteriorate even further. So when you have someone coming in and making a bid, if you're providing the financing, you have certainty that, that deal is going to get done in a short period of time. If you don't provide the financing, and they have financing they're bringing to the table, we've seen sometimes with that financing walks, now it's 30 to 60 days later. Things are marching on, things are getting worse and then you to the market. So the certainty of execution is something we let a lot. As far as how we're lending, I know it looks like when you look at the disclosures that the sale prices are pretty much on top of the loans. But you've got to look at it a little differently, and we've beefed up our disclosure this quarter to help people with this analysis. When someone is buying an asset, they're buying it for the purchase price, they're paying closing costs, they're bringing in CapEx, they're bringing in reserves. So the total capitalization is much higher than the purchase price in a loan and carry. So when we look at it, we're probably lending on average anywhere from 75% to 85% loan to capitalization. That's the loan to value. We're looking at some as high as 88%, some as low as 70%. But in general, we're targeting 75% to 85% of the total capitalization of that deal to be our loan.

Got it. Okay. And look, we're a month into the third quarter. Gain on sale margins have fluctuated a great deal between first and second quarter. Can you talk about that dynamic? And can you help us think about what -- where we stand quarter-to-date so that we can all refine our models around that assumption as well.

Sure. It has a lot to do with the change in profile of our business line and a lot of it's been directed by the agencies. I think if you go back to the prior administrations. There was a real push towards small balance loans, towards B&C properties, towards affordability. And we did a lot of small balance loans, and that was what was encouraged by the agencies. In the current administration, that is not the case. So we've shifted our business dramatically, and our average loan size is probably going to be more than double what it was last year, and we're doing a lot of large transactions. In the larger transactions, the fees are less and the margins are less, but also note that the labor is less and the commissions are less as well. So we are working on a significant number of larger transactions. The gain on sales will be smaller, but the expenses affiliated with those will also be significantly reduced. But that's definitely the shift in our business line.

Yes. And I would say just to guide you, guys, Rick, is that I would say the margins are probably in the range that you saw this quarter going forward, maybe a tad lower in some quarters, maybe a tad higher. But I would say the 186 margins are not here for the next few quarters as when I look at our forward pipeline, we have -- as Ivan said, we have a lot of larger deals. We're upscaling to a better borrower, a better asset class. And we think even though the margins are in and the servicing fee is in as a result, from a risk-adjusted return perspective, it's a better deal.

Got it. And I apologize to my peers for asking one last question. But interesting dynamic here. Obviously, on the agency side, you guys have an incentive to increase the loan size. Historically, the business has been make and hold in order to make and sell. Does that mean that going forward, we should assume on the structured side, balance sheet side, loans are going to be bigger as well. And can you give us a sense of sort of what the new normal loan size will be in that case?

Yes. There's no question about it that the balance sheet side has to match the agency the execution. That's correct. And that there was a big push 5, 7 years ago to a lot of CI assets interment to B or a lot of B interment to A. And that thesis was not as successful and the agencies are encouraging. So without a doubt, we are adjusting our balance sheet business. We are working on larger loans. I do want to point out that this, to me, is the most competitive market I've ever seen. I haven't seen a more competitive market on a bridge lending side of the business. I think [ '21 and '22 ] were competitive. I'm finding this more competitive because it's not just competitive on spread. It's not competitive on proceeds. It's competitive on structure as well. So what we're having to do is work on bigger loans and really weigh in, in terms of where we want to compete and put a lot of executive management into almost each and every single loan that we do. So I would say that our average loans on our bridge has been significantly higher. And you'll see a much larger loan balance. And Paul, do you have what...

So just for the second quarter, Rick, we did 5 balance sheet bridge loans totaling $160 million. So obviously, the average is is over $50 million. We had one at $50 million, one at $100 million and I think one at $20 million. So -- and in the prior quarter, we had, I think, $100 million and maybe even a $200 million loan. So I would say that the loan size is anywhere from $50 million up right now, right, Ivan, that's what we're saying?

Yes. I would say our minimum loan size is probably $25 million, and I wouldn't be surprised to have a $50-plus million average loan on the bridge.

We'll take our next question from Jade Rahmani with KBW.

Could you talk about what drove the increase in GSE risk sharing? And if there's been any loan repurchase requests from the GSEs?

Sure. So we have seen, and I think all lenders have seen in the [ Fannie ] world, an increase in the delinquencies and the loss share needed to handle those delinquencies. I think delinquencies on the agency side and the [ Fannie ] side are about 3.3% of our book. We have $80 million -- $82 million in reserves tucked away. We have $51 million of specific reserves. We took another $9 million this quarter. So we have seen an increase in the delinquencies. And this is what's to be expected when you're hitting the bottom of the cycle when you're at the bottom of the cycle, this is what you normally see. It should level off here at some point, but it's about 3.3% of our portfolio. As far as buybacks, we have not had anything material brought to us from the agencies to require us to buy back. I think we had to buy back one asset, right? I mean it was -- it's a small asset, and we don't predict we're going to have any loss on it. I think it was $4 million. So we have not seen any substantial significant material buyback request at this point.

That's good to hear. Turning to the REO side, what do you expect the cumulative amount of CapEx spend to be on the remaining REO assets?

It's tough to predict because this quarter, I think CapEx was around -- having in front of me, this quarter, the CapEx was about $8 million on the assets, but it should come down because we are liquidating these things quickly, Jade. So we're not looking to -- if we have something lined up that we're brokering and have good bids on, we'll look to turn and sell that quickly. But we did $8 million for the quarter. I don't know if it stays there. We'll have a couple of new ones. We'll have some runoff. It all depends on [indiscernible] a tough number to [indiscernible] our hands around.

I think the real comment that I have on that is on a go-forward basis, we're looking to dispose of loans that go from delinquent to REO, not taking them on management and invest in them. There were a lot of assets we took back earlier that were really, really got destroyed. And we felt it was best to put the CapEx and bring them up to speed. We think it's better to transition those assets even if we bring in a partner or maintain an interest who's more adept than we are. But we're not looking to build up an inventory of heavy CapEx REO.

SP1 We'll take our next question from Crispin Love with Piper Sandler.

First, Paul, can you share your net interest income outlook and trajectory going forward off of the second quarter levels and just some of the puts and takes there?

Sure. So I think as we said in our commentary, we are making a very big push and being very aggressive at resolving or delinquencies as quick as possible and also the legacy book that I've had this commentary. We're trying to bring that down to a very nominal number as a portion of our total portfolio. So we do think -- and with rates being elevated, we do think things are slower, it's taking longer, and it will put some pressure on our short-term earnings. I think that -- the things that offset that are the significant expense reductions we mentioned today on the call, in the cuts we made in staffing and also the fact that buying back a significant amount of stock, which we think is 1 of the best investments we could make especially where it's trading relative to book is very accretive going forward to diluted EPS and distributable earnings. So I think all those things weighing together, I think we're expecting distributable earnings to be in this range, probably in the 15 to 17 range over the next 2 or 3 quarters until we get a lot of this behind us. And then like I said in my commentary, we expect that we'll start to see some movement up in 2027 as we make more meaningful progress. On the net income side, we could see some losses over the next couple of quarters just because we are being more active in resolving things and taking more reserves. I think I said in my commentary, we expect -- we think, given the market we could take similar reserve levels going forward. Now general CECL was a little higher this quarter due to just the way the models work. I don't know if that continues. But on the specific side, we are expecting to take similar specific reserves going forward over the next few quarters until we can get this behind us.

Great. I appreciate that. And then just on agency originations, definitely strong in the quarter despite the rate moves we saw. But can you discuss what drove that? Was it just because of the larger loans or anything else? And then just relatedly, I might have missed this in the prepared remarks, but just the origination outlook and agency near term, just given break those with treasury yields trending higher?

So I think that we've developed a good pipeline and good pipeline management. What we've been very effective to do with our team is to put every single loan in the system in a rate lock position as quickly as we can and as rates were volatile and go up and down, if there's an inch day or into a weak drop of 10, 20 basis points we're able to really step up with that borrow and get them to move along. So it's really getting the pipeline in a great position. That's the goal. That's a different management technique that we've really instituted over the last 90 days. A new management team is really adept at it. So it's been very beneficial to us. We do have a lot of larger loans, so you can really pay attention on a larger loan basis and really get geared off. We have shifted our customer profile. We've done a great job with us and the pipeline is pretty sizable. And as rates continue to be volatile, I think you'll see in our estimation, the opportunity to match what we did last year in volume.

Yes. And I think it's just hard to predict the timing of those loans with where rates are. Some loans are rate sensitive, right? So in July, we did $305 million of volume. I think we had targeted over $400 million in some of those loans pushed into August given where rates are. So we're hopeful that given the size of the pipeline, that we have on the back half of the year, we can get to similar numbers, maybe within 10% of what we did last year. We just don't have the exact timing of when things could close given the rate. We did do $305 million in July, if that helps you kind of figure out where we've gone.

This concludes the time we have for our question-and-answer session. I would like to now turn the conference back to you, Ivan Kaufman, for any additional or closing remarks.

All right. Thank you, everybody, for participating. It's been a long downturn. We're extraordinarily well positioned to work through the rest of this downturn. Everybody, have a great weekend. Take care.

Thank you. This brings us to the end of today's meeting. We appreciate your time and participation. You may now disconnect.

Hello, and welcome to the Arbor Realty Trust Annual Meeting of Stockholders. Please note that this meeting is being recorded. [Operator Instructions] The meeting is about to begin.

Okay. Thank you, Victor. This meeting will come to order. Good morning, ladies and gentlemen. I'm Paul Elenio, Chief Financial Officer of Arbor Realty Trust. As we have in each of the past several years, this meeting is being conducted virtually. Unfortunately, our Chairman and Chief Executive Officer, Mr. Ivan Kaufman is unable to be with us today, so I will be chairing the meeting. Mr. John Bishar, the Corporate Secretary of Arbor Realty Trust, Inc. will serve as Secretary of the meeting.

Let me welcome all of you to the 2026 Annual Meeting of Stockholders. As we begin, I'd like to introduce the following directors who are able to join us today, Mr. Edward Farrell, Chair of our Audit Committee; Mr. Elliot Schwartz, Chair of our Corporate Governance Committee; Mr. Melvin Lazar; Ms. Caryn Effron; Ms. Carrie Wilkens.

Written notice of this meeting was mailed on or about April 16, 2026, to all holders of record of our common stock and our special voting preferred stock and the Secretary has presented affidavits certifying to the mailing of such notice. April 2, 2026 is the record date for the voting of shares at this meeting. The Secretary will file such notices and certificates with the minutes of this meeting.

Equiniti Trust Company, LLC, represented by Ms. Christine Pino, has been designated the Inspector of Elections and is participating today. The Inspector has taken the oath of office, which I direct to be filed with the minutes of this meeting. The Secretary will now report the number of shares of common stock and shares of special voting preferred stock entitled to vote at this meeting.

As of the close of business on the record date, April 2, 2026, Arbor Realty Trust, Inc. had outstanding and entitled to vote 192,361,203 shares of common stock and 16,170,218 shares of special voting preferred stock, each share of which is entitled to 1 vote. The Inspector of Elections has determined that a quorum is present. Therefore, this meeting is duly organized for the transaction of business.

We will now proceed with the business agenda. Each of the following proposals is described in detail in the proxy statement distributed in connection with this meeting. I am pleased to announce that based on the preliminary report from the Inspector of Elections, each of the proposals presented at the meeting has been approved. The preliminary voting results are subject to final certification and the company will report final voting results in a Form 8-K to be filed with the SEC within 4 business days after the meeting.

The first item of business on the agenda is the election of 4 Class II directors, each for a 3-year term to serve until the 2029 Annual Meeting and until their successors are duly elected and qualified. The Secretary has advised that under the company's bylaws, the only persons who have been properly nominated are those nominees listed in the company's proxy statement.

Ms. Carrie Wilkens; Mr. Ivan Kaufman, Mr. Melvin Lazar, and Mr. John Natalone as the Class II Directors. I therefore declare that the nominations for directors are closed. The following resolution is deemed duly presented at this meeting. Resolved, that the stockholders of Arbor Realty Trust, Inc. hereby vote to elect each of Ms. Carrie Wilkens, Mr. Ivan Kaufman, Mr. Melvin Lazar, and Mr. John Natalone, each to a 3-year term, each to serve until the 2029 Annual Meeting and until their successors are duly elected and qualified. The Board of Directors unanimously recommends that the stockholders vote for this proposal.

The second item of business on the agenda is the amendment and restatement of the company's 2024 amended Omnibus Stock Incentive Plan to authorize the issuance of an additional 8 million shares of the company's common stock for grants of restricted stock, restricted stock units or underlying stock options under the new stock incentive plan and to extend the term of the plan to May 20, 2036. The following resolution is deemed duly presented at this meeting. Resolved, that the stockholders of Arbor Realty Trust, Inc. hereby adopt and ratify the amendment and restated of the company's 2024 amended Omnibus Stock Incentive Plan and extended the terms of the plan to May 20, 2036. The Board of Directors unanimously recommends that stockholders vote for this proposal.

The third item of business on the agenda is the ratification of the appointment of Ernst & Young LLP as the company's independent registered public accounting firm to examine the report on the financial statements of the company for the 2026 fiscal year. The following resolution is deemed duly presented at this meeting. Resolved, that the stockholders of Arbor Realty Trust, Inc. hereby ratify the appointment of Ernst & Young LLP as the company's independent registered public accounting firm for 2026 fiscal year.

Board of Directors unanimously recommends that the stockholders vote for this proposal. Representatives of Ernst & Young LLP are participating in the meeting today and are available to respond to appropriate questions from stockholders during the question-and-answer period.

The fourth item of business on the agenda is the adoption of a nonbinding advisory resolution to approve the compensation of the company's named executive officers. The following resolution is deemed duly presented at this meeting. Resolved, that the stockholders of Arbor Realty Trust, Inc. hereby ratify a nonbinding advisory resolution to approve the compensation of the company's named executive officers. The Board of Directors unanimously recommends that stockholders vote for this proposal.

The polls are now open for voting on these proposals. If you have previously voted by proxy, you do not need to vote during the meeting unless you wish to change your vote. Voting at the meeting revokes any prior proxy you may have submitted. If you're attending this meeting as a stockholder of record or a beneficial owner and you have logged into the meeting by following the instructions given in the proxy statement, you can vote your shares by clicking the Vote My Shares tab at the top right of your screen and then following the prompts. We will close the polls promptly following the conclusion of the question-and-comment period.

The floor is now open for appropriate questions or comments from stockholders of the company. Being respectful of the time and interest of all stockholders questions and comments are appropriate to the extent that they relate to the 4 proposals on which the stockholders are being asked to vote. Any other questions should be directed to Investor Relations at [email protected]. If you're attending this meeting as a stockholder of record or a beneficial owner and you've logged into the meeting using your control number, you can ask a question by clicking the question box to the right of your screen, typing your question into the text box and clicking the submit button. The floor is now open for stockholder questions or comments.

Secretary, do we have any questions?

There are no questions, Paul.

Thank you. That concludes our question and comment period. If you intend to vote your shares during the meeting, please do so now. Again, I emphasize that if you previously voted by proxy, it is not necessary to vote during the meeting unless you wish to change your vote. Voting during the meeting revokes your prior proxy. I will pause a moment for any last votes to be submitted.

John, are we waiting for any additional votes?

No, there are no additional votes.

Great. Thank you. I now declare the polls closed. All votes and proxies are now in the custody of the Inspector of Elections. Mr. Bishar, will you please report on the voting results?

The Inspector of Elections has preliminarily determined that each of the 4 proposals has been approved by the necessary votes. Therefore, all of the resolutions have been duly adopted.

The Secretary will file the master ballot and the final report of the Inspector of Elections with the minutes of the meeting of the company's stockholders. We are pleased by your attendance at this meeting. We are grateful for your interest and support of Arbor Realty Trust, Inc. This concludes the 2026 Annual Meeting of the company's stockholders. I declare this meeting adjourned.

Good morning, ladies and gentlemen, and welcome to the First Quarter 2026 Arbor Realty Trust Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

[Operator Instructions] I would like to now turn the call over to your speaker today, Paul Elenio, Chief Financial Officer. Please go ahead.

Okay. Thank you, Stephanie, and good morning, everyone, and welcome to the quarterly earnings call for Arbor Realty Trust. This morning, we'll discuss the results for the quarter ended March 31, 2026. With me on the call today is Ivan Kaufman, our President and Chief Executive Officer.

Before we begin, I need to inform you that statements made in this earnings call may be deemed forward-looking statements that are subject to risks and uncertainties, including information about possible or assumed future results of our business, financial condition, liquidity, results of operations, plans and objectives. These statements are based on our beliefs, assumptions and expectations of our future performance, taking into account the information currently available to us. Factors that could cause actual results to differ materially from Arbor's expectations in these forward-looking statements are detailed in our SEC reports.

Listeners are cautioned not to place undue reliance on these forward-looking statements, which speak only as of today. Arbor undertakes no obligation to publicly update or revise these forward-looking statements to reflect events or circumstances after today or the occurrences of unanticipated events.

I'll now turn the call over to our President and Chief Executive Officer, Ivan Kaufman.

Thank you, Paul, and thanks to everyone for joining on today's call. As you're all aware, our stock has been subject to attacks by short sellers in recent years. Some of those short reports appear to have provoked investigative interest from regulators as well as class actions and derivative claims from plaintiffs law firms. We have steadfastly maintained that these attacks and claims they made against us were baseless and misleading. We are pleased to report in that regard that we believe that any pending investigations that were initiated in the wake of the short reports have now been closed without any action against us.

Additionally, and very recently, our motion to dismiss the class action lawsuit against us was granted and the claims dismissed without prejudice. We are very pleased with these developments. Although our management team never lost sight of our shareholders and their interest during this challenging period, we are happy to put this chapter behind us and to focus on creating shareholder value free of these costly and unwarranted distractions.

On our last earnings call, we discussed at length we feel we are at the bottom of the cycle and have ring-fenced the majority of our nonperforming and subperforming loans and are working exceedingly hard at accelerating the resolution of these loans into performing assets, which will allow us to start to build back our run rate of interest and income for the future. This is our top priority as these loans are having a tremendous drag on our earnings. We also mentioned that if rates went down, the process would accelerate. And if rates increase, it would lead to a longer period of time needed to resolve these loans.

Unfortunately, given the geopolitical landscape, the 5- and 10-year have actually increased roughly 50 basis points in the first quarter, which is certainly pushing our timetable out a little bit. Despite these challenges, we continue to make progress in working through our assets. And again, we believe we have a clear line of sight on resolving a bulk of these assets over the next several quarters. We ended the first quarter with approximately $500 million in delinquencies and around $500 million of REO assets for total nonperforming assets of roughly $1 billion. These numbers are down approximately $100 million from the last quarter or a 9% reduction. This is steady progress.

And again, our goal is to continue to accelerate the resolution of our noninterest-earning assets and redeploy the capital into performing loans and grow our run rate of income. We had $200 million of new delinquencies in the first quarter and $300 million of resolutions, which is consistent with our goal of continuing to shrink our total delinquencies each quarter. Additionally, we have line of sight on roughly another $200 million to $300 million of delinquencies we expect to resolve in the second and third quarter in addition to another $100 million, we believe we have the potential to resolve by the end of the year.

We also remain optimistic that we can reduce our REO assets to around $250 million to $300 million by the end of 2026, even after adding an additional $100 million of REO assets over the next few quarters, which were already reflected in our delinquency numbers at March 31. We have been actively marketing several of these assets for sale, which will go a long way towards helping reduce the drag on earnings and increase our run rate of income for the future. As we discussed in detail in our last quarter, we continue to focus heavily on our legacy portfolio, which currently sits at approximately $5 billion.

$500 million of these loans are delinquent and we are working through very aggressively and $1.5 billion continue to perform in accordance with their original terms. The other $3 billion have been modified to pay and accrue features, of which only half of these loans we are accruing the full rate of interest on. We continue to make progress in reducing the amount of accrued interest outstanding on certain loans at the subset by resetting the rates to today's market spreads and requiring that the borrowers pay down a large portion of the outstanding accrued interest as part of the modified terms.

In fact, we are currently working on several loans totaling approximately $400 million that we think we can modify in the second and third quarter that will result in receipt approximately $19 million in back accrued interest and reducing the loans outstanding and accrued interest down to around $1.1 billion. This is a very effective strategy that will also put these loans in a much better position to cover our debt service from property operations and is resulting in improved terms from our line lenders. This, combined with having the proper guarantees and requiring the borrowers to commit significant additional capital to support their deals gives us comfort about how these loans will perform going forward and will greatly limit the potential risk of future losses.

As Paul will discuss in more detail, we produced distributable earnings of $0.18 a share in the first quarter. Clearly, our earnings are being greatly affected by the significant drag from our noninterest-earning assets as well as from resetting legacy loans to today's market rates. This is something we believe we will improve in the next several quarters. We continue to make progress in resolving our legacy issues and grow our business volumes. Our first quarter numbers were also affected as we expected, by a normally slow start in the agency business from the seasonal nature of that platform, which was also impacted by the increase in rates.

On our last call, we mentioned that we would continue to evaluate our dividend policy based on how quickly we think we could resolve our delinquent loans and subperforming loans and reduce that drag on earnings. With the recent increase in rates as well as the expectation that rates can continue to remain volatile, we are now predicting a slightly longer time line in resolving these loans. As a result, the Board has decided to reset our quarterly dividend to $0.17 a share.

We believe this is the dividend we will be able to cover from earnings for the rest of the year and the potential for growth in the later part of the year and in 2027 as we work aggressively to reduce the earnings drag from our legacy assets and improve our run rate of interest income. We also believe it is very prudent in this current environment to retain our capital to fund the growth of our platform and to buy back stock where appropriate, which generates strong risk-adjusted returns on our investment.

Turning now to the production numbers for the first quarter in our different business lines. In our agency platform, we originated $708 million in volume. In addition to our first CMBS brokerage transaction of $88 million of the total first quarter volume of $795 million. These numbers were in line with our previous guidance as we normally experience a lighter first quarter due to the seasonal nature of the business. Despite the challenging rate environment, we are seeing an influx of new opportunities that are increasing our current pipeline significantly. We're off to a good start for quarter 2 with $350 million of volume closed through the first week of May, and we still feel we can produce similar volumes as last year and a strong second half of the year, which is obviously rates dependent.

In our balance sheet lending business, we originated $400 million of volume in the first quarter. This business continues to be incredibly competitive. And as a result, we are being highly selective and are focusing our attention on larger deals with high-quality sponsors. The bridge lending business is a very important part of our overall strategy as it generates strong levered returns on our capital in the short term while continuing to build up our pipeline in future agency deals.

And with the significant efficiency we continue to see in the securitization market, and with our line lenders, we are able to produce strong returns on our capital despite the competitive landscape. In fact, in the first quarter, we issued another CLO with very attractive pricing and terms. We priced the deal at 1.73% over and 88% leverage with a 2.5-year replenishment feature. This was an incredible accomplishment, especially in light of the fact that we priced the deal during the height of the Iran conflict.

We continue to have access to this market and are a leader in this space, which allows us to finance our new originations with nonrecourse, non-mark-to-market debt and drive higher returns on our capital. In our single-family rental business, we experienced an unusually slow start to the year, which was primarily driven by the noise surrounding the housing bill that is being considered. This bill in its current form surprisingly does not have a full carve-out for the build-to-rent business as initially expected and definitely keeps folks on the sidelines due to this uncertainty.

There's been a tremendous amount of talk lately that this bill will not get passed in its current form and will be serious considerations to building in the appropriate carve-outs for the build-to-rent business, including removing the for-sale provisions in year 7 that currently exist in the proposed legislation. As a result, things are starting to loosen up now that people believe this will occur, and we expect to see a real uptick in our new originations in this platform going forward.

We originated approximately $125 million in the first quarter and expect we will see a significant increase in these volume numbers over the next few quarters. This is a great business as it offers us returns on our capital through construction bridge and permanent lending opportunities and generate strong levered returns in the short term while providing significant long-term benefits by further diversifying our income stream. In our construction lending business, we continue to see our share of high-quality deals with very experienced developers. We closed 1 deal for $113 million in the first quarter and are expected to close another $250 million in the second quarter. And our pipeline continues to grow each day, give us comfort in our ability to hit our target of between $750 million and $1 billion of production in 2026.

In summary, we are laser-focused on resolving our legacy book as quickly as possible, which will reduce the significant drag that these assets are having on our earnings. We believe we have a clear path to resolving the majority of the assets over the next several quarters, which will set us up nicely and build our earnings base heading into 2027. We also continue to focus on growing the many different verticals we have and generate strong returns on our capital that are being enhanced by the significant improvements and efficiencies we continue to create on the right side of our balance sheet. We will continue to work exceedingly hard through the bottom of this cycle. And as always, we remain focused on maximizing shareholder value.

I will now turn the call over to Paul to take you through the financial results.

Thank you, Ivan. In the first quarter, we produced distributable earnings of $37.4 million or $0.18 per share, excluding onetime realized losses of $23 million from the resolution of certain delinquent and REO assets. On our last quarter earnings call, we guided to around $10 million of realized losses in Q1, all of which we had previously reserved for. We had some success resolving some loans ahead of schedule, resulting in additional $13 million in losses in Q1. We will continue to do our best to give guidance on expected resolutions, although it is a very fluid process and often hard to predict the exact timing of these resolutions. Having said that, our best estimate is a range of approximately $15 million to $25 million of realized losses a quarter for the balance of the year that we will continue to reserve for as we receive more price discovery on these assets.

As Ivan mentioned, our first quarter numbers were in line with our expectations, especially given the light first quarter we usually experience in our agency business. We also expect that it will take a little longer to work through our legacy book given the current rate environment, which will likely keep our earnings in a similar range for the next few quarters before we start to see an increase in our run rate towards the end of the year as we reduce the drag on our earnings from our underperforming assets. This should put us in a position to start to show growth in our earnings in 2027 as we realize the full benefit of converting our delinquent assets into performing loans.

With that said, the second and third quarters of this year are likely to be our low watermark and hover around $0.17 a share as we continue to reset certain subperforming loans to lower rates that will affect our earnings run rate for the next few quarters. We do expect this number to grow in the fourth quarter with further upside potential in 2027 as we're working diligently to resolve nearly all of our nonperforming assets over the next several quarters. We're estimating the second quarter will actually come in around $0.15 a share as there is roughly $0.02 a share of unusual drag from some inefficiencies related to our financing costs that are resulting in a temporary overlap of interest for a few months.

This includes the $100 million ramp feature in our new CLO that we expect to be able to fully utilize by the end of May and the timing of redrawing on our repo lines to pay off our 4.5% unsecured notes last week as we use some of the proceeds from the December bond issuance to temporarily pay down higher cost repo debt until the April notes came due. And given the nonrecurring nature of this expense, combined with the expectation that we will resolve the bulk of our delinquent loans by the end of the year, we believe we'll be able to start to grow our earnings in the fourth quarter with additional upside expected in 2027 as well.

In the first quarter, we recorded an additional $12.5 million of impairment on our REO book to properly mark these assets to where we think we can effectuate a sale. We have engaged brokers to sell the bulk of these REO assets quickly and create interest-earning loans for the future. As Ivan mentioned, we're expecting to take back roughly another $100 million of assets as we work through the bottom of the cycle, $50 million to $75 million of which will likely happen by the end of the second quarter. Most of these assets are already reflected in our delinquent numbers.

And again, we are working very diligently to dispose of these assets quickly with an estimated $100 million to $150 million of sales scheduled in the second quarter and another $200 million to $250 million expected in the third and fourth quarter. This should put our REO assets between $250 million and $300 million by the end of '26 and greatly improve our run rate of income for the future. We also booked another $9 million of specific reserves on our balance sheet loan book for total REO impairment and specific reserves of $21.5 million in the first quarter.

We expect to book similar level of reserves and impairments over the next few quarters, which is consistent with our strategy of accelerating the resolution of problem loans as we look to mark certain loans that we are marketing for disposition to where we think we can execute a sale. In our GSE agency business, we originated $708 million in volume and had $671 million in loan sales in the first quarter. The margins on these loans were very healthy at 1.86% this quarter compared to 1.36% last quarter, which was mostly due to a shift in product mix and loan size with some larger deals in Q4 that contain lower margins.

We also recorded $10 million of mortgage servicing rights income related to $734 million of committed loans in the first quarter, representing an average MSR rate of 1.32%. Our fee-based servicing portfolio of $36.3 million (sic) [ $36.3 billion ] at March 31 with a weighted average servicing fee of 35.5 basis points and an estimated remaining life of 6 years and will continue to generate a predictable annuity of income going forward of around $129 million gross annually.

In our balance sheet lending operation, our investment portfolio was $12 billion at March 31, with an all-in yield on that portfolio of 7.03% compared to 7.08% at December 31. This was mainly due to resetting rates on certain legacy loans and from the slight decline in SOFR. The average balance in our core investments was $12.04 billion this quarter compared to $11.84 billion last quarter from the full effect of our fourth quarter growth. The average yield on these assets increased to 7.5% from 7.38% last quarter, mainly due to significantly more back interest and default interest collected in Q1 on loan resolutions, which was partially offset by a decline in SOFR in the first quarter.

Total debt on our core assets was approximately $10.7 billion at March 31. The all-in cost of debt was approximately 6.4% at 3/31 versus 6.45% at 12/31, mainly due to a reduction in SOFR, along with a lower rate on our new CLO issuance in March. The average balance on our debt facilities was approximately $10.4 billion for the first quarter compared to $10.1 billion in the fourth quarter, mainly due to funding our fourth quarter growth and from a full quarter of the new unsecured debt issued in December of last year.

The average cost of funds in our debt facilities was 6.52% in the first quarter, down from 6.66% for the fourth quarter, excluding interest expense from leveraging our REO assets, the debt balance of which is separately stated in our balance sheet and therefore, not included in our total debt on core assets. This decrease is mostly due to a reduction in SOFR, which was partially offset by the unsecured debt we issued in December. And our overall spot net interest spreads were flat at 0.63% at both March 31 and December 31.

So in summary, we continue to make steady progress in resolving our delinquencies and are extremely focused on completing the process as quickly as possible, which will significantly reduce the drag on our earnings. This, combined with growing our origination platforms will go a long way towards allowing us to increase our run rate of income in 2027.

That completes our prepared remarks for this morning, and I'll now turn it back to the operator to take any questions you may have at this time. Stephanie?

[Operator Instructions] We'll take our first question from Jade Rahmani with KBW.

Could you comment on the outlook for SFR originations picking up? And also, if you can give any color on the types of borrowers that you are dealing with, the number of properties they hold, what their intended hold period is and how the financing terms from counterparties are changing the cap rates and return profile of that business?

Can you repeat the first part of that question? It didn't clearly...

Yes. Sorry about that. Could you comment on the outlook for the single-family for rent originations business? If you could provide some color on the types of borrowers you're dealing with, whether they're institutional or whether they're smaller number of properties they hold and their hold period. Just about your comments regarding the housing legislation and how that's changing that business.

Sure. Let me respond to that thought first. Let's talk about the legislation because I think the business got frozen a little bit initially with the concern and the fear. But the consensus now, a very strong consensus is those prohibitions that we're putting into that bill restrict closing the sale is not going to be put in the bill.

And as a result, we've seen a real momentum over the last couple of weeks in that business. I think we're already at $200 million, and we expect to exceed $300 million for the quarter. So we're back in line and back in pace and the enthusiasm is back in the business. Most of the people we're dealing with, a lot of their investors are institution based. A lot of them have anywhere between 5 and 30 assets. That seems to be the typical profile of what we're dealing with. Some have high net worth families, but a lot of them are institutional based.

If you can refresh me on your second part of the question?

I think it was cap rates, returns and how we're seeing the financing side of that business, which I think has been really strong, right.

Yes. Listen, the credit markets are extremely aggressive right now and the cap rates are very aggressive. It's a very, very well-liked business, and we think there's a lot of momentum in the business. So it's still viewed very, very favorably. And anything that's completed and goes to a bridge loan is priced extraordinarily competitively and the agencies, Fannie and Freddie as well as the CMBS market, they love this product.

Great. And that's really good to hear in terms of the resiliency of that asset class. Just turning to the outlook on credit. I think you touched on it that the 5- and 10-year move this year is kind of slowing the pace of resolution. But my main question would be if there's any new delinquencies or new defaults you would expect as a result of where the 5- and 10-year. I imagine that there's at least some cohort of borrowers that have been kind of on the fence as to what they're going to do and the outlook for rates makes a huge difference in their consideration. So if you could just comment on how the 5- and 10-year move this year has affected the credit outlook.

Well, I think it's very clear from management standpoint that we've taken a look at the change in the rate environment. And in the fourth quarter, we clearly had a drop in rates, and there was a lot of liquidity flowing into the sector and a lot of enthusiasm. And now with the Iran situation and rising rates and with the approach that rates will remain a little bit higher, we've adjusted our philosophy, and we're getting ahead of where we think the market is. And that's why we've adjusted our dividend to reflect a more difficult environment. We don't want to be sitting here in the second and third quarter making the adjustments.

We think that this rate environment is going to slow the resolutions. It's going to slow liquidity into the sector and it's going to slow where these resolutions go. And in fact, as Paul has guided in his comments, we're expecting to continue to have reserves going in the second, third and fourth quarter, and it's reflective of where this new environment is. So we've made the adjustments. I'm not sure everybody else has, but we do think that this new rate environment is going to affect the balance of the year, and that's what we're reflecting in our comments.

We'll take our next question from Chris Muller with Citizens Capital Markets.

I was having some connection issues, so apologies if you already hit on some of this. But looking at originations in the bridge portfolio, average loan size looked to be about $128 million versus $38 million in the fourth quarter. And I think Ivan touched on this a little bit. But was this more opportunistic? Or are you guys intentionally moving up the loan size spectrum, and we should expect to see more of this going forward?

Well, I think it's a great question. And we are definitely going in a larger loan size, but the market is extremely, extremely competitive. And it's to the point where on each individual loan, you have to make certain credit decisions in order to bring those loans on. So we've chosen to go to larger sponsor, larger deals and be more selective in that sense and put more management attention on each and every loan that we do and the larger loans gives us the ability to do that.

Got it. That makes a lot of sense. And then I guess, gain on sale margin stepped up quite a bit in the quarter to 1.86% from 1.36%. Can you just remind me if there was a large deal in 4Q numbers or something else driving that dynamic?

No, that's exactly right, Chris. So a couple of things happened. If you go back and look at our margins, you may want to go back and look at 3Q, 4Q and even 2Q of last year. If you look at 1Q and 2Q of last year, the margins were actually very strong. Similar, 1.86% is a very healthy margin. We did 1.70% -- I think we did 1.75% in the first quarter of last year, 1.70%, I think, in the second quarter of last year. Then in the third and fourth quarter, you saw that dip of 1.15% and 1.36%. So in the fourth quarter and third quarter, we had some really large off-market deals that we were able to get over the line.

And we also had a lot more Freddie Mac business in the fourth quarter, which is a different type of business. In the first quarter, we had a lot more Fannie Mae business, and we had a lot more smaller deal size. So we were able to extract a higher margin. So it all depends on what's in our pipeline. We do have a lot of large deals in our pipeline that we're working through. Our pipeline is growing each and every day. So you could see that number dip a little bit in the second quarter and the third quarter depending on deal size. But it's deal size and mix. And to your point, the fourth quarter did have some really large deals in it.

[Operator Instructions] We'll take our next question from Rick Shane with JPMorgan.

A couple of different things. In prior calls, you had talked about some fairly substantial capital investments in REO properties. I'm curious how much you guys have spent life to date in terms of that CapEx and what you expect going forward given your sort of expectations for additional REO?

Sure, Rick. So I think we look at it a couple of different ways. We break down the REO book. As I said in my commentary, we've been in the process recently of engaging brokers and really trying to find people that are interested in these assets that are experts in that particular market with that particular asset. And we're doing a really nice job, I think, of getting a significant amount of bids. There's certainly more capital out there now chasing deals.

So we've seen a real influx in opportunities to dispose of the assets quicker, which is why we are guiding to getting our REO book down to roughly $250 million to $300 million. I would say that from a CapEx perspective, there are certain assets that we expect to hold on to. There's a subset of assets in that $250 million to $300 million that we expect to hold on to a little longer and stabilize, and we are feeding those assets with CapEx. Let me see if I can get some numbers for you of what we've done. In the quarter, I think we put about $8 million to $10 million in CapEx in certain assets. That's what happened in the first quarter. And let me see if I can find for you what we've done life to date because that was your question, right?

Yes, and actually, it's interesting. I appreciate you referencing the comment about working with the brokers. I was curious, that's actually what precipitated -- that comment is what precipitated my question. I'm curious if there's a little bit of a change in strategy here, which instead of sort of investing and trying to potentially optimize outcome on a longer time line, whether you're sort of taking a first loss, best loss approach here and accelerating the disposals.

I think a lot of it's loan specific. If we feel we can get to the market with an asset fairly quickly without putting CapEx, and we will do it. Early on in, there were certain assets that really required CapEx to put them in a better position. So it's really an asset-specific situation.

It is asset specific, Rick, but I would say we're leaning towards the side, as you referenced, if we can resolve the asset on an accelerated basis at our mark, we're certainly looking to do that. So we've had a few of those this quarter in my commentary as part of that $23 million of realized losses, and we continue to push that way. It is asset specific, but we are definitely leaning towards if we can resolve them quicker, we will.

Got it. Okay. And then that actually relates to something that someone pinged me about, which is during the quarter, you sold a property for $25 million, provided $24.5 million bridge loan, which seems like a fairly aggressive financing structure. Is -- as you are resolving the REO, are you -- is part of the intention to provide financing for those transactions? And is that type of advance rate going to be typical of how you're approaching things? And how should we think about that from a credit perspective?

I think once again, it's asset specific, but a lot has to do with loan structures as well. So while it may be a high advance rate, there are capital commitments and guarantees that are required on those loans from the people who are stepping into those transactions. So we'll look at our recoveries on our returns and look at it on each particular case. Many of the times, as borrowers we've done a lot of business with have strong balance sheets. And while we may give them a high level of leverage going in and create a very seamless process, their commitment to maintain that asset with the right guarantees of CapEx and interest guarantees offset that high leverage.

Got it. Okay. And then last question, and I know I've asked several. But as we think about dividend policy going forward, and again, it really -- you guys have clearly trued the dividend up to distributable earnings. I know different commercial mortgage REITs talk about distributable earnings ex realized losses, not always our favorite metric. I'm curious, as we are looking at our models, what do you think we should use as the guidepost for dividend? Is it distributable earnings? Or is there something else, particularly, obviously, we know that you guys have an incentive ultimately to increase the dividends aligned with shareholder interest. I want to make sure we're looking at the right metrics so that we catch any inflections either up or down going forward.

Sure. Good question. So we clearly look at it distributable earnings ex the realized -- onetime realized losses that we've provided for already that have reduced book value already. That's how we look at it. What are we earning from a cash perspective, and that's how we look at it. So in this quarter, we put up $0.18 ex the losses. What I've guided you guys to is a little bit of a low watermark in second and third quarter.

Absent the $0.02 onetime drag that probably puts me at $0.15 for the second quarter, I think we're really at $0.17 if you add that back in Q2 and $0.17 in Q3. And then what I've guided to is if we can execute our business strategy very effectively, which we're laser focused on and really start to turn a lot of these nonperforming assets into performing assets, we'll start to see growth in the fourth quarter in that distributable earnings number. So we've set the dividend where we think we can earn it for the rest of the year, and we've set it to where we think distributable earnings will be ex those onetime losses.

We'll take our next question from David Farnum with Raymond James.

So roughly 40% of your loan portfolio is in Texas and Florida, where there's quite a bit of housing supply across multifamily, SFR and single-family housing. Can you please provide some updated commentary on what you're seeing on the ground in those geographies?

Sure. What we're really seeing is being at the bottom of the market. I think the last 24 months, there's been an extreme amount of softness, but we're seeing [indiscernible] month. I think some of the issues that we faced in the Texas market and in the Florida market, in particular, and also in the Atlanta market, the issue with immigration and the issue with the ICE rates has really had a real negative impact on the portfolio and has really accelerated some of the delinquencies. We've had assets that were 90% occupied for years and years and years and occupancy dropped to [ 50% ] overnight.

So over the last 12 months, I think with the ICE rates, it had a negative impact in those markets. But that's kind of getting behind us at this point. And we are seeing a reset of rental rates, growth in occupancy rates. But we also saw for a period of time, there's a real slowness and a real issue with respect to credit on our tenants and also the inability to remove them from occupancy. That has changed, the court system has sped up, and I think that the software that's been put in place and the discipline that put in place to catch fraud and put the right tenant base in place, that's improved dramatically as well.

The other thing we're seeing is we're accelerating our efforts in terms of assets that are not performing properly. We are requiring management changes, and we're taking control over these assets. It's generally the case when assets are cashed off that they get poorly managed. So we've taken very aggressive steps to make those corrections. And that's why it's reflected a little bit in our forecast because we're taking control over those assets either directly or indirectly. And during that period of time, we're going to have a little bit of a drag on earnings while we're doing it. But we're seeing the benefit of our effort by seeing a real stabilization in these assets and our growth back in occupancy and operating income.

We'll take our next question from Jade Rahmani with KBW.

I wanted to ask you about the CECL reserve or the credit loss reserve, which currently stands at $131 million, which is 1.1% of the portfolio. You said you expect realized losses of about $15 million to $25 million a quarter for the next 3 quarters, so that's $70 million. Assuming that comes out of CECL, there would be a remaining $60 million of CECL, which is 0.6% of the portfolio. So I think the question is if you're going to be taking additional CECL reserves in future quarters and if there's a normalized CECL reserve ratio that should be on this portfolio. You mentioned that there's about $1 billion of nonperforming assets, including REO and nonaccruals.

Sure. So Jade, I think you can't look at it just on the nonperforming assets on the delinquencies. You got to look at it with the REO assets. So yes, we have $130 million of CECL on the balance sheet book. And obviously, we have $500 million or $481 million of delinquencies on the balance sheet book. But we also have $520 million of REO assets that we took another $12.5 million of impairment this quarter, took $20.5 million the prior quarter. And before those loans were transferred to REO, we had booked CECL reserves on those. So there's about $85 million of reserves sitting in the REO book. So that REO book has been written down by $85 million. So you've got to take that $85 million, you got to take the $130 million and you've got to divide it over the REO and delinquency book, which puts your ratio more like 1.7, 1.8x, and that's probably the right ratio.

To answer your second question, a couple of parts of your question I want to address. One is, yes, we've guided to $15 million to $25 million realized losses going forward, but not all of those will be delinquencies. Some of those will be REO. So you've got to look at those buckets together. That's how we look at it. And then third, yes, we are guiding that in this market, given the interest rate environment, given the fact that we've engaged brokers and we're getting more price discovery on assets that it's hard to sit here and tell you exactly what the numbers will be. But if past performance is an indication of future events, given this rate environment, we think the range of CECL reserves we'll be booking, including impairment on REO, is probably in the same range for the next few quarters. Does that help answer that question?

Yes, definitely. Lastly, I think I missed the weighted average of the interest rate on the portfolio, which is 6.49%, could you parse out the weighted average cash pay rate or current pay rate?

Sure. I can do that for you. So yes, 6.49% is the pay rate, but another, call it, 25 basis points of that is origination and exit fees that we accrete over time. So that is cash. And then the other $25 million is PIK. But I want to talk about the PIK a little bit because we had some commentary on the call that I think is helpful for you guys. So during the quarter, we booked just about $5 million of PIK interest on our bridge loans, okay? We have about $2 million of PIK interest on our mezz and PE, but that's standard. That's how mezz and PE operates. You put on a mezz and PE behind an agency, you get a certain pay rate and the rest of the PIK. So that's always happened since the beginning of time on our mezz and PE.

But on the balance sheet business, the bridge business, the PIK for the quarter was down to $5 million. If you remember going back about a year ago, that PIK was probably about $18 million. So certainly, it's come down a lot. It's a lot smaller portion of our earnings for a few reasons. One, SOFR has dropped. Two, we've worked out a lot of the loans, and we've reset them at current rates and the PIK has now been paid or recovered and doesn't have PIK going forward.

And as Ivan mentioned, we're working on a bunch of deals now, some pretty big ones that we're going to get a fair share of that PIK paid back and then it won't have PIK going forward. So I'm thinking that in that all-in rate that we gave you of 7.03% was probably $5 million-ish of balance sheet loan PIK and maybe $2 million-ish of mezz and PE PIK. I think the balance sheet PIK will actually go down because when we work these loans out, there won't be PIK. So I'm thinking that $5 million a quarter goes down to probably like $4 million a quarter on the balance sheet loans.

I'm showing no additional questions at this time. I'd like to now turn the conference back to Ivan Kaufman for any additional or closing remarks.

Thank you, everybody, for your participation today, and everybody, have a great weekend.

Thank you. This brings us to the end of today's meeting. We appreciate your time and participation. You may now disconnect.

Good morning, ladies and gentlemen, and welcome to the Fourth Quarter 2025 Arbor Realty Trust Earnings Conference Call. [Operator Instructions] I would now like to turn the call over to your speaker today, Paul Elenio, Chief Financial Officer. Please go ahead.

Okay. Thank you, Angela, and good morning, everyone, and welcome to the quarterly earnings call for Arbor Realty Trust. This morning, we will discuss the results for the quarter and year ended December 31, 2025. With me on the call today is Ivan Kaufman, our President and Chief Executive Officer.

Before we begin, I need to inform you that statements made in this earnings call may be deemed forward-looking statements that are subject to risks and uncertainties, including information about possible or assumed future results of our business, financial condition, liquidity, results of operations, plans and objectives. These statements are based on our beliefs, assumptions and expectations of our future performance, taking into account the information currently available to us.

Factors that could cause actual results to differ materially from Arbor's expectations in these forward-looking statements are detailed in our SEC reports.

Listeners are cautioned not to place undue reliance on these forward-looking statements, which speak only as of today. Arbor undertakes no obligation to publicly update or revise these forward-looking statements to reflect events or circumstances after today or the occurrences of unanticipated events.

I'll now turn the call over to Arbor's President and CEO, Ivan Kaufman.

Thank you, Paul, and thanks to everyone for joining us on today's call. Today, we will focus on how we closed out 2025 as well as our outlook for 2026, especially in the first half of the year.

As we discussed in the past, we believe we are at the bottom of the cycle and are working very hard to accelerate the resolution of our nonperforming and sub-performing loans into performing assets and improve our rate of income for the future. This is a top priority for us as these loans are having a tremendous drag on our earnings.

In fact, as Paul will lay out in more detail, we estimate, as we resolve these loans, we will add back as much as $100 million of income to our annual run rate or about $0.48 a share. This is an important point, and we believe we have a clear path to resolving the majority of these loans over the next few quarters, which will put us in a position to start to build back up our run rate of interest income.

Additionally, we also have a very active originations business with several diverse platforms and will continue to contribute to growing our income streams and increase our future earnings.

We ended the year with approximately $570 million in delinquencies and around $500 million of REO assets, for total nonperforming assets of roughly $1.1 billion. These numbers are down by over $130 million from the last quarter or an 11% reduction.

This is strong progress in 1 quarter. And again, our goal is to continue to accelerate the resolution of our noninterest-earning assets and redeploy the capital into performing loans and grow our run rate of income. In fact, we have a line of sight in roughly $100 million to $150 million of delinquencies that we expect to resolve by the end of March and another $100 million to $150 million we believe will resolve in the next 90 days.

We're also very optimistic we can reduce our REO assets to around $250 million to $300 million by the end of 2026, even after adding an additional $100 million to $200 million of REO assets along away, mostly all of which are already reflected in the $570 million of delinquent loans reported by year-end.

This estimated pace of resolutions will go a long way towards significantly reducing the drag on earnings and increasing our run rate of income for the future.

We continue to focus heavily on legacy assets, which currently sit around $5 billion. $570 million of these loans are delinquent that we are actively working through, and $1.5 billion is performing in accordance with the original terms. The other $3 billion have been modified to pay in accrued structures. On roughly half of these loans, we are currently accruing the full rate of interest. On the other half, we are being more conservative and only recording the pay rate of interest.

We generated around $2 billion of runoff in 2025 in our balance sheet loan book, approximately $1.5 billion of which is related to legacy book. Given these -- given where rates are today, we believe we will experience similar runoff in 2026, which will continue to reduce our legacy book down to a much smaller number by year-end.

Certainly, if rates come down even further, we could accelerate the runoff process as well as further reduce the legacy book. We are closely monitoring the performance of these assets. And while we believe we will experience some additional delinquencies as we work through the bottom of the cycle, we are seeing steady progress on the bulk of this portfolio, which we believe indicates that the worst is behind us.

One of the ways we are resolving our legacy book is by resetting the interest rates on certain loans to today's market spreads, which puts these loans in a position to positively cover debt service from property operations without a shortfall. This, combined with having the right guarantees and requiring the borrowers to commit significant additional capital to support their deals, gives us comfort about how these loans will perform going forward.

This strategy does temporarily affect our earnings, and we are resetting these loans to lower rates. However, this will result in improved terms from our line lenders and greatly limit the potential risk of future losses, which is very important and will allow us to preserve our book value.

As Paul will discuss in more detail, we produced distributable earnings of $0.22 a share in the fourth quarter. As stated earlier, our earnings are being greatly affected by the significant drag from our noninterest-earning assets, which is something we are working very hard to chip away with over the next several quarters, and we continue to make progress in resolving our legacy issues and grow our new business volumes.

It is very important to highlight that despite the significant drag we are currently experiencing, we still produced strong earnings of over $200 million last year and we have managed through this very long elevated rate environment without a material decline in book value, unlike the rest of our peers who have experienced significant book value deterioration.

Despite these accomplishments, we are trading at significant discount to book value. We believe our stock is substantially undervalued, especially in light of our extremely valuable Agency Business and diverse and growing business platforms compared to most of our peers, who have monoline businesses.

One of the opportunities available to us right now is to liquidate nonperforming assets and use a portion of those proceeds to buy back stock at a significant discount to book value. This is a tremendous trade for us as it allows us to actually grow our book value and generate mid-teens returns on our investment.

We have approximately $120 million left in our buyback plan. And in the fourth quarter, we entered into a 10b-5 plan that allows us to purchase stock in a blackout period. We purchased roughly $20 million of stock in the last few months under this program at an average price of $7.40 or 64% of book value. We will continue to evaluate the strategy, and the future is highly accretive to both earnings and book value at these levels.

Turning now to the production numbers of 2025 in our different business lines. We had a very active strong fourth quarter in our Agency platform with $1.6 billion in origination volume, which puts us at a $5 billion for the year. This is a 13.5% increase from our 2024 production numbers in what was a very challenging rate environment for the majority of the year.

We're extremely pleased with these results and believe this is a real testament to the value of our franchise and the resiliency of our originations network with a loyal borrower base that we have cultivated over many years.

We have a large pipeline, which combined with the current rate environment and the fact that the agencies have increased their caps by 20% for 2026, gives us confidence in our ability to produce very strong volume numbers again in 2026.

As we talked about in the past, our servicing portfolio, which is now over $36 billion and grew another 8% in 2025, generates a very predictable and growing annuity of over $128 million a year of income. This annuity, combined with the earnings on our escrow balance, generates around $200 million a year in annual cash earnings.

This is addition to the strong gain on sale margins we generate from our originations platform, and it is extremely important to emphasize that our Agency Business generates approximately 50% of our net revenues, the vast majority of which occurs before we even turn the lights on every day.

In the balance sheet lending business, we originated $340 million of volume in the fourth quarter, closing out 2025 with $1.2 billion of production. This business continues to be incredibly competitive and with consistent concessions being given on credit and structure.

This is not something we will sacrifice to win a deal. And as a result, we are being very highly selective and are focusing on attention on larger deals with higher quality sponsors. This will likely result in originations similar to volume of 2026 of approximately $1 billion to $1.5 billion, which we can easily scale up if the landscape becomes more constructive throughout the year.

The bridge lending business is a very important part of our overall strategy as it generates strong levered returns on our capital in the short term while continuing to build up the pipeline of future agency deals. And with the significant efficiencies we continue to see in the securitization market and with our line lenders, we're able to produce strong returns on our capital despite the competitive landscape.

We continue to do an excellent job in growing our single-family rental business. We originated approximately $580 million in new business in the fourth quarter and $1.6 billion in 2025. We also had a very strong pipeline, giving us confidence that we will be able to produce approximately $1.5 billion to $2 billion in volume again in 2026.

This is a great business as it offers returns on our capital through construction, bridge and permanent lending opportunities and generate strong levered returns in the short term while providing significant long-term benefits by further diversifying our income streams.

And again, with the enhanced efficiencies we are seeing in the securitization market and in our bank lines, we are generating mid- to high returns on our capital, which will contribute to increased future earnings, especially as we continue to scale up this business.

Over the last several months, we received a lot of questions from investors how the President potential ban on institutional single-family home purchase would affect our SFR business.

First of all, we are not entirely sure what, if anything, transpires and that is just the political noise ahead of the midterm elections. Clearly, this is a serious housing shortage in the country, and we applaud any effort to address this very important issue.

We want to make it clear that we do not traffic in scattered site single-family businesses like Invitation Homes. We focus on build-to-rent business, which we believe is actually being excluded from this proposed ban. These are 200 to 300 homes and communities that are built by experienced developers and are commercial properties more akin to multifamily.

Therefore, we believe we will not be affected by any efforts to ban large institutions from buying and aggregating single-family homes and that this build-to-rent business will continue to be a very attractive solution when dealing with supply issues in the market.

In the construction lending business, we are having great success in growing out this platform with a real influx of new opportunities that we're seeing due to larger loans on high-quality assets with very experienced developers. We closed out the year with around $500 million of production and also have a very large pipeline, giving us the comfort that we can meaningfully grow this platform and produce between $750 million and $1 billion of production in 2026.

And so between our Agency Business, bridge lending business, SFR and construction platform plus our mezzanine PE businesses, we originated $8.5 billion in volume in 2025 in a very difficult environment for the majority of the year. And again, we are confident in our ability to produce consistent volumes in 2026 and potentially even some room for growth, depending on how the market conditions evolve over the balance of the year.

In summary, we had a very active and productive year with many notable accomplishments. Clearly, the outlook for the interest rate environment has improved from where it was at this time last year, and we are feeling more optimistic as a result. We believe this will allow us to continue to grow our origination volume and generate strong returns on our capital from the significant improvements in efficiencies we continue to create on the right side of our balance sheet.

We've also done a great job on preserving our book value even in the face of unprecedented elevated rate environment and reductions in property values where we have experienced over the last several years. We believe we have ring-fenced the majority of our delinquencies and have a clear path to resolving these assets over the next few quarters, which again will allow us to significantly reduce the drag on earnings and grow our future run rate of income.

I will now turn the call over to Paul and take you through the financial results.

Thank you, Ivan. In the fourth quarter, we produced distributable earnings of $46.3 million or $0.22 per share, excluding onetime realized losses of $12.4 million from the resolution of certain delinquent and REO assets that were previously reserved for and $7.3 million of income we generated through reduced tax expense in the fourth quarter from the sale of the Homewood asset.

On our last quarter's earnings call, we guided to $15 million to $20 million in realized losses for Q4, depending on how quickly we could liquidate certain assets. We had $12.4 million in the fourth quarter and have liquidated 2 other assets in January for approximately $10 million in losses for the first quarter of 2026, all of which have been previously reserved for.

On our last call, we discussed how our decision to accelerate the resolution process of certain loans resulted in a temporary increase in our delinquencies. We guided that this would reduce our fourth quarter earnings by approximately $0.05 to $0.06 a share.

This, combined with a few new delinquencies as well as some reduced rates on modified loans, resulted in an additional $0.02 of drag for Q4, which was in line with our expectations. And although we expect to have some new delinquencies as we work through the bottom of the cycle, we are working very hard to continue to resolve more delinquencies than new ones that come on.

This process takes time, which could temporarily affect our earnings. However, we are making meaningful progress and have a clear line of sight to resolving the bulk of these assets over the next few quarters, which will increase our run rate of income for the future.

As disclosed in our press release, we have roughly $600 million of delinquencies and $500 million of REO assets on our books at December 31, 2025. Currently, these assets are not producing income and some of the REO assets are actually generating negative NOI as we work through the process of stabilizing these assets and improving occupancy.

We estimate that these assets are creating a temporary drag of between $80 million to $100 million annually or roughly $0.40 to $0.48 a share, which translates into $0.10 to $0.12 a quarter of additional income that we are optimistic we will be able to create as we resolve the vast majority of these assets over the next several quarters.

With respect to accrued interest on modified loans, in the fourth quarter, we reversed approximately $4 million of previously accrued interest on new delinquencies during the quarter, which, as we discussed earlier, is reflective of where we are in the cycle.

This adjustment, combined with $7 million in back interest collected on a loan payoff in the fourth quarter, has resulted in total accrued interest on modified loans remaining relatively flat quarter-over-quarter, even after accruing an additional $10 million in interest on modified loans that are performing in accordance with their terms.

In the fourth quarter, we recorded additional $20.5 million of impairment on our REO book to properly mark these assets to where we think we can effectuate a sale as we look to dispose of certain of these assets quickly and create interest-earning assets for the future. This puts our reserves at roughly $75 million life to date on our REO book.

As Ivan mentioned, we're expecting to take back roughly another $100 million to $200 million of assets as we work through the bottom of the cycle, $50 million to $75 million of which will likely happen by the end of the first quarter.

Most of these assets are already reflected in the $600 million of delinquencies we reported at year-end, and we're working very diligently to dispose of these assets quickly, which we believe will put our REO book at between $250 million and $300 million by the end of 2026.

We also booked another $3 million of specific reserves on a new delinquency in the fourth quarter, which was offset by a $9 million recovery of a previous reserve on the Homewood asset that we sold in the fourth quarter. We expect to book similar level of reserves and impairments over the next few quarters, which is consistent with our strategy of accelerating the resolution of our problem loans as we look to mark certain loans that we are marketing for disposition to where we think we can execute a sale.

In our Agency Business, we had another outstanding quarter with $1.6 billion in originations and $1.5 billion in loan sales, which generated $21 million in gain-on-sale income in the fourth quarter. The margin on this business was 1.36%, which is up from the prior quarter due to some large off-market portfolio deals we were able to capture in the third quarter, which contained lower margins and smaller servicing fees.

We also recorded $20 million of mortgage servicing rights income related to $1.6 billion of committed loans in the fourth quarter, representing an average MSR rate of around 1.24%.

Our fee-based servicing portfolio grew 8% in 2025 to approximately $36.2 billion at December 31 on very strong 2025 originations. This portfolio has a weighted average servicing fee of 35.6 basis points and an estimated remaining life of 6 years and will continue to generate a predictable annuity of income going forward of around $120 million gross annually.

In our balance sheet lending operation, our investment portfolio grew to $12.1 billion at December 31 from originations outpacing runoff for the fourth straight quarter. Our all-in yield on this portfolio was 7.08% at December 31 compared to 7.27% at September 30, mainly due to a decline in SOFR.

The average balance in our core investments was $11.84 billion this quarter compared to $11.76 billion last quarter from growth in our portfolio. The average yield on these assets increased to 7.38% from 6.95% last quarter, mainly due to the significant nonrecurring adjustments we booked in the third quarter, including reversing $18 million of accrued interest, which was partially offset by a decline in SOFR in the fourth quarter.

Total debt on our core assets was approximately $10.5 billion at December 31. The all-in cost of debt was approximately 6.45% at 12/31 versus 6.72% at 9/30, mainly due to a reduction in SOFR, which was offset slightly by the new unsecured debt we issued in December. The average balance on our debt facilities was approximately $10.1 billion for the fourth quarter compared to $10 billion in the third quarter, mainly due to funding our fourth quarter growth.

The average cost of funds in our debt facility was 6.66% in the fourth quarter compared to 6.88% for the third quarter, excluding interest expense from levering our REO book, the debt balance of which is separately stated on our balance sheet and therefore, not included in our total debt on core assets. This decrease is mostly due to a reduction in SOFR. And our overall spot net interest spreads were up slightly to 0.63% at December 31 compared to 0.55% at September 30.

So in summary, we had a productive year and have made considerable progress and have a clear line of sight to resolving the vast majority of our delinquencies over the next few quarters, which when completed, will significantly reduce the drag on our earnings. This, combined with the growth in our origination platforms will go a long way towards allowing us to grow our run rate of income in the future.

That completes our prepared remarks for this morning, and I'll now turn it back to the operator to take any questions you may have at this time. Angela?

[Operator Instructions] And we'll take our first question from Chris Muller with Citizens Capital Markets.

Congrats on a really solid quarter here. So I guess on the GSE business, your guys' full year originations in 2025 were about $5 billion, but the FHFA increased caps pretty substantially, as Ivan mentioned in his prepared remarks. So I guess, how are you guys thinking about 2026 GSE originations relative to that $5 billion number?

I think that a lot will be dependent on interest rates and also the GSEs increase in the cap. They also have a parallel match on affordability. So you have to be able to originate at least overall certain affordability percentages to be able to get to that cap.

I think we're feeling fairly comfortable. Our pipeline is fairly strong. And I think targeting similar levels that we had last year, if these levels remain in this space, would be something that we feel comfortable with.

Got it. And then I guess looking at the servicing portfolio, fees have compressed a little bit there. Do you expect that dynamic to continue into 2026? Or are servicing fees starting to bottom out? And can you just talk about what's driving that compression?

Sure. Chris, it's Paul. So a couple of things. There's two components that are driving the compression in the servicing fees.

One, obviously, where rates have gone over the last 2 years, there's a lot more 5- and 7-year product being done in the agencies versus traditionally what was 10-year, 9.5-year yield maintenance product. So it's a little bit shorter on the curve and therefore, the servicing fees are changing over that period of time.

Secondly, during the COVID era, servicing fees were very, very high from the agencies. They've cut them back pretty substantially to more in the range of 40 to 45 -- 40 to 50 basis points on Fannie. But I think it's just a matter of we've got some higher servicing fee loans back from the pre-COVID days running off and then the new products coming on shorter part of the curve, and it's also coming on at the more normalized servicing fees.

We've run a model. We think this starts to bottom out towards the end of the year and then kind of levels off. So I think it compresses a little bit for the balance of the year as you flush out some of those higher servicing fee loans that are older and you're putting on the product that's more in today's market. But I think by the end of the year, we'll start to see that level off and then you don't have that compression anymore.

And we'll take our next question from Gabe Poggi with Raymond James.

Can I ask a little bit about just your -- you talked about the SFR book. You obviously made a lot of loans in the fourth quarter. Who knows what happens out of DC. Have you seen any credit issues in your build-to-rent borrowers? And if so, kind of any geographic color? Or has everything kind of just been home and pretty good on that front? And then I've got a follow-up to that.

I mean I'll give you the overall view. Our SFR book is really outstanding. It's probably the best performing book we have. The loans generally have institutional backup behind the sponsors. They're not syndicated loans and they're usually lower leverage.

Paul, I don't know if you have any stats in terms of performance. But from my knowledge, this book is performing exceptionally well.

Yes, I do have some stats. So as Ivan said, it's been an exceptional book for us. Obviously, the levered returns are real high, Gabe, on that book. We've done a nice job of putting together the first-of-its-kind securitization on the build-to-rent business earlier in the year. So we've been at the cutting edge of how to lever these things appropriately and get strong returns. I mean the returns are mid-teens on this business.

And as I look at the book, not a single delinquent loan or a loan on our watch is one of this type of product. So this product has been spotless from a credit perspective. And like we said, it's a great business that we're able to scale.

And then a quick follow-up. Just on the delinquent/REO book, is there any kind of geographic color that you can provide where you may be seeing more pockets of weakness? I mean I know the Sunbelt has gone through a lot. But is there any state, city, MSA, et cetera, that is weaker or stronger than another? Any color there would be helpful.

Yes. I would say that there are certain markets that are soft like Houston, and that's a factor of the history of Houston being boom and bust, but more significantly by being very adversely affected by the issues with immigration that got hit on both sides.

First, under the prior administration, you had a significant number of immigration centers next to a lot of properties, took over the properties, damage those properties severely. And then under the current administration, you're seeing a lot of ICE rates in those areas. You're seeing that in areas of Texas. Houston primarily, San Antonio a little bit and even in Dallas, which is shocking.

So you're seeing properties that were 90% occupied. Next day, they're back down to 65% or 70%. We're also seeing a little bit of that in the Atlanta area, too, in certain pockets of Florida. But those would be the markets that have significant softness.

[Operator Instructions] We'll move next to Jade Rahmani with KBW.

Thinking about credit, you mentioned you expect another tough couple of quarters, but the worst, generally, seems to be behind the company. At the same time, in this year-to-date economy, multifamily fundamentals have been pretty weak, looking at new lease growth, and then the economy is sending a lot of mixed signals with weak trends on the hiring side and the K-shaped recovery with the low end of the consumer suffering.

So are you seeing any additional headwinds year-to-date? How have the trends been so far in 2026?

So we've seen tremendous headwinds over the last couple of years. due to higher interest rates and distress on borrowers and due to the softness in the economy and bad credit and just poor operations. We think we're at the bottom. We're seeing a firming up of economic occupancy. We're seeing the properties stabilize. And in our case, what we've made a decision is any time we're not pleased with an operator or they don't have enough capital, we're stepping in and we're seeing good results with our efforts.

We're also seeing liquidity return to the market. And when we do market a distressed asset and when we get it back, we're seeing multiple, multiple bids, and we're seeing pretty good price discovery. So we think we are in the bottom. We're being very aggressive, and we're pleased with the results once we get our hands on these properties.

In terms of the current earnings level, you mentioned there's $0.48 in untapped earnings trapped inside of these NPLs. Yet earnings remain below the dividend. Could you give any thoughts? I saw that the dividend was maintained, but that was for the fourth quarter of 2025. What are your thoughts around the likelihood of the dividend being maintained in 2026?

Yes, I'll let Paul answer most of that, but our goal is to facilitate the resolution of the drag on earnings. The quicker we do that, the quicker we'll have line of sight and every day and every month matters. What used to take us 90 days to resolve a loan, a little bit of a drag, it's taken more like 120 days. So it takes us about 90 days to get on site for the property. And then the marketing process takes us around 90 days.

We're marking our book, and we took additional marks because we're trying to market to a level where we think we can get rid of these assets quickly. So the whole concept is how quickly we can get to that resolution and how fast we can take that drag on our earnings and start having returning.

Paul, do you want to give some comments?

Sure. Jade, so Ivan is correct. We look -- just to level set, we look at the dividend and the Board looks at the dividend from a more long-term perspective. So obviously, we have purposely accelerated the resolution of these delinquencies to put them behind us and get our run rate back up. That's temporarily certainly hit our earnings. And I laid out in my commentary, it's about $80 million to $100 million.

It will all determine how long it takes us to get that resolved. We have clear line of sight right now. We did resolve $350 million of loans of delinquent loans and REO loans in Q4. We put on another $270 million, which was the wave, the one more wave we said we were going to have in the fourth quarter.

We think the lion's share of this is behind us. And now our job is to continue to resolve those assets at a quick pace. If we can resolve them very quickly, the run rate will get up quicker. If it takes a little longer, it takes a little longer. But again, we look at it more long term.

A couple of other things I'll mention is, yes, we put up $0.22 today. The good news is we've seen our net interest income level off here. So we've gotten good net interest income for Q4. I'm kind of projecting that to probably stay in that range for Q1. And then hopefully, it starts to run up in Q2, 3 and 4 when we start resolving a lot of these loans.

The first quarter could be our low watermark though, as we've mentioned in the past. The Agency Business is very seasonal. We did do a pretty large number of volume in Q4 of $1.6 billion. The first quarter is normally much lighter. If you go back and look at last year's numbers, we did $600 million in Q1. I think we'll do $750 million to $800 million in Q1 this year and then run back up.

So we'll see a couple of penny drag from the gain on sale number being lower in Q1 just because of seasonality, but nothing to do with further delinquencies. We think we've ring-fenced everything, and we think that we've got a clear path to resolution.

So again, we look at it long term. We expect to be able to resolve these things quickly. If we're right, our numbers will get up quicker. If it takes a little longer, it will take a little longer, and we'll evaluate it. But again, it's too early for us to say that we're not going to be able to earn that back in a reasonable period of time.

At this time, there are no further questions in queue. I will now turn the meeting back to Ivan Kaufman for any closing remarks.

Well, thank you, everybody, for your time today. It's been a bouncy road a little bit, and we do feel that we've ring-fenced our issues and have clear line of sight to how to resolve those issues and get back to getting rid of the negative drag of earnings. It's just a matter of what the timing is. We are aggressive, and we'll look to facilitate that. And once again, thank you. And everybody, have a great weekend.

Thank you. This brings us to the end of today's meeting. We appreciate your time and participation. You may now disconnect.

Good morning, ladies and gentlemen, and welcome to the Third Quarter 2025 Arbor Realty Trust Earnings Conference Call. [Operator Instructions]

Please be advised that today's conference is being recorded. [Operator Instructions]

I would now like to turn the call over to your speaker for today, Paul Elenio, Chief Financial Officer. Please go ahead, sir.

Okay. Thank you, David, and good morning, everyone, and welcome to the quarterly earnings call for Arbor Realty Trust. This morning, we'll discuss the results for the quarter ended September 30, 2025. With me on the call today is Ivan Kaufman, our President and Chief Executive Officer.

Before we begin, I need to inform you that statements made in this earnings call may be deemed forward-looking statements that are subject to risks and uncertainties, including information about possible or assumed future results of our business, financial condition, liquidity, results of operations, plans and objectives. These statements are based on our beliefs, assumptions and expectations of our future performance, taking into account the information currently available to us. Factors that could cause actual results to differ materially from Arbor's expectations in these forward-looking statements are detailed in our SEC reports. Listeners are cautioned not to place undue reliance on these forward-looking statements, which speak only as of today. Arbor undertakes no obligation to publicly update or revise these forward-looking statements to reflect events or circumstances after today or the occurrences of unanticipated events.

I'll now turn the call over to Arbor's President and CEO, Ivan Kaufman.

Thank you, Paul, and thanks, everyone, for joining us on today's call. As you can see from this morning's press release, we had another active and productive quarter with some very significant accomplishments that are worth noting. First, we continued our strong progress of creating substantial efficiencies on the right side of our balance sheet with a new $1 billion CLO that we issued in the third quarter with tremendously accretive terms. This deal was priced at 1.82% over, contained 89% leverage on a 30-month replenishment feature and generated an additional $75 million of liquidity. The CLO securitization market is incredibly constructive, and we're very pleased to be such a prolific multifamily originator and a seasonal well-recognized securitization platform, which allows us to consistently access the market and grow our platform. And with the pricing levels we are seeing in the market, which has only gotten tighter since our last deal. We were able to compete very effectively in today's extremely competitive market and generate a strong levered returns on our capital. We also successfully [ called ] one of our legacy CLOs in October, refinancing these assets within our current banking facilities, the bulk of which have replenishment features that allow us to substitute collateral as these loans run off. This was a great trade for us as a CLO was past this replenishment period and with delevering and becoming less efficient. After the unwind, we are now financing these assets through these loans at similar prices to the CLO we redeem and very significantly, we're able to pick up another $90 million of liquidity to enhance leverage.

One of the more significant accomplishments we had this quarter was the realization of a $48 million gain from the sale of a portion of the assets in the Lexford portfolio. If you recall, our equity investment in its portfolio was a focus of the first short seller report, which claimed that this was a fraudulent transaction and that we were misleading our shareholders as to the value of these assets. In fact, the reality was exactly the opposite. This was by far 1 of the best restructurings we have ever accomplished. Including the large game we recorded this quarter, this investment has generated over $100 million of income over its life span, the return of $67 million of preferred equity on top of us receiving all of our invested capital back. And there is still a portfolio of assets that we expect to be able to liquidate in the near future. We're also very successful in selling our interest in the legacy assets that we expect to close at the end of the business today. This transaction will generate approximately $7 million of additional income in the fourth quarter, which combined with the gain from the Lexford transaction, totaled $55 million of income that we generated from 2 of our legacy investments. This gives us a tremendous amount of flexibility to be very aggressive in addressing our legacy issues. Our goal is to resolve these noninterest earning assets, which are creating a tremendous drage on earnings as quickly as possible. We believe it will take until the second quarter of next year to accomplish our goals. When completed, we will have effectively resolved a significant amount of our troubled assets and set up with a much better improved run rate of income, which will go a long way towards increasing our earnings and being able to grow our dividend again sometime in 2026. And very importantly, we will accomplish all of this with a very minimal impact on our book value, which is something no one else can say in our space.

As I mentioned on the last call, the prolonged elevated rate environment has put certain borrowers in a position where they are running out of steam and are having differently raising additional equity to continue to manage their assets. As we've stated before, we believe that the third and fourth quarter of this year will be the bottom of the cycle. And again, we are working very hard to quickly through our loan book and redeploy our capital into the performing assets and improve our run rate of income for the future. Certainly, with the 2 recent interest rate cuts and the likelihood that we could potentially see 1 more cut this year, we are starting to feel more optimistic about the rate environment moving forward which we believe will provide some need -- some much needed relief for our borrowers. This positive trend gives us some wind at our backs for the first time in a while. As this trend continues, we believe we will be able to meaningfully grow our origination volumes and start to move more assets off our balance sheet, which will increase our earnings run rate and position us well for the future. As I mentioned earlier, we've taken an aggressive approach to resolving our legacy assets through a myriad of different strategies, including modifying loans, taking back assets as REOs to own and operate and bringing in new sponsors to take over assets and assume our debt and create a more current income stream. We continue to examine all loans that are showing signs of distress and are accelerating the process of taking control of the real estate and working quickly towards an accretive resolution. This has resulted in a temporary spike in our delinquencies. And again, we look to expedite the resolution process of these loans, the number of which we have targeted to take back as REO and flip to the sponsors. This will take a few quarters to complete. And in the interim, we'll temporarily reduce our net interest spreads until we complete the resolution process. However, when the smoke clears, we will have cleaned up the vast majority of our legacy book and create a more stable and growing run rate of income in the future.

Turning now to our third quarter performance, as Paul will discuss in more detail. Our quarterly results included the large gains from Lexford investment, as I mentioned earlier. This provides us with the unique ability to be able to be very aggressive in accelerating the resolution about problems loans without materially impacting our book value. The timing of these resolutions will take place over the next few quarters, which created some lumpiness in our quarterly earnings going forward. However, we are committed to continue to make our quarterly quarter dividend for the balance of the year. And if we accomplish our goals effectively, we'll be able to improve our earnings run rate and put us in a good position to consider an increase our dividend again sometime in 2026. And I can't stress this enough, we are accomplishing all of these goals without a material change in our book value unlike the rest of our peers who've experienced significant book value deterioration. As I mentioned in our last call, the balance sheet lending business is incredibly competitive right now. There is a tremendous appetite for deals and a significant amount of capital out there chasing each transaction. As a result, we are being highly selective and have closed about $400 million in the third quarter, putting us just around $850 million of volume for the first 9 months of the year. The guidance we gave at the beginning of the year of $1.5 billion to $2 billion of bridge production for 2025 was reflective of our views that market would become overheated. And as a result, we [ dialed ] back our production numbers for 2025 to a more conservative level. We do have some large highly -- high-quality deals in our pipeline, and we think we are likely to close by year-end, which gives us confidence that we may be able to come in with our original guidance despite the extremely competitive landscape. The bridge lending business is an important part of our overall strategy as it generates stronger returns on our capital in the short term while continuing to build up our pipeline of future agency deals. And with the significant efficiencies we are seeing in the securitization market, we're able to continue to produce strong leverage returns on our capital despite this extremely competitive landscape. In the Agency business, we had a tremendous third quarter, originating $2 billion of loans, which is the second-highest production quarter in our history. We had a very strong October, originating $750 million, which puts us -- puts our 10-month volume numbers at $4.2 billion, making us very comfortable that we will easily surpass our origination guidance of $3.5 billion to $4 billion and the best year production number of $4.5 billion as well. This is a tremendous accomplishment, especially given the rate environment, which we were in for the better part of the year. This is a tremendous testament to the value of our franchise and the resiliency of our originations network with a loyal borrower base that we have cultivated over the many years. We can continue to do a strong -- we continue to do an excellent job in growing our single-family rental business. We originated approximately $150 million of new business in the third quarter and another $200 million in October, bringing our 10-month numbers to $1.2 billion. We also have a strong pipeline, giving us a comfort that we were able to meet our internal guidance of between $1.5 billion to $2 billion of production for 2025. This is a great business as it offers 3 turns on our capital through construction, bridge and permanent lending opportunities and generate strong level of returns in the short term while providing us significant long-term benefits by further diversifying our income streams. And again, with the efficiencies that we're seeing in the securitization market and in our bank lines, we are generating mid- to high teens returns on our capital, which will contribute to increased future earnings, especially as we continue to scale up this business.

In our construction lending business, we are having great success in growing out this platform with a real influx of new opportunities that we're seeing to do larger loans on high-quality assets with very experienced developers. In the third quarter, we closed $145 million of deals and another $65 million in October, bringing our 10-month numbers to around $500 million. We also have a very large pipeline of roughly $185 million on the application and another $675 million of additional applications outstanding and $900 million of deals we are currently streaming, given us confidence that we can up our guidance for the year from an initial $250 million to $500 million to $750 million to $1 billion for 2025. And Additionally, and very significantly, the size of our current pipeline gives us real visibility into how we will be starting our 2026 which, by all indications, we expect will produce meaningful growth over 2025 numbers.

And so between our Agency business, bridge lending program, SFR and construction platform plus our mezz and PE business, we expect to originate between $8.5 billion and $9 billion in volume this year and what was -- which was a very difficult environment for the vast majority of the year. And again, we have started to feel more optimistic about the rate environment, which we believe will lead to more robust origination volumes in the future.

In summary, we had a very active and productive quarter with many notable accomplishments. Clearly, the outlook for interest rate environment has significantly improved from where it was in the beginning of the year, and we are feeling more optimistic as a result. We feel we now have some wind at our back and will allow us to continue to grow our origination volumes and generate strong returns on our capital from the significant improvements and efficiencies we've created on the right side of our balance sheet. We also feel that the gains we have generated from our legacy investments puts us in a great position to accelerate the resolution of our legacy book and create a much improved run rate of income going without materially affecting our book value. And these improvements, coupled with the growth we are expecting in our originations platform will go a long way in allowing us to achieve our goals by being able to grow our earnings and dividends in the near future.

I will now turn over the call to Paul to take you through our financial results.

Okay. Thank you, Ivan. In the third quarter, we produced distributable earnings of $73 million or $0.35 per share. As Ivan mentioned, a portion of these earnings were associated with the large gain we recorded this quarter from the Lexford exit investment, which will partially offset some of the accelerated losses and reduced interest income we will experience temporarily from the resolution of some of our problem loans over the next few quarters. As discussed earlier, the resolution of these loans will likely occur over several months, which will cause our quarterly earnings to fluctuate during that time period. We will do our best to give as much forward guidance as we can as it relates to any realized loss we will incur as we continue to make progress in resolving the bulk of our legacy book.

As to the income Ivan mentioned, that we will generate in the fourth quarter from the sale of our legacy asset, this is related to a land development deal that we have -- we've had on our books since 2006 called Homewood. We had $128 million of loans in this asset that were written down to $50 million many years ago, again, on an asset that was originated prior to the great financial crisis. We sold the note for $59 million, which should result in a $9 million reserve reversal and a $1 million distributable earnings charge in the fourth quarter. Additionally, some of these loans were in our TRS as they were not considered good REIT assets. And as a result, we will receive a $20 million tax deduction in the fourth quarter, which will reduce our tax expense by approximately $7.5 million and increased distributable earnings.

With respect to accrued interest on modified loans, in the third quarter, we reversed approximately $18 million of previously accrued interest, the majority of which were related to new delinquencies during the quarter which, as we discussed earlier, is reflective of where we believe we are in the cycle. This adjustment, combined with $6 million in back interest collected on a loan payoff in July, which we referred to in our last call has resulted in us reducing the total accrued interest on modified loans by $13 million this quarter, even after accruing an additional $11 million in interest on modified loans that are performing in accordance with their terms. We have also started the process of accelerating the resolution of problem assets by reworking some of our loans at lower interest rates in and efforts to create a higher run rate of future earnings going forward on loans that were likely to default in the near future. This resulted in a reduction in interest income of approximately $8 million this quarter and will affect interest income going forward by about $4 million a quarter. As stated earlier, we have taken an aggressive approach in dealing with our troubled assets. And as a result, our delinquencies have risen this quarter to $750 million at September 30 compared to $529 million at June 30. This peak stress reflects where we are in the cycle. And again, we're working very hard to take control of these assets quickly and we rework these loans with higher quality sponsors and provide a more consistent run rate of income going forward. In fact, in October, we took back $110 million of these assets as REO and expecting another $40 million to be foreclosed on in November and December. We're working very hard at bringing in new sponsors to take over assets and assume our debt over the next few quarters, which will create a longer-term benefit of creating a more predictable run rate of income. This will come with a short-term temporary effect of reducing interest income, as I mentioned earlier, as we accelerate the process. This strategy will also result in a quicker ramp-up in our REO book as we look to accelerate the process of bringing in new sponsors to take over the real estate and assume our debt. In the third quarter, we took back $122 million of new REO assets, putting our REO book at $470 million at September 30. And as I mentioned earlier, we're expecting to take back another roughly $150 million of loans in the fourth quarter. And again, we're working hard to dispose of some of these assets very quickly, which should keep us in the range that we previously guided to of between $400 million to $600 million of REO assets that we will own and operate.

In the third quarter, we recorded an additional $20 million of net loan loss reserves on our balance sheet loan book, $15.5 million of which were specific reserves with the remaining $4.5 million being general CECL reserves. The additional specific reserves that we recorded this quarter are consistent with our strategy of accelerating the resolution of problem loans as we look to mark certain loans that we are marketing for disposition to where we think we can execute a sale. This could result in roughly $15 million to $20 million of realized losses next quarter if we're successful at liquidating all these assets quickly, which will be partially offset by the $7 million of income we generated from the sale of the Homewood asset. And again, this will very importantly allow us to increase our run rate of income going forward and create a more predictable earnings stream. In our Agency business, we had an outstanding third quarter, as Ivan mentioned, with $2 billion in originations and $2 billion in loan sales, which generated $10 million more in gain on sale income this quarter. The margins on this business were 1.15%, which are down from prior quarter due to some large off-market portfolio deals we were able to capture, which contain lower margins and smaller servicing fees. We also recorded $15.5 million of mortgage servicing rights income related to $2 billion of committed loans in the third quarter, representing an average MSR rate of around 100.78%, which again reflects the larger deals we closed this quarter with lower servicing fees. Our fee-based servicing portfolio grew 4% this quarter to approximately $35.2 billion at September 30 on record third quarter originations. This portfolio has a weighted-average servicing fee of 36.2 basis points and an estimated remaining life of around 6 years, and will continue to generate a predictable annuity of income going forward of around $127 million gross annually.

In our balance sheet lending operation, our investment portfolio grew to $11.7 billion at September 30 from originations outpacing runoff for the third straight quarter. Our all-in yield in this portfolio was 7.27% at September 30 compared to 7.86% at June 30, mainly due to stopping the accrual of PIK interest on certain loans, modifying some loans with rate reductions, new third quarter delinquencies and a reduction in SOFR. As mentioned earlier, we do expect this run rate to improve meaningfully over the next few quarters as we look to accelerate the resolution of our delinquencies. The average balance in our core investment was $11.76 billion this quarter compared to $11.53 billion last quarter from growth in the portfolio. The average yield on these assets decreased to 6.95% from 7.95% last quarter, mainly due to the significant nonrecurring adjustments I spoke about earlier, including reversing accrued interest and lower rates on modified loans, combined with the additional delinquencies we experienced in the third quarter.

Total debt on our core assets was approximately $9.9 billion at September 30, the all-in cost of debt was approximately 6.72% at 9/30 versus 6.88% at 6/30, mainly due to a reduction in SOFR. The average balance in our debt facilities was approximately $10 billion for the third quarter compared to $9.5 billion in the second quarter, mainly due to funding our third quarter growth and the addition of our new senior notes in July. The average cost of funds on our debt facilities was 6.88% in the third quarter compared to 6.87% for the for the second quarter, excluding interest expense from levering our REO assets, the debt balance of which is separately stated on our balance sheet and therefore, not included in our total debt and core assets.

Our overall spot net interest spreads were 0.55% at September 30 compared to 0.98% at June 30 from stopping accrued interest on certain loans and from new delinquencies and modifications at lower rates. We estimate that this new run rate will temporarily reduce our quarterly earnings by $0.05 to $0.06 a share that again, we believe we will be able to improve upon meaningfully as we resolved our troubled assets over the next few quarters.

So in summary, we're very pleased to have produced significant gains from a few of our legacy investments, which will put us in a good position to be able to accelerate the resolution of our legacy book. This will take us some time to complete. In the interim, we will experience some fluctuations in our quarterly earnings. When completed, we will resolve the lion's share of our legacy assets and built up a more predictable and growing run rate of income for the future which will complement the growth in the origination platforms that we're expecting from the improved rate environment and go a long way towards helping us achieve our goal of growing our earnings and dividends in the future.

That completes our prepared remarks for this morning. I'll now turn it back to the operator to take any questions you may have at this time. David?

[Operator Instructions] We'll take our first question from Steve Delaney with Citizens Capital Markets.

Well, thanks for a very detailed layout. I mean, you guys have your hands full and kind of a multi-front war, but you seem to understand the game plan very well. I'm trying to understand it. It seems to me that you have 3 steps as you're looking at your portfolio, to modify what you can where you have a viable partner and property, it's not to foreclose and then obviously to eventually move those foreclosed properties to a sale. I guess in terms of -- let's start with the loan [ mezz ]. Can you look at the portfolio now and guess or can you give us an estimate of like what inning we're in terms of the loans that are left in the portfolio that are performing have not needed to be modified. Do you have a sense for how stable those properties and those loans might be over the next 6 to 12 months? Or do you have concerns that there could be slippage in that? So I guess -- are you down to solid borrowers with improving market conditions or this triage process? Again I'm trying to get a sense for how many additional potential loan mods might come up in the next couple of quarters. And I apologize for that long-winded question, trying to express what I was getting at.

Sure. Let me walk through from a macro standpoint where we are. And in terms of what inning we're in, it also depends on what market we're in. Different market reacts a little differently. But generally, when we do a modification, people are bringing capital to the table. And remember, this is a full long period of time that people had to continue to bring capital to the table with the elevated interest rates. So it's 1 thing when you bring capital to the table and rates remain elevated and you keep on having to bring capital to the table. So most modifications have a duration to them. And then if rates remain elevated and the market doesn't recover, they have to bring more capital to the table or there is a future modification potentially that will happen if the asset doesn't improve. Where we become very aggressive is if the asset is not improving because the people are not managing effectively or don't have the capital to maintain that asset. That's where we've become extraordinarily aggressive. So we don't want a deteriorating asset, and that's where we stepped up our efforts. We're at a different point in the cycle. And we also view that these assets with the right amount of capital and right amount of management, they will stabilize. And that's why you see a little bit of a peak. There are certain markets that have been extraordinarily hard hit and kind of been whiplashed. I would say markets like San Antonio, like Houston. Those are markets that under the prior administration, the Biden situation really took over these assets and really negatively impacted these assets. And they are quite a few of them. And we're not the only one who's had these happened to our assets, it's across the board. And then, of course, with the change in administration, that began to change and we got control back on those assets. Unfortunately, with the new administration,and what we've seen is if they've had massive [ ICE ] rate and emptying of properties that were perhaps 88% or 90% occupied, and they're stripping down 20% to 25%. And that's why we've seen a little bit of an acceleration on our part now in Houston and in those markets that going to take control over those assets and restabilize those. So that's a little unique to those markets. But what we are seeing, which is a great interest is when we do take back assets where in the past, maybe we had 1 or 2 people who are interested in buying these assets. We probably have 3 or 4 people, and the appetite for these assets is growing and growing and with the drop in rates and a little bit of return of liquidity, we're seeing some real strength. We're also seeing a lot of benefit right now to good management, where if you have good management and you have the right CapEx to improve your units, where we've had a lot of economic issues in terms of economic collections on some of these assets, we're seeing if these assets are run correctly, they'll fill up and they'll be the right paying tenants, and that's why we're stepping in as well. So the peak in our delinquencies are reflective of our attitude that somee of these assets are better in our hand or even transition to new people. Paul mentioned, we have about $750 million in delinquencies. We have resolution and clear sight that within the next 45 days, about $500 million of those will be resolved either through recapitalization of modifications or finding new borrowers and stepping into those deals. So we are being aggressive. We're very optimistic that we'll get through those. We do expect 1 more wave until in the fourth quarter, which we're prepared for and we're taking an aggressive approach to try and resolving the majority of these legacy issues. So we're hopeful that by the end of the fourth quarter, we would have really addressed each and every 1 of these transition, ones out of ownership that's not doing a good job. And well towards my comments that by the end of the first quarter, we should be in a real position to change our run rate and get these either resolved income-producing or dispose-off appropriately. I want to point out that there's 2 aspects that we focus on maybe a little different than everybody else, which are important to note. We focus not only on our dividend, which, of course, has outperformed everybody else and our guidance was to maintain our dividend for the year because we believe that this is just a little lumpy right now, but more significantly, Steve, if you look at the company's performance, we really have a growth in book value of 23% over the last 5 years, and everybody else has declined 27%. There's a 50% change in book value relative to us and everybody else. So we're managing through this with a nominal change in book value, we're maintaining a tremendous dividend. And I think that is where we're looking at as business operators to manage both of those items. It's a long-winded response, but it wasn't an easy question.

Paul, I know it wasn't an easy question. And we do note that the stability in the book value because we've seen some a big drop in some of your peers with additional 5 rated loans. Just to wrap it up, and I really appreciate that thoughtful response. It seems like to me we're 1 to 2 quarters away of having the problem loans on your books as REO. The first step is to get it into REO. And it seems like by mid -- certainly by mid-2026, you're going to be in a very active REO sales process. As far as -- to finish the loan mod, take back what you have to take back in 2026, hopefully sell a lot of real estate.

Yes. But I want to point this out, and I think this is important because we're a little different than everybody else. We have a very diversified business with a lot of skill sets. And if you take a look at the Lexford transaction, which was something everybody said, "Oh my God, you're taking that back." We had $67 million of preferred equity on that portfolio. Not only do I get back to $67 million, not only do we make hundreds of millions of dollars on these and get a good income stream going forward. We took assets that were in the mid-80s, low 80s in occupancy, improved these assets, got them in a great position, had a great asset. And now look at the impact of creating another $50 million. We still have assets on our books. So we have to be guided as good operators as producing what we need to produce, and we have very diversified income streams. And yes, we can generate gains, and we can have the patience to have the right kind of gains. And what we will do is we look at the REOs and see, is it quicker to dispose of them now at the right levels? Or do we need to give them a little TLC and get them a little bit better. We can be patient unlike other people.

We'll take our next question from Jade Rawani with KBW.

The $18 million accrued interest reversal that was booked a quarter, does that mean that the current level of interest income for the third quarter is a reasonable baseline to use. It was $208 million, reflects most recent interest rates as well as interest rates on most modified loans. And then I think it sounds like you expect originations potentially to drive either a flat to slightly higher portfolio in 4Q. So do you have any comments on the run rate of interest income?

Yes. Jade, it's Paul. It's a good question. And the answer is no to the first part, and let me give you some guidance. As I laid out in my prepared remarks, we did see, as Ivan mentioned, some elevated defaults this quarter given where we are in the cycle and how aggressive we're being to be able to resolve things quickly and take back assets before they deteriorate. So we did reverse $18 million of accrued interest, mostly related to delinquent loans and loans we remodified, but on a run rate, that's a onetime adjustment in my mind. On a run rate, that's going to cost us $4 million going forward because we reversed interest as of June 30 that we had accrued of $18 million, but we're losing interest on those accruals going forward of about $4 million. So the $18 million is onetime, the $4 million is recurring. Additionally, we had modified loans during the quarter, some we had to remodify where we gave additional relief. Due to the structure of some of those mods, it cost us $8 million for the quarter in interest, but it's only going to cost us $4 million going forward on the rate reduction. So we have $4 million going forward in a run rate reduction from stopping PIK interest on certain loans. We have $4 million in reduction from lower interest rates on mods. And on the new delinquencies, right now, and I want to go through this, that would affect the run rate going forward by $8 million. If you add up those 3 numbers, that $16 million in reduced interest income on a run rate going forward. However, that's before we resolve anything. And we've already agreed to on $225 million of loans. We've already agreed to reposition and sale with new interest rates of like SOFR plus 250 for the most part and that's picking us up $3 million in the fourth quarter alone before we even get into the items that Ivan mentioned, getting our [ 750 ] down even further to like [ 250 ]. Some of that will happen at the end of the quarter. It won't impact the fourth quarter as much but it will impact the run rate in the first quarter going forward. So the long answer is interest income went down by about $34 million this quarter, roughly due to $18 million of reversed accrued interest, another $4 million from stopping the accrued interest, $8 million from modified loans, $5 million from delinquencies. That $34 million was offset by about $7 million or $8 million in back interest and fees we collected and about $5 million in growth in our portfolio for a reduction in interest income of $22 million. I see the reduction of interest income going forward, starting at $16 million for the reasons I talked about the $4 million of PIK, the $4 million of mods and the $8 million of delinquencies, but we've already improved that by $3 million on loans we know we've executed. So now I'm down to $13 million of a reduced interest income going forward. And that's my $0.05 to $0.06 I was referring to. However, having said that, that $3 million that we improved already in the fourth quarter was just the fourth quarter impact, and those were done like mid-quarter. So the first quarter will be impacted by $5 million instead of $3 million. So that brings that $13 million down to $11 million. And then everything Ivan talked about, we're working on will likely bring it down further. So we look at this as the third quarter got hit by some reversals and some elevated delinquencies. The fourth quarter will be hit by a lot less. And in the first quarter and second quarter, you should see drastic improvements to our run rate to get us back to where we were and then above.

Jade, I want to make 1 comment. Paul mentioned that we reduced our spread on a long to 250. I want to be very clear that the market spreads today on new originations are 225 to 275. So some of the relief that we're giving is really reflective of where the market is and adjusting the borrowers down to today's market and giving them consideration for where lending spreads are. So while we're seeing spreads come in a little bit, it's really reflective of the current market spreads. It's not really a discounted spread to the market.

That's helpful. I'll have to go through the numbers in more detail after this. But on interest expense, similar question. I went in the wrong direction despite the positive developments on financing. So were there any onetime items in interest expense causing it to be elevated for the structured business at that $176.2 million for the third quarter?

Yes. So Jade, there was a couple of things. One, so interest income, you saw it went down by $22 million. I just laid out all the pieces for you there. The other side of it is interest expense went up about $10 million, $5 million of that -- some of that is growth in our portfolio, right? So I had $6 million or $7 million of income from growth in portfolio and then I had debt increase, obviously, as I leverage the growth of that portfolio. But we also issued the $500 million of senior bonds in early July, so that full effect was in the third quarter as we laid out in our last quarter for our run rate. And also, we had a little bit of double interest in the third quarter, which won't repeat because we paid off our converts in August but we issued the senior bonds in July. So we weren't allowed to pay our converts off until August. So we had like a 1-month of double interest on those 2 bonds that will go away. But the big reason that interest expense is up is because we issued $500 million of senior bonds to grow our portfolio.

Got it. That's clear. Lastly, if I could, GAC credit it has been pretty benign. I don't know if you follow Greystone, which is owned in joint venture by a CRE services company. but it looks like they had a big spike in credit loss provisions in their agency multifamily business. Now I'm seeing Arbor's results and the provision was for risk sharing $8 million, double what it was last quarter and higher than the recent run rate. So can you comment on what drove the increase, if there were any loan book putbacks, any cases of fraud or if it's just broader credit deterioration and if that's something that will be a headwind going forward?

Yes. I think if you look at Fannie Mae and if you look at our all-in numbers, They were in [indiscernible] peak part of that delinquency. You run that at a very low delinquency and of -- constant right now, it increased a little bit. Like all lenders who had a little bit of a portfolio in New York City, you've had some distress on the rent control and rent stabilization, not material, but it does move the needle a little bit. We're all working through a change in the economic climate, which shows in across the board delinquencies at the agencies. It's just where we are in the cycle.

Yes, Jade. And given Ivan's commentary, what we said with everything else, we think peak stress is Q3, Q4 leak into Q1 a little bit. So we are expecting, I guess, similar reserves for Q4, maybe a little bit less in Q1, given where we are, but it's just based on where the peak stress is.

We'll take our next question from Rick Shane with JPMorgan.

First, if we could talk a little bit about Homewood and the sale there. it looks based on the disclosures that you're selling that property basically at the reserve value, maybe a slight uptick versus the reserve value. But the original carrying value of the loans was $112 million. I'm curious if there is some realized -- or actually the original cost base is well above that because there's a $71 million reserve. Is there going to be a realized loss in the quarter that we need to think about in terms of distributable income?

Yes. So Rick, it's Paul. Great question, and I'm glad you asked it, and I'll give a lot of color on it. So we did have between all the Homewood pieces, we had about $128 million of UPB on the loans. We wrote it down to $50 million by taking almost 80 million reserves many, many years ago. All but $10 million of those reserves that brought it to a net carry of $50 million, we're taking way before COVID when AFFO and FFO reflected reserves as realized losses. So the only portion that has not been brought into AFFO, FFO, which has been superseded by distributable earnings was $10 million of reserves we took around the time of COVID. So as you pointed out, we have a carry value of $50 million, we sold it for $59 million. When the accounting gets done, we believe what's going to happen is we're going to have a reversal of a reserve of $9 million, right, because we have more reserve than we sold it for, but we'll have a distributable earnings hit of $1 million because technically, the distributable earnings carrying value is $60 million, and we sold it for $59 million. So we'll have only a $1 million realized loss, we believe, on that sale because all the other reserves had already been taken through AFFO and FFO before we went to distributable earnings because this is an asset that goes way before the great financial crisis, as you know. And then we're going to pick up $7.5 million of real GAAP income and distributable income and book value growth from the savings on the tax side because a lot of these loans are in our TRS, and we are a taxpaying entity in the TRS and we get to save the taxes on that.

Got it. Okay. That makes sense. And thank you for walking me down GAAP accounting memory lane. I did not actually remember that. So that's very helpful. The other thing is that contractually, that was a loan that had about a 10% coupon, you did not accrue interest on that. Is there any accrual reversal associated with that? Or should we just look at this as you sold a loan that was effectively generating a 0% interest rate and there was no cash flow associated with that. So there's no nuance there.

Yes. So there's a couple of nuances we should talk about. You're exactly right. Per disclosures, which you read correctly. This loan had an accrual rate for many, many years. But when it was struggling, we did not accrue any of that interest, none of that interest is accrued in my books. So there's no reversal of interest. And in this deal, as Ivan will point out, we are ending up -- we are providing a little bit of seller financing, and we are ended up with a performing loan now at 10%. So not only are we going from alone that was 0%, we're getting about $6 million or $7 million of cash in the door, we're saving $7.5 million in taxes with an offset of about $1 million distributable earnings hit, but we have like a $53 million loan that's going to earn 10% now and be current. So we took an asset that was earning 0%. We created some cash. We created some income, GAAP and distributable, and we have a performing asset going forward that's adding to our run rate.

Got it. And was that loan unlevered on your balance sheet?

Yes.

Okay. That's really helpful. And the other thing I'd just like to talk about is when we look at the income from REO, property income fell 20-plus percent sequentially and property operating expenses went up modestly. And that's despite the fact that there is more REO. So trying to understand what is going on there? And I would like to understand potentially the implications for lower cash flows in terms of recovery values.

Okay. I'm going to let Ivan talk globally about the REOs. Let me give you some numbers and then he's going to go into where the market is. You have to remember, some of the loans we're taking back is REO, we're taking them back because we want to make sure that we won't see significant deterioration. Some of them are coming back and we've been emptying some parts of the building. Their occupancies are super low, probably 40% occupancy. So now our job is to go in there, do what we need to do to get to manage these assets and get the occupancies up. So right now, some of those assets are, as you said, throwing off negative NOI, and I think it was about $3 million for the quarter. in the third quarter and that spiked up because we took some more assets back. We're getting our hands around those assets. We are taking back more REO. I have it probably projecting around the same number despite bringing more REO back. But that number will improve quickly. And I'll let Ivan talk about it as we get in there and we rework these assets, you're going to see occupancy go up a lot. And when it does, that NOI weren't positive at some point, correct, Ivan?

Yes. And I think, Rick, you're hitting on something very important. When we take back REOs that need a lot of work. We will empty those building out and do the work and get the occupancy up. When the occupancy gets up, we're going to sell those assets. So you'll see a little bit of the spike. You'll see them all down. And then as they get leased up, those will be disposed of, and then new ones will come in. So expect that to go up and down. We track that. We track the occupancy. We track the CapEx. And we've had a few on our books that we've actually stripped down to 0. We have all the units, and now they're going to start to lease up very quickly, and we're targeting those which were deeply damaged to dispose of in the third quarter of this -- next year. And right now, they're going from 0 and they're gaining like 10 to 20 points in occupancy. And when they get to 70, 75, we'll look to dispose of those. So it's going to be very lumpy, but the philosophy is strip them down, take the pain and then move them on.

Okay. That's very helpful to understand what's going on. And then just 1 last housekeeping question associated with it. It sounds like when you were doing this, you're obviously making significant investments in the properties, are you capitalizing those investments? Or are you expensing them?

Yes. Most of the times, if you're making improvements to the real estate, you're capitalizing them, it's increasing your basis. Most of the work is done to improve the assets value correct.

We'll take our next question from Crispin Love with Piper Sandler.

Paul, just based on your comments on an earlier question on interest income, there was a lot there, so I just want to make sure I'm thinking about it correctly. It seems that you're confident that the third quarter is trough in NII, just given the onetime hits being larger than the go-forward impact. Is that accurate or prior to potential forward stress that could impact that run rate?

Yes, I think that's accurate. So the way we look at it is we do expect we'll have some more delinquencies, but we're going to resolve a lot more than we're going to have new. So we think this is probably peak. There were some onetime adjustments and reversals of the accrued interest I mentioned on some delinquencies that we accelerated. But yes, we expect this to be -- we do not expect the fourth quarter to have the same run rate as the third quarter. In fact, if you look at the numbers that we gave in the press release today, and I spoke about in my commentary, Crispin, we're showing a spot rate of $7.27% going forward of 9/30. That's an all-in yield on our $11.7 billion book. It was 7.86% spot rate as of June 30. If you do the math, that's $16 million. So what I'm telling you is net interest income came down $22 million after growth, but it really came down $34 million for the adjustments I laid out. Now we're saying, it's going to spot go down $16 million, but that spot has already been improved by $3 million, as I mentioned, on loans we've already worked out and sold. So now I'm down to $13 million. And yes, new ones will come on, but we're also, as Ivan mentioned, real-time conversations, which aren't in my numbers, we're resolving a lot more of the [ 750 ] than we originally planned, we do quickly. So by the end of the fourth quarter, I hope to have a lot more of that resolved and that run rate will improve drastically in the first and second quarter of next year. So that's why when we laid out our commentary, we said it will take us to probably around the second quarter of next year to accomplish our goals. And when we accomplish all of our goals, our run rate will be back up, plus the growth in our portfolio from the improved environment and our origination platforms, we'll start to see that turn.

Great. And then just on the $48 million gain on the Lexford portfolio. Can you just give a little bit more detail on that transaction. How long has that been in process? And who was the buyer there?

So I guess, it's an asset that we took back from the great financial crisis, and we've been managing that [indiscernible] creating significant gains. And I guess we made a decision based on where we are in the cycle that it would be good to monetize those gains and especially in light of the criticism that we got that they were really degraded and caused a lot of the short interest in the short reports as bad transactions. So for us, wanted to be very clear that they were good transactions. And that not only were they good, there was substantial profitability. As an operator, we felt by creating those gains, it gives us a lot of flexibility in trying to manage our business and accelerate these legacy issues. And keep in mind, we still have a good portion of those assets left as said in my script. So we just thought it was good timing. It spoke about the credibility of our performance on those assets. And it really negated some of the accusations that have been thrown at us with clarity, and that's kind of the reason we did it. In terms of the buyers, we don't really disclose. We -- very heavy interest. We have 5 buyers for them. This is very competitive process. And it was a good portfolio. We had a great experience with the buyer. It's a public record. I'm not really exposed who they are. but they were well bid and there's a deep audience for that portfolio as well as for the balance of the portfolio that we have.

Okay. Great. And then just a final question. You mentioned the $48 million gain that you did talk about in the prepared remarks, I believe, an additional $7 million you expect to come through. Should that come through in income from equity affiliates as well?

No, that $7 million, Crispin, was referred to the Homewood transaction. So what we said is that we had a $48 million gain. What we said in our prepared remarks is we had a $48 million gain from Lexford. We still have a portfolio of assets that we'll look to dispose of in the near future. So whatever comes out of that, comes out of that when it happens. And then in addition to that, in October, which should close today, we're closing on the sale of the Homewood note, which will generate about $6.5 million, $7 million of GAAP income and distributable income in the fourth quarter.

We'll take our next question from Lee Cooperman with Omega Family Office.

An observation and a question. So the observation is I think that your experience when you take control of this real estate show that the real estate was well underwritten when was originally underwritten. But when you foreclose and default the property, your experience has been very good. So should you not take a bow for that? Number one. And number two, with the stock is now trading below book value, with you having an optimistic outlook in the second half of next year, do you -- are we willing to [indiscernible] some capital? Or would you want to hold on to the capital?

With respect to -- we have a share buyback program out there. We'll look at the best use of our capital. As you know, a lot of our insiders, including me, continue to buy stock as it goes below book. And when the company has the right capital, we'll make that evaluation. So we do view that as an opportunity.

With respect to the REOs, it's very interesting. If I said to everybody on this call that the loans that we're originating today, I would say that they could withstand as I said back 4 or 5 years ago, they could stand this, underwritten to a certain rate spike. Keep in mind that when we took these loans on, SOFR was quarter, it went up to 5%. Nobody could have underwritten a 4.75% increase in SOFR. We always said we can underwrite a 200 to 250 basis point spike in SOFR. This is unprecedented. Also, keep in mind that the treasury was 150, 125, 150. The treasury went up to 5, these are unusual moves and they prolong. So there has actually been a very, very difficult environment, but we do feel that when we get these REOs back, if we improve them and put them back in order, we're doing very well. We're getting very close to where we mark them. Sometimes a little above, sometimes a little around it, but our marks are pretty accurate. And we've been doing a really good job. So we have the skill set to do it. And we're very effective at it. And we're very pleased with the ones that we've taken back and as I said earlier, we'll look to dispose those probably in the third quarter, and we're comfortable with how those are going.

And there are no further questions on the line at this time. I'll turn the program back to Ivan Kaufman for any additional or closing remarks.

Okay. Thank you, everybody, for participating. We are at a very difficult point in the cycle. We're prepared for it. We've actually anticipated it. We'd rather not [ be here ]. But as operators, we're addressing it. We have a lot of tools to address it. Thanks for your participation, and we look forward to next quarter's call. Take care. Have a great weekend.

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