MFA Financial, Inc. (MFA) Q2 2020 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by. Welcome to the MFA Financial, Inc. Second Quarter Earnings Conference Call. [Operator Instructions]. As a reminder, this conference is being recorded. I would now like to turn the conference over to your host, Hal Schwartz. Please go ahead.

Thank you, operator. Good morning, everyone. The information discussed on this conference call today may contain or refer to forward-looking statements regarding MFA Financial, Inc. which reflect management's beliefs, expectations and assumptions as to MFA's future performance and operations. When used, statements that are not historical in nature, including those containing words such as will, believe, expect, anticipate, estimate, should, could, would or similar expressions are intended to identify forward-looking statements. All forward-looking statements speak only as of the date on which they are made. These types of statements are subject to various known and unknown risks, uncertainties, assumptions and other factors, including those described in MFA's annual report on Form 10-K for the year ended December 31, 2019, and other reports that it may file from time to time with the Securities and Exchange Commission. These risks, uncertainties and other factors could cause MFA's actual results to differ materially from those projected, expressed or implied in any forward-looking statements it makes. For additional information regarding MFA's use of forward-looking statements, please see the relevant disclosure in the press release announcing MFA's second quarter 2020 financial results. Thank you for your time. I would now like to turn this call over to MFA's CEO and President, Craig Knutson.

Thank you, Hal. Good morning, everyone. I'd like to thank you for your interest in and welcome you to MFA Financial's Second Quarter 2020 Financial Results Webcast. Also dialed in with me today are Steve Yarad, our CFO; Gudmundur Kristjansson; and Bryan Wulfsohn, our Co-Chief Investment Officers; and other members of senior management. Before we begin, I'd like to give a shout out to our entire MFA team. The last 5 months have obviously been extremely challenging and made exponentially more so by the fact that all of our efforts have been remote as the company fully implemented our business continuity plan during the third week of March. The effort and commitment put forth by our entire team over the last 5 months has been extraordinary, and I have been awed by their dedication. Although the bottom line earnings per share results for the second quarter of 2020 might, at first glance, appear to be a return to normal for MFA. The second quarter was anything but normal. After a COVID-19-induced mortgage market meltdown that began in mid-March, we were in the middle of negotiating a forbearance agreement with our significant lenders as the second quarter began. These negotiations resulted in our first forbearance agreement, which took effect 10 days into the quarter on April 10, although our lenders who were party to this agreement had essentially been granting us forbearance since March 23. Forbearance agreements were extended on April 27 and again on June 1, and we exited forbearance on June 26, so we spent nearly the entire second quarter under forbearance. And while these forbearance agreements were expensive and required a massive effort to manage, they did afford us the time necessary to delever our balance sheet, generate liquidity and conduct a thorough and competitive process to source third-party capital. Please turn to Page 4. Our second quarter financial results were overwhelmingly dominated by unusual events and transactions. Sales of residential mortgage-backed securities in the second quarter generated $177.5 million of net realized gains versus their March 31 marks. The sale of a large Non-QM whole loan pool generated a loss of $127.2 million. However, $70.2 million of this loss was booked as an impairment in the first quarter in anticipation of this sale, so the second quarter recognized loss was $57 million. We booked a $49.9 million loss related to swap hedges that were terminated during the first quarter. High forbearance interest expense and $14.2 million of amortized swap losses generated very high interest expense of $82.1 million for the period that resulted in no net interest income for the quarter. We also recorded $40 million of expenses related to forbearance and portfolio restructuring. So although we earned $0.19 per share in the second quarter, this was the result of many large and unusual items. GAAP book value was up primarily due to GAAP earnings that were not paid out in dividends. Economic book value was up additionally as we saw continued price improvement in our carrying value whole loans. We elected the fair value option to account for all new and reinstated financing. This allows us to expense upfront fees and other costs associated with these transactions. Thereby allowing us to present a more true economic go-forward cost of these financing arrangements. Our leverage ratio at June 30 was 2 -- was 2:1. And our investment mortgage assets consisted of $5.9 billion of residential whole loans and approximately $400 million of mortgage-backed securities. Please turn to Page 5. As previously announced, we closed our capital transaction with Apollo and Athene on June 26. This transaction included a $500 million senior secured term loan, a warrant package to purchase 7.5% of MFA common stock, over $2 billion of new non-mark-to-market financing provided by Apollo and Athene, together with Barclays and Credit Suisse. In addition, Apollo and Athene have committed to purchase the lesser of 4.9% or $50 million of MFA common stock in the open market over the next year. And Athene has committed to purchase a portion of MFA's first Non-QM securitization. I cannot stress enough that this transaction was about a lot more than a $500 million check. It is very much a holistic solution and a strategic and collaborative partnership. Please turn to Page 6. With the pause afforded to us through forbearance and with assistance from Apollo and Athene, we have also been able to profoundly restructure our liabilities to a much more durable form of financing. Of the $4.7 billion of borrowing that is asset-based or secured, $3 billion is non-mark-to-market and 2-year term, and another $785 million is intentionally underlevered by approximately $55 million, thus creating a margin cushion of approximately 6 points. In addition, we have $330 million of unsecured debt with our senior note and convertible bond. So all told, over 80% of our borrowing is either non-mark-to-market or under levered. As you can see in the last bullet point on this page, the cost of the aggregate secured financing away from the Apollo, Athene senior loan and our existing securitized debt is approximately 3.6%. Replacing some of this borrowing with securitization, particularly for Non-QM loans at current new issue levels could lower this cost by about 100 basis points depending on how deep in the capital stack we sell. Please turn to Page 7. As previously mentioned, we exited forbearance on June 26, with substantial liquidity, no outstanding margin calls and all lenders repaid in full. Although an expensive and time-consuming process, we believe that we were able to protect hundreds of millions of dollars of book value by liquidating assets in an orderly and negotiated fashion rather than through a fire sale liquidation during the last 2 weeks of March. In addition, we were able to conduct a robust and competitive third-party capital process which, again, we believe, led to a much better transaction than we could have achieved in a compressed and rushed time frame. Please turn to Page 8. While MFA is admittedly a smaller company than we were in February, we believe that we are on very solid footing during what remains a very uncertain economic environment. Not only do we have very durable financing, but we also have substantial liquidity. The securitization market continues to improve amidst dwindling supply, offering us the ability to realize considerable cost savings. Additionally, we have reinstated the payment of all accumulated but unpaid dividends on our Series B and Series C preferred stock issues, and we have declared a common stock dividend of $0.05 per share payable on October 30, 2020, to stockholders of record on September 30. Please turn to Page 9. As most of you know, as a REIT, we are required to pay 90% of our taxable income in the form of dividends, and those dividends can be both on preferred stock and common stock. To the extent that we pay 90% or more, but less than 100% of our taxable income, we are required to pay income taxes on the amount by which our dividend distributions fall short of 100% of our taxable income. Finally, what we generally have until October of the year following the tax year to make these distributions, we are required to pay an excise tax to the extent that we do not declare at least 85% of the required distribution during the current year. The excise tax is paid on the amount of shortfall below the 85%. So preferred dividends for the year at the contractual dividend rate will total $29.8 million or about $0.065 per common share. So the preferred dividends will satisfy the distribution requirement on the $0.05 of 2019 taxable income with $0.085 approximately to apply to taxable 2020 income. With an estimated $0.14 of taxable income through June 30, the 85% distribution required to avoid the excess tax -- the excise tax is about $0.12 or $0.105 after the preferred dividends. Pages 10 and 11 present pie charts to illustrate how our portfolio composition and how our secured liability structure changed during the second quarter. Our investment portfolio is lower by 1/3 and is now comprised of 94% whole loans, up from 74% at March 31, and our secured liabilities are lower by 43% and are nearly 2/3 non-mark-to-market versus nearly 100% mark-to-market at March 31. I will now turn the call over to Bryan Wulfsohn to provide more details on the investment portfolio.

Thank you, Craig. Turning to Page 12. The pandemic has had a material impact on the origination of Non-QM loans. Production basically ground to a halt coming into the second quarter but is gradually -- has been gradually increasing with the stabilization of the securitization market and markets more generally. Underwriting standards have tightened a bit on new production, but given that standards were already conservative pre-COVID, the change hasn't been dramatic. Part of our deleveraging process to rightsize the balance sheet, we executed on a sale of approximately $1 billion of loans in the second quarter. The loans in the portfolio sold at higher LTVs. This resulted in a reduction of the remaining weighted average portfolio LTV at the end of the second quarter by 2 points. Through the pandemic our relationships with our originators remain strong. We are working with our partners to be a stable source of liquidity moving forward. Approximately half of the portfolio is now financed with non-mark-to-market leverage. And we expect to be a programmatic issuer of securitizations, which will further increase the percentage of non-mark-to-market funding in addition to lowering our cost of funds. Turning to Page 13. Significant percentage of borrowers in our Non-QM portfolio have been impacted by the pandemic. Many of our borrowers are owners of small businesses that were affected by shutdowns across the nation. We instituted a deferral program in March with our servicers in an effort to help our borrowers manage through their crisis while maintaining value for our shareholders. Any borrower that was current leading up to the pandemic and raised their hand during the months of March, April and May, saying that they were impacted were granted a deferral until the June 1 payment and were reported as current through that time period. Looking at the table on the right-hand side of the page, you can see the cumulative percentage of loans impacted by COVID increased over 30%. As of June 30, we are happy to report that over 2/3 of the borrowers that received COVID related deferrals have made one or more payments post deferral conclusion and continue to be current. Turning to Page 14. Our RPL portfolio of $1.2 billion has been impacted by the pandemic, but continues to perform well. 83% of our portfolio remains less than 63 days or 60 days delinquent. Although the percentage of the portfolio of 60 days delinquent and status was 17%, over 20% of those borrowers continue to make payments. Prepay speeds in the second quarter slowed down a bit from elevated speeds seen previously. However, they have maintained within our expectation and we could see them tick up a bit in the coming months given the mortgage rate landscape. This portfolio exhibited a similar percentage impacted by COVID as a non-QM portfolio, we are working with our servicers to ensure positive outcomes post forbearance. Turning to Page 15. We are lucky to have an exceptional asset management team to manage through our portfolio of nonperforming loans. The team has worked through the pandemic in concert with our servicing partners to maximize outcomes on our portfolio. This slide shows the outcomes for loans that were purchased prior to the second quarter ended of 2019, therefore, owned for more than 1 year. 34% of loans that were delinquent at purchase are now either performing or paid in full. 46% have either liquidated or REO to be liquidated. And we have significantly increased our activity, liquidating REO properties, selling 90% more properties versus the second quarter a year ago. 20% of loans are still in nonperforming status. Our modifications have been effective as 3 quarters are either performing or paid in full. We are pleased with these results as they continue to outperform our assumptions at the time of purchase. And now I'd like to turn the call over to Gudmundur to walk you through our business purpose loans.

Thanks, Bryan. Turning to Page 16. The COVID-19 pandemic negatively affected fix and flip borrowers as project completion and home sales slowed down dramatically in April and May, and borrowers in general were negatively impacted by the severe contraction in economic activity caused by the COVID-19 pandemic. As a result, principal paydowns slowed down and delinquencies increased in our portfolio in the months of April and May, for stabilizing and/or recovering slightly in June. MFA's Fix and flip portfolio declined $116 million to approximately $860 million in UPB at the end of the second quarter. Principal paydowns were about 65% of pre-COVID levels in April and May, but recovered almost fully in June for a total of $135 million of principal paydowns. We're equivalent to about 45% CPR on an annualized basis. The third quarter looks on track to continue this trend with principal paydowns in July exceeding $50 million. We advanced about $26 million of rehab draws and made no new investments in the quarter. The average yield on the fix and flip portfolio in the quarter was 5.73%. As part of our exit from forbearance, we refinanced our existing secured financing and borrowed against previous untouched loans. All of our fix and flip loans are currently financed at 2-year term non-mark-to-market debt with the cost of funds of around 4%. Delinquencies in our portfolio increased as borrowers struggle with delays in costly transactions, slow down of rehab work and impaired personal financial conditions due to the COVID-19 pandemic. We implemented the payment deferral program with multiple services, where borrowers were assessed on a case-by-case basis confessed with our short-term payment deferral would be beneficial to the borrower and loan outcome. Borrowers will require to document the negative impact from COVID-19 on the profits of the project. The granted payment deferrals usually for 3 months of about 1.5% of outstanding loans in the quarter. Turning to Page 17. As previously noted, we saw delinquencies increase in the quarter with 60-plus day delinquent loans, increasing by approximately $70 million or 10% to approximately $182 million or 21% of UPB at the end of the second quarter. 7.5% of the increase was due to new delinquencies, while 2.5% of the increase was caused by lower portfolio balance due to paydowns. MFA is fortunate to have a highly experienced and talented asset management team and has for years, consistently improved outcomes in loss mitigation efforts on our residential whole loans, in particular, our nonperforming loan portfolio. The team consists of people with extensive experience in designing loss mitigation strategies dealing with title issues, bankruptcy filings and other common issues that can emerge in dealing with seriously delinquent loans. We work closely with our services and are confident that our hard work will lead to acceptable outcomes. Approximately 2/3 of the seriously delinquent loans are at a completed project where bridge loans were limited or no work is expected to be done, meaning these properties should be in decent condition. In addition, approximately 1/3 of the seriously delinquent bonds are already listed for sale. These reserves of $30 million were recorded as of June 30 for expected credit losses. This is modestly down to $35 million at the end of the first quarter. Loss reserves are recorded in earnings, and these numbers have therefore already been reflected in our first and second quarter earnings. As mentioned previously, we believe our asset management team provides us with a huge advantage in loss mitigation. In addition, we believe it's because we have secured term non-mark-to-market financing on our entire fix and flip portfolio, we will be able to patiently work through our delinquent loans to achieve acceptable outcomes. An additional loss mitigating factor is lower level reps and warranties made by originators at the time loan purchase. We currently have 8 purchased claims outstanding of $7 million of seriously delinquent loans, where we agreed with services that these loans will be purchased over time. Finally, on a massive level, we do the fiscal and monetary policy are supportive of home prices. As the Fed has lowered rates and purchase assets, mortgage rates are an all-time low and home sales will recover through depressed levels in April and May. Turning to Page 18. Our single-family rental loan portfolio held up reasonably well in a tough quarter. The size of the portfolio was relatively unchanged at $490 million as prepayments remained low due to strong prepayment protection, and we made no new investments in the quarter. The portfolio yield was relatively unchanged in the quarter at 5.8%. 60-plus delinquencies increased to 5% in the quarter. We primarily saw delinquencies increase in the month of April, May, we were stabilizing in June, and we have seen some positive delinquency trends post quarter end. Similarly to the fix and flip portfolio, we worked closely with our services to implement a payment deferral to forbearance program that borrowers will assess on a case-by-case basis to assess the benefit of a deferral of the brands. We granted payment deferrals for forbearance usually for 3 months and about 3% of outstanding loans in the quarter. As mentioned before, we believe the fiscal and monetary policy has been helpful from a credit perspective for our business purpose loans. In addition, short-term push for borrowers to new ratable apartments in densely populated cities to single-family homes to more space as well as long-term trends toward increased single-family rentals is supportive to our SFR assets. And now I will turn it over to Craig for some final comments.

Thank you, Gudmundur. The second quarter of 2020 for MFA was unprecedented in so many ways. It was through an unwavering determination and a herculean effort on the part of all of our dedicated team that we have repositioned and restructured our company so that once again, we can focus on creating value for our shareholders. While we are still somewhat cautious about the macroeconomic environment today and the still unknown post COVID-19 world, we believe that we are positioned to weather future storms while we evaluate investment opportunities, manage our capital structure and look to grow earnings. Greg, would you please open up the line for questions?

[Operator Instructions]. Your first question comes from the line of Steve Delaney from JMP Securities.

And first, just congratulations on all your work over the last 5 months now to position MFA to be able to move forward. And it's exciting to hear -- so I'm hearing that given with your financing, your package with Apollo and Athene and everything, and it sounds like you guys are kind of chomping at the bit to put some of that $600 million of cash to work. And if I'm hearing you right, it's probably going to be the NQM loan product and with an eye towards securitizing that. First, I guess, am I hearing you correct on that? And can you just give us an update? Have you actually started buying any loans here in August? And is that effort kind of well underway? And then I guess, attached to that question, you've got almost $6 billion of loans. Should we expect that this focus on securitization that some of which you already have that you'll be working to structure securitizations on those existing loans as well?

Sure. Thanks for the question, Steve. So I guess the first thing I would say is as much as we feel good about having a high cash balance. I wouldn't say that it's necessarily burning a hole in our pocket, and we need to go invest it tomorrow. As Bryan said, most of the originators, if they didn't shut down, they almost shut down for several months, and they've started to originate again. But I think it'll probably be another month or so before we see any sort of volume of closed loans to purchase just because the underwriting process takes some time. So it's certainly on the agenda, but it hasn't happened yet, but I think we're certainly close on that. In terms of securitization, with securitization spreads, they're not quite in as tight as they were back in early March, but they're pretty close, there's a deal of the price this week. And I think AAAs were or maybe 125 over swaps. Swaps are still low right now. So the securitization execution is really compelling. And ironically, because to your question about the existing portfolio, because originators have not been active or that active over the last few months or so. If you talk to some of the dealers that are active in securitizations, they're a little worried where their volume will come from over the next month or 2 while everybody ramps back up. So I think we have a great opportunity to be able to securitize the existing portfolio. And as I said during my remarks, the execution was securitization. It is substantially more attractive than non-mark-to-market financing, it's probably on the order of 100 basis points or a little bit more, and it probably provides almost as much as double the amount of leverage even if only selling down through single A. So it's very compelling right now. There's not a lot of supply. So I think the supply demand dynamics are good. And so I think that's where we stand on securitization. Also in terms of the cash balance, I think there are a variety of investment considerations that we can make, but one of those is -- obviously the obvious one to look at is we have that note from Apollo and Athene at a fairly high interest rate. So it's pretty easy to figure out what the ROE is if you pay down a portion of that note and unlike most of our other investments where the ultimate return depends on things that we don't really control like prepayments and interest rates and so on and so forth. It's pretty simple to calculate the ROE on that. So I think we have an array of things that we'll consider deploying cash on, but it's across the spectrum.

Great. And that paydown, it's interesting you would say that. Some of the -- I'm going to use the term rescue financing that we've seen, the facilities have had a minimum required interest payment over 1 or 2 years. It sounds like you might have more flexibility where you're not locked into having a certain amount outstanding for a period. Am I hearing you right on that?

You're correct. It's callable right away, and I'd like to take credit for that, but there was actually someone else that broke the ground on that in getting callable debt. So we were able to take advantage of that as well.

Our next question comes from the line of Henry Coffey from Wedbush.

And let me add my congratulations. An enormous amount of work has been done and a lot of value preserved. If you look through your existing assets to home values, which are actually holding up fairly well, I don't know on a zip code by zip code basis, but on a national basis they are. What is the -- what would the outcome be if you had to go to foreclosure on a lot of these properties? Would you be able to recover par? Would it create a manageable loss? Maybe that's obviously not the solution you want to pursue, but what would it look like if you did pursue that avenue?

Sure. So I'll sort of give an overview and let Bryan and Gudmundur talk about some of the specifics. But obviously, we did take some allowances for credit losses in the -- at the end of the first quarter. And I guess, we reversed some of those slightly at the end of the second quarter. So it's hard for me to say we're going to get that par on everything because we've taken allowances for loss reserves. But you're right, home prices have held in there pretty well. Mortgage rates are at all-time historic lows. And then also consider what the LTV was of the loans, right? So the Non-QM portfolio, those LTVs are typically in the mid or high 60s. And on the fix and flip portfolio, those LTVs, again, are in the mid-high 60s generally. So we think we've got a pretty good cushion. It takes some time to work through those. More time probably on Non-QM than fix and flip only because those are business purpose loans, so they're not owner occupied. But that's part of what this solidified financing is about as well. Gudmundur said, virtually our entire fix and flip portfolio is financed with 2-year non-mark-to-market financing. So we have the time to work through those to get good outcomes. And as Gudmundur also pointed out, we have a very experienced asset management team that's been doing this for 5 or 6 years. So I think we have the ability, we have the flexibility given the financing to get the best outcomes that we can.

Right. And as it relates to the Non-QM portfolio, right? When you're talking about a weighted average LTV of 63.5%, right? The most likely outcome of the borrower really gets into distress is going to be just a borrower sale of the property on their own, right? We really won't even have to necessarily go through the foreclosure process. And if, unfortunately, we had to that, when you get to that option, most likely, it would go to a third party, resulting in no loss. And as you look at the 3 years in which we've held loans, I think we've taken 2 loans or 1 loan so far has really gone to REO, and that property is not expected to take a loss. So far, so good. Obviously, when you looking to project things out, you have to think about sort of downside scenarios as well, but so far, so good.

I mean, mainly, it sounds like the issue is time and you have time.

That's right.

That's exactly right, Henry.

I would just add, look, I mean, you're right, home price would help as well. Probably better than people expect that it is certainly better than we expected probably late March, early April. Obviously, more of that has to do with fiscal [indiscernible] where rates are very low, and we don't see the headline mortgage rates, right? That overall supports home prices. But foreclosure is still expensive. It's foreclosure costs, basically, throughout the entire process, you might have to deal with delinquent taxes, insurance, other things. At the end of the day, also we could take over the property, when you sell it, there's agent and reality cost and so on and so forth. So I think you're right, home prices have held up well, so that's positive, but it's still costs money to take things for foreclosure, and so our preferred outcome is to get to a solution before that. But we're confident in our ability to get to accessible solutions in most of these cases.

Your next question comes from the line of Doug Harter from Credit Suisse.

You talked about how the forbearance financing was more expensive and a drag on the second quarter. Can you just talk about kind of the new financing levels where they are relative to what you were paying in forbearance? And then just help us also understand what the pacing could be of securitization to achieve that cost saves that you talked about on that?

Sure. So forbearance interest, suffice to say was very expensive. I would say LIBOR plus 500, maybe even a little bit more. So it's not really indicative going forward. We do, and we actually have a couple of places in the press release where we laid this out by product type. But I mentioned in my comments, Doug, that the secured financing away from the Apollo, Athene, large loan and securitizations, the average cost is about 3.6%. And it will vary a little bit by product type, but I would say generally speaking, it's probably in the vicinity of LIBOR plus 300 to, in some cases, maybe LIBOR plus 400. And you asked about securitization. So not many people probably remember this, but we were one day away from pricing a Non-QM securitization on March 16. So suffice to say, that's probably our single biggest priority right now is to move forward on securitization. And as I said before, I think the timing is very good because there aren't a lot of deals in the market and the executions have continued to improve.

I guess how can we think about the sizing? I mean, is kind of the size that you were looking to do before the right size? Obviously, you have a lot of loans. Could you do bigger? Just kind of how should we think about that potential for sizing?

Sure. So it's a good question. Obviously, we're not a first-time securitizer. But for Non-QM loans, it will be our first securitization. And so I think at least the first deal that we do we probably don't want to bite off too much. You want to make sure that the deal is an orderly deal and gets a good execution. But I think over time, we could look to grow those. So I think when we -- the deal that we were close to doing back in March, I think, was $400 odd million or so. I think it's probably safe to assume that's sort of a good place to start. But depending on demand, we could grow that. I think the existing Non-QM portfolio is about $3 billion. So there's a lot of securitization to do.

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

Can you hear me?

We can hear you, Rick.

I wanted to talk a little bit about the allowance on the Non-QM portfolio. And I realized that there was a significant reduction related to the sales, which makes sense. But when we look at the characteristics, and understandably, there weren't sort of the normal slides when we look at the -- for first quarter. But if we compare the December 31 metrics on the Non-QM portfolio versus the 6/30 there's not a huge change. It doesn't look like there's a big change in terms of portfolio quality in terms of FICO, LTV, coupon. Yet the reduction of the portfolio is about 30%, and the net reduction of the reserve was close to 70%. I'm curious sort of what drives that discrepancy I know there's a comment in the footnotes related to economic outlook. I'm also curious what time frame you're comparing economic outlook from March 30 to June 30 reserves?

Okay. Steve, I don't know if you want to talk about that. Rick, I would point out that we implemented CECL in 2020. So anything that you compare to December 31, and it was a different accounting standard back then. But Steve you want to talk about it?

No, no. I understand. I understand. I'm looking at allowance March 30 to June 30. I was just trying to figure out if by looking back to December, if the portfolio characteristics would have changed since I don't have the March data, but it doesn't look like you sold something lower quality in the second quarter to explain the decline in the reserve.

So Rick, it's Steve. I think I'm not entirely sure which numbers you're looking at, whether it's in the press release and the deck. But maybe if you have a look on Slide 23 of the deck, you can see a bit of a roll forward by product, right? And backing out the reversal of the adjustment that we made at March 31 on the loans that were being sold, that $70.2 million adjustment. The reduction in the Non-QM reserve was about $2.3 million for the 3-month period. And you're right. The -- if you look at the statistics, we have a table in the press release, Table 6 on Page 8. If you look at the statistics and even on the delinquency levels, certainly the FICO and the LTV statistics on the Non-QM portfolio aren't that different. So what's driving the CECL between March and June is primarily some assumptions that we made back in April around unemployment and future unemployment. And we've modified those slightly, and that's really driving that reduction in the CECL reserve. So you -- we -- when we did our CECL reserves back in the end of the first quarter, we had a little bit more time to think about economic outlook than say, some of the bigger banks who were doing their CECL reserves and announcing earnings before we did. And so we had what I thought were appropriate, but fairly conservative assumptions for unemployment, for example, and we've sort of just dialed those back a little bit at the end of the second quarter. Obviously, there's still a lot of uncertainty around what might happen if there's a double dip in the -- with the slowdown with the states closing and whatnot. But we still feel that our unemployment assumptions are appropriate into the second quarter, but that did drive some of that reduction in the Non-QM CECL reserve.

Just to be clear. I think it's a little -- it may be a little bit confusing just with the presentation. We had a $70.2 million valuation adjustment on our Non-QM loans at 3/31. That was that large pool that we sold that I referenced in my comments that we had a total loss of about $127 million on. That $70 million was not a credit adjustment. It was a valuation adjustment on loans held for sale. We knew at March 31 that we would be selling those. So basically, that $70 million write-down was to take that down to where the mark was on March 31. And then the sale occurred, so it gets reversed out. So it was never really part of a credit loss allowance, if you will.

Got it. Okay. So that makes sense. What you're basically saying and we understand all the geography on this, that the fair value mark was, in fact, greater than the CECL reserve for the remaining portfolio. So that's really what's driving the differential. The follow-up question to this is that with a relatively large, as you look at the migration into July, and we appreciate those metrics. When you look at the migration of a high number of 30-day delinquencies as loans come off forbearance and a relatively low 60-day delinquency. Should we expect CECL reserve to continue to build as a relative -- a percentage of those loans cascade from 30 to 60 or actually, by the time we have this next conversation, 90 or 120 days?

Sure. So Steve, do you want to talk about that? Because we have a separate process for delinquent loans than we do for just CECL reserves in general.

Yes. When we -- so when we look -- when loans go 60-plus delinquent, we -- our process looks at those loans and looks at them somewhat separate, depending on how certain we are going into foreclosure. So when we think they're going into foreclosure, we'll look at them more on an individualized basis, and it will be more of an analysis of where we are from a security position, the value of the underlying collateral versus the loan balance. But as loans are inside that level, the CECL reserve will look at macroeconomic assumptions and look at peer data, historical loss estimates and peer data to drive that season reserve. So the process sort of flips over as loans get more seriously delinquent.

Got it. And look, I realize my questions are a little bit pointed, but -- and let's all acknowledge that the normal probabilities of a loan going from 30 to 60 or 60-day loan defaulting is very different and it's going to be challenging going forward. I need to recognize that as well.

Right. And just to be -- just to try to provide some additional color. So we're now in August. So the 30-day population for that Non-QM portfolio was somewhere between 3% and 4% delinquent in July, right? So a piece of it rolls to 60, so you're at like 7% 60 in July. So you had basically half the guys that -- or 1/3 of the guys sort of stayed where they were, half of them went to 60 from that 30 bucket, and then another piece of them went to current. So you're not seeing that 30-day bucket sort of continue to build. It's sort of a onetime item, the guys who had experienced that the COVID impact. So the concern was you have 30% that's impacted by COVID, there's 30% of your portfolio going to be 30 and then 60, but really, no, that's not the case. It's a much lower number. And sort of that's what we're trying to sort of explain here.

Got it. And thank you for the update on July. I misspoke when I said -- looking at so many slides.

[Operator Instructions]. We have a question from the line of Eric Hagen from KBW.

I had a similar question as Rick. You guys gave a good answer. But the net earnings reported through -- as a result -- to the nonperforming loan portfolio, I think it was $20 million last quarter. What's the outlook for returns there? I know you don't typically ascribe a yield to that portfolio, but how should we think about the earnings potential there? And then as a tap on to that, is your portfolio of REO pledged as any -- as collateral for any of your funding loans or it's essentially just held with equity?

Eric, so I'll let Bryan talk a little bit about the FBO loans -- sort of the REO portfolio, that is not encumbered at the present time. So we own all those with equity. I think we have said over time that our expected over time returns on our fair value loans are nonperforming loans. We've generally said, we see a range of 5% to 7%. And I don't know that, that's really changed at this point. Obviously, it varies quarter-by-quarter. As you know, the income that we show on those fair value loans is not just the change in the fair value mark, but it's also cash that we received. Because many of those loans that we purchased as nonperforming loans and elected fair value accounting on, we've been able to modify and they're now performing loans. But once we designate them as fair value, they're fair value forever. I will say that the mark change -- the fair value mark change on the nonperforming loan portfolio for the second quarter was only about $2 million. So I think of that $20 million, about $18 million of that was cash. So there was very little change in that mark. So again, not to speculate what happens in the future, but there's probably some room in the valuation side. The market June 30 was not really that much higher than it was in March.

Great. And then I think maybe Steve Delaney asked a kind of similar question, but I feel like just to clarify around any runoff that might be reinvested or kind of shored up to potentially reinvest later. What's the current status in the third quarter? Have you guys been able to reinvest any paydowns? Or is -- are you really just retaining that liquidity for the, I guess, kind of near future?

Sure. So we're in conversations with originators. So I would say we're close to start putting out some money again. But I think we're proceeding somewhat cautiously. We don't -- I mean we don't know what the world looks like in a post COVID world. We don't know for certain that we won't see another downturn. So I think we feel really good about where we are from a liability standpoint, from a balance sheet standpoint and from a liquidity standpoint. And I think you've heard a similar theme on pretty much every other mortgage recall that I don't think anybody says, well, I have a lot of cash. I have too much cash. Obviously, it's a drag to have that much cash. But I think given what we've all lived through over the last 5 months or so, it's probably understandable that we're going to deploy cash, maybe somewhat conservatively, at least for the short term.

Right. And how about loan sales, Craig, I mean, could you -- as the market continues to recover and your fair value mark gets back to where you guys think it should be. Could you entertain selling a portion of the Non-QM portfolio from here?

I suppose we could. I think it's certainly not our intention. We did sell some in April, I guess probably closed in May. But at this point, I don't really think that's on our -- I'll never say that we'll never do it, but I really don't think that we're thinking that loan sales make all that much sense. I think securitization makes a lot of sense, particularly right now given where rates are and where spreads are. So I think that's probably the more likely route to go.

And at this time, there are no further questions.

All right. Well, thanks, everyone, for your interest in MFA Financial. We look forward to our next update after the third quarter in early November.

Ladies and gentlemen, that does conclude your conference for today. Thank you for your participation and for using AT&T teleconference. You may now disconnect.