Ladies and gentlemen, thank you for standing by, and welcome to the MFA Financial Incorporated Second Quarter Earnings Call. [Operator Instructions] As a reminder, this conference is being recorded. .

I would now like to turn the call over to our host, Ms. Danielle Rosatelli. Please go ahead. .

Good morning. 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, 2015, 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 and the press release announcing MFA's second quarter 2016 financial results. Thank you for your time. .

I would now like to turn the call over to Bill Gorin, MFA's Chief Executive Officer. .

Thank you, Danielle. I'd like to welcome everyone to MFA's second quarter 2016 financial results webcast. With me today on the line are Craig Knutson, MFA's President and Chief Operating Officer; Gudmundur Kristjansson, Senior Vice President; Bryan Wulfsohn, Senior Vice President; Steve Yarad, CFO; and other members of senior management. .

In 2016, we continue to execute our strategy of targeted investment within the residential mortgage universe. We have many years of experience in analyzing and investing in such assets, and thanks to our permanent capital REIT structure, we have the staying power to hold these assets throughout fluctuations in market value. .

Turning to Page 3. Despite this period of extremely low interest rates, we remain well positioned to generate attractive returns. Interest rates are being held down on a worldwide basis by accommodative monetary policy, low economic growth rates and a bias towards low inflation.

In this environment, we continue to identify and acquire attractive, credit-sensitive residential mortgage assets such as 3-year step-up RPL/NPL securities and credit-sensitive residential mortgage loans. In the second quarter of 2016, we generated net income of $75.2 million or $0.20 per common share.

The dividend was again $0.20 per share, and book value per common share at June 30 was $7.41. .

Turning to Page 4. MFA began operations nearly 18 years ago and has generated strong long-term returns to investors through volatile markets and through various interest rate and credit cycles.

Since 2000, we've generated annualized shareholder returns of approximately 15%, and over the last 10 years have generated annualized shareholder returns of 13%. .

Turning to Page 5. We'd lay out MFA's strategy for 2016. We continue to focus on high value-added, credit-sensitive residential mortgage assets. The credit assets we've acquired continue to perform well and tend to have less interest rate sensitivity.

These assets are typically purchased at a discount, so they actually benefit from increases in prepayment rates. Our strategy does require staying power, which gives us the ability to invest in and hold long-term distressed less liquid assets. .

We have permanent equity capital. Our debt-to-equity ratio of 3.3x is low enough to accommodate potential declines in marks. MFA is able to invest significant amounts at advantageous prices, while other investors may be facing capital outflows. Just as a reminder, we are internally advised and our compensation is not tied to size. .

Turning to Page 6. Our mortgage assets run off due to amortization, paydowns or sale, allowing reinvestment opportunities in changing interest rate and credit environments. In the second quarter, we were a buyer of 3-year step-up RPL/NPL securities, loans, CRTs and legacy non-agency MBS.

In the quarter, our 3-year step-up portfolio, our reperforming/nonperforming loan portfolio and our holdings of credit risk transfer securities each grew. We did not acquire any agency MBS in the quarter. .

Turning to Page 7. As you can see, MFA's yields and spreads remain attractive and relatively consistent despite the interest rate environment. .

Turning to Page 8. We present yields and spreads for our more significant holdings. Given the leverage we're utilizing or may utilize in the future, each of these asset types are generating attractive returns to MFA shareholders. .

Turning to Page 9. Our undistributed REIT taxable income stood at $0.22 per share as of June 30, 2016. This amount per share grew in the second quarter due to the taxable impact of the Countrywide settlement. .

Turning to Page 10. Gudmundur will present an update on MFA's interest rate sensitivity and, on the following 2 pages, present the impact of prepayments on MFA's portfolio. .

Thank you, Bill. On Slide 10, we review the interest rate sensitivity of MFA's assets and liabilities. MFA's asset duration declined 10 basis points in the quarter to 122 basis points at the end of the second quarter.

The decline was due to continued growth in our holdings of credit-sensitive loans, RPL/NPL securities and CRT securities, all of which exhibit limited interest rate sensitivity as well as a modest decline in duration estimates for our agency MBS due to lower interest rates.

The notional amount of our swap hedges remains unchanged at $3 billion at the end of the second quarter, while our hedge duration declined 30 basis points to minus 3 at the end of the second quarter as our swap hedges shorten naturally over time.

In aggregate, MFA's maturation declined 5 basis points in the quarter to 50 basis points at the end of the second quarter.

MFA continues to pursue a strategy of maintaining low levels of interest rate risk in our portfolio by keeping net portfolio duration below 1 and limiting the impact of changes in long-term interest rates by preferring shorter credit-sensitive assets. .

Let's turn to Page 11. 10-year Treasury rates have declined by approximately 75 basis points in 2016 and reached a new all-time low of 136 basis points in early June before settling in around 150 basis points recently. In fact, interest rates haven't been this low since 2012.

Even though primary mortgage rates have fallen slower than Treasury rates, it is reasonable to assume that prepayment risk has increased if rates stay at these levels. .

premium amortization on our agency MBS that were acquired at an average premium of 4% and discount accretion on our legacy non-agency MBS that have an average purchase discount of approximately 27%.

MFA currently holds about $4.3 billion of agency MBS with a remaining premium amount of $155 million and about $3.5 billion of legacy non-agency MBS with a total purchase discount of about $1 billion. .

Interest rates have been on a downward trend since the first quarter of 2014, but as you can see from the graph on this slide, prepayments of MFA's agency MBS have been relatively stable, rising in the summer and declining in the winter.

However, prepayments of MFA's legacy non-agency MBS have trended higher during the same period as positive home price appreciation has allowed more and more non-agency borrowers to take advantage of lower rates. In addition, we have observed that prepayment of agency and legacy non-agency MBS have tended to rise and decline at the same time periods.

And as you can see from the graph, the prepayment rate on agency MBS and legacy non-agency MBS are currently very close to each other. .

Now let's turn to Page 12. On Page 12, we show the magnitude of agency MBS premium amortization and legacy non-agency MBS discount accretion since 2013. As we can see, discount accretion has been twice the premium amortization in almost all quarters over this time period.

Keep in mind that premium amortization is a negative for our earnings, while discount accretion is a positive for our earnings. .

In summary, since the size of MFA's legacy non-agency MBS purchase discount is much larger than the premium amount on MFA's agency MBS, and the fact that prepayment rates on the 2 asset classes trend together and are now close in magnitude, we believe that an increase in prepayments would not negatively impact MFA's earnings. .

With that, I will turn the call over to Craig, who will talk about our credit assets and credit fundamentals in more detail. .

Thank you, Gudmundur. On Page 13, although economic data continued to be mixed, the U.S. residential mortgage credit market enjoys unequivocal, fundamental and technical support. The labor market and employment numbers continue to improve. Interest rates and mortgage rates remain very low.

Sales of existing homes rose 3% year-over-year to 5.57 million, the highest level since February of 2007, and median existing single-family home prices are up 4.8% year-over-year. According to a recent CoreLogic report, foreclosure inventory is down 24.5% in the last year.

Clearing out foreclosure inventory is an important step in producing home price appreciation, as distressed inventory tends to hold prices down. Also, 268,000 residential properties with a mortgage that were previously underwater regained equity in the first quarter of 2016.

Borrowers with equity in their homes are obviously less likely to default in the future, and they can also refinance their mortgage at low current rates, which generates a prepayment on our existing loans. .

Moving to Page 14. We closed on close to $90 million of credit-sensitive residential home loans in the second quarter, growing this portfolio to almost $1.1 billion. The supply picture for the balance of 2016 looks promising with continued selling expected from GSEs, large banks and other market participants. .

either the loans begin to reperform, perhaps after some loan modification, or we take possession of the property and eventually sell the property. We currently have borrowings through 3 warehouse lines with aggregate borrowings of approximately $584 million, and we're in the process of opening up a fourth warehouse line.

Once completed, we'll look to add some leverage to the reperforming loans that we have not financed with leverage to date. Our asset management team is actively managing delinquent loans and REO properties together with our servicers in order to enhance outcomes on these property liquidations. .

Turning to Page 16. We purchased approximately $466 million of RPL/NPL mortgage-backed securities in the second quarter, outpacing our runoff and growing this asset class by about $143 million in Q2.

We continue to like these assets due to their low sensitivity to interest rates and what we believe to be low credit risk, while at the same time providing attractive ROEs. And despite heavy supply, spreads on new issued deals actually tightened during the quarter as more buyers have become attracted to these assets. .

Turning to Page 17. The credit metrics on the loans underlying our legacy non-agency portfolio continue to improve. 83% of the loans underlying our legacy non-agency portfolio are now amortizing. This principal amortization, together with home price appreciation, continues to drive LTVs down. Delinquencies are curing.

60-plus day delinquencies as of June 30 for this portfolio have declined to 12.6%. .

On this page, we illustrate the LTV distribution of current loans in the portfolio. The red bars on the right-hand side represent at-risk loans where the homeowner owes more on the mortgage than the property is worth. These are the loans we worry most about transitioning to delinquent and defaulting in the future because the borrowers are underwater.

As you can see, these red bars are disappearing. .

Please also note the increasingly large green bars on the left side. These are loans with LTVs below 80%, and these are attractive refinancing candidates.

A combination of low interest rates available today and a 30-year amortization on a new mortgage versus the 20-year remaining term on existing loans can offer homeowners substantially lower monthly payments. And of course, given our deeply discounted purchase price for these assets, we're very happy when the underlying loans prepay. .

And with that, I like to turn the call back over to Bill. .

Thanks a lot, Craig. So as you can see, despite the low interest rate environment, we continue to identify and acquire attractive credit-sensitive residential mortgage assets. Our credit-sensitive assets continue to perform very well. We believe we're well positioned for changes in prepayment rates, monetary policy and/or interest rates. .

This completes today's presentation.

Operator, could you please open up the lines for questions?.

[Operator Instructions] And we will go to our first question coming from Steve Delaney with JMP Securities. .

A lot of chatter in the market lately about the money market fund rule changes going into effect in October. Early comments have been that it might be a positive for agency repo, and I was curious if you have thoughts as to how those changes might affect your non-agency repo borrowings. .

Thanks for the question, Steve.

Gudmundur and Bryan, would you guys like to address this?.

Yes, so, Steve, you're right. What we've seen is that there's been no impact on agency repo in general and, in fact, probably the difference between LIBOR and agency repo is strong.

So it appears, at a minimum, it will have no impact on agency repo, but probably, it might actually be a positive because it might force some of the prime money market funds to basically invest more in kind of U.S. government-related securities, and repo will fall under that. .

And do you know, Gudmundur, I mean, in terms of your -- I've always heard MFA and others tell us that the -- the repo market for non-agency assets is just a far different market in terms of counterparties and terms than the agency repo.

But do you know or -- if money market funds, especially these retail and government money market funds, previously, were they involved at all in non-agency repo, to your knowledge?.

So I'll just make a quick comment and let Bryan answer in more detail. But what we've seen over the last few years is that we've seen more counterparties on the non-agency repo side. And it's been also not just the broker dealers, or kind of the trading desk, but it's been more directly with the banks themselves.

So to that extent, where they're using deposits to fund repos, it really shouldn't impact that because they're not necessarily getting that money from their money market funds. .

Got it. So this sounds like, if anything, maybe a slight benefit on the agency side and probably no damage on the non-agency side, so. .

Yes. .

On your CRT investments in the second quarter, the $51 million, should we assume those are more normal, the stackers and the Connecticut Avenue [ph], the GSE-type offerings? Or have you ventured into any of these private risk transactions that some of the large banks are looking at doing?.

So, Steve, they're all GSEs, at least at this point. We've looked at all those deals, but at least so far, we've had the most luck with the GSE securities. .

And just one last thing. I was touched by your Slide 17. Obviously, a tremendous shift, if we were to look at that a year ago or 2 years ago. I'm curious. You're buying -- we have companies that own -- that have some servicing or they buy the call rights and the control bonds.

You acquired, obviously, I guess what we still call born AAA bonds, by and large, I believe.

As I look at that slide, I guess, my question is sort of from an operational standpoint, is there any way that you as the owner of the senior bonds, is there any way that you can work proactively with servicers to somehow or another accelerate whether it's through attractive refi offers or discounted payoffs? Other words, is there a sort of an operational strategy around there to accelerate repayment if, in fact, you see that as a benefit to you?.

So Steve, it's a good question. Unfortunately, the fact is that the cleanup call rights typically exist somewhere else, and we don't frankly know where those cleanup call rights are.

So we do see deals that get called from time to time, but unfortunately, because of the nature of those securities, we really don't have visibility into who owns that and how that decision is made.

But the one thing that makes that call a little bit tricky is when they execute that cleanup call, they have to buy all the loans, right? They have to buy all the loans at par. And so typically you'll still have delinquent loans.

So the current loans are clearly probably in most cases, if not all cases, worth more than par, but the delinquent loans obviously are not. So it's also a tricky timing issue as to what the current state of the underlying loans is before these deals get called. .

Our next question will come from Douglas Harter with Crédit Suisse. .

As the flip side to the strong book value performance you guys showed this quarter, how do you see the opportunity to continue to invest in some of the NPL deals that you guys have been relatively early into?.

Are you talking about the securities or the loans?.

I guess I'll make it a little broader and just, if you could contrast both of those opportunities. .

You're right. We don't have to buy every day. Certain days, we're buyers, certain days, we're sellers. But I will tell you, in this quarter, we've still been able to identify assets at attractive yields.

But you're right, pretty much since the Brexit vote, there's definitely been a somewhat of a decline in available yields, but they're still available at attractive levels. .

And our next question in line will come from Joel Houck with Wells Fargo. .

A couple of related questions, if you can reference Slide 6. So obviously, you've talked about the increase in RPL/NPL and the step-up. Looks like your funding that with runoff of legacy and agency.

I guess, the first question is, absent any meaningful change in economic landscape, is that -- is the second quarter kind of a good run rate in terms of us modeling how the composition of your portfolio will change over time? And then the second part of the question is, can you maybe help us frame up the delta between the IRRs you look at in the top 2 categories that you're growing versus what's running off? I'm assuming it's a positive delta, but any attempt to try and put a band in terms of quantifying it would be helpful.

.

All right. So in terms of declines in growth, it's hard to predict the runoff in a quarter. So I guess, a good starting point might be the prior quarter. But I guess, we would not be surprised if prepays increase in this quarter. So you might see a little more runoff each of the categories.

In terms of acquisitions, we continue to be focused on the same asset classes, but like I say, a buy or sell decision depends on the market day to day. So unfortunately, I can't help you predict the acquisitions we're going to make in the third quarter, but we are finding opportunities in the quarter.

In terms of the delta of the ROE, the reperforming and nonperforming -- well, first of all, in terms of the 3-year step-up, RPL/NPL securities.

Craig, you want to go through those economics on ROEs?.

Sure. So the most recent bolt [ph] deal probably came at a 3.5% type yield, which is in from -- those were wider than 4% at one point in the first quarter, so they've tightened in a lot. And so those ROEs are going to be sort of high single digits.

In terms of legacy, those yields are also in, and it really depends whether it's hybrid or fixed rate, but I would say yields on those are probably, best case, meaning widest, probably 4% these days. And in many cases, they could be as tight as 3.5% where the RPL/NPL securities trade.

So ROEs on that are similar because the funding is pretty much the same, similar haircut, similar rate.

In terms of reperforming and nonperforming loans, and to go back to your question about what to expect in the future, the big difference between buying loans and buying these securities, especially these new issue RPL/NPL securities, is those deals come and a bunch of people put submit bids and maybe they're oversubscribed so you get an allocation, you get cut back on your allocation.

When we buy loans, only 1 guy wins. So if we are fortunate and we buy a large loan pool, then obviously, we're the only winner. But by the same token, if we don't buy it, then we don't get any of those loans.

So I think, it's much more difficult to predict that and the additions are much more lumpy than what you will typically see on the rest of the portfolio. .

And is it fair to say that there's a strong correlation in pricing with the loans versus securities? Or is there somewhat of a lag, given what happens with spread widening and contraction in the market?.

You've seen some contraction, but you've seen it more on the reperforming side versus the nonperforming side, especially the assets that are more stable, just because rates are valleyed and people are looking for duration. .

And then lastly, I mean, how should we think -- or how do you guys think about the agency portfolio? It's -- the recent history is you hadn't added much. It's kind of been running off. It doesn't seem like it's an asset class that you guys view as attractive as your other opportunities. I'm just kind of curious as to how you're viewing it currently. .

It's -- I think, it's been a couple of years since we actually have acquired agencies, and we probably haven't had as strong a view on interest rates as you have. We try to have the company portfolio not be sensitive to calls in interest rates. But it's the absolute yields that have kept us away.

And you look at the absolute yields in our assets, which are in excess of 4% and with low leverage. I have more comfort there than buying very low-yielding assets, even with the government guarantee and increased availability of funding.

Gudmundur, what would agency assets yield today?.

Well, the absolute yields on the short end of the curve, if you talk about hybrids and 15-years, the yield's probably between 1.5 to 190 basis points, something like that. That's the absolute yield. .

So Joel, hopefully, I don't -- we haven't bought any recently, and unless things change, we're probably not going to be very active in that marketplace for now. .

And our next question will come from the line of Bose George with KBW. .

Just a follow-up on pricing. Curious what you think about pricing now on the legacy non-agency MBS.

And to the extent that valuations go up there, I mean, could we see any culling of the portfolio into stuff that might be less rich?.

So Bose, your question is what's happened since the end of the quarter? Is that basically your question?.

That is, and also to the -- if we could see more movement away from that portfolio, just thoughts on the valuation there?.

Craig, you want to take it?.

Sure. So Bose, I think, you've actually seen -- yes, that portfolio has declined sort of on a regular basis, I think more so due to paydowns than to sales. But actually, in the last couple of quarters, we've actually both bought and sold those. So I wouldn't say necessarily as they get tighter that you should expect more sales.

I think we've approached that very opportunistically, and we've been able to find bonds that we think are, on a relative basis, are cheap, and we've been able to sell some that we think on a relative basis are a little bit rich. .

[Operator Instructions] And we'll go to the line of Rick Shane with JPMorgan. .

Obviously, given the commentary about the expectations of a little bit of a pickup in CPRs on the non-agency portfolio, there's going to be increasing focus on the relationship between credit reserves and accretable discounts.

Can you just talk a little bit about the methodology and how you update that? So for example, as you guys have highlighted some pretty significant underlying HPA, how we should be thinking about this going forward and where the potential is. .

Craig?.

So with the review process, that credit reserve review process, if you will, is a pretty well-established process here. And basically, every bond gets a deep dive analysis at least every 9 months. And the things that we really focus on are the things that we typically talk about, right? It's really LTV.

We actually use 2 methodologies to come up with liquidation assumptions, and one is primarily LTV based. And the other, which is somewhat independent, is based on payment history. And typically, those -- the 2 conclusions are pretty similar.

But it's a bond-by-bond analysis, and we're not taking projections of future home price increase into effect when we do that analysis. We're looking at sort of current state. And I think, you've seen that, that credit reserve has steadily declined over the last few years.

That said, there's still 20 years to go, so I think, like I say, it's a pretty well-established process that we have to do that. .

Great, that's helpful. So and it actually raises a couple of questions for me. One is, are you actually looking on a loan-by-loan basis and basically doing a bottoms-up analysis? And you made the comment that every bond is reviewed at least every 9 months.

Do you then take the sort of generic assumptions each quarter that you develop on the bonds that you review and apply them to the overall portfolio? Or do you wait until that review? And the reason I'm curious is, obviously, in the second quarter, there were some very favorable trends, and what I'm trying to figure out is, was that applied -- what percentage of the overall portfolio was that applied to?.

Okay. So I think I understand the question. And the answer is, if something happens in the second quarter and it affects the bond that we are examining specifically in that quarter, we don't then extrapolate that to the other bonds that we didn't look at in that quarter, because every bond is modeled and has its own set of assumptions.

So if we didn't look at a bond in the second quarter, we're not likely changing assumptions with respect to that bond. But you also asked about, do you look at individual loans. And in theory, yes, we look at individual loans because we look at -- and we do it statistically, but we look at underlying loans at a ZIP code level.

So you could argue we're not exactly looking at every single underlying loan, and that would be a true statement. On the other hand, for instance, if we own a bond and there's a couple of large loans that are in foreclosure or might even be REO properties, there are some cases where we might actually be able to figure out where that property is.

And so in those cases, we actually are looking at specific loans and specific information. But it's difficult to triangulate and to actually get that information. But to the extent we can, we do. .

And Rick, this is Steve Yarad.

In addition to what Craig outlined around the review of the bonds that takes place every quarter and we typically cover each bond every sort of 6 to 9 months in the portfolio, to the extent that bonds that weren't sort of on the regular time line for review, if they have significantly appeared to have outperformed, or even underperformed, during a quarter, they will also be picked up for a special review.

So we get significant coverage every quarter of bonds on a routine basis, but anything that also looks like a bit of an outlier based on its performance for that past quarter. .

Right. Steve's right. We actually have bonds that are called scheduled review bonds for the quarter, but then we also have certain flags, where we'll flag unscheduled bonds, and they will also get reviewed during that quarter. So for instance, if some bond gets a really large prepayment, then that might trigger an unscheduled review during the quarter.

.

Got it, okay. And I'm going to ask one last follow-up, and I apologize to everybody else. So you have just under $3 billion in amortized cost. What you basically described is that once every 3 quarters, every bond is reviewed, so that suggests about 1/3 of the portfolio.

Perhaps, because of the special reviews, 40% of the portfolio is reviewed for second quarter marks and -- correct me if I'm wrong there. And how -- what was the mechanism this quarter? Because again, we'd love to extrapolate what that implies for the remainder of the portfolio. .

So you said they're reviewed for marks. Mark is a completely different... .

I'm sorry. I meant accretable discount. .

Okay. So, again, it's really -- look, it's really bond by bond. So the home price appreciation is looked at on a ZIP code level basis, so I think it would be very difficult for us, so I would imagine it would be at least as difficult for you and maybe more so, to try to extrapolate that.

And again, I'm not sure that -- I mean, the results of that extrapolation, I'm not sure it's all that helpful because when we change the credit review, all that really does is it changes the yield prospectively. So it's not like that credit review change flows through our income statement and would drive something for the quarter.

It's all going to be realized over the remaining life of those securities. .

Understood. But again, given the trend on the yields, that's something to look at. .

And currently, no further questions in queue. .

All right. Well, I'd like to thank everyone for joining us today, and we look forward to speaking to you again next quarter. .

And ladies and gentlemen, this conference will be made available for replay after 12:00 today and running through Thursday, November 3 at midnight. You can access AT&T executive playback service by dialing 1 (800) 475-6701 and entering the access code 399029. International parties may dial 1 (320) 365-3844. .

That does conclude our conference for today. Thanks for your participation and for using AT&T Executive Teleconference service. You may now disconnect..