MFA Financial Inc (MFA) 2013 Q1 法說會逐字稿

Ladies and gentlemen, thank you for standing by and welcome to the MFA Financial, Inc. First Quarter 2013 Earnings Call. For the conference, all participants are in a listen-only mode. There will be an opportunity for your questions, and instructions will be given at that time. (Operator Instructions).

As a reminder, today's call is being recorded.

With that being said, I will turn the conference now to Ms. Danielle Rosatelli, Financial Analyst and Investor Relations. 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, plan, continue, intend, should, could, would, may, 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, but not limited to, those relating to changes in interest rates and the market value of MFA's investment securities, changes in the prepayment rates on the mortgage loans securing MFA's investment securities, changes in the default rates and management's assumptions regarding default rates on the mortgage loans, securing MFA's MBS, MFA's ability to borrow to finance its assets, implementation of or changes in government regulations or programs affecting MFA's business, MFA's ability to maintain its qualification as a real estate investment trust for federal income tax purposes, MFA's ability to maintain its exemption from registration under the Investment Company Act of 1940, MFA's estimates regarding taxable income and the timing and the amount of distributions to stockholders, and risks associated with investing in real estate-related assets, including changes in business conditions and the general economy.

These and other risks, uncertainties and factors, including those described in MFA's annual report on Form 10-K for the year ended December 31st, 2012, and other reports that it may file from time to time with the Securities and Exchange Commission 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 first quarter 2013 financial results.

Thank you for your time. I would now like to turn this call over to Stewart Zimmerman, MFA's Chief Executive Officer.

Good morning. Welcome to MFA's First Quarter 2013 Earnings Call. Joining me this morning on the call are the senior officers of MFA.

Today we announced financial results of the first quarter ended March 31st, 2013. Recent financial results and other significant highlights for MFA include the following; our first quarter net income per common share was $0.21 and core earnings per common share was $0.20.

Book value per common share was $8.84, as of March 31st, 2013, compared to $8.99 at December 31st, 2012. Excluding the impact of the $0.50 per share special dividend declared March 4th , 2013, book value would have increased in the quarter due primarily to continued appreciation within the non-agency MBS portfolio. Relevant to that, as you might notice on the front page of the "Wall Street Journal" today, it mentions housing market accelerates and, again, we believe that's something will be very positive -- a continued positive for MFA and our non-agency portfolio.

On April 30th, 2013, we paid our first quarter 2013 dividend for $0.22 per share of common stock to stockholders of record as of April 12th, 2013. A combination of both home price appreciation and mortgage amortization has led to a decrease in the loan devaluate ratio for many of the mortgages underlying our non-agency portfolio. Due to this lower LTV, we have reduced our estimate of the future losses within our non-agency portfolio. As a result, in the first quarter, we transferred $34.5 million to accretable discounts from credit reserve, bringing the total transferred over the last nine months to just short of $170 million. This increase in accretable discount prospectively increases the yield on our non-agency mortgage-backed securities and will be realized in income over the life of the assets.

For the first quarter ended March 31st, 2013, we generated net income allocable to common stockholders of $75 million or $0.21 per share of common stock. Core earnings for the first quarter were $72.3 million or $0.20 per share of common stock. We continue to provide stockholders with attractive returns to what we believe to be appropriate leveraged investments in both agency and non-agency residential mortgage-backed securities.

At quarter end, our debt-to-equity ratio was 3.1 to 1. Our agency portfolio had an average amortized cost basis of 103.4% of par as of March 31st, 2013, and generated a 2.42% yield in the first quarter. Our non-agency portfolio had an average amortized cost of 73.2% of par as of March 31st, 2013, and generated a loss of adjusted yield of 6.8% in the first quarter.

We believe MFA, as an (inaudible) managed REIT continues to be a very efficient vehicle for delivering the benefit of residential mortgage-backed security investment to our stockholders. I thank you for your continued interest in MFA Financial. And at this time, I would like to open the call for questions.

(Operator instructions). We first go to the line of Cheryl Pate. Please go ahead. Cheryl, your line is open, possibly take yourself off mute.

Hi, good morning.

Good morning.

Just a question, first on the -- I'm wondering if you can give some color on weighted average coupon rate on the agency portfolio this quarter? I was -- with the CPR sort of flat, I was surprised that the yield compressed to the degree it did this quarter. So any color there would be helpful.

Hold on. We are actually looking up the information as we speak. Do you have another question and we'll -- ?

Sure. I guess just to build on that, if you could give us a sense of sort of where the more attractive opportunities are on the agency portfolio? And what sort of yields or all-in costs you are looking at in the current environment?

The answer on the agency coupon -- the coupon for the fourth quarter was [356]. And the coupon for the first quarter was [343]. So I think that explains the change.

Okay. Great. Thanks. And then on the -- on sort of the attractiveness of opportunities in the current market?

On the agency side or the non-agency side?

Both, actually, would be great.

Why don't we start on the agency side? Goodmunder, will you?

So on the agency side, in the first quarter, we were predominantly investing in 15 years loan balance securities. Our view has been and continues to be that we need to be cognizant of prepayment risk here. And as we put money to work there, we have favored -- you can call it better (inaudible) securities, which will have lower prepayments -- lower prepayment volatility. The majority of the dollars that were put to work in the first quarter was in those prepayment protected securities on the 15-year side.

And on the non-agency side?

On the non-agency side, I think the investments that we have made are pretty similar to the investments that we have been making for the last year or more. So no real change in that strategy.

Okay. Great. Thank you.

You're welcome.

Our next question is from [Dan Alcher] with FBR. Please go ahead.

Hey, thanks. Good morning, everyone.

Good morning.

A question on the recent preferred issuance. I'm going to take out, call it the 96 or so, out of the 200 that you raised -- but for that remaining portion, can you give us a sense as to where you are putting that to work? What you think the relative returns are versus the incremental -- or the coupon pickup on the 7.5s?

It's basically the same ratio that we have had in the past. Again, as you know, our portfolio consists of both agency and non-agency securities -- and it will be a similar balance, as you have seen before.

Okay. And then also just a question on the credit release. Was that the total -- the $34.5 million, was that the total that was moved out of the credit reserve? Or was there any that was actually earned in the quarter through the loss-adjusted yield?

So that was released from credit reserve to accretable discount.

Right.

Those are assumption changes.

Yes.

The other things that occurred during the quarter, there were realized credit losses. So those reduced the credit reserve. That was about $50 million. We increased the credit reserve from purchases by about $23 million. And then we actually had some small sales that decreased it by about $6 million.

And the other thing I would say is when you release that credit reserve, it impacts prospectively over time. So very little of that would have been actually realized during the first quarter, but the benefit of that will be realized going forward over remaining life of the assets.

That's a good point, Steve. When we do that, we do that typically at the end of the second month of the quarter. So it shows up for this quarter in March. But it would not have shown up in January or February.

Right. Okay. Good point. Yes. That explains the delta -- the $74 million between two cores of credit reserve. And a quick one, if I can. Can you give us a sense of where you think the book value is [murking] now that we are into May, on a fair-value basis -- now that we have seen the prices come back?

I will talk about what we have seen in non-agency pricing during the month of April and they are certainly up -- prices have been up. It's more difficult to buy securities now than it was a month ago. That being said, the volume has been somewhat light in April. So I guess I would say prices are probably up a good point or so. It's possible that they are up more than that, but I think it will take a little bit of testing, a little bit more volume to confirm these market levels.

Okay. That's great. Thanks so much.

Sure.

Our next question is from Henry Coffey with Sterne Agee.

Good morning, everyone. Thank you for taking my question. As you look forward, obviously there are a lot of wins helping you a lot, home pricing, the demand for securities; as you look forward for new reinvestment opportunities, can you give us some comment on what's coming out of the GSE world? They have been told to off-load illiquid assets. I know they have put our tapes recently to get people to start looking at the potential for credit risk assets. Make you can get a sense of what kinds of return yields could come out of that conglomeration?

Henry, I think they are still ways off on deciding how they are going to lay off some of the credit. So basically, the state of the art was the old treasury white paper and now we have a new Secretary of the Treasury, and there will be a new white paper. And they fall back to this option three where somehow they will lay off part of the credit risk to private entities with some sort of -- some sort of crisis insurance behind that. But really, the thinking there hasn't gone much beyond that. So it's hard to say. But we are completely prepared and we have the capabilities to analyze mortgage credit risk and to get paid for it and to accept the credit, if it makes sense.

I think one of the things -- just to add to Bill's remarks, Henry -- has been the fact that we have been immersed in credit for years now. We are not new to the game. So in terms of being able to analyze credit on residentials, I think we are at the top of the list. So I think anything that comes down the pike, there will be opportunities for us. And we do continue to look.

But it really is a waiting game, given the state of affairs in DC?

Yes, it is.

They have also been directed to, quote, off-load illiquid assets, which could include their non-agency holdings. Have you seen inventories coming out of Fannie or Freddie?

No, not to date, we haven't.

Great. Well, thank you very much, and good quarter.

Thank you.

And we'll go to Steve DeLaney with JMP Securities. Please go ahead.

Thanks. Good morning, everyone.

Hi, Steve.

Guys, in the past you have actually commented to -- you alluded to the fact, when you were talking about book value, that the non-agency portfolio gained value. I know in the past, you have actually kind of tried to quantify that in your press release. We did a little math this morning, just based on the disclosed information, your cost and your fair value, and using the 73.2, we backed into par. But it looks like we are getting an average fair value relative to par of about 85.4 at March 31, versus 82.8. So that suggests an average fair value gain of maybe 2.5 points in first quarter. Does that sound reasonable or realistic to what you saw?

I think it's in the ballpark.

Yes, I'm just -- I'm just checking. I think your numbers are actually spot on, Steve. As of December 31st , the waited average mark was 82.8.

Yes.

And as you will see in our Q that comes out today, it's 85.4. So you hit the number right on the head.

Those are exactly what we calculated. So about 2.5 points. And on $6.4 billion, 2.5 points, $160 million and you have 358 million shares. So it locks looks like $0.45 or so. So you would have had a pretty meaningful book value increase, were it not for the special dividend.

That is correct.

Okay. And I think we can take Craig's comments of a point or so and apply the same math going forward. One of the things I would like to just compliment you guys on -- we are so used to seeing yield compression and spread compression. And I just -- this method of being conservative on the reserve and having the ability to release, we don't see many earnings reports where the non-agency yield went up from 670 in 4Q to 680 in 1Q. You don't see that in mortgage REIT land these days. So congratulations on that.

And I guess the second part of that is if the $35 million release in 1Q really didn't benefit 1Q very much, I assume we should expect that, all things the same, that 680 has some room to move higher if we look out the next couple of quarters?

Yes, Steve. That's a good point. There's two things that happen. So we -- when we readjust those yields at the end of the second month of the quarter, we take into account any changes in our credit reserve and accretable discount. We also reset the yields based on the forward curve. And for the first time probably in more than two years, we actually got a small increase in the yields due to the forward curve. So I can tell you that the non-agency yield in the month of March was 692.

692 for the -- just the one single month?

Correct.

Excellent. Okay. Appreciate that. And one final thing and I will let somebody else hop on. The FHFA -- there's new reports this morning that Obama is going to nominate Mel Watt, and we'll see where that process goes. There's been a lot of chatter and I've talked to people who have been in the West Wing in the last two weeks and they are still focused on refis. But it seems like there is this idea of figuring out a way the GSEs could help with refis of private label (RMB) loans and private label securities -- is definitely on the table. I don't necessarily know how we can transfer that risk to the GSEs. But do you have any concern looking at your RMBS portfolio, your non-agency portfolio, that there could be some program put in place where those private label loans could be refied, but you as the bond holder would be asked to receive something less than par when the loan is refied? I guess what I'm referring to is something that would be like a short refi as akin to a short sale. I appreciate if you all are talking to anybody about those various proposals and have any views. Thanks.

The only concept we have seen of anyone forcing us to take less than the face amount has been these eminent domain discussions, which really haven't gotten far. So, no, Steve. Then to the extent that Fannie and Freddie wanted to make available to people with high LTVs, government-guaranteed mortgage, it would probably only be a positive to us -- sort of HARP for people under water on private labels. But we have heard no discussions about people forcing us to take any haircut.

All right. Great. Appreciate that.

One thing I want to add, Steve. You made a very good point when you calculated our mark-to-market, which is about 85% of par. So that's important when you look at how much book value does change, to know that these are assets that trade in the mid-80s, not assets that trade in the mid-60s.

Yes, got it. Because we do see big price differences, obviously, between the more sub-prime stuff and the more near-prime stuff. So we'll try to be attentive to that going forward on our book value estimates.

Thank you.

Thanks.

Our next question is from Douglas Harter with Credit Suisse. Please go ahead.

Thanks. I was hoping you could talk about how you see the opportunities between still investing in legacy non-agency versus new jumbo securitizations -- how you view the return profile of those two assets?

Sure. So when you talk about investing in the new securitizations of new clean originations, there are two sides to it. You could buy the loans and securitize it on the bottom piece or buy the top AAA pieces, which have lower yields. Doug, which part are you asking about?

I was thinking more of owning the credit piece.

So when we solve for the ROEs on the credit piece, depending on the execution of the last securitization deal that's happened, we see the ROEs comparable, but with a lot more implicit interest rate risk. If you are long to the fixed rate, and you are the bottom piece and you are locked out, it could be a very long asset, where the underlying mortgages might only yield you 3 or 3.25. So we still see comparable ROEs, but we still like the seasoned non-agency trade better.

Doug, the other thing I would add, is you know, in terms of upside from your base case yield, the upside on that bottom piece on the new securitization, there really isn't an upside, because most of those are pretty much priced with no credit losses. So you assume that you get back par. And as Bill pointed out, it's a very long asset. On the non-agency side, there's certainly a lot more room for credit improvement, home price appreciation, loans are amortizing, there could be various settlements from various banks that pull through these trusts. So we like the potential upside on the legacy non-agency as well.

That makes sense. And then I was just wondering on the dividend, have you -- could you tell us what portion of the current dividend was related to kind of prior excess taxable income?

So, Doug, this is Steve Yarad. The $0.22 dividend that we declared for the first quarter, we think that's added back to 2012. And you will see some disclosure of this in our 10-Q which will be filed later today. But we still think based on the dividends we have declared to date, we think that our estimate of taxable income, which might be finalized now until September, but we still think that taxable income will exceed for dividends declared to date.

So would that would imply that there could be a further special dividend coming?

You can't make that implication. It would say that we would have to be distribute yet our 2013 taxable income by September 2014.

Got it. Thank you.

The question is from Joel hulk with Wells Fargo. Please go ahead.

Thank you. Just a question regarding the comment about projected defaults, approximately twice the amount of the underlying 60 days delinquent. Where is that run historically? I guess the implication here is obviously as home prices continue to appreciate, there's more accretion into earnings. But I'm just trying to get a sense for the dynamic between that ratio over time and how that impacts the -- not necessarily the reserve releases but I guess accretion of that reserve back into the yield?

So it's hard to say. It's hard to answer that, because, remember, that our liquidations -- if we are liquidating approximately twice our 60 plus day delinquency bucket, that those are liquidations over the life of the security. So it's over the next 20 odd years or so. But what I will tell you is when we make adjustments to our credit reserve, we are really focused primarily on those loans that I would categorize as high-risk loans. So they are current loans that are high LTV loans. And as we continue to see those loans amortize and the home price appreciation, so LTVs decrease, the bucket of those loans get smaller.

And as each quarter and each year goes by, we have another year of pay history. And we added a table -- or we added a couple of lines to the table, what we call our FICO table, in the Q which show you the percentage of loans that are always current, so lifetime, never missed a payment, and then the percentage of loans that have not missed a payment over the last year. Those are things that we obviously we take into consideration. It's a somewhat slow process, I realize that. But we are looking for evidence that we can reduce the future loss assumptions.

Okay.

Joel, just so we're clear; it's not that we double the delinquent loans to come up with the assumption of what will default. We calculate them separately and then for illustration purposes we compare them. But they are calculated separately. And the calculation of what will default is based on loan to values, characteristics of the home borrower. It's just a coincidence that it's a two-to-one and its there for illustration, but the ratio will change over time.

No, that's a good point. Another question, maybe a better question is; of the projected defaults, you guys in your Q break-out that total of how much is current versus delinquent because then we can get a sense for where directionally you think -- as you said a lot of these loans are paying but they have high LTVs. And as they come down, that's what triggers a lot of the reserve accretion.

Well, we show you -- in that FICO table, we show you the percentage of loans that are 60 plus days delinquent. And that number has come down pretty smoothly over the last year and a half or. So I'm sure it was probably up around 21% a year ago. It was 19.3% at December 31st and it's currently 18.5%.

Okay. But if I look at your total projected defaults, can I -- am I able to tell how much of -- what bucket comes from the delinquent bucket versus the currently paying bucket?

I don't believe we publicize our total projected default. There's a table in the investor deck that we file periodically before equity conferences where we show our 20 top holdings. And we do show the projected defaults for the 20 largest holdings. And, again, it's probably safe to assume that it's roughly similar for the entire portfolio. So I think that will get updated the next time we update that slide. But for the most part, loans that are 60 plus days delinquent, we don't assume that any of those loans get saved.

(inaudible-sidebar conversation).

Yes, so if you -- and we will be filing the 10-Q later today, but if you look at the prior filings, we do bucket it. We show you the credit reserves by vintage and FICO score. We show you the always current. So I think you could interpolate into it and if it's not clear, we can change the disclosure going forward. But hopefully the 10-Q today will help with this question.

Yes, no, we will look at the Q later today and see if we can interpolate that. And just a clarification of a previous question, regarding the comparative opportunity between seasoned non-agency and the jumbo securitizations; I think what I heard you guys say is the risk adjusted returns you find more attractive in the seasoned non-agency?

Absolutely. That's correct.

Joel, think about it this way. So the collateral would be considered pristine, correct? So it's low LTV, high FICO.

Yes.

Therefore, when you buy the loan, you are not getting paid a lot to accept credit risk. You are getting paid to accept interest rate risk, in our mind. Because theoretically there's almost no credit risk there. That's how people look at the market.

Are we talking about the same thing? I'm talking about the discount -- the subordinated piece in the jumbo.

Okay, the subordinated piece. Yes, we still prefer the seasoned non-agency. Yes.

Okay. I think that's more apples to apples comparison, comparative to the credit piece. I agree that the AAA portion is all interest rate risk.

Well, no, I'm saying the bottom pieces are too. Because in extent all the collateral doesn't have much credit risk. There's very low loss assumptions. Otherwise, you wouldn't be buying these mortgage loans.

Right.

You are selling the AAA, you own the bottom. So you have magnified the loss assumptions but you have very low loss assumptions. I think you are getting paid for interest rate risk. That's our view.

And most of those 30-year fixed rate -- most of those deals.

My calibration is based on deals currently being done in the marketplace -- if you look at a large issuer in the marketplace, for example, you are getting, with some loss assumptions, you are getting an IRR of somewhere between 15%, 20% with a 15 CPR assumption. But, again, that's their production. What's available to you guys to buy, you still -- it's still more attractive in the seasoned non-agency portion, non-jumbos, is what you are saying?

Yes.

Yes, that's correct.

And our analysis wouldn't necessarily give you the same answer (inaudible). We don't necessarily sell for 15 to 20.

Okay. No, I understand. People have different loss assumptions and things like that. Okay. So thank you very much.

Thank you.

Our next question is from Mike Widner with KBW. Please go ahead.

Good morning, guys, I will follow-up on some of the questions that were asked from Joel and Steve and some others. Can you talk a little bit about the agency side and what you see there today as far as the most attractive opportunities, whether it's continuing to buy seasoned collateral or seasoned bonds or -- and the new issue market has gotten tight but the seasoned market is pretty tight too. So just what do you like today on that side?

Well, so you were asking about the agency side, right?

Specifically on the agency side and I will come back to the non-agency side. Again, I think you guys like your existing portfolio very much. But it's very difficult to replicate that. So for incremental capital or for putting capital back to work when something pays down, what do you like buying tomorrow?

Yes, this Goodmunder. On the agency side -- and I guess I alluded to this earlier -- in the first quarter, predominantly where we put money to work was on the 15-year side and in prepayment projected collateral, predominately lower loan balance, but also some higher LTV paper. And the philosophy there is the same as it has been over the last couple of years, is that we feel that on the agency side, there's a significant amount of prepayment risk. And it makes sense to try to minimize that by acquiring securities that we'll prepay at a lower rate than kind of generic securities. So that's what the philosophy has been. I don't think that will change a lot going forward. I mean, we had a little bit of a back-up in the first quarter. So that type of paper fell out of favor. But we rallied right back and I think it's probably prudent to stick to that strategy and try to minimize the prepayment side volatility.

You also mentioned some of the prepayment -- yes, we are happy with some of the seasoned hybrids that we own. We own them close to par and the yields on them are obviously higher than the current yields in the marketplace. And what that allows us, as well, to do is basically, we have a very short portfolio in terms of duration. So on the agency side, our interest rate risk is very limited. And we think the mix we have, in terms of some of the prepayment protected securities on the 15 year side and the short duration on the hybrids is a very powerful mix.

And so, specifically on the 15-year pre-pay protected -- that stuff is not cheap anymore, either. And what we see some of your peers doing are things like focusing on dollar roll markets, whether it's 15-year or 30-year, And then there are some others that are continuing to press forward in the hybrid ARM markets, 7/1s mostly, but it's gotten to be a tough market. So, I don't know, maybe just a quick comment. Do you really feel that that's the best place to be for your strategy right now on, like you said, the 15-year pre-pay protected? How do you feel about the other options?

Things change pretty quickly, and it really depends on price at any given time, and whatever the market is offering to you at any at any given time. The pre-payment protected securities, they are not by any means always the best vehicle of choice. Sometimes the hybrid will be the better choice. And sometimes even just generic security if the pay-up for the pre-payment protected security goes too high. So it really depends on the pricing, in terms of the way we allocate capital. But I do believe, though, kind of fundamentally, that the aggregate, if you think about the big picture that the government, the Fed, they are all kind of leaning lean towards increasing refinancing. So over time, it is going to be the best strategy to try to minimize your exposure to that and kind of minimize prepayment risk.

Mike, we are just replacing run-off. We are certainly not growing that portfolio at the moment.

No, I understand. So then let me ask you a related question going back to what you said about jumbos. If you are buying the residuals on the jumbos, you are still effectively buying an interest rate risk security. Have you contemplated how you view the attractiveness of that as an interest rate security versus a new issue, 15-year, whatever you want to look at on the agency side? So if we intellectually split your portfolio between interest rate risk and credit risk -- agencies are no longer the only thing that you can do on the interest rate risk side. Why not some of those jumbo subordinates instead of agencies to replace runoff on the agency side?

Good out-of-the-box question. But for us, they are not at all substitutes because that would not -- for the [40 act], we think it's cleanest to own the whole agencies. And also, the bottom piece is much, much longer, because you are locked out on -- you are locked out and it's 30-year collateral and very, very low coupon, if it's new origination with a high-quality borrower. We do look at that. But right now we prefer the agencies.

It would be tough to put 7.5 times leverage on those anyway. But it's an interesting thought. One final one I would ask is, as you look at pre-payments on your non-agencies so far -- across different parts of the country, we have seen pretty different home price appreciation trends. And if I look at your portfolio, it looks like, on the non-agency piece, we saw CPRs step down this quarter. I'm wondering if you have seen any market differences across different geographies? And if that has any implications for both the yields and the CPRs on your non-agencies, as we go forward, given that home prices keep moving higher.

So, you are right, they ticked down slightly. Again, the CPR on non-agencies is a combination of the voluntary pre-pay rates and then the default rates, times one, minus the loss severity. I think as far as refinancing on non-agencies, the thing you have to keep in mind is for people to refinance, they need to be able to get the new mortgage. So we might see that their LTV has come down from 110 to high 90s or so. But it's unlikely that those jumbo borrowers can refinance that 97% LTV unless they pay it down to an 80% LTV. I think we need to see a little more LTV improvement before we look to revise our pre-payment assumptions on the non-agencies. But the trend has certainly been helpful so far.

Great. Appreciate all the comments and color, guys. And good quarter and good performance.

Thank you.

Thank you.

Our next question is from Chris Donat with Sandler O'Neill. Please go ahead.

Hi, thank you for taking my question. It's a small question relative to a lot of other things. But just within the scope of the $34.5 million that moved from the credit reserve to the accretable discount; I thought you mentioned that $6 million of that was related to sales. I would just be curious what you sold and why in your non-agency portfolio?

That wasn't related to the $34.5 million. That was related to all the changes in the total federal reserve, which would go up and down by realized losses, new purchases versus --

Got it. Yes.

I don't recall. It wasn't a lot. There were maybe two securities that we sold early in the quarter. Again, it's just looking at market pricing and our assumptions versus market assumptions, implied by market prices. When we see an opportunity where it looks like that our credit assumptions, for instance, show us a 3% yield and the market is trading at 4%, we reexamine those credit assumptions. And if we stand by our assumptions and really do believe that at that market price, the security yields 3%, then we are inclined to sell that and buy one that we believe yields 4%.

But it's usually exception rather than the rule. Again, we have tried to build a portfolio and we feel very comfortable with it, but it doesn't mean we don't take advantage of the opportunities when we see them.

Got it. I'm trying to get at the level of discipline you are using with the price signals you get versus your strategy. So that helps me understand it.

I would say it's very disciplined.

Yes. It sure looks that way. Got it. Thank you.

And we'll go to Rick Shane with JPMorgan. Please go ahead.

Hey, guys. Thanks for taking my questions this morning. I just wanted to follow up on the comments about speeds on the non-agencies. One of the other elements of your portfolio is it's pretty concentrated in -- or not concentrated but it's skewed towards hybrid securities. Is there any sense that one of the reasons the speeds have not picked up there is those customers have been playing rates down? And as you see any back up in rates, some of those customers will come back in? Or some of those borrowers will come back to having try to bottom tick the rate cycle?

It's a good point, Rick, and it's certainly possible. Again, I think the LTV is really the biggest consideration. And while we may believe they are much less likely to default if they have some equity in the home, some equity doesn't necessarily translate into an easy refinance.

The other thing to keep in mind, relative to that is, you are right, those hybrid coupons have reset down. So they are probably 3ish on post reset hybrids. At some point, and most of them are now amortizing, so they had a five-year IO and now they are amortizing -- and at some point, they could actually refinance to a similar coupon 30-year fixed and lower the payment. Because the 30-year fixed will have a 30-year amortization and they might have, say, a 22-year amortization at the same rate. And that will lower the payment. But, again, it will really be driven by LTV.

Got it. So -- and it's clear at this point, you guys really think it's a structural -- and I didn't understand that coming in -- it's more of a structural issue, at this point, than it is a rate issue?

I think so, yes.

Okay. Great. Thank you for the clarification.

One other thing to point out, when you look at this CPR for non-agencies, a portion of that is defaults. So not all of these -- so the fact that the CPR goes up or down is not necessarily a good or bad thing. Voluntary pre-pays are very good. But a large portion of this 15%, just a little less than half, were defaults. So the CPR could be dropping and that could be a very good thing.

Got it. Yes. I started to pick up on that when Craig went through the math on the default impact on CPRs as well.

Right, and you will see this in Q. The three month voluntary pre-pay rate on the non-agency portfolio in December was 7.9%, and it's 8.3% in the first quarter. So the voluntary actually did tick up slightly.

Okay. Thank you, guys.

Sure.

Our next question is from Aaren Cyganovich with Evercore. Please go ahead.

Thanks. Could you tell us what the current LTV is of the non-agency portfolio?

I think we disclosed the original LTV. I can't really give you our current estimated LTV because it's a somewhat subjective number. We make assumptions real estate prices, we make assumptions about original appraisals that may or may not have been inflated. I'm looking at the FICO table. I think we show what the original LTV was.

We do. That was about 70.

Yes, I was just thinking about in terms of if you had a current LTV, it would make it easier for us to look at your projections of defaults and loss severity relative to changes in HPI. So if you don't have it, that's fine, but --

So I think we have said in the past that we believe with amortization that's occurring in home price appreciation that the weighted average portfolio LTV is probably a little bit below 100 right now. The thing -- and I'm not sure that really helps you with those loss assumptions, because really what you need to know for loss assumptions is what percentage of the loans are still above 100, or for that matter, above 120, right? Because those are your more at-risk securities. Those are loans that are current today but still have very high LTVs.

Got you.

You can almost use the headline numbers and get close. Because basically the LTVs at origination in 2006 and 2007 were 70% LTV. So you can assume home prices went down 35% and now they are up 9%. And you are basically going to solve for the right LTV number for the whole portfolio. But as Craig points out, it's the outliers that drive the defaults.

Right. Okay. Thank you.

You're welcome.

(Operator Instructions). And we do a follow-up with Douglas Harter with Credit Suisse. Please go ahead.

Craig, I just wanted to follow up on your comments about the yields for non-agency, going forward. So if it was 692, I think was the number you gave, should we assume that that falls, just given that the interest rates have fallen in the second quarter to date?

No, not necessarily. So 692 was our one-month yield for the month of March.

Okay.

Okay? And that number changed from the prior numbers because the credit reserve release occurs at the end of February, beginning of March, and we also reset all of those yields based on the forward. And we did actually gain a few basis points from the forward, which is the first time in several years. So I think the way to look at it is it's 692 in March, you start off April around 692 -- what changes? It's not so much what interest rates are today that change that, because we are using the forward curve. But we won't reset those yields based on a new forward curve until the end of May. And so the first month that you would see those new yields would be the month of June.

Got it. So if rates are where they are today, June could be reset lower because of the forward curve. But for May and -- April and May, it's based on the March?

That's right. And obviously, to the extent we have new purchases, that go lower the yield somewhat --

Right.

You don't know how loans -- which loans pay off. So if 9% yielding loans have big pre-payments, that could lower it. If lower yielding bonds pre-pay, it could raise it. So there are a lot of moving pieces.

Right. That makes sense. Thank you.

Sure.

And to the presenters, there are no additional questions in queue.

We thank you for continued interest in MFA Financial. We look forward to speaking with you next quarter.

Ladies and gentlemen, that does conclude your conference for today. Thank you for your participation. You may now disconnect.