MFA Financial Inc (MFA) 2014 Q3 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to the MFA Financial, Incorporated, third-quarter 2014 earnings call. At this time, phone lines are in a listen-only mode. We will have a question-and-answer session later on. (Operator Instructions)

And as a reminder, today's conference call is being recorded.

At this time, I'd like to turn the conference over to our first speaker,

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 operation.

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 those described in MFA's annual report on form 10K for the year ended December 31st, 2013, 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 third quarter 2014 financial results.

The discussion today also contains certain non-GAAP financial measures. Information relating to comparable GAAP financial measures may be found in the third quarter 2014 earnings release and earnings presentation slide, each of which has been filed with the SEC and posted on our website at mfafinancial.com. Thank you for your time.

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

Thanks very much, Danielle. I'd like to welcome everyone to MFA's third quarter 2014 financial results webcast. With me today are Craig Knutson, MFA's President and Chief Operating Officer, Gudmundur Kristjansson, Senior Vice President, Steve Yarad, CFO, and other members of senior management.

Tuning to page 3, the executive summary, you see that in this low interest rate environment, we've continued to identify and acquire attractive investments within our targeted residential mortgage credit universe.

In the third quarter, we generated net income of $75.1 million, or $0.20 per common share. The dividend per share was also $0.20, and that's consistent with the first and second quarters. Book value per common share declined approximately 1% to $8.28 per share as of September 30th.

Based on our continuing improvement in the loan-to-value ratio, loans underlying our non-agency MBS portfolio and other factors, we again transferred a significant amount, approximately $20 million, from credit reserve to accretable discount.

We remain positioned for more flexible monetary policy by the Federal Reserve that will be driven by measures of labor markets, inflation, and other economic data.

Our net duration is 0.62. Our leverage ratio as determined by GAAP is 2.7 to 1. And 69% of our MBS are adjustable rate, hybrid, or step-up.

Turning to page 4, you see that in the third quarter we continue to identify attractive investment opportunities across the residential mortgage asset universe. We are positioned to grow our holdings of securities backed by re-performing, non-performing loans, by 91% to $943 million, while increasing our holdings of credit-sensitive residential whole loans by 89% to $113 million.

We acquired $42.5 million, and opportunistically sold $61.6 million of non-agency MBS issued prior to 2008, realizing a gain of $13.9 million dollars. This is the ninth consecutive quarter we have realized gain through selected sales of non-agency MBS based on our projections of future cash flows relative to market pricing.

In this quarter, as in the prior quarter, we did not acquire any agency MBS.

Turning to page 5, you see again, despite the change in interest rates (inaudible), our key metrics remain generally consistent. In the quarter, we generated a yield on interest-earning assets of 4.12%. If you go back to the third quarter of 2013, our yield on assets is actually higher. At the same time, we generated net interest spread of 2.3%, which is, again, higher than it was in the same quarter in the prior year.

Now, looking at this page, which is GAAP, you see the debt-to-equity ratio is 2.7%. That's due primarily to the fact that the RPL/NPL we're acquiring, in many cases we're financing them with the counter party where we bought the assets. As a result, they're considered to be linked transactions and don't show up on the balance sheet. But if they were on the balance sheet, our debt-to-equity ratio would remain at three times. So we are basically replacing our assets as they run off.

Gudmundur Kristjansson will now present the next two slides, which are you an update on changes in book value and an update on MFA's interest rate sensitivity.

Thanks, Bill. On slide 6, we show changes to MFA's book value in the third quarter. Overall, our book value was fairly consistent in the quarter, but ended down 1% on September 30th.

Moving to slide 7, on slide 7, we show the duration of our assets and liabilities, as well as the maturation of our entire portfolio. The duration of our assets declined modestly to 1.8 at the end of the third quarter, from 1.9 at the end of the second quarter, as more of our hybrids got closer to reset, and our new acquisitions primarily consisted of RPL and NPL MBS and have limited sensitivity to interest rates.

The duration of our hedges also declined modestly to minus 3.7 at the end of the third quarter, or minus 3.9 at the end of the second quarter, as $46 million of swaps matured in the third quarter and our existing swaps got three months closer to maturity.

In aggregate, MFA's net duration declined modestly to 62 basis points at the end of the third quarter, versus 65 basis points at the end of the second quarter. MFA continues to maintain net portfolio duration below one. And we believe that by limiting interest rate risk in our portfolio, we are well positioned for uncertain monetary policy after the end of quantitative reasoning.

Now I will turn the call back over to Bill, who will discuss our asset allocations.

Thank you, Gudmundur. Turning to page 8, which is the page where we set the record for most footnotes, we present our assets, yields, and spreads broken out into what is now five categories. These are agency MBS, seasoned non-agency MBS, MBS backed by re-performing loans and non-performing loans, credit-sensitive residential whole loans, and cash and other.

In the quarter, we continued to find a wide range of residential mortgage asset investment opportunities, allowing us to maintain a level of assets consistent with the prior quarters. While our agency and non-agency MBS holdings declined due to runoff and some non-agency MBS sales, our holdings of RPL/NPL MBS and credit-sensitive residential whole loans grew.

By pursuing our wider residential mortgage strategy, we continue to find opportunities to generate attractive net interest rate spreads without increasing interest rate exposure. It's important to remember that equity sensitivity to changes in interest rates is impacted by both duration and the leverage utilized. We continue to maintain a leverage of approximately three times.

Asset allocation and selection is what drives both our income and our interest rate sensitivity.

And I would again like to talk a little bit about our holdings. We hold $5.3 billion face amount, or $4.8 billion market value of non-agency MBS, with an average amortized cost of approximately 75% of par, and a credit reserve of approximately $930 million. These assets generated a loss-adjusted yield of approximately 7.7% in the third quarter.

We continue to maintain our historical preference for adjustable rate, hybrid, and step-up MBS, and 69% of all our MBS fall into these three categories. We are actively acquiring residential mortgage credit assets, both securities backed by RPLs or NPLs, or credit-sensitive residential whole loans that we believe deliver attractive yields relative to interest rate exposure.

As usual, none of our agency MBS are 30-year fixed rates.

I'd like to now and the presentation over to Craig Knutson, our President and Chief Operating Officer, to give you some more detail on these RPL/NPL MBS, and update you on MFA's credit fundamentals.

Thank you, Bill. As Bill mentioned previously, we have substantially grown our holdings of RPL/NPL MBS during the third quarter. We particularly like this asset class for several reasons. First, loan duration, minimal interest rate risk, and price volatility, these securities are callable after 12 months, and typically they have a 300 basis point coupon step-up after three years if they have not been called at the end of the third year. So we expect that these funds will be called before the step-up date, and, hence, the loan duration and minimal interest rate risk.

Second, low credit risk, these structures have significant credit enhancement, generally approximately 50%. The subordinate bonds receive no cash flow. And therefore, the transaction de-levers very quickly.

Third, with leverage, we're able to produce approximately 10% ROEs with these securities. And then finally, there's a larger size available. So 25 to 100 or even more than 100 million trade sizes are relatively easy, versus legacy non-agency, where, at times, even a $10 million trade size is difficult to access.

Moving to page 10, the credit metrics on the loans underlying our legacy non-agency portfolio continue to improve. LTVs continue to decline due to home price appreciation and principal amortization.

Delinquencies have declined as fewer current loans become delinquent and foreclosure pipelines are slowly liquidated. Loans are, on average 102 months seasoned. That's eight and a half years. And in addition, since over half of these loans were refinancings, so they were not purchase money mortgages, we know that the homeowners have actually been living in these houses for more than that eight and a half years.

As Bill mentioned previously, we again lowered our estimates of future losses in the portfolio and we transferred almost $20 million from our credit reserve to accretable discount. All else equal, this increases the yield we will recognize over the remaining life of the bonds.

Moving to page 11, these graphs illustrate the LTV improvements since 2012. On the left, we display the average portfolio LTV. Note that the green line shows that the total portfolio LTV has declined from over 105% in early 2012, to nearly 75% today. And the gray line indicates that even the LTV of delinquent loans has dropped from over 90% one year ago, to below 85% today.

On the right, we depict the percent of loans with LTVs over 100%. While the lower average LTVs shown on the left are obviously a good credit metric, we still worry about the underlying loans with LTVs above the average, and, in particular, the so-called underwater loans, or those with LTVs over 100%, where the loan amount is greater than the value of the property.

Note the orange line on the graph on the right. This portrays the percent of current loans that are underwater. While not delinquent at present, in some cases, these are loans with past delinquencies that may have been modified. These are the loans that we worry about becoming delinquent in the future. At the beginning of 2013, over one third of the current loans were underwater. Today, they've declined to approximately 10%.

Even the delinquent loans, which is the gray line, have seen significant LTV improvement. At the beginning of 2013, half of the delinquent loans were underwater. Today, it's less than 25%.

Starting on page 12, on page 12 we illustrate the LTV distribution of the current loans in the portfolio. Again, we focus on the at-risk loans where the homeowner owes more on the mortgage and the property is worth. As of September 30, less than $200 million face amount of current loans had LTVs over 110%. This is only about 4% of the current loans in the portfolio.

Finally, page 13, on page 13, we illustrate the historical losses experienced on our legacy non-agency portfolio. These losses are and were fully expected. Our $929 million credit reserve is established in anticipation of $929 million in expected future losses.

Losses occur through realized losses, from liquidations of defaulted loans, where the net amounts realized through these liquidations is less than the principal amount owed on the defaulted mortgage. In addition, losses occur and will occur in the future when principal is used to pay interest on fixed-rate bonds that we own, where the underlying mortgages have received rate reduction modifications.

Realized losses in both 2012 and 2013, were approximately $163 million in both years, which is purely a coincidence. We show the first nine month losses for each year, respectively, in blue. In 2014, losses for the first nine months are running much lower than in the prior two years, and our annualized loss run rate is less than $100 million for 2014.

Finally, please note this $929 million credit reserve has already been reduced for the $326 million of realized losses in 2012 and 2013, as well as the $70 million of realized losses incurred thus far in 2014. This $929 million credit reserve is for future expected losses.

And with that, we'd like to turn the call over to questions.

Thank you. (Operator Instructions) Dan Altscher with FBR.

I wanted to take a look at the CPRs in particularly the agency ARM book. We had a bit of a tick up, looks like 18.3 versus 15.9. Now, I understand there can be some seasonality as we roll through the third quarter there, so I get that.

But do you think there was anything else going on there? I guess maybe across the board, thoughts about [fed] funds or short-term rates increasing, folks looking to maybe refi ahead of what they think might be higher rates.

No. Hi. This is Gudmundur. I mean, in general, prepays pick up a little bit during the summer months, and this is just regular seasonality, particularly to the short reset hybrids. The four-month LIBOR is still very low and one year LIBOR is still very low. So the reset coupon that they're resetting to is below where they can refinance a new five or a new seven one. So we're not really seeing kind of a refi-driven pickup for these people to extend their mortgages.

Okay.

And I actually expect as we move into the winter months, that we'll see the reverse of the seasonality. And in fact, the aggregate agency CPR for October was 12.7, so down from the 15.1 average for the third quarter.

Got it. Thanks. That's helpful. As we look at the RPLs and NPLs and the credit sensitive, as that continues to become a bigger portion of the pie, can we talk a little bit about the competition for that business? I mean, the yield on the assets seem to be trending down slightly. But I'm wondering whether a reflection of that is competition verse maybe just a continued improvement in the underlying credit profile of those borrowers, and, hence, the yields compressed there, as opposed to maybe competition-driven.

So I would say on two fronts, obviously, on the home loan front, there is a lot of competition for those assets and they trade, at times they trade very tight and then at times they trade a little bit cheaper. Most of the acquisitions that we've done have been smaller pieces which are probably a little bit more off the radar screen. So we think we've been able to pick those up a little bit cheaper.

But there's clearly a lot of competition for loans, especially some of the well-publicized, very large auctions, like for instance, the HUD action.

On the RPL/NPL space, there certainly is competition for those assets. But I wouldn't read into lower yields. I think if yields are down there, it's probably because we had more NPLs, which typically trade at a little bit lower yield because they're shorter securities than RPLs. But I think both yields have been fairly consistent. If you go back three or four months ago, they tightened -- they actually tightened quite a bit, and we sat out a couple of those deals. The yields have now come back in, so 3.5ish or so is probably a good proxy there.

Dan, if I might add, yes, you're correct, there's more demand. But also, there's a lot of supply. And that's why -- another reason we're very attracted to these assets. Besides the fact we like the yield [of] relative interest rate risk, there is a good supply that's coming to market over the next couple of years and this is a business that could be important to us and grow over time.

Yes. Thanks. I really appreciate that, Craig and Bill. Maybe just one other thought. As the credit-sensitive portion continues to get bigger, bigger piece of the pie, how does the opportunity to maybe put a little bit more leverage on those or maybe look to re-securitize them or do something maybe a little bit more structured to get the yield profile, I guess the ROE profile - excuse me - up a bit?

On the loans, you mean?

Yes, on the loans. I'm sorry. Yes, on the loans.

Yes, at this point, it's $100 million. There certainly is leverage available there, both warehouse type of leverage and also a securitization. So we'll certainly look to that in the future. But for the time being, we're just trying to amass some critical mass.

Got it. Thanks so much.

Jason Weaver with Sterne, Agee.

My questions have actually been asked already. I just had two. But I'll say congratulations on a solid performance in what was a really wild quarter for the space. Thank you.

Thanks a lot, Jason.

Thanks, Jason.

Thank you.

Operator, next question? Thank you.

Steve DeLaney with JMP Securities.

We calculated a return on equity of about 10.9% on the RPL/NPL securities. Just curious, one, do you agree with that calculation? And, two, Craig, how would you just roughly compare that return opportunity on equity to what you might be able to obtain in any purchases of new legacy RMBS, senior RMBS?

So, I guess I would agree with your ROE calculation, between 10% and 11%.

Okay.

And I would say that generally for, if you look at legacy, legacy probably right now yields about 4.5% if you call financing, call it 150. So a spread of 300 basis points. If you put four times leverage on that, you'd end up with 12 plus the yield on the asset.

So I think the ROE would certainly be higher if you fully levered legacy non-agency. That being said, as I said previously, it's difficult to acquire size in legacy non-agency. So while the ROEs might be a little bit higher, it's much more difficult to put money to work there.

Isn't there also a longer duration there as well, that at these higher price levels that you have to be concerned, so it would seem the NPL/RPL, being such short duration, you would expect a lower ROE.

Well, you're right, although it depends on the type of mortgages in legacy. So if it's a post reset hybrid, it's a pretty --

Yes.

-- short duration --

Oh, sure.

-- instrument as well. But, yes, your point's a good one.

No, I hear you. Yes, exactly. And, as Bill said, you guys have well over 60% are in hybrids or ARMs.

So just following the same theme of this new expanded asset class. So Bloomberg put a list out Monday morning, and they're showing -- it's just their usual weekly what's in the pipeline as far as non-agency RMBS, both legacy and the new NPL/RPL class. And they're showing seven new NPL/ RPL deals in the market. And there's two or three names on there that I've never seen before, people like treatment, [Truman], Gold Creek.

I guess I mention that, are you and, Brian, are you all seeing a pickup in the number of people that are issuing? And has there been some recent pickup in activity? This just seems to be a pretty long list for this asset class.

Yes. This is Brian. There has been a slight pickup as more participants kind of had their loans on warehouse lines, and then they look to term out their funding. Those names that you're listing, our issuers that kind of have participated in the past and they are -- there are larger names behind that. So I don't know if you know Truman and Gold Creek are very large money managers that kind of sit behind those two.

Got it.

So they're not sort of like fly-by-night new entrants. These are well-established players in the non-performing loan space.

Appreciate that color. And just one final thing. You guys, as you've mentioned, you've been harvesting gains, understandably so, on the legacy RMB. And that's generating taxable gain on sale revenue. And you've been pretty steady here at $0.20 GAAP EPS.

Are you in a position where you can make any general comments as to how taxable EPS is tracking your GAAP EPS, say for 2014? Thanks.

Sure. So one, as of the end of the third quarter, I believe we have $0.05 per share undistributed. Is that correct, Terry? And as you point out, Steve, our taxable income is running somewhat less than our GAAP income, but we're still undistributed for the year.

Okay. Great. That's important information. Thank you, Bill.

Congrats, guys.

Thank you.

Thank you. Joel Houck with Wells Fargo.

A lot of good questions here. So maybe switching gears more on a higher level. So far this earnings season, I wouldn't say it's unanimous, but most of your peers are claiming that there's more relative value in the agency MBS space relative to the non-agency. And understanding the non-agency market is obviously very dynamic.

I don't want two put words in your mouth. But you guys haven't purchased any agency MBS since the first quarter of 2014. Are you guys taking a different approach? Or maybe talk a little bit about the uniqueness in your model and perhaps why you see more relative value on the non-agency side if that's the case?

Thanks for the question, Joel. So we've been fairly consistent saying that we prefer non-agency's to agency. Even in a quarter when we bought agencies, we typically bought a lot more non-agency. So we definitely have a preference. And the preference is based on the yield relative to interest rate risk. So we think we can solve for a comparable ROE with less interest rate risk.

As to what decisions other mortgage REITs make, I can't tell you. But I can't tell you, our decision is pretty consistent over time, although usually it's not zero in a quarter. But it just so happens the last couple of quarters it has been.

Okay. And you have shorter duration, it's a higher quality borrower. So the credit risk would seem to favor your non-agency book relative to some of the other guys which are deep subprime. How do you view the potential volatility of just the pricing of the bonds or securities in light of what we're kind of seeing in the marketplace, which is, there's a slowdown in housing formation or at least new purchases, oil's hitting new lows seems like every day or every week. The 10-year is signaling what we could possibly go into a recession next year.

How do you feel about the pricing of your securities? And how much volatility and book are you willing to accept, if, in fact, we do go into a recession?

All right. Those are questions we think about all the time. It's kind of fundamental to why we got into the business when we get. That with permanent capital and with low leverage, we can accept some price volatility over time. We can accept book value going up and down, quarter to quarter.

The point is to make sure we're comfortable with the purchase price and that we're comfortable that our projections of cash flow over time are correct. Because if those things are correct and you do have the staying power, based on the fact you have permanent capital and low leverage, we are fairly indifferent to volatility. And in fact, and I know it's trite to say, we actually do welcome volatility. If we could acquire the same assets at a lower price and we're still comfortable with projected cash flow and the quantity asset, we prefer to pay at a lower price.

Now, you're saying, gee, or should we not be as optimistic because of what the 10-year is doing in the United States and everywhere. And we continue to believe that our quality forecasts are conservative. And I think Craig's data and analysis shows you that, that we're still assuming that a lot of people that are current, even though the fact that these loans are now eight, nine years old and the fact that unemployment is not going up, we continue to think there will be delinquencies in the future.

So we still think our assumptions are appropriately conservative, despite the fact that that 10-year here and everywhere might indicate a slowing down in the economies.

All right. Thanks, Bill. Thanks for your answer.

Thank you, Joel.

(Operator Instructions) Mike Widner with KBW.

I'm going to ask a couple follow-ups just because I think you answered most of the things I was thinking of. You mentioned a 12.7 CPR for October. And I just wanted to clarify, is that for the total agency MBS portfolio, so relative to the 15.1, or was that just the ARM portfolio?

Yes. This is Gudmundur. It's for the entire agency book. So 12.7 versus an average of 15.1 for the entire agency book in the third quarter.

Okay. So little 2 point, 2.5 CPR drop. So that's good. Hopefully that continues. Also, just following up on the undistributed taxable income for the year. When you say you have $0.05 undistributed taxable, I mean, that's basically the total overall as opposed to the question, I think the question that Steve, I believe asked was sort of how is taxable running relative to the GAAP. So I just wanted to make sure that you're basically saying, including carryover from 2013, and so on and so on, the total outstanding undistributed taxable is $0.05, which is sort of a different, I guess if the question was year to date, how is GAAP running relative to taxable.

Mike, you're right, that is a cumulative over time number.

So I guess the real question I guess is really how is taxable tracking right now. Because as we think about the 2015 dividends, I mean, it seems that there's been sort of a steady, I don't know, steady state that taxable has been running below GAAP and below the current dividend level. And just wondering if there's implications of that and if you can comment on that.

Well, it's hard to forecast what taxable income will be in the future because our portfolio does turn over. So while you're right, the tax is lower than GAAP for a lot of the seasoned non-agencies that we've owned for a while, that portfolio is declining somewhat and we're adding new assets where we don't have a differential.

So I know, Mike, you're trying to forecast. So that would be difficult. So I'll continued -- I'll repeat that our GAAP income is somewhat higher than our tax. On a cumulative basis, we still are in an undistributed position. But in terms of forecasting 2015, it's somewhat difficult because our assets are turning over. So if taxable income tends to be front loaded with non-agency assets, to the extent we're buying new assets, like credit-sensitive loans, for example, it's hard to project what's going to be happening in 2015.

So maybe I'm not giving you the detail that'll help you, but I'm telling you the factors that will impact 2015.

Yes. No. I mean, I don't expect you guys to give the answer that taxable's going to continue to run lower, and, therefore, there's an implication that the dividend's going to get trimmed.

But, I mean, it just, it seems that, sooner or later, that's the trend that we're running on, and that's -- I guess that's a question that you're not going to answer. But --

I would just like to add one other factor, Mike. As you know, taxable income is a floor, it's not a ceiling. You must distribute 90% of your taxable income. But if your economic and GAAP earnings are higher than your taxable income because of strange quirks in the tax laws that front loaded our taxable income and caused us to pay a very large special in prior years, doesn't necessarily mean it's a ceiling either.

We will take into account GAAP, economic, and taxable earnings when declaring dividends.

Well, I mean, I guess that leads to a follow-up. I mean, you guys don't report what everyone else sort of calls core. And by our measures, core has also been running below GAAP, stripping out the gains on sales. And so as you do think about how to set that dividend level, I mean, most of us have sort of become accustomed to not really thinking about GAAP at all in the mortgage REIT space, just because there's so many distortions in it. Most mortgage REITs tend to talk along the lines of core, and for a lot of them, that tends to track closer to taxable.

And by our estimates, your core has been $0.18 the first two quarters of the year, a little over -- I might need the Q to be precise about it, but closer to $0.16 this quarter.

And so again, how does that -- and if you think about economic returns as well, I mean that's not one that a lot of the guys talk about in the sector. But if you factor in mark-to- market changes and all that sort of stuff, so changes in book, that would be, I think this quarter a lower number as well, so I don't know.

You opened up that question. So how do those things play relative -- again, you mentioned GAAP, and GAAP's just not one that many people in the mortgage REIT space think much about just because of all the distortions with regard to setting dividend policy.

Yes. Well, again, I can't answer for everyone. The reason that we're comfortable with GAAP over core is that we still have effective hedge accounting. And that, I think is the big differential in most cases, and that's why we don't have a special defined term as you're talking about. So I think that's the difference.

Yes, that's fair enough. That's a good point. It's a pain in the ass, but you guys still do it. All right. Well, I think that's all my questions. I appreciate all the comments and color as always, and another solid quarter and nice job with the portfolio.

Thank you very much.

Thank you.

Doug Harter with Credit Suisse.

This is Matt Freeman asking on behalf of Doug Harter. I was wondering if you can give us a sense of what the underlying assumptions are in the credit reserve for legacy non-agency. I'm trying to get a sense for future transfers to accretable yield.

I'm sorry. Can you repeat the question?

I was wondering if you could give us a sense of what the underlying assumptions are in the credit reserve for legacy non-agency in order to get a sense for future transfers to accretable yield.

Well, I can generalize about the assumptions. But the assumptions are truly on a bond level. So there are probably 500-odd line items, and the assumptions are built on a bond-by-bond basis. But generally speaking, I think the legacy portfolio has about 15% of the loans that are 60-plus days delinquent. And I'm guessing that our assumptions assume that approximately 26% or so, of the loans will ultimately default. So if you will, all the delinquent loans and then about another 10% of loans that are current today.

But the assumptions are, like I say, they're on a bond by bond basis. And it's not just defaults, but it also depends on what prepayment speeds are assumed over time. So there's really a lot that goes into it. And I think it would be difficult to take those broad assumptions and then infer some future credit change.

Okay. Thank you.

Operator?

There are no further questions in queue at this time.

I want to thank everyone for joining us today. And we look forward to speaking to you all again next quarter. So thank you very much.

Thank you. Today's conference call was recorded and will be available for replay beginning today at noon and running through February 4th at midnight. You may access the AT&T playback service by dialing 1-800-475-6701, and using the access code of 341281. International callers may use 320-365-3844. Those numbers again our 800-475-6701 or 320-365-3844, with a common access code of 341281.

With that, that does conclude our conference call for today. We thank you for your participation and for using AT&T executive teleconference. You may now disconnect.