Ready Capital Corporation 7.00%

Joseph Lloyd McAdams - Chairman, Chief Executive Officer and President Joseph E. McAdams - Chief Investment Officer, Executive Vice President and Director

Daniel K. Altscher - FBR Capital Markets & Co., Research Division Steven C.

Delaney - JMP Securities LLC, Research Division Douglas Harter - Crédit Suisse AG, Research Division Jason Arnold - RBC Capital Markets, LLC, Research Division Michael R. Widner - Keefe, Bruyette, & Woods, Inc., Research Division Howard Henick

Thank you very much. Good morning, or good afternoon, ladies and gentlemen.

I am Lloyd McAdams, and I welcome you to this conference call in which we will summarize the company's recent activities and answer your questions about the past and future during our question-and-answer session. First I will briefly describe recent results and events, and then I will comment on current plans.

During the third quarter of 2013, Anworth earned net income to stockholders of approximately $16.6 million, which is $0.12 per diluted share. This amount includes a net realized gain of approximately $2 million or $0.01 per share.

Stockholders equity available to our common stockholders at quarter end was approximately $83.5 million -- $834.5 million, my apologies, which equates to a book value of $5.89 per share based on approximately 141.6 million shares of common stock outstanding at September 30. This represents a decline of approximately 2% from our book value of $6.01 per share at June 30, 2013.

The unrealized loss component of our AOCI in our book value, or unrealized losses, is approximately $100.6 million, which consists of the unrealized loss on our mortgage-backed security portfolio of $32.3 million and unrealized loss on our swap portfolio of $68.3 million. This means that the balance of our book value excluding this $100.6 million component is $935 million or $6.60 a share, which could loosely be defined as our book value estimate if all of our government-backed mortgage-backed security's repay and our swaps mature.

I'd like to discuss for a moment our -- the income in detail from third quarter. During this quarter, our net interest rate spread was 0.82% or 82 basis points compared to 1% during the quarter ending June 30.

This decrease in the third quarter interest rate spread occurred primarily as a result of the increase in our interest rate swap positions that occurred during the quarter. At September 30, the average borrowing rate on that day after adjusting for interest rate swap transactions was 1.44% compared to 0.98% at the end of the prior quarter, June 30.

During the quarter, we entered into 21 new interest rate swap agreements with an aggregate notional amount of $1.75 billion, with a weighted average interest rate of 2.17%. Also at quarter end, the weighted average term to reset on our adjustable rate agency mortgage-backed securities was 42 months, and after adjusting for interest rates paid through our swap transactions, the average term to interest rate reset of these liabilities was 1,024 days or about 34 months.

During the quarter, the CPR of our portfolio of agency mortgage-backed securities that occurred was approximately 23%. Now the average amortized cost of our portfolio of agency mortgage-backed securities was 103.26%, which is a small increase from the prior quarter.

A bit more detail about our mortgage-backed security portfolio itself. The fair value of our portfolio of agency mortgage-backed securities at quarter end was approximately $8.77 billion, of which we issued no sign into 4 broad categories of agency mortgage-backed securities.

The first is those adjustable rate mortgages whose interest rates will reset or are resetting within 1 year, and that is 18% of our portfolio. The second category is hybrid adjustable rate mortgages whose interest rates reset between 1 and 3 years.

That is approximately 23% of our portfolio. The third category is hybrid adjustable-rate mortgages whose interest rates reset after 3 years, and that is 39% of our portfolio.

And finally, the fourth category is 15-year and 30-year fixed rate mortgage-backed securities whose interest rates of course never reset, and that is 20% of our portfolio, and I might add that the predominant component of that 20% portfolio is the 15-year fixed rate mortgage-backed security category. As to the financing which we used to acquire additional mortgage-backed securities, a bit more detail.

The repurchase agreement financing of our agency mortgage-backed securities was approximately $7.57 billion at September 30. This is about 8x our total equity, and we describe total equity as our common stockholder equity plus all of our preferred stock equity and our junior subordinated notes, which are related to trust-preferred equity.

At September 30, our floating to fixed rate interest rate swaps were $5.18 billion, which represent approximately 68% of our outstanding repurchase agreement balance as of September 30. As you know, as we have discussed in the past, since 18% of our portfolio consists of ARMs which will reset their interest rates within 1 year and, therefore, contribute very little to our portfolio's interest rate sensitivity, we measure the swap position this way as being 80% of our portfolio which excludes these assets which will reset in 1 year.

And from our risk management perspective, I view this later 87% percentage as a more formative information measure of the margin of safety which we maintain in our portfolio. We should note that when determining the nature of a prudent hedging strategy, we do take into account the specific characteristics of each of the 4 categories of assets which I have discussed earlier.

During the third quarter, the $5.18 billion of swaps decreased in value by approximately $26.2 million. That would be defined as the unrealized loss that occurred during the quarter.

This is approximately 3% of our beginning common equity at the beginning of the quarter. And as you may recall, the book value of our company declined by 2% during the quarter, which includes this component of our swaps.

Current plans and things that we're looking at and thinking about. I believe there are several important subjects.

During the third quarter, as we noted in our press release, we materially increased our long-term floating rate to fixed rate swap positions in both size and maturity. The primary reason behind this significant decision, which will reduce our income going forward in the near future, are as follows: The first is as is happened regularly during the past 15 years we have been managing this portfolio, when mortgage-backed securities decline in price, as they have in 2013, this price decline reduces the safety margin of our available unencumbered collateral to support our future financing activities and needs.

Our response to this reduced safety margin is to reduce our portfolio sensitivity to interest rates so that we can better manage additional interest rate stress should it occur. Second reason, for me in 2013, an important issue has been the speed with which interest rates have moved and, more particularly, they spiked during the second quarter.

This indicates to me that the market is less liquid than probably expected, more sensitive to specific events, and that means that I believe a similar spike is likely to occur again. So when we stress tested our portfolio, as we regularly do, using this information and conclusions, we did reach the conclusion that a significant increase in our hedging instruments was needed to maintain the appropriate degree of safety which we have maintained in the past.

As to when this level of hedging may no longer be needed to maintain the desired level of safety, I can say that I now visualize more scenarios in the future requiring it's being maintained than scenarios where I see that it's not required. So while it is our current plan to remain in a capital-preservation mode, maintain a low asset liability mismatch and accept the accompanying reduction in net interest income, which these will create, there are additional income expected during this period from other sources.

The most nearly certain, I would say, is that our $2.18 billion short-term swap position, which we usually describe here as being of less than 3 years, currently has an unrealized loss of approximately $56 million or $0.40 per share. These swaps all mature over the next 3 years.

And as this happens, this $0.40 per share will predictably accrue through our book value during this period. We do have a choice here.

If we were to choose to liquidate these short-term swaps at these prices and recognize a realized loss of the same $0.40 per share, our book value would not recover as I just described, but our interest income would increase by the same amount since our short-term swaps would have been replaced by the current market rate for short-term swaps, which is currently much lower than the rate we have on the existing swaps. So as you can see in one scenario where we liquidate our swaps, our interest income predictably increases while our book value does not recover and is unchanged.

And the other scenario where we hold our swaps to their maturity over the next 3 years, our book value does increase predictably while our income is unchanged. In both of these scenarios, the total economic return and benefit to shareholders is clearly the same.

So right now, at this moment, we do not intend to realize the unrealized swap positions and -- which will result in, as I mentioned at the beginning of my comments, we expect that the interest income level will decline. I should point out that the interest income level will decline, notwithstanding whether we realize these or not, but clearly would not decline by as much.

We look forward to taking your questions now. There's a lot to talk about weighted to asset and liability mismatches and how we manage those.

With me here today are Joe McAdams, our Chief Investment Officer and Director of the Company; Thad Brown, our Chief Financial Officer; and Chuck Siegel, our Senior Vice President, Finance. Now we'll turn the call over to Chad, our conference operator, to begin the question-and-answer period.

Lloyd, a question on leverage. Obviously you've come down now from about 8.7x to 8x on debt leverage.

Do you think that's kind of the right amount on a go-forward basis, or maybe how do you think about that for going forward?

Dan, this is Joe McAdams. I think -- going back to our last conference call in July, we have been operating towards a target leverage ratio of approximately 8x, which we felt was an appropriate level given our perception of market risk going forward, as well as haircuts on repo and the right amount of excess unencumbered collateral we wanted to have available to meet potential margin calls.

After the decline in -- substantial decline in book value during the second quarter, our leverage did increase to about 8.7x as you discussed. And as we indicated then, we were looking to take steps both to bring our leverage back down towards our target leverage ratio of 8x, as well as looking to narrow our asset liability mismatch.

So between some sales of securities as well as the utilization of the majority of our principal pay-downs on our portfolio during the quarter to reduce our repo balance, we were able to bring the leverage ratio back down to 8x as our target. And while Lloyd mentioned that, certainly, we foresee a market environment in the near term where we could have higher than average volatility, I believe we're still comfortable with a leverage ratio of around 8x.

I don't think we'd be looking to move it significantly higher or lower in the near term.

Got it. That makes sense.

And also a follow-up on, I guess, all else being equally, a lower net interest spread on a go-forward basis since the long-duration swaps portfolio was certainly increased in the quarter. Just thinking about that go-forward, would it be wrong to think about the 1.44%, I guess, average cost adjusted for swaps.

Is that kind of the right number to be thinking about on a go-forward basis in terms of the average cost of funds because there's probably some noise there just on timing, but is that kind of like the right way to be thinking about this from a spread basis go-forward?

You're correct, 1.44% is the average cost of funds at September 30. We don't foresee any significant events during the current quarter that would cause that to move substantially higher or lower.

We do not have any swaps maturing during the fourth quarter, and again we're not anticipating a significant increase in the amount of swaps from where they were at September 30. As Lloyd had pointed out, one of the significant factors that we have dealt with, really, for the last several quarters, and really since the middle of last year, has been a number of our short-term swaps that were originated -- that were initiated in 2010, 2011, have significantly higher rates than where the market would be for 1-, 2- and 3-year swaps today.

And at the same time, the assets that were purchased relative to those swaps back in that period of time, experienced pretty substantial prepayments and still have a very significantly lower duration than at the time of their purchase. So when he talked -- I think the issue there is while our average cost of funds is 1.44%, that is higher than where what our market rate would be for our swaps and, in particular, I think if you look at our entire swap position, the average cost of funds on a market basis would be about 95 basis points.

So there is a pretty significant difference between the cost that we're paying in terms of interest expense versus these economic costs, which as Lloyd pointed out, would be the interest expense we pay, particularly on our swaps, but then also a corresponding reduction of our negative AOCI position relative to those swaps. So from a GAAP standpoint, the way things pan out, 1.44% is our cost of funds.

But sort of as a, I guess as an illustration, every quarter the mechanics of what takes place with our swap position, given that it is at an unrealized loss position, is we have a net interest payment that we make relative to the swaps. Again it's at a high rate.

It's higher than the current market rate for similar maturity swaps. And then there's a corresponding reclassification of a portion of our neg AOCI or I guess other comprehensive loss relative to our swaps back in to the book value.

And relative -- just to the short-term swaps that Lloyd discussed, the 3-year maturity, and that -- and for the third quarter that reclassification was approximately $9 million. So it is a fairly significant economic difference between the sort of the GAAP cost of funds and this sort of net economic effect of both the cash payments on our swaps and the reversal of the negative AOCI.

Lloyd, I was prepared to ask you about sort of the thought process behind the derisking of the portfolio, but you, in your opening comments, you covered that perfectly and especially with respect to the way you viewed the continued need for that protection. So I guess I'll ask you this, I have one housekeeping thing, but big picture.

Do you think the Fed has missed its best and least disruptive opportunity to begin taper, and is it your view that when that eventually comes in the spring or whenever, that we could be exposed to even more volatility than what we saw with the 3% 10-year on September 5?

That's a short question with a very complicated answer.

I know, I know it is. It's an unfair question.

I'm just -- I'm asking for your -- kind of your gut -- yes.

All right. I'm happy to give you our thought.

In its most straightforward way, putting $85 billion a month into the U.S. economy is quite important to maintain the standard of living that the American public has gotten used to over the last decade.

What we would certainly hope is the economy would pick up, and it would create the same $85 million being pumped into the economy -- $85 billion, yes, I lose track of zeros every now and then. And until that happens, if that $85 billion is not put back into the economy by printing money, of course, it could cause disruptions amongst people's standard of living and that goes wherever it goes.

So withdrawing the $85 billion is not some sort of political event that it's good for the country. I think withdrawing the $85 billion would be bad for the country in a sense that a lot of people's standard of living would decline.

We have a new Federal Reserve Chairman, and if that decision is going to be made, it's going to -- I'm almost certain it's going to be made by the new Federal Reserve Chairman. There is really no precedent for an old Federal Reserve Chairman going out of office doing anything dramatic.

So we're close enough to the finish line for Mr. Bernanke, but I'm sure he will work on like having a smooth transition.

That being the case, eventually we see that at some point in time there will be a fear, just a literal fear about do you make money owning debt? Would you rather have some sort of tangible asset since so much of the economy is, right now, based on people who save money and who hold dollar deposits?

And if that moves -- goes away from being a holder of savings, whether it's treasury bonds or FDIC-assured accounts, and it starts moving more toward tangible assets like real estate and stocks and whatever else there is, I think interest rates could make a move upward. And as I mentioned earlier, I know that a lot of people have said that the speed with which interest rates moved up in second quarter had something to do with hedge fund managers borrowing a lot of money to buy assets and doing very short-term trading, and as soon as they got a whiff that Bernanke might do something differently, they all ran for the door.

The question is, I don't think anyone knows how much of that type of relying on low-price money is keeping the economy where it's going, but as long as we keep scaring people into believing that savings money is not a good thing, and you should be an investor in more real assets, that could trigger the increase in interest rates. And as soon as you stress test that -- and I should point out, when I say stress test, some people think of stress testing as like how will this affect the dividend?

Some people think of stress test is how will it affect book value. The mother of all items that needs to be monitored is something I made reference to in my remarks earlier and Joe has made reference to, and that is your unencumbered assets.

It's quite...

Liquidity, right?

Well it's your ability to meet a margin call, and that number is always spread in on most mortgage rates. It's the second line of the financial statement on the balance sheet.

And when that number goes to 0, I guess, they play the funeral march because it has to stay above 0, and what you're stress testing is the ability to maintain that adequate amount of liquidity to meet a margin call. When interest rates move up by 1%, it's not much of a problem for anybody, but as you know, in our financial statements, we actually do publish what we think will happen if interest rates will move up 2%.

And for the last 15 years, we've always shown people what we thought would happen if rates went up or down 2%, and that is the stress test that we're looking at. And when we think that the asset liability mismatch that we currently have may not deal with that the way we want to, to maintain the margin of safety, we believe we have to basically reduce the asset liability mismatch.

And that's how we think about it, and I think rates will go up, but I'm not sure what day, but I'm quite convinced that what I've seen in the last 6 months tells me that the rates will go up sooner rather than later.

I appreciate the additional comments and explaining the way that you and Joe evaluate risk in the Anworth portfolio. So Joe, I have one housekeeping question for you, if you would.

Can you give us, given the work that you've done primarily with the hedges and some reduction in the portfolio, could you give us an estimate of your net duration in the portfolio at September 30 post the hedging as it would compare to June 30.

Sure. The duration of our portfolio at September 30 was approximately 2.9 years on the asset side.

And on our liability side, the taking into account the effect of swaps and our repo position was approximately 2.75 years, almost 2.8 years. So net -- and again, just to walk back through the timeframe on March 31 of this year, our duration of our portfolio was 1.5 years with approximately 1.3 years on the liability side.

So approximately quarter year mismatched at March 31 when rates went higher...

Are these comparably weighted for notional? Are these dollar-weighted figures?

Yes, yes.

Okay. So it's effectively a net duration?

Okay.

Right. We do have a different amount of assets versus liabilities because there is a certain portion of our assets are funded with equity.

But I guess what I'm comparing is the net duration of our assets versus the net duration of our liabilities.

Okay. Very good.

So essentially equal at September 30 is what you're saying?

Slightly positive. Right.

But really back to the position we were at, at March 31. If you recall at June 30 while we had begun taking on some additional hedges at the time of our last conference call, at the end of the second quarter, our duration on our portfolio was approximately 2.8 years with the net duration of our liabilities still right around 1.3, Not substantial change for the first quarter.

So our GAAP moved up fairly significantly from about -- from approximately 1/4 year to almost 1.5 years and, now it's back to approximately 1/4 year.

I was wondering if you could give us an update as to where you think hybrid ARM prices have been moving, quarter-to-date.

Sure. Quarter-to-date, we have seen some relative improvement in the price of hybrid ARMs.

When we look at back at the third quarter, it certainly wasn't a reversal of the underperformance we saw in the second quarter, but Agency MBS, on balance, did have a good quarter, outperforming their hedges. But some of the hybrid ARMs, particularly some of the longer hybrid ARMs, were lagging in that area, but we have seen some decent performance quarter-to-date.

We would estimate our book value even though as just discussed with Steve's question, we have a very small asset liability mismatch. And net-net, there's not a lot of change in interest rates in the quarter.

Our book value quarter-to-date is up about 3%, which would reflect the relatively good performance of both our hybrid ARM and our fixed rate position.

I was just wondering if you could talk about prepayment speeds, perspectively. Here, in the last couple of months, I guess, we've seen some slowing in ARM speeds.

So maybe just talk a little bit about that.

Sure, Jason, This is Joe. Our average CPR during the third quarter was 23% CPR.

The most recent prepayment report we have seen, which was released in October, and those prepayments would not have been included in that average for the previous quarter of 23%, was 15% for the portfolio as a whole. So that is a fairly substantial reduction.

This really is the time that we would've expected to see the impact of the higher mortgage rates that happened late in the second quarter and into the third quarter to begin to show up in the prepayment reports. And I think we expect to see sort of a continued slower rate as we go forward.

There are -- obviously, month-to-month, can be some noise. We have seen rates come down some off their high in September, but I don't think we're at the point where we would yet see the reflection of that effect yet.

But we do see -- expect a significant reduction in our portfolio CPR as we move into the fourth quarter.

Okay. And would you say then, kind of all else equal, 15% to 18% CPR might be kind of what you might think would be likely for the quarter?

I know it's reliant upon interest rates but just kind of your impression.

I think the majority of the prepayments we see in the fourth quarter will be driven by this -- by the higher rate environment we saw last quarter. There was a few sort of technical day count issues around the number of business days of 1 month versus the others, but I would -- I think you're correct that 15% or, possibly, a few CPR higher would be, right now, where we would expect to see the fourth quarter CPR come in.

So you talked a fair bit about managing duration both now but in a significantly higher rate environment, and with Steve, you kind of talk about your view of the duration gap. What do you think you're duration gap would change to in that up 200 kind of scenario and how does that compare to where you kind of felt like what you were back in March?

I guess, 2 points I would make relative to that are, one, from a model standpoint. We saw that when mortgage rates moved up, we -- almost 100 basis points on our portfolio from March to June.

But we saw an increase in that GAAP of approximately 1.5 years, which was a very significant increase. And additional moves up from here would -- our expectation would be that it would result in a similar but slightly less of an extension because we've had a number -- the majority of the extension in our portfolio came from securities that had been trading at a pretty substantial premium, and now have moved back down to par or below.

And as those securities were potentially to move to a more substantial discount, there is less negative convexity, less potential for extension in those bonds than in sort of your first move when they extend. So I think our expected level of sensitivity is less as we move forward from here because of that reason.

The other point I would make that was, I think, really underscored in the second quarter was that we have a portfolio of assets whose price is tied to mortgage rates, and we have a portfolio of hedges whose price is tied to swap rates. And there is a basis between those 2 rates.

They don't necessarily move hand in hand. And I do think that part of what resulted in such a significant change in the net value of our portfolio during the second quarter was, in a good part, due not just to rising rates but also to the underperformance of mortgage-backed securities relative to swaps.

And as I think about another move higher in rates, another 100 basis points or 200 basis points higher, thinking about how mortgages would perform versus other interest rates, I think is a very important question in trying to establish how much your gap would widen out. It's my belief that given -- that while we've had some retracement of the underperformance that mortgages had in our portfolio of predominantly hybrid ARMs we've had during the quarter, we certainly haven't gotten back to the position we were on March 31 in terms of the relative rates.

So I think we have less likelihood of more significant underperformance at mortgages versus swaps, especially if we start thinking about a more significant cyclical move higher in rates as opposed to a sort of tapering-related underperformance we had in mortgages going forward. But we do have -- that's the long answer.

The short answer is, I don't think we would see the level of extension we saw during the second quarter, but I still think it would be significant to definitely be in an order of magnitude of a year of extension if you were to have a significant move higher as you were hypothesizing or formulating.

And then I guess just one other question related to the prepay speeds, as I recall last quarter, you guys had made an amortization adjustment -- an adjustment to your lifetime CPR expectations. I noticed that you didn't make one this quarter.

So I guess, even though prepays are clearly coming in slower now, you guys stick a little closer to the level yield method than a lot of others. Does this actually change your lifetime CPR expectations -- so in short, should we assume that just because CPRs fall that it's actually going to change your effective yields?

You're correct. We've been in an environment the last quarter or 2 where our lifetime assumptions on CPRs that we used for the level yield method of amortizing our purchase premium have been below the actual CPR of our portfolio.

And now we expect the actual CPR of the portfolio to come down. As we mentioned, the portfolio CPR was 15% in October, which is below our long-term lifetime CPR in our portfolio.

So I do think -- we'd certainly like to get a few more data points during the fourth quarter, but it's my expectation that we were comfortable with leaving our assumptions on -- our long-term CPR assumptions relatively unchanged September 30 versus June 30, but I do think, on balance, that those assumptions are probably more likely to decrease in the near term than increase.

Buybacks. I see you guys finally did do some buybacks, and a little bit of size -- obviously, you issued some stock again, but the buybacks were greater.

I think that's a good thing. Obviously I'd encourage you to do it.

I'm curious what motivated you to do it this time, and what do you think of it going forward given that you guys still trade even with the reduced book value, I don't know, in the neighborhood of 80% or below price to tangible book?

It's Lloyd. Let's see, one of the things we did to, as much as we could, to discourage investors from exercising their dividend rights to repurchase shares of stock because we eliminated any discount that we offed it at, though we "do pay the brokerage commission."

So that hopefully reduced the number of shares that are sold on dividend reinvestment. As to buying shares back so that whatever shares are bought back will be a larger net number, as you pointed out, I think of the company as just like any other investor.

I try to figure out what is the price to buy low and can we sell it at a higher price later. It's almost no different than buying any asset that goes up in value and you sell it.

We don't obviously, when the stock is trading below book value, the more trades below book value, the more likely I think it'll probably go back up because I just got through mentioning in our comments that we have a swap position that's going to, somehow, add 40 basis points to the value of the company over the next 3 year -- I'm sorry, $0.40 a share over the next 3 years. So these are all good things, and so that helps reach the conclusion that we think the stock's going to go up.

If the book value were way above the paid in capital, and the stock was trading below that book value, but I thought the book value was going to go down back toward the paid in capital, I would think buying the stock back then would probably cost us money in the long run. So you say what caused us to purchase more shares?

I like the price, the stock, it was trading in. I thought the stock would go up.

It turns out a lot of other people think the same thing when we're doing that, so it's quite competitive to buy the stock when it gets down to the levels that we think is attractive that we will have an unrealized gain at the end of the quarter. But that's what does it.

So how do you feel now about the stock price versus attractiveness or not?

Where's the stock price now? We started the phone call 40 minutes ago, and we've been saying things, so where is the stock price now?

I'll tell you exactly, right. I believe it was $4.75.

It's $4.75. Down [indiscernible] the day.

I don't really know, but clearly, we are in a level -- we definitely showed you, and we will show you -- we show you exactly where prices we bought this year is at, and clearly you can reach the conclusion that we thought those were attractive prices the stock would go up from. Hopefully we'll be right all the time.

I guess, some time we'll buy the stock back, and at the end of the quarter it will be lower, and we'll be disappointed that we bought the stock back.

All right. A few quick other questions.

You said about $0.40 of the AOCI is the swaps 3 years in end, is the other 31% -- the other $0.31 of the $0.71, is that on the longer swap book or is that on the asset side or where is that $0.31?

Both together.

A little bit of both?

Both.

If you look at our -- this is Joe, if you look, we had about $8 million of positive AOCI relative to -- obviously it was a positive change in the quarter of $8 million of positive AOCI, but we do have, still, slightly negative mark-to-market on our assets in general. And then on the net balance of our swap position, even though some of our swaps are gains and some of our swaps are at losses, those 2 together make up the balance of the AOCI --

And given this slightly, why am I saying slightly, substantially more hedged balance sheet, assuming you don't see big moves in rates, do you think the dividend is not sustainable at either the $0.12 level going forward? Because obviously this is a ceteris paribus assumption, which is not the case, but how do you see it, in particular in light of lower prepays, how do you see the earnings and dividends going forward with a more hedged balance sheet?

This is Joe again. As we've discussed on some of the other questions before, our average cost of funds at September 30 was 1.4%, which was significantly higher than the average cost of funds during the third quarter.

So that reduction alone will put some significant pressure on our GAAP earnings for the quarter. We do think it's likely that some portion of that will be offset by lower prepayments going forward, but we are at a position where we would expect the net interest margin, at least for the next quarter or 2, to continue to decline.

You're not giving an estimate how much?

No. But, I mean, that would imply the dividend would probably need to go down.

I understand that. I'm just trying to figure out what the earning asset assumption is so I can figure out the spread.

The average cost of funds during the third quarter was 1.1%, and the average cost of funds at September 30 was 1.4%. So that's a 30 basis point reduction, all else being equal, I would not think it would be unreasonable at December 31 to see a decrease in our rate of premium amortization by, potentially, 10 to 15 basis points.

But I would not expect the premium amortization, which was 68 basis points for the quarter in the third quarter to decline enough to offset that 30 basis point increase in the cost.

Okay, fair enough. And in the -- and the last question is you also alluded to the fact that there's a lot of negative -- there's less negative convexity in the portfolio than there was before, but there's still negative convexity in the portfolio.

Do you -- have you considered doing anything to hedge the negative convexity either in addition to or in lieu of duration matches.

That is something we -- we talked about this in the past as well. It's something that we consider relative to our portfolio, just very briefly, the concept of negative convexity has to do with [ph] the idea that, as rates rise, your portfolio duration will continue to extend and in some cases, start to extend by an even greater and greater amount.

That -- you can measure the negative convexity at a point in time, or you can look at how the negative convexity will evolve either based on changes in interest rates or through the passage of time. It's been our belief that given that such a substantial portion of our portfolio consists of ARMs and hybrid ARMs, that while you do have some significant negative convexity in the near term, and especially on newer securities, to the moves in interest rates during the fixed portion of the ARMs' life, that we are moving towards a point where ARMs reset to be of either to prepay at par or to become a fully indexed ARM that will trade with very low price volatility.

So to a certain degree, when you think about a fixed-rate portfolio and the use of swaptions or out of the money options or mortgage options as a way to, at some point, rein in the negative convexity and limit your duration exposure, it's been our feeling from an investment standpoint that, that is one of the benefits of ARMs, that knowing that we have, whether it's 2 years or 5 years or even 7 years down the road, there is an event that is going to pull the price of that security back to par significantly limits the problem of sort of the spiraling negative convexity event that certainly could happen with 30-year mortgages. So our belief is -- has been to date, that our portfolio, which consists predominantly of ARMs and most of our fixed-rate position are in 15-year mortgages, has less exposure to that sort of event and the sort of -- at the significant out of the money event that causes substantial portfolio extension and losses.

And so that's the reason why we have not entered into any sort of alternative methods of managing that convexity going forward. The sales of securities we made during the quarter were securities that we felt would contribute substantially to increased duration and more negative convexity relative to our portfolios as a whole.

We sold some of our 30-year mortgages while they were at premiums. We're still at a fairly short duration that did have a potential to extend going forward as well as some of our 15-year mortgage positions.

So that's been our view on how our portfolio on the asset side mitigates some of the need to have more than just a swap on the other side of the balance sheet.

And my very last question, and I appreciate what time it is, where are current spreads? In other words, if you -- where's fewer adding marginal assets, as you get prepayments, you have to add -- what spreads are you able to garner now on the fully hedged or to the extent your hedging now basis?

Are they only 60 basis points or, obviously, are they 82 or are they 100?

They're 100. For the first several months of the quarter we had utilized the principal pay-downs in our portfolio to bring our leverage back to its target level.

But we did add assets and put on additional hedges relative to those assets late in the third quarter, and there is a net spread of approximately 100 basis points on those new purchases.

So new assets, as you put them on, seem to be a better spread than what you're anticipating your current overall portfolio of spreads to be at?

That's correct. And to get back to sort of Lloyd's point in his comments, this idea of where market prices are and where the yield on our existing portfolio is at market versus at cost, the yield -- the spread on our portfolio at market, and everyone marks their portfolio to the same market price, is 90 to 100 basis points, pretty similar to where our spread was a quarter or 2 ago.

So I do think that new investments are attractive versus the sort of economic spread, ROE on book value as opposed to GAAP earnings on costs. But yes, they are attractive versus -- the new investments are relatively attractive versus our existing portfolio as a whole.

Well we very much appreciate your attendance today, and we appreciate your very thoughtful questions, and we're glad we had a chance to answer them. I hope they benefited everybody who is on the call, including all the people who didn't have a question to ask.

If you need more information about the company, please visit our website, and thank you again for your participation. We look forward to visiting with you either in person or, again, during our conference call about this same time next quarter sometime in January.

All the best.