Dynex Capital Inc (DX) 2011 Q3 法說會逐字稿

Good morning and welcome to the Dynex Capital Inc. third quarter 2011 earnings call. All participants will be in listen-only mode. (Operator Instructions) After today's presentation, there will be an opportunity to ask questions. (Operator Instructions) Please note, this event is being recorded.

I would now like to turn the conference over to Alison Griffin, VP, Investor Relations. Please go ahead, ma'am.

Thank you, Maureen. Good morning. Thank you for joining Dynex Capital's third quarter earnings conference call. The press release associated with today's call was issued yesterday and filed with the SEC today, November 2, 2011. You may view the press release on the Company's website at WWW.DYNEXCAPITAL.COM under Investor Relations.

So before we begin, we would like to remind you that this conference call may contain forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. The words believe, expect, forecast, anticipate, estimate, project, plan, and similar expressions identify forward-looking statements that are inherently subject to risks and uncertainties, some of which cannot be predicted or quantified. The Company's actual results and timing of certain events could differ considerably from those projected in or contemplated by the forward-looking statements as a result of unforeseen external factors.

For additional information on these factors, we refer you to our annual report on Form 10-K for the period ended December 31, 2010, as filed with the SEC. The document may be found on our website at www.DynexCapital.com under Investor Relations.

This call is being broadcast live over the internet with a streaming slide presentation and can be found at out webcast link on the Investor Relations page of our website under IR Highlights. The slide presentation may also be referenced by clicking on the Q3 earnings conference call link on the presentation page of our Investor Relations website.

In this earnings presentation, we use certain non-GAAP financial measures which we refer to as [X] items to explain our results for the third quarter of 2011. Slide number 3 of the presentation slides for this earnings presentation gives additional information about these financial measures and the amount that has been excluded from the related GAAP measures.

So with that, I'd now like to turn the call over to Chairman and CEO, Thomas Akin.

Thanks, Alison. With me today, as usual, is Byron Boston, our Chief Investment Officer, and Steve Benedetti, our Chief Operating Officer and Chief Financial Officer.

The core earnings for the third quarter was $0.32 and our reporting earnings was $0.04 due to several excluded items that I'd like to discuss in more detail. First of all, we took a $0.20 charge related to an old business activity that took place from 1997 to 1999. This was prior to the current administration and is charged with a business decision related to litigation that could have dragged on for years and at a substantial cost to the Company. Our previous insurance policy limits had been exhausted in the second quarter and we took a charge related to the actual charge itself and the ongoing litigation expenses. We have now substantially increased our D&O insurance.

The second charge is an additional $0.05 non-cash charge related to an upgraded tranche of CMBS collateral that we originally issued at a discount. When we called the bonds at par, we incurred the related charge versus the amortized issuance cost. Assuming that financing of these bonds remains available, Dynex will save $2 million a year in interest and generate $0.05 in additional earnings to the Dynex bottom line annually.

Finally, we also adjusted our prepayment assumptions going forward to reflect record low interest rates and changing government policy. The $0.03 adjustment reflects, in our opinion, a conservative prepayment assessment. Our current portfolio continues to prepay at rates much lower than our expectations and we continue to focus on prepayment-protected securities. Given this increase in forecasted prepayments, we do not expect a similar charge in the fourth quarter.

Moving on to our book value, Dynex reported book value of $9.15 a share versus $9.59 in the previous quarter. This $0.44 change I'd like to go into a little more detail as well. We did earn $0.32 in core earnings. Offsetting that is our 27th dividend, $0.28 in excluded items that I just previously discussed and then finally, $0.21 in fair value adjustments to our portfolio. These changes add up to the $0.44 change in our book value. The historic move in treasury yields in the quarter, mortgage spreads widened and affected a 2% move in our portfolio fair value.

Our net interest spread also declined slightly a couple of basis points, but we continue to find attractive investment opportunities.

We declared a $0.27 dividend in the quarter and while our Board of Directors has the last say, our forward core earnings should easily support a continuation of this policy.

Our core investment strategy of high-quality, low-duration assets has not changed if we see substantial opportunity to add accretive investments at attractive high-mid-teens [turns].

Now, I'd like to turn the call over to Byron Boston for a deeper dive into why our portfolio has performed so well.

Good morning. 2011 continues to be a challenging time within the global financial markets. This is not a new thing; this is simply a continuation of a long, corrective process that started in 2007. These developments have led to heightened concerns regarding several risk factors that impact the mortgage REIT industry.

We at Dynex recognize and share your concerns regarding the future outlook and hence, this morning, will address our thoughts on each of these issues. We will fully understand your concerns regarding both prepayment risk and credit risk and we want you to understand we have carefully constructed our portfolio to minimize prepayment risk, credit risk and extension risk.

The essential theme of our entire strategy is diversification such that we do not have outsize exposures to any one specific type of borrower or sector. In addition, since 2008, we have carefully constructed our portfolio based on several large macro themes, while meticulously choosing each bond based on the individual characteristics of the underlying loans.

Despite having a high average dollar price portfolio, we have continued to deliver solid results through three successive refinance events in 2009 and 2010 and we fully expect our portfolio to continue to perform in the future.

So let's first go to slide 5 and let me just summarize our main key points that we want to leave you with today. Our portfolio continued to perform well in the third quarter, despite volatile markets. Our portfolio is constructed, as I mentioned a second ago, to minimize prepayment risk, credit risk and extension risk. Despite the high dollar price, we have solid built-in loan level characteristics that will work off that trend in faster prepayment speeds.

In the third quarter, we rotated further into securities in the CMBS sector, which all have explicit prepayment protection and higher expected ROEs. In addition, we have focused most of our CBS exposure in the GSE multi-family program, which historically demonstrated superior credit performance. We have also added CMBS IO exposure which are bonds constructed with explicit prepayment penalties.

And to minimize extension risk, which has not been discussed as much in the marketplace, but it is a large concern here at Dynex, we have focused on portfolio on securities that either reset or mature within 10 years. Our average monthly maturity or reset within our book of business happens to be 46 months.

So let's go to slide 6 and let's consider the composition of the portfolio. Our portfolio reflects a diversified strategy that includes both residential and commercial securities issued by government agencies and non-agency entities. Since 2008, we have steadily built this portfolio on specified pools. Each purchase has been based on selected criteria of loan level characteristics that we find attractive -- loan size, year of production, origination channel, originator, ownership status, credit, LTV and loan restructure.

As we mentioned in our last conference call, during the third quarter, we rotated our new investments into CMBS securities where our ROEs were more attractive and prepayment risk is limited. Now, note the actual changes. You can see that our agency hybrid book went from 59% of the portfolio as of June 30 to 51% as of September 30. Within the quarter, we actively sold approximately $100 million of what we consider to be our higher risk agency hybrid paper which is really what we consider our (inaudible) of the newer paper. And you can see where we put our capital, you can see our CMBS exposure move from 44% of the portfolio to 32% of our overall exposure.

In addition, please note that our CMBS IO position, which we increased in the third quarter -- these are credit IOs that have explicit prepayment protection. Although not vulnerable to spikes in voluntary prepayment fees, returns will depend on credit performance of the underlying bonds, but most importantly, you can see that we decreased our res exposure from 76% of the portfolio to 68% and we increased our CMBS exposure from 24% to 32%, and as we've moved on into the fourth quarter, this trend has continued. We continue to find attractive opportunities in the CMBS sector.

Let's go to slide 7. For the last three years, we have pointed out that we're building a high-quality portfolio. As reflected in the chart as of September 30, 90% of our portfolio is rated AAA. And please note that the material portion of the non-AAA assets represent assets that were originated by Dynex prior to the year 2000.

Let's move to slide 8. This is one of the largest things in the marketplace is that structural changes in the mortgage origination process has changed the ability and the desire of the average American to refinance their home loan. As you can see in this chart of the Mortgage Refinance Index, mortgage prepayments peaked at the large spike in 2003. Since 2003, we have seen successively smaller spikes in prepayments. These structural changes are limiting the demand for, and supply of, mortgage credit. Even with the current government proposal for HARP-2, we should not see a massive spike in prepayments as we saw in 2003.

Although we expect speeds to increase due to the low rate environment, we are not expecting speeds to approach the levels seen in 2003. We have built our portfolio, as I mentioned a second ago, carefully by adding specified securities to withstand expected increases in speed over the next few months. In fact, our current forecasted speed built into our model more than incorporates the prior spikes in speed seen over the last several years. Our dividends and earnings per share have remained relatively stable over the past two years despite these three successive refinance events.

So what are the facts? Let's move to slide 9. Over the past year, our age CPR for the portfolio has steadily declined and remained below our expectations. IN fact, despite having a premium portfolio, as I mentioned a second ago, we have withstood three major finance events, and even with the drop in rates at the latter part of last year, our portfolio really peaked. As you could see, our average -- one-year average is (inaudible) 22.7% for the agency book of business and 18.5% for the portfolio as a whole.

As of October, we did see a slight increase in speed, but the trend of our speed remained well below expectation is intact. Total portfolio speeds increased from 17.1 -- to 17.1% from 15.1% and that's a one-month speed for October and our agency book of business increased to 21.7% from 18.4%.

So will HARP impact the portfolio? First, the 30% of our book that is backed by commercial loans will not be impacted by HARP. Second, HARP is expected to have less of an impact on the overall marketplace than originally feared. Third, HARP-2 is expected to have less of an impact on hybrid ARMs than on 15-year and 30-year borrowers. Currently, as the rules are written, hybrid borrowers have less options to refi under the HARP-2 plan.

Finally and most importantly, our portfolio has certain characteristics, as I continued to say this morning, that should work to reduce the overall impact of HARP and we will now go into it in more detail.

If you go to slide 10, let's look at our prepayment exposure from a couple of perspectives. Since inception of this portfolio, we have maintained an intense focus on purchasing pools backed by borrowers that have a lower propensity to refi. Within our portfolio we have layers of prepayment protection. First, if you look to the left chart, under the 30% that has exclusive prepayment protection, we are expecting CPRs to remain at zero.

Second, 16% of our portfolio comprises post or near-reset ARMs. These homeowners have less incentive to refi since their mortgage rate has either already reset or is post-resetting lower. And third, we do not buy TBA securities at Dynex. Hence, 54% of our portfolio is comprised of selected specified pools that have favorable characteristics that should be correlated with slower speeds in the universe of hybrid securities.

exposure is within the CMBS sector and has explicit prepayment protection, while another 8% supports the post or near-reset paper. That's a total of 59% of our paper in the lowest expected prepayment sectors, while the other 41% is supporting the selected specified pools.

Let's go to our next slide and let's look at our overall prepayment metrics. What characteristics make our portfolio special? Simply, we have chosen pools with lower loan balances, with less third-party originated loans, more investor properties and more interest-owned payment structures. Along with these characteristics then, they have lower -- they have borrowers with less incentive to refinance even with our new HARP guidelines.

And let me be real specific. I've heard the word prepayment protected used a lot in the marketplace these days. I want to be very clear in the definition. A true prepayment-protected security is similar to our CMBS portfolio where there are actual legal obligation, or prepayment penalties, attached to the underlying structure of the loan. What we're showing you here in terms of selected specified pools are characteristics that will work to mute or reduce the overall prepayment of that pool versus the universe.

So first, in terms of average loan size, Dynex's average loan size is less than the overall ARM universe and especially smaller than the universe of ARMs with a higher probability of faster speed, which happened to be ARMs originated within the last two years.

The second characteristics -- third-party originator. These are both brokered loan originations, channels, of the largest originators. We only have a 41% exposure versus 58% and 61% of the universe. This construction did not happen by chance. It's important for you to note that we raised capital in December, we raised capital in March of this year. It took us some time to put that capital to work. We were very, very open about the fact that it takes us time to put that capital to work because we take our time in examining the specified pools.

The hybrid ARM sector is not like the 30 and 50-year sector. In those marketplaces, you can specifically ask for low loan balance pools, pools originated in certain states, high LTV pools. Within the hybrid ARM space, you simply examine the pools that are created because the volume is smaller and you look to, as I mentioned earlier, minimize the loan size, minimize the third-party originated pools. We want to purchase more pools with more non-owner occupied. Why? Investor properties or investors have less options to refi than a normal homeowner and under the current HARP-2 program, if it's an investor property, you cannot refinance that loan via HARP-2.

And then third, one of the more important characteristics are interest-only loans. So if you look at our book of business, [67]% of our book happens to be in interest-only loans versus 48% and 45% for the universe of hybrids. More importantly, if you look at our book of business that was originated prior to 2009, 75% of those loans are -- have interest-only features to them. Those borrowers have less of an option to refinance.

Here's the math. Let's just use 6% as an example. If you're a homeowner with a 6% mortgage and an IO payment, it is equitant to having a 4.375 equivalent payment, such that when you consider your refinance options, you must be refinancing into an instrument that has less than a 4.375 rate, or you could have a difference in your monthly payment.

As I mentioned earlier, under the HARP-2 program, hybrid borrowers cannot refinance into a new hybrid ARM unless they have an LTV that's less than 105. If your LTV is greater than 105, they will have to refinance into a 30-year instrument. And as many of you know, current 30-year rates would not allow a borrower with an IO payment to refinance and subsequently end up with a payment that's lower than his current monthly rate.

Let's move to slide 12 and let's discuss our overall credit exposure. Similar to prepayment, that we're managing prepayment risk, we meticulously manage our credit risk. We selected the specific pools that are in our portfolio. As you can see in this graph of non-agency CMBS vintage portfolio, you can notice what's not here. We have no exposure to the 2006 and 2008 vintage. We've avoided this vintage because we considered this to be some of the worst vintages ever created in the history of the CMBS marketplace.

We have layers of protection and as you notice, we've avoided the 2006-2008 vintage. We have emphasized the 2009 and newer vintages. These loans were all -- these pools were all put together starting in December of 2009 and later. We believe these deals are better constructed. The loans are better underwritten.

In addition, as I mentioned earlier, prior to the year 2000, a substantial amount of our portfolio was originated and securitized by Dynex Capital. We have tracked, managed and followed these borrowers and properties for the last 15 years.

Let's go to slide 13. Our next metric from a credit perspective is property type. As you can see in our chart, we have emphasized the multi-family sector. Over 55% of our portfolio is backed by multi-family properties. More importantly in the third quarter, 77% of all of our investments in CMBS were backed by multi-family loans.

So why the multi-family sector? This is one of the largest trends in the real estate world. There is a growing population of renters. The American dream of owning a home has faded for a growing part of our country. The home ownership rate has declined and is expected to decline further. Hence, the demand for rental housing has grown and is expected to continue. Vacancy rates have been declining over the past two years

SO let me summarize. The overall credit support behind the multi-family sector is relatively stronger than the other sectors within the CMBS universe.

Furthermore, here at Dynex, we have a core competency in multi-family. As many of you may already know, but some of you may not, prior to 1998, Dynex was an originator of loans, both residential and commercial loans. A large percentage of those loans that were originated were multi-family properties. Hence, are attraction to the sector, is a natural outreach of our business.

Move to slide 14 -- even though it's not discussed much, one of the risks that we have not been wiling to take to date at Dynex is extension risk. Whenever rates rise, these newly created mortgages will be vulnerable to large duration extensions as the homeowners will be locked into very low rates. This will help slow prepayment speeds and hence (inaudible) the average realized durations of these mortgages.

We've used some very elementary examples here and I apologize to those of you who are very experienced in the space, but as you notice within the chart, the average life (inaudible), the Fannie Mae [34]% at a 15 CPR assumption is a 5.4-year average life. At a 2 CPR, the bond extends out to a 14, almost 15-year average life instrument. Within a 15-year sector today at a 15 CPR, you'd assume that this 15-year has a four-year average life and at 2 CPR, this bond will extend three years, and in fact, I believe some of this paper in the 15-year sector will extend a little further than three years.

And just an example of a hybrid ARM, a new hybrid ARM at 15 CPB, which is similar to CPR, but with the entire principal being paid back after five years, the bond will extend from three -- close to three and a half years to four and a half years and that's a minor extension of one year.

We're concerned at Dynex of taking this risk within our portfolio and as such, if you move to slide 15, you'll see that the majority of our portfolio will mature or reset within 10 years. That's approximately [93]% of our portfolio. More importantly, 85% of our portfolio will mature or reset within seven years.

Extension risk is of great concern to us even though it appears that rates are low and will remain low according to the Federal Reserve for some time in the future. But at this point, it takes very little rise in rates. Let's say mortgage rates move from 4% to 5% to 5.5%. Many of the lower coupon mortgages will be out of the money and as such, subject to large, extreme extensions.

Let's move to slide 16. What are the opportunities today? We continue to see attractive return opportunities in the marketplace. What has taken place in 2011? Non-agency spreads have widened out, with the largest widening taking place in non-agency RMBS space; secondary to that has been non-agency CMBS.

We wisely invested our new capital from our December raise and our March raise in agency securities and as we pointed out to you, we would take our time and invest it in non-agency securities. We've methodically increased our exposure to non-agency CMBS paper. We will continue to do so as the fourth and first quarter unfold in front of us. There continues to be, across all sectors, mid-teens to low 20% returns.

And let me summarize. On slide 17, our portfolio has performed well since 2008 and the earnings power today remains relatively intact. Prepayment risk is litigated by superior portfolio construction and HARP-2 should not have -- should have less of an impact on hybrid ARMs.

Credit risk is mitigated by highly (inaudible) rate securities, superior loan origination years and a concentration in multi-family collateral. And finally, extension risk is mitigated by the short duration investment portfolio with 71% of the investments maturing or resetting within five years.

And one risk factor we have-not included in our slides is counterparty risk. As many of you know, a primary dealer went bankrupt this week. We actively manage our counterparty risk and as such, we had eliminated all exposure to MF Global prior to the day we (inaudible) their bankruptcy filing.

We believe financing will always be our number one risk factor and hence, we will continue to actively manage our exposure. We currently have 26 counterparties with whom we are approved to trade or borrow money. And as I mentioned earlier, there continue to be attractive opportunities to deploy capital despite the volatile financial markets.

Tom, I'll turn it back over to you.

Thanks, Byron. You went into a very substantial dive into the portfolio because we think it's very important for people to realize that there are a lot of crosscurrents in the market right now, and when we put this portfolio together in 2008, we went into it with our eyes wide open and our portfolio strategy has not varied since then. We're focused on short-duration, high-quality bonds that should minimize volatility particularly in a levered strategy -- portfolio strategy.

In conclusion, I know we had several items this quarter, but we believe the core performance of Dynex continues to be quite strong. The current environment of low rates gives Dynex a unique opportunity to refinance several of our outstanding investments, including the previously mentioned CMBS, and quite likely, our (inaudible) investments in the future. Our core earnings will support the current dividend and we continue to find quality, high, mid-teen investment opportunities that fit our conservative strategy.

We expect continued volatility in the marketplace. We feel our strategy will provide the best risk-return opportunity for our shareholders. We look forward to discussing our fourth quarter results in January and we'd like to open up the call for questions.

Thank you. We will now begin the question-and-answer session. (Operator Instructions) Our first question is from Trevor Cranston, JMP Securities. Please go ahead.

Hi, guys. Congratulations on a solid quarter in a pretty volatile environment. I was wondering if you guys could expend a little bit on some of the investment activity in the non-agency CMBS space, specifically kind of where are you guys finding the most value in the capital stack right now, and what the financing for that non-agency CMBS paper has been like throughout the course of the third quarter? Thanks.

Let me go through and give you just a little more detail. As I mentioned, we have definitively focused our attention on the CMBS sector during the third quarter. The majority of the purchases, 77%, were within the multi-family sector. The issuer of most of that paper happens to be either Freddy Mac or Fannie Mae. There are two different programs. Fannie Mae -- we call it Fannie Mae [dumps] paper. It has an agency guarantee wrapped around it.

The Freddy Mac program is a little different in the sense that it is a structured deal where Freddy Mac -- it comes through the Freddy Mac system, comes through the Freddy Mac in origination process, comes through the Freddy Mac underwriting process, but they only wrap the most senior bonds within [any structure] and they lave the private capital for bonds beneath the Freddy Mac guarantee. We have been willing to purchase CBS IO off of that most senior bond. We've also even willing to purchase the bond right below the Freddy Mac guarantee.

Financing for non-agency securities have increased despite the volatility that's taken place in Europe. Haircuts, pricing levels, they're volatile and we see that they've been stable throughout the quarter, but it is clear that the dealer community -- and this is also -- we include agencies -- is willing to adjust prices in haircuts appropriately. The availability of capital, however, continues to be really solid, to be honest with you, especially for program such as the CMBS securities that we're investing in.

Why did we choose these programs? One, the Freddy Mac program is the only consistent non-agency-ish flat agency program in the marketplace. Every dealer is involved in their program, every dealer has had an opportunity to do one of those deals and the financing tends to be very fluid. The Fannie Mae guaranteed paper speaks for themselves and we get agency financing on that paper.

So again, we love the multi-family sector. We think it's one of the strongest themes in the marketplace and I stated earlier, we built this portfolio based on a few macro themes and that's one of the strongest ones.

Does that give you everything that you wanted on that question?

Yes, that was very helpful. And just one more -- you guys commented on the Freddy Mac [K] program there. A few weeks ago, there was news that they were thinking about expanding a similar type of program to the residential space. I was wondering if you guys had any thoughts about where they are in that process and the likelihood that it'll eventually come to fruition or whether or not you guys would be (inaudible) spend there.

I've got a personal opinion that I'm going to share with that. I can't tell you how close they may be within Freddy Mac or Fannie Mae on that issue. I think it's a structure that should be used. When I was at Freddy Mac some 10 years ago, we used that structure. I think it's very viable. It's an opportunity for private capital to be involved, in effect, in reassuring -- or ensuring mortgage credit.

So it's an opportunity for the agencies to take the most senior risk, but more importantly, you create the security from a very structured, disciplined regulated process. And so from my perspective, I think it is a structure that will, and can, work in a residential space on a larger scale. So it's done today, but I think it can be used on a larger scale and I do think it will be an instrument that will be used in the future.

Okay. That's helpful. Thanks, guys.

Thanks, Trevor.

The next question is from Jason Weaver, Sterne, Agee. Please go ahead.

Good morning, guys. Thank you for the commentary.

Good morning, Jason.

First, I was hoping you could talk a bit about where you're seeing spreads on new investment across the various products.

Yes, I would -- let me give you some examples. If you go to slide 16, I think we have given some ranges in terms of where we think the spreads happen to be on the agency RMBS between -- let's call it 1.8 to 2.3, with the non-agency securities having wider spreads, but we carry lower leverages against those securities. So you can see, there's a pretty broad range here, but we have a diversified portfolio, but we're considering a broader range of securities.

In those yields that you're calculating from the AT RMBS you show there, is that based upon your 28% CPR expectation?

Yes, it'll be -- and that average CPR expectation is a summary across what I call -- we think about homeowners here when we think about our portfolio.

Right.

And so we think about which borrowers and so that is an average across a variety of borrowers based on the percentages within the portfolio. So some homeowners that we were buying certain securities will have a higher prepayment fee attached to them. So for example, a new 5-1 hybrid, which I'm not crazy over, but if we were to buy something like that, we'd attach a much higher prepayment expectation to it than we would to some of the really seasoned paper that's about to reset.

That's fair. And also, regarding the settlement, do you expect any following-on legal expenses continuing into this or later quarters?

Jason, this is Steve. We tried to accrue for those this quarter, so we do not expect that.

And finally -- this is more conceptual than anything else. I know you talked about HARP and the to extent of whether that works is debatable, but it really doesn't touch on the larger issue of negative equity out there. Now, what else do you see on the horizon as possibly being proposed either by the administration or the fed as far as further intervention goes in the mortgage market?

I do have an opinion on that too. Again, this is going to be a Byron opinion. One, I think there is a recognition now that the mortgage issue has created some problems. However, I think as many -- as much in the academic circles that if you do refinance, let's say you go through HARP-2, you take income from one hand and you put it into another, but net-net, the overall broad impact of that may not be as large as you think. I think the largest impact of HARP-2 happened on the first day and that was the announcement. Barack Obama was allowed to be in Arizona, make a huge announcement about what they were doing for homeowners. The actual numbers impact, I don't think, will be as large.

From a negative equity perspective, there are academicians who would say that they believe that the only way out of this situation fast would be to forgive homeowners -- forgive debt. I don't think that type of proposal would fly through our US Congress, so I think that's a remote possibility. I do think that Washington had some complications in terms of how much they actually can do at this point in time.

This is a long-term corrective process. You're correct, there are losses in the system with real estate in the United States and it's going to take some time to work out. I'm not sure there's an enormous amount that the government can do to speed that process up without creating additional problems someplace else. Do you have any other thoughts on that, Tom?

Jason, I think that we have an election year next year and I think they're going to pull out all the stops to be elected, and so I think that's why we keep our portfolio as volatility-resistant as possible because I think anything could happen.

Well, I appreciate the color, guys. Congratulations on a good performance in what's been a real tough quarter.

Thanks.

Your next question is from Douglas Harder, Credit Suisse. Please go ahead.

Thanks. Byron, as you were talking more about the potential for extension risk, obviously, other than portfolio selection, which is important, are you changing anything on the -- that you're doing on the hedging side?

We've got a real basic hedge strategy in terms of plain vanilla interest rate swaps and we consider also swaps as part of that strategy. No, I think the more important part about hedging is what exactly are you hedging? If you're hedging -- as we showed in that chart, you're hedging a four or five-year duration on an instrument that may not be four or five years, at some point, it'll show up, but that might be out in the future.

It'll be -- it may not be far out in the future as it may have been before because again, if you stay at these rates for six months, you have an enormous amount of new paper at 3%, 3.5%, 4%, and rates rise simply to 5% to 5.25% or 5.5% mortgage rates, that paper will be out of the money. And I'm hard-pressed to believe that many of those homeowners being locked into these low rates, that you'll see really fast prepayment on that paper.

So the hedging strategy is more what are you hedging? And we've put together this portfolio with the shorter duration paper of really to minimize what I call the overall hedging risk of the portfolio. So mortgage schedules are not certain. Therefore, your hedging process always has some risk to it as we saw -- as we've already seen over the last couple of months with some of the other information that's come out in the marketplace.

Yes, Doug, this is Tom Akin. I think the best hedge you can make in a portfolio, particularly this levered portfolio, either 5X, 6X, 7X, 8X, whatever, is to buy assets that don't need a lot of hedging. It's almost intuitive, but we're in an extremely volatile environment and the outcomes are really unknown. And therefore, I think it's really important to provide as much protection on the asset sides because the basis risk between your assets and your hedges are unknown.

I guess sort of along those lines, do you see moving more of the agency portfolio to the agency CMBS where there's more predictability there away from sort of the hybrids?

That's going to be a relative value decision, Doug, in terms of where spreads happen to be at any point in time, to be honest with you, but we like that trade, but it will -- these spreads are very volatile. We've watched these CMBS spreads, both on agencies and non-agency, move in large increments.

So it will depend on -- I hate to tell you, oh, yes, we're going to do that today, and then find out in mid-November that spreads are tightening so much that this rate really doesn't make sense. So it really is a relative value decision process that we're very disciplined about, but it will depend on where spreads are.

And today, I guess today, obviously, understanding that things do change, but today, is that an attractive relative value trade?

Yes, into -- out of agencies into some of the CMBS product continues to be a very attractive trade.

And the momentum that we have that we mentioned from the second quarter that we acted on in the third quarter is continuing on in the fourth quarter and I think you could expect us to take the prepayment risk out of the portfolio and add yield, and any time you can do that, that's a win-win.

All right. Thank you, guys.

The next question is from Matthew Howlett, Macquarie. Please go ahead.

Hello, guys, thanks for taking my question. Just a follow-up on sort of the portfolio strategy, but you guys today, I think, have one of the more well diversified, seasoned portfolios, most likely almost impossible to replicate in today's environment.

When you look at it, I know you're going to be just rotating stuff around a little bit here, but could we see you introduce any new type of assets, I mean, going forward? I mean, there's been REITs buying servicing assets to whole loans to first-loss pieces (inaudible) the continuation market. I mean, you guys have always been innovative. Could you see yourself sort of trying or adding a new component to the portfolio over the next four quarters?

We've looked at a lot of different things and we've been quite actively thinking outside the box, so I can't divulge exactly all the areas we're looking at. We look at a pretty diverse set of asset classes when we take a look at where we're going to end up investing our money in the final analysis. I will say that the opportunities in the CMBS market right now, with the pullback that we've seen in that sector in this third quarter, remain outstanding and remain our focus, but there are other things, Matt, that we are looking at the same time. And I wouldn't rule out anything at this point. Byron, you could finish up that as well.

No, I think that's a great explanation. I mean, if you look at our balance sheet, it includes agency residential securities, agency commercial securities, non-agency residential securities, non-agency commercial securities. It includes some AAA assets, it includes some loans and it includes some first-loss pieces and some pieces in between.

So as you mentioned, Matt, there's a broad variety of assets and our opportunity set is broad and we do relative to value analysis across the set. We don't try to get too far away from what our core competencies happen to be, but in today's world of technology, you do have the ability to look at a lot of asset classes. And most important to us is diversifying risk and can I diversify my leverage risk? Can I reduce my leverage risk? I'd love to be able to do that. So we're always looking at opportunities to continue to diversify our overall risk.

Great.

And Matt, I think the point is, is the government needs to get a non-Freddy-Fannie or non-GSE mortgage origination, mortgage portfolio environment going. I think that's one of the reasons that the SEC is looking into the [40 Act] because they see that as expanding and growing and they want to clean it up a little bit.

At the same time, I think these non-agency opportunities, while small right now, are going to be more and more interesting as they execute on that plan. We're quite interested in remaining involved in that. We look at ever one of these new opportunities as they come along and as they are value-added to our portfolio and fit within our conservative investment style, we'll be looking at them.

Great. Well, we'll certainly watch for new developments, but great quarter. Thanks for answering the question.

Thanks, Matt.

Your next question is from David Walrod, Ladenberg. Please go ahead.

Good morning, guys.

Good morning, David.

Just a couple -- you talked about in your prepared remarks about a $0.03 charge due to increasing the forecast of CPR. So that was not a function of prepayments this quarter, but more so your outlook for prepays in the agency portfolio to be 20% in the fourth quarter?

Steve, I know you've got all the numbers on that. Do you mind answering that question?

Yes, David, that's exactly right. That's a cumulative catch-up of premium amortization from an increase in our forecasted prepayment speed from this point forward.

Okay. So that $1.3 million, assuming that the prepays come in similar to what you're forecasting, will not be there in the fourth quarter, so -- and then if the prepays come in below what you're expecting, you'll be able to reverse some of that out. Is that fair to say?

Yes, I think it's fair to say that this quarter would have reflected amortizations based on the entire speed to the extent we either would -- well, I'd say change the forecast in the fourth quarter, which we don't anticipate doing, certainly not up, then you would have potentially that adjustment, but to the extent that the prepayments actually come in lower than forecast, then you would have some, in a sense, reversal of some of the amortizations in the quarter.

Okay. That's helpful. And then when you -- the early reduction of the collateralized financings, is that something you'll be able to do more of in the next couple of quarters or is this more of a one-time thing?

That particular bond was a bond that we issued in the late '90s. That is the one of the last remaining bonds that we have to be able to redeem off our issuances. We did make reference to an additional $50 million of collateralized financings that we have that we could redeem or retire, if you will, and refinance. Those don't relate to our issuances, but relate to some counts that we did several years back, retiring that. And I think we said that would save us roughly $600,000 in interest expense going forward if we were to do that.

On an annualized basis.

On an annualized basis, yes.

All right. Thank you much.

Your next question is from Jeff Porter, Porter Capital. Please go ahead.

Hi, great quarter, guys. I was just wondering if you could comment on the status of some of the legacy assets, in particular, I believe, we've got a situation with the Denver Merchandise Mart that is a non-performer. Can you just give us an update on where those assets stand?

Steve, do you want to take that?

Sure, I'll take that. Yes, hey, Jeff, the -- we did have a decline quarter-to-quarter in our 60-plus day past-due loans. We had some liquidations. I think sort of overall, you're seeing fewer and fewer and fewer assets were coming in and more and more assets were going out of that box.

So I would say on a whole, the portfolio is performing quite well and we've sort of got the problems in that portfolio circled off. So unless there's some change from what we see now going forward, I think kind of the loans that we have sort of circled off now seem to be the ones that remain the challenging loans in that pool.

With respect to the Merchandise Mart, I think as we might have talked about last quarter, that is a $9 million loan in our balance sheet. That borrower has filed BK. We believe that to be a strategic filing. We do have -- would have included that in our analysis of reserves, so we think we're adequately reserved for any event there. That's just an ongoing situation where you're in a strategic situation. The borrower is trying to get the bankruptcy court to take some action, to give it some relief on the debt. We're obviously challenging that and/or working with the borrower on some sort of plan to meet the borrower's needs in that regard.

So I'd say on that one, it's going to take some time. It's a complicated situation, but we feel like the reserves that we might have against that loan and the balance sheet are adequate to cover us for that contingency.

I think we can also add, Steve, that we had some appraisals done on that property which were well above our loan to the borrower.

Okay. I've just got one follow-up. I know that you guys have an authorized a share buyback in place and we're in a volatile market and we've seen some dislocation in the share prices across the board in the mortgage REITs. What's your disposition owards using some of our capital to buy back stock if we trade significantly below book value?

I think the problem with buying back stock for a mortgage REIT is that it's only a one-time earnings hit, whereas our ROE for the quarter was substantially greater than -- well, it was right around 14%. So if we were to buy it at, let's say, a 14% discount, we're trading currently at, let's say, a 10% discount. So that's a one-time earnings gain, but it -- and it doesn't recreate itself.

We like to be in the business of recreating a recurring revenue stream, so unless the stock gets dramatically -- I was going to say a major dislocation, I would say the best use of our capital is to continue returning 14% to our shareholders on an ongoing basis.

Okay, thanks.

Your next question is from Jay Weinstein, Highland Wealth Management. Please go ahead.

Hey, Good morning, guys. Byron, thank you for that extra detail on the investment portfolio. It was helpful.

Okay.

Hey, Jay.

Hey, how are you, Tom? A couple of questions -- actually, I was about to, of course, ask the share buyback question. Somebody beat me to it. You briefly touched on the SEC comment which I know kind of came out of the -- I think came out of the blue. It surprised everyone. Tom, if you could elaborate a little on your thoughts as to what that is and I'm sort of curious if you've had any discussions with other mortgage REIT executives as to what it's all about and kind of what they see happening?

This would be just simply my opinion and we joined NAREIT, the National Association of Real Estate Trusts, and we're going to be down at the annual meeting in Texas in a couple of weeks and we'll have a lot more discussion on that. There is some -- there is a lot of movement within NAREIT to discuss this whole situation with the SEC, but I think that you have to take a look at the service that the mortgage REITs provide.

And in terms of portfolio'ing mortgage securities or mortgage loans, which the government really wants to see happen, they're trying to execute a move from the GSEs to the private sector and we are the private sector and we are portfolio'ing tens of billions of these things for the government. So we are in line with what the government is trying to do.

The fact that they have not had any review over this group over the last -- well, 50 years, indicates they want to look at that. I mean, I think the three big questions are how much leverage should a proper mortgage REIT employ? Should it be 5, should it be 10, should be as high as [Carlyle] was up at 30? Maybe there is -- maybe they should have some say on that.

At the same time, I think that the way we issue capital and the prices we issue capital are a question. Should you be able to issue capital substantially below the market cap -- or I'm sorry -- the last price or even the net asset value? As you know, closed-in funds and mutual funds have really strict limitations on how they issue capital.

And then finally, management compensation, we know that that's a very big touch point for the current administration and some of the recent REITs that had been offered to the public have almost a hedge fund-like compensation structure. And is that the right way to administer compensation in a REIT?

So I think a lot of these questions are very, very positive. I think they're good for the growth of this sector because I think it does need to grow and I think answering these questions cleans it up so that we can continue to attract capital and become viable -- a more viable business model.

We're encouraged by this. As you know, Dynex has always taken a pretty conservative approach to how we raise capital. We are internally managed. We don't have a compensation structure that's based on performance and we utilize a light amount of leverage vis-à-vis our competitors. We're still right around 6X leverage and that would be on the lower end of the mortgage REIT universe. So we stand by what they're doing and we look forward to putting in our -- opining ourselves as we get to the end of the comment period, which I believe is November the 9th.

One question -- this might be a little trickier one for you to answer. So you guys have sent out pretty confident signals that your dividend at least could be sustained at the current level. Do you feel that way about some of your other competitors that they'll be able to do that, or do you think that with HARP-2 and the other spread pressures in the business and Operation Twist and all the other things, that they're going to have a harder time?

Well, we'll let those guys worry about their business models. We're busy working on ours, so you're right, we're not going to answer that question.

Actually, I thought I'd try. One easy question -- on the balance sheet, I assume that the drop in the collateralized securities line was just the result of the refinancing that kind of went into the other line on the balance sheet? There was a pretty sharp drop, I think, in collateralized securities.

Collateralized financing? Yes, that's right.

That was just for the refinancing, correct?

Yep.

Okay. All right. I think that's all I have today. Thank you.

All right. Jay, thank you much.

Okay. Thank you for all your work.

Your next question is from Jason Stan (sic), Clayton Partners. Please go ahead.

Jason, you've shortened your name.

Yes. I guess those "ski's" at the end are a little tough for people to get. A lot of the questions I had have been answered. Just curious if -- when you talk about diversifying your financing sources, where you guys see the market and if you've contemplated anything? We had a preferred out a while back and is the market healthy enough now to where -- for some of our non-agency paper, there are viable, longer term kind of match-funded, whether it's preferreds or other type of financing out there, and if you could explain kind of what those are and how close or far away you think they are from us being able to avail ourselves of any of them?

We're looking at all sort of opportunities that come along. We, as you know, are very sensitive. We are -- our Board of Directors and myself own a substantial amount of the stock. We're very sensitive to issuing stock below book. We do have some extraordinary investment opportunities that we have right now, so there is use for the capital. So we would love to have more capital right now. Obviously, [Gary Kane] over at Agency felt that way and he issued some -- a substantial amount just recently.

We do -- we will look at other opportunities. I don't think a preferred is outside the question. I think it might make some sense and we'll keep an eye on it as we -- as opportunities present themselves. But it's really based -- our capital needs are really based on the investment opportunities out there and we see quite a few of them right now.

Yes, and I was just speaking mainly to the financing risk of some of our non-agency stuff being on repo as opposed to match-funded and kind of how you think about that.

Hey, Tom, let me chime in on one point of that.

I was going to say, Byron, why don't you --

We are looking at it and have been and continue to look at opportunities to finance that on a longer term basis. The market is -- I mean, there have been pockets and it'll change with the overall risk (inaudible) fears in the marketplace, but generally, there's at least one lender out there that may lend money against -- for a long-term basis against non-agency.

There are other ways to potentially do a structured deal against the non-agency paper and we continue -- we agree with you in the sense that we want to think about how can we get longer term financing on the non-agency instruments? We think about that more than so on the agency side of the book.

Okay. And lastly, I guess, since the quarter has ended, given where your swap book is at and kind of what's -- can you give us an update of kind of what's happened with prices and directionally, which way book values have headed here after the quarter, given the continued kind of volatility we've had?

There really hasn't been that much. I mean, if you notice, a [little more] rates are gone. You had a little spike up in rates and then you spiked back down and you're in the same range that we've been in since this large drop in rates. So -- and in addition, Steve just mentioned in terms of overall spreads, we have seen, as the HARP information has come out, which is kind of interesting, premium prices, both in ARMs and in 30-years and 50 years, all lagged in the third quarter due to a fear of HARP.

As that information has been digested, while we have seen a much, much better pricing from a premium perspective, we've given it a bit more clarity around what may actually come out of the HARP program and the expectations -- or the last (inaudible) fear. So, see, on some of these, it's the premium agency paper seeing better spreads. Some of the CMBS paper has some tightening in spreads, but then rates drop somewhat directionally and a fear on (inaudible) see the spreads widening back out, but there hasn't been an enormous change throughout the month of October to really speak of at this point.

Okay, great. Thanks for doing a great job, guys. Appreciate it.

Thanks, Jason.

Having no further questions, this concludes our question-and-answer session. I would like to turn the conference back over to Thomas Akin for any closing remarks.

It's been a long call, over an hour. We appreciate all the good questions. We appreciate the confidence of our shareholders and we look forward to updating you on our progress in the fourth quarter call next year. Thank you much.

The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.