Capstead Mortgage Corporation
Q2 2019 Earnings Call Transcript

Published: 25 Jul 2019

[Operator:]

Good morning, and welcome to the Capstead Second Quarter 2019 Earnings Conference Call. All participants will be in a listen-only mode. [Operator Instructions] After today's presentation there will be an opportunity to ask questions. [Operator Instructions] Please note that this event is being recorded.I would like to now turn the conference over to Lindsey Crabbe. Please go ahead.

[Lindsey Crabbe:]

Good morning. Thank you for attending Capstead's second quarter earnings conference call. The second quarter earnings release was issued yesterday July 24, 2019 and is posted on our website at www.capstead.com under the Investor Relations tab.The link to this webcast is also in the Investor Relations section of our website. An archive of this webcast and a replay of this call will be available through October 23, 2019. Details for the replay are included in yesterday's release.With me today are Phil Reinsch, President and Chief Executive Officer; Robert Spears, Executive Vice President and Chief Investment Officer; and Lance Phillips, Senior Vice President and Chief Financial Officer.Before we get started, I want to remind you that some of today's comments could be considered forward-looking statements pursuant to the Safe Harbor Provisions of the Private Securities Litigation Reform Act of 1995 and are based on certain assumptions and expectations of management.For a detailed list of all the risk factors associated with our business, please refer to our filings with the SEC, which are available on our website. The information contained in this call is current only as of the date of this call July 25, 2019. The company assumes no obligation to update any statements, including any forward-looking statements made during this call.With that, I will turn it over to Phil.

[Phil Reinsch:]

Thank you, Lindsey. After a few brief remarks Lance will give a recap of the quarter and then we'll close in the call after questions.Our core earnings held up well this quarter in the face of significant market volatility. As the market recalibrated from expecting 125 basis points fed rate cut, to now expecting as much as 75 basis points in cuts this year, to another 25 basis points or so in 2020.For the second consecutive $0.12 quarterly earning sprint, we increased our common dividend by 50% to $0.12 per share this quarter. This reflects our belief that we can produce strong risk adjusted returns this year and the next, regardless if the fed reduces the fed funds rate at the pace anticipated by the market or take more of a one and done stance at its meeting next week.On the one hand should the fed reduced the fed funds rate a number of times as expected by the market, we will benefit significantly considering that our $7.5 billion in swap balances at quarter end represented 70% of our outstanding repo balances, with another $1.25 billion in swaps maturing by year end.This will leave us with plenty of room for our borrowing cost to benefits from rate cuts. The resulting lower borrowing costs will help insulate earnings from many effects of higher mortgage prepayment activity spurred by lower pertaining mortgage rates.On the other hand, should the fed not reduce the feds fund rate as much as expected, we remain well managed at a reasonable cost, having effectively banked rate cuts through the liberal use two and three year swaps with lower fixed pay rates relative to unhedged repo rates.This has improved our net interest margins while helping to insulate us from a more hawkish threat fed. Additionally in this scenario mortgage prepayment pressure should subside with the market recalibrating to expect higher rates.The negative of our hedging activities is that we lost value in our swap book on a mark-to-market basis with a slight decline in rates this quarter. This occurred because many of these positions were put on prior to these declines and the agency MBS pricing underperform due largely to rising expectations for a higher mortgage prepayment.At quarter end book value has been relatively stable. These negative marks will dissipate over the terms of these swaps to the benefit of book value, and should the fed disappoint the market by not reducing the fed funds rate as much as expected, valuations could improve more rapidly.With that, I'll turn the call over to Lance.

[Lance Phillips:]

Thank you, Phil. We incurred a GAAP net loss of $63.5 million this quarter or $0.80 per diluted common share. Our core earnings were $14.8 million or $0.12 per diluted common share. The difference between our GAAP net loss and core earnings primarily reflects realized and unrealized valuation losses on our swap portfolio as Phil described, as well as certain other amounts excluded from our core earnings mentioned. We include a reconciliation of these differences on page 8 of our press release.Portfolio yields averaged 2.82% during the quarter, an increase of 7 basis points from the 2.75% we reported in the first quarter. Yields directly benefited from higher cash yields on acquisitions as rates on the underlying mortgage loans in our portfolio reset higher, while absorbing the effect of higher mortgage prepayment levels.Our portfolio related borrowing costs increased 12 basis points over the prior quarter. This increase was primarily due to higher hedging costs as the interest rates swaps with lower fixed rates matured, new swaps were entered into at higher rate and variable rate swap receipts were negatively impacted by declines in three months LIBOR. Unhedged borrowing rates were relatively unchanged from the previous quarter.Book value increased $0.50 per share during the second quarter, ending at $8.93 per common share. The decrease reflects $1.05 decline associated with our hedging activities that was only partially offset by a 55% increase in portfolio related pricing changes.With that, we will open the call up to questions.

[Operator:]

[Operator Instructions] The first question comes from Eric Hagen of KBW. Please go ahead.

[Eric Hagen:]

Thanks, good morning. Two questions on prepayment speeds. Number one, when do you expect speeds to slow; and number two, the slowdown that you're expecting which corresponds to the amount of premium amortization that you are currently booking, what is that prepayment speed expectation? Thank you.

[Phil Reinsch:]

So prepays are going to remain elevated this summer and then they will start to recede in the fourth quarter we assume, and any amount of decline will stand up on how aggressive the fed is in cutting rates and then we would expect some moderation next year.

[Eric Hagen:]

Okay.

[Phil Reinsch:]

But we don't actually make a disclosure that – of what our estimated prepayment should be for the future life of our portfolio. I know some guys do, but we have not made that disclosure to-date.

[Eric Hagen:]

Okay, I can appreciate that maybe an exact number isn't something that you want to provide, but you know the portfolio paid down at 26 CPR in the second quarter, that expectation, is that something in the low 20’s, is it in the high teens, is it in the mid-20’s. I mean there has to be a range that you can sort of guide us to. Thank you.

[Phil Reinsch:]

Well I think Eric, ARM speed generically, if you look at where we are right now in the mid-20’s, so we would see earlier I think that a few more factors will kick that number up somewhat in the third quarter before declining in the fourth. But right now you’ve a an interesting time in the ARM market where you have longer reset securities that were originated in the last year or two where speed assumptions may have gone from 18 to 25 generically, and at the same token you have very seasoned ARM securities that are resetting down now and estimated life speeds have gone from the mid-20s to the high teens. So you are starting to have convergence where new issue on the [ph] reset paper is ramping up, at the same time their receiving bonds, life expectations are going down.So to take all that in context I would think you know the natural progression would be from mid-20s to upper-20s back down to lower-20s and then over time depending upon what rates do, I think it probably settles in, in aggregate somewhere, you know low-20s to 20 CPR, something like that. I think that would be a – we paid a projected ramp of ARM speed in the entire market. I think that's kind of what you'd be looking at.

[Eric Hagen:]

Okay, okay, so we should figure they'll go back to the maybe low-20s over time. But I want to hammer in kind of a focus in a little on the timing right, because ARMs is such short duration securities, and you are presumably booking at higher yield today with the expectation that speeds will eventually slow into that, call it low-20s range, then after how much time or at what point would you be required to take you know what’s effectively a catch up charge if the slowdown that you are expecting doesn't actually materialize?

[Phil Reinsch:]

Well that’s, I don't know if we can really answer that question, whether a catch-up charge would actually be required. We look at our life speeds, we keep an eye on that all the time and we’ll adjust them as we need to and I wouldn't necessarily anticipate that we would have a sharp change in speeds required.While conservative with our speeds, we've just – you know I know the concern is that amortization is a little lower this quarter relative to the pick-up in speeds, but as we discussed last quarter we've been working to improve our estimation process for prepayments to lessen the impact of cyclicality and seasonality, and we have a large diverse portfolio of ARMs. They are doing different things in two different cohorts. Premium levels on recent acquisitions and average outstanding balances were also in part a bit lower at this time, and that also leads to lower acquisition costs.

[Eric Hagen:]

Right. I understand that, but I guess I'm just trying to understand like the timing around actually realizing you know what the effective amortization was on the bond relative to your expectations, and when that true-up needs to take place?

[Phil Reinsch:]

It’s happening every quarter, the true-up of things, so I don't, I don't get the question.

[Eric Hagen:]

Okay.

[Lance Phillips:]

Eric the one thing I think I might say is we do book the actuals and that rolls through our core earnings. We do not make an adjustment on premium amortization and catch-ups, but every premium dollar we spend does eventually come through timing on core earnings as the bond either is paid, whether prepaid or through it’s scheduled. So I wouldn't expect the cache-up like you're describing. I know others adjust that catch-up plus or minus on their core earnings, but we have traditionally been and continue to be conservative with that. We are just – whatever dollar we spend on premium will be amortized through core earnings.

[Eric Hagen:]

Okay, yeah maybe we can follow-up. [Cross Talk]. Yeah, well may we can follow-up offline on kind of the timing aspect of amortization, but on the swap side, the $550 million in swaps that are rolling over this quarter; I assume that's already been rolled. Where along the yield curve did you replace those swaps?

[Phil Reinsch:]

We are looking at every swap we put on in the last however many months or whatever, and we don't necessarily just replace swaps as they roll-off. It really depends upon what we're trying to accomplish with our duration GAAP. We are looking at potentially locking in lower financing costs in the future, etcetera, but we have been staying – all the swaps that we’ve been putting on have been the two to three year part of the curve. We haven’t gone any longer or any shorter for the most part.But you know obviously we increased our swap position somewhat and a lot of that at the time was we were looking at swap rates that were cheaper than our implied repo financing and so obviously in hindsight the market rallied since then, but if you look at our average outstanding swap coupon it’s around 2.25 or whatever, which is still cheaper than swap repo. And obviously the new markets pricing and through fed cuts and most of that rally came from May to June and so, yeah a do over if you wouldn’t want as many swaps on, but we thought that was a, still a fairly conservative 70% of our liabilities hedged effectively and then have upside on the remaining 30% the fed does cut.Now we will monitor that over time and we may not be at 70% of our liabilities hedge, it could be less, it could be more, but that's kind of what happened this quarter and essentially the book value decline was if you just kind of do the simple math on those swaps and assume the swaps had roughly a two year duration, we will offer a little over a point in value on our swaps and our bonds only improved about a half a point, because as mortgage spreads widened. Our margin spreads widened 20 to 25 basis points and we had roughly 70% of our liability hedged, so if you the math, it kind of backs into the book value numbers.

[Eric Hagen:]

Yep, but just confirming that that $550 million was rolled into a new two year swap this quarter, and that previous swap had a pay rate of 140, and I assume that you know roughly speaking the new swap has a pay rate of 170, 180.

[Phil Reinsch:]

Well, I think what Robert was saying was it wouldn't necessarily be rolling over the swaps just because they are maturing. We are looking at the total picture and we may be adding some swaps, we may not be. I think in my remarks I pointed out that with the fed looking to reduce rates, we can enjoy more of that benefit with fewer swaps and we do have swaps maturing by the year end.

[Eric Hagen:]

Okay, and where are you guys rolling the one month repo today?

[Phil Reinsch:]

Well, that's kind of a moving target with most recent stuff that we rolled a week or so ago. It’s pricing in partially the fed needs [ph] and so kind of the mid-240’s, we would think that if the fed does these 25 next week, we’ll be rolling new repo in the 230 to 235 area.

[Eric Hagen:]

Okay. Alright, thank you for the comments.

[Phil Reinsch:]

Sure.

[Operator:]

[Operator Instructions]. The next question comes from Steve Delaney of JMP Securities. Please go ahead.

[Steve Delaney:]

Thanks. Hey, good morning everyone. You're pretty busy down there in the second quarter on both sides of the balance sheet. So definitely sense that you are looking at the environment and trying to shift you know to the greatest advantage possible. I guess where I'd like to start is, you know I'm looking at your reported interest income on the portfolio of $85.1 million, you know up a little over $1 million and I'm looking at my model and obviously we had higher pre pays and you know we had a figure closer to $82 million in our model.So Phil, you guys used to have a rule of thumb where a 1% increase in CPR would cost about $1.50 on EPS and from the comments that you've made this morning, I'm hearing that maybe that we should kind of throw that figure, you know out the window. It certainly didn't look like it was applicable you know when you had almost a 6% increase in CPR in the second quarter. Is there a figure that you can give us to quantify the EPS impact of a 1% shift in CPR either up or down at this time?

[Phil Reinsch:]

I don't think we can point to a specific figure, because it is going to be, it is going to morph with the level of prepayments we see, higher prepayments are going to result in more amortization. But we have taken some of the cyclicality and seasonality out of the ARM amortization equation with how we estimate the prepays.But we are definitely – you know you could argue we're not being as conservatives, but you could also argue that we were very conservative in the past with how we amortized.

[Steve Delaney:]

Well you were, and it led to obviously quarter-to-quarter volatility, right, in terms of just the seasonality regardless of rate moves you know and our model is kind of crazy, but you know it had that you know down in 1Q, up and down in 4Q, and it really is, makes it you know kind of crazy to follow, especially for people that aren’t close to the story as Eric and I, you know might be.But I do appreciate your comment. You are saying in your earlier response about your methodology for premium and I'm hearing you clearly convey that you are on an – and I'm not trying to put words in your – these are my words, not you're Phil, that you are trying to smooth out the impact of quarter-to-quarter CPR volatility to get something, some sort of a smoothing if you will that is kind of true-up on an annual basis as opposed to the big quarter-to-quarter.So I think that what, from a modeling purpose we need to do is to come to more of a – if the range of your CPR, just to be simplistic, if it was going to be 20% at the low and 26% or 27% at the high, I probably would take the midpoint around 23-ish or something like that and just change my quarterly CPR to something along those lines. And I'm curious, is that type of approach sound like the advice that you would give to new analysts that decide to come in and pick up coverage on CMO?

[Lance Phillips:]

Yeah Steve, this is Lance. I would say that’s fair. You know I think one of the things we are real careful on, we look hard at our estimates and as Phil described a couple times, we have tried to not move those estimates with the latest news and looking more of a lifetime estimate. In that lifetime estimate I think ultimately you could model generically something similar to what you described, because we have a longer horizon in those estimates and…

[Steve Delaney:]

Well, that's helpful and I'm glad you used - I didn't want to use the term lifetime, because life can be long, but I appreciate - that is the context I think that many of your peers use as they – you know quarter-to-quarter, but it's a lifetime assumption. Some disclose it some don't, but that concept, I appreciate the fact you used that phrase, because it helps me understand kind of where you guys are on this.

[Phil Reinsch:]

No, that's good Steve and you know ARMs are unique, so I will use that. I also will caveat to say that those lifetime estimates continue to be a challenge, you know and something as Phil mentioned, we looked at on a regular basis to try and do the best we can with it, but hopefully that gives you some help for modeling purposes.

[Steve Delaney:]

Yeah, and Robert thank you for your comments on repo. You know we had assumed that it – the market had not really fully priced in a cut, but it sounds like you are getting some relief from where we may be worth you know 30 days ago already, in your mid-240s.

[Robert Spears:]

Yeah, it’s kind of best. They never price it and anticipated [Cross Talk].

[Steve Delaney:]

Yeah, of course. LIBOR prices it all in, right, because LIBOR is a real market and the dealers just sit there and say oh, we don't owe you anything, the fed hadn't done anything, so I get it, but we will work with that 230, 235 you know going into the - towards the end, you know August-September timeframe. It sounds like we are pretty much definitely going to get something here on the 31st.My last question guys is, you know I totally support the dropping the hedge accounting and going to core EPS, it’s a very positive move in my mind for you to have made that. I just need some comfort, that was just last quarter and this quarter you terminated, I want to say it was an $800 million swap in June.Looks like there's a – you know you have in your GAAP EPS, you have derivative expense of $74.8 million and then I’d see three different items in the core reconciliation and I guess my question is, it doesn't look like it adds up exactly, but is there anything that is in the core reconciliation that you were adding back that is not already reflected in your GAAP EPS, like a one-timer or something like that.

[Phil Reinsch:]

Well, we did. We did sell about $300 million of bonds and so we had a $1.4 million of realized portfolio loss that ran through earnings this quarter. Ordinarily that just would have been an unrealized loss running through book value.There was more realized swap loss, because we terminated early $800 million of 30 month to term out swaps and replace them with new 24 month swaps at significantly lower rate. So that kind of distorted the relationship between the unrealized swaps loss and the realized swap result.

[Steve Delaney:]

So I guess, that’s honing on it Phil. The 24.3 million add back for the realized loss, I’m just – I guess my question is, is that add back relate – is there a unrealized – are you reclassifying unrealized loss to realized loss?

[Phil Reinsch:]

Well effectively we are, because you’ve terminated those positions. So you book those loss as realized and then the new position that came on puts us in the same position effectively from a hedging perspective, but we’ve locked in the lower fixed pay rates on the new swaps. Now that duration, you know we actually reduced our term on those swaps which was very important to doing a trade, so we went from 30 months to 24 on that component.

[Steve Delaney:]

Okay and you refer – go ahead Lance.

[Lance Phillips:]

Phil’s answer is absolutely correct. I just want to make sure that was realized in GAAP and we noted as realized in core earnings. So there is no swap between unrealized – you know switch between unrealized and GAAP and realized and core earnings. I want to make sure that if that’s what…

[Phil Reinsch:]

Yeah, core continues to include the cash flows off the active swaps, and that 6.7 million positives is just running off the gain in those swaps that were there when we seized hedge accounting at the end of February.

[Steve Delaney:]

Correct, I remember that, that we would have that each quarter for a while, so that is run off. So okay guys look, thanks for the comments. If I have anything else as we're going through the model I will reach out. Thank you.

[Phil Reinsch:]

You bet.

[Operator:]

Our next question comes from Gabe Poggi of Shoals Capital. Please go ahead.

[Gabe Poggi:]

Hey guys, thanks for taking the question. I appreciate that you guys are going to provide you know to Erick’s and Steve’s point of what’s the estimated. How you guys are thinking about estimated pre pays going forward and Lance you know it’s kind of a lifetime adjustment. Can you talk about what assumptions go into that assumption, specifically how you guys are thinking about the forward shape of the mortgage curve?And what I'm getting at is, talk about the fed lowering rates and obviously we probably get 25 basis points next week, but I'm trying to get a gauge as what happens if the long end of the curves follows suit and we just continue get a big flattener and how you guys think about that in conjunction with those prepayment estimates. Thank you.

[Lance Phillips:]

Sure, I mean we’re on estimated prepaid speeds on three or four different models and this is far out. There is not a uniform consensus on our prepayments out there and there is not much time spent modeling on prepayments as there is fixed. So there are a lot of moving parts and you can get three different answers on three different models.And so when we're looking at things, we will model them to the forward curve and to the static curve and at a given point in time. Obviously, the long and adverse, those models are going to kick up speeds, for instance more on certain bonds than others. But right now what's going on in ARMs as I said earlier, very seasoned bonds. Some of these guys are seeing their mortgage rate drop 75 to a 100 basis points from where they were reset last year.And because of that and where the forwards are right now, it assumes those bonds, if you look at the life speeds on those bonds now versus six months ago, they are a lot slower. Conversely if you take it, you know call it a new issue [inaudible] with a 3.5 coupon, it may have been modeled at 15 CPR six months ago and it might model at 20 to 25 CPR now.So when we look at it, we look at it versus static and the anticipated forward curve at that point in time, and that’s what goes into our lifetime assumption. We don’t really – you know, there’s so many potential scenarios out there, we don't model everything, but those are what we look at when we look at life speeds on our bonds.

[Gabe Poggi:]

Hey guys, I'm trying to get a gauge of what – I understand there's a lot of models out there, but what specifically you guys are focused on to come up with that estimate? And just because you look at the statics speeds of the last few months and they've really ramped and obviously you know you, Steve mentioned it’s up you know 6% or 26 CPR, but just thinking about that in conjunction of the new estimate.Kind of how you are looking at things now the last two quarters with that change and I appreciate you being conservative in the past, but kind of how that – how you ebb and flow with that depending on the shape of the mortgage curve. I guess the answer is you just taking what the forward curve is showing you today and basing it off of that and if things change and I assume you have to make a change. Is that correct?

[Phil Reinsch:]

Yeah, and we are at a point right now for instance where we have bonds that are prepaying at 15 to 17 CPR and models have life speeds as high as 25. We have bonds that are pre paying at 25 right now, where models have life speed going down the 15 and so you have a very interesting point.I mean basically as I said earlier, newer issue, longer reset paper, life speeds are projected to go higher, very seasoned, post reset paper where they are adjusting now and are projected to go lower, and so you are going to have kind of a convergence between those, you know whatever that speed is, but that’s kind of what the trend in the market is right now. And you know it's totally different than where we were six months ago, and it could be totally different six months from now.

[Gabe Poggi:]

Correct me if I’m wrong, having a lot of the seasoned stuff prepaid pretty fast though in the last couple of years or last couple of 18 months or so.

[Phil Reinsch:]

Yeah, that’s what I’m saying, and so a lot of the [Cross Talk] – yeah, a lot of the season paper, its sped up because these borrowers were seeing their rates increase and so they sped up beyond the life speeds would have expected and right now you are actually seeing for instance on our very seasoned ARMs last month. Their drop was much more precipitous than other cohorts. They declined as much as 20% from June to July and you know the aggregate ARM cohorts declined 6%. So you are starting to...

[Gabe Poggi:]

But they were also coming off a really high prepays, right. That 20% decline was off of very high speeds.

[Phil Reinsch:]

Yeah, well it depends on the land, the law, etcetera, but you saw as much as a 5 CPR drop. So bonds that were prepaying at 30 with the 25, bonds were prepaying at 25 with the 20 on some seasoned cohorts.

[Gabe Poggi:]

Got it, okay, thank you. That’s helpful.

[Phil Reinsch:]

Sure.

[Operator:]

This concludes our question-and-answer session. I would like to turn the conference back over to Lindsey Crabbe for any closing remarks. Please go ahead.

[Lindsey Crabbe:]

Thanks again for joining us today. If you have further question please give us a call. We look forward to speaking with you next quarter.