Armour 住宅房產信托 (ARR) 2016 Q4 法說會逐字稿

Ladies and gentlemen, thank you for standing by. Welcome to the ARMOUR Residential REIT Incorporated Fourth Quarter 2016 Conference Call.

(Operator Instructions)

As a reminder, this conference is being recorded Thursday, February 16, 2017.

I would now like to turn the conference over to Jim Mountain, Chief Financial Officer. Please go ahead.

Thank you, operator, and thank you all for joining ARMOUR's fourth quarter 2016 earnings call. By now, everyone has access to ARMOUR's earnings release and Form 10-K, which can be found on ARMOUR's website.

This conference call may contain statements that are not recitations of historical fact, and therefore constitute forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. All such forward-looking statements are intended to be subject to the Safe Harbor protection provided by the Reform Act. Actual outcomes and results could differ materially from the outcomes and results expressed or implied by the forward-looking statements due to the impact of many factors beyond the control of ARMOUR.

Certain factors that could cause actual results to differ materially from those contained in the forward-looking statements are included in the Risk Factors section of ARMOUR's periodic reports filed with the Securities and Exchange Commission. Copies are available on the SEC's website at www.sec.gov.

All forward-looking statements included in this conference call are made only as of today's date and are subject to change without notice. We disclaim any obligation to update our forward-looking statements unless required by law.

Also, our discussion today may include references to certain non-GAAP measures. A reconciliation of these measures to the most comparable GAAP measures is included in our earnings release, which can be found on ARMOUR's website. An online replay of this conference call will be available on ARMOUR's website shortly and will continue for one year.

ARMOUR's Q4 GAAP net income was $94.1 million, or $2.46 per common share. Core earnings were $30 million, or $0.71 per common share. That represents an annualized return on equity of 9.8% based on book value at the beginning of the quarter.

Differences between GAAP and core income are mostly due to the treatment of TBA drop income and unrealized gains on our interest rate contracts. ARMOUR does not use hedge accounting for GAAP reporting, and fluctuations in the fair value of our open interest rate swaps is a dominant factor in GAAP income, while the inversely related mark-to-market on our agency securities flows directly into shareholder's equity.

We paid dividends of $0.22 per common share during each month of Q4, for a total of $24.3 million, or $0.66 per common share. We also paid common dividends of $0.19 per share in January 2017 and have announced February common dividends of $0.19 per share to shareholders of record on February 15, 2017, payable on February 27, 2017.

At December 31, 2016, ARMOUR's book value was $24.39 per common share, down 12.5% over the quarter. As a reminder, we include updated estimates of book value per share in our monthly Company Updates, available on our website, www.armourreit.com.

Our December monthly Company Update described our plans to prepare future updates based on portfolio and other information as of the end of the first two months of each quarter and post them to our website after our regular monthly closing process is complete. Yesterday we posted the monthly Company Update with January 31, 2017 information, including estimated book value of $24.59 per common share.

Our next monthly Company Update will be based on February 28, 2017 information and should be posted in mid-March. The first quarter 10-Q will serve as our March 31, 2017 monthly Company Update, and therefore we are not planning a separate presentation.

ARMOUR's quarter-end portfolio consisted of over $6.5 billion of agency securities, plus another $2.85 billion of agency TBA positions on a gross basis. Our non-agency positions increased slightly this quarter, to $1.1 billion at December 31.

The majority of the purchases have been credit risk transfer securities and NPL/RPL deals. We have a number of counterparties expressing strong interest in financing our non-agency positions. During the quarter we reduced our agency interest-only securities in our portfolio to $33.6 million.

We've continued to reduce ARMOUR's interest rate hedge positions to $4.2 billion of notional coverage at the end of December, as the borrowings under repurchase agreements came down. The decline in repo balances is a natural consequence of taking TBA positions and diversifying the portfolio with non-agency securities.

Now let me turn the call over to Co-Chief Executive Officers Jeff Zimmer and Scott Ulm to discuss ARMOUR's portfolio position and current strategy.

Thanks, Jim. Good afternoon.

In addition to the customary SEC filings, we also continue to provide a Company Update, which is furnished to the SEC and available on EDGAR as well as our website, www.armourreit.com. The Company Update contains a considerable amount of information about our portfolio, our hedge book and our repo financing book.

These Company Updates, along with the comments we made during our last earnings call, provide our shareholders and analysts with substantial information on the state of the Company. Thus, the annual financial report that we filed last night should contain very few surprises for any of our listeners today.

ARMOUR realized total shareholder return for 2016 of 15.2% despite the volatility during the fourth quarter. Additionally, core income exceeded dividends declared and paid for the second quarter in a row.

Our book value declined by 12.5% for the quarter because of the rapid increase in rates after the presidential election. In our update for January, released last night, our book value on January 31 was $24.59, up 0.9% so far this year.

Our net balance sheet duration as of January 31 was 1.23.

The prepayment rate on our agency assets rose modestly during the quarter, from 10.93% CPR in the third quarter to 11.15% CPR in the fourth quarter. With the substantial change in the rate environment since the election, we expect that prepayment rates will decline and reduce amortization expense over the next few months. Accordingly, the CPR rate for January 2017 was 10.7% CPR and February 2017 was 7.3% CPR.

Please note that a majority of our agency portfolio is composed of assets with prepayment protection through lower loan balances or contractual prepayment lockouts.

Our notional swap position has been reduced from $4.9 billion on September 30 to $4.2 billion at the end of quarter four. We have mitigated rate risk while reducing swaps by changes in our portfolio composition, primarily by reducing 30-year agency positions in favor of 15 years.

Repo financing remains consistent and reasonably priced for our business plan. ARMOUR maintains MRAs with 42 counterparties and is currently active with 28 of those, for total financing of $6.8 billion at the end of the fourth quarter. We carefully analyze opportunities for longer term financing and will add this to our book as it proves attractive.

The agency portfolio is comprised of six major components, not including TBA dollar rolls. As of December 31, 50.6% of our agency portfolio consisted of 15-year final maturity pass-throughs with a weighted average seasoning of 47 months. 62.9% of those 15-year pass-throughs have loan balances less than $175,000. These are great convex assets for us.

22% of our agency portfolio was comprised of Fannie Mae multifamily bonds, or DUS, stands for Delegated Underwriting and Servicing bonds. The DUS bonds we purchased are generally locked out from prepayments for the first 9 and a half years of their 10-year expected maturity.

Any prepayment penalties received due to early payoffs can enhance the yield on the bonds. At the end of the fourth quarter our DUS bonds had a weighted average maturity to the end of the lockout period of 7.6 years.

The bullet-like maturity of these assets means they roll down the curve over time, much like a corporate bond or Treasury note, providing great potential to trade tighter, particularly as they approach benchmark areas like the five-year and seven-year Treasury notes.

18% of our agency portfolio was comprised of 30-year maturity fixed rates, currently maturing between 241 and 360 months; 88.1% of which have $875,000 loan balance or less. 5.4% of our agency portfolio consisted of 20-year fixed-rate assets, maturing between 181 months and 240 months, with a weighted average seasoning of 117 months. The seasoning provides great convexity performance, as well.

Our 10-year final maturity or shorter agency assets represent 2.2% of our portfolio. Our $83.6 million ARM position, 1.3% of our total agency assets, has a weighted average reset of 11 months.

$33.6 million of our agency portfolio was in interest-only securities. Late in the second quarter of 2016 we started purchasing very seasoned high-coupon I/O securities. These act as interest rate hedges and can have positive carry.

At the end of the fourth quarter we had dollar rolls with a net notional value of $2.1 billion. We continue to see certain TBA dollar rolls at very attractive levels versus owning and financing bonds. We actively monitor the attractiveness of risk and return in dollar rolls and may increase or decrease this position, depending on market conditions.

We believe that our current investments in non-agency assets will provide attractive and stable returns going forward, enable us to operate at a lower leverage multiple and reduce the risks associated with swaps.

We currently see our allocation of non-agencies as approximately 45% of our capital base. We define that equity allocation as the percentage of our equity tied up in haircuts for repo.

While gross portfolio allocations will show a much larger quantum of agencies on our balance sheet, we think the purest way to think about equity capital allocation is the amount committed to financing haircuts in each sector. Equity is not tied up in financing haircuts, is our liquidity, and that liquidity is available to support any part of the portfolio.

We began accumulating non-agency assets in the first quarter of 2016 with a focus on credit risk transfer, or CRT, securities, a position which was valued at $820.3 million at year end. We have been rewarded both by the spread tightening that has occurred in this sector over the last three quarters and by the attractive carry.

Our weighted average CRT coupon as of the end of the fourth quarter was 5.26%, with a weighted average margin of 4.49%. The weighted average purchase price of our CRT positions is 98.3. As of January 31 of this year, the weighted average market price of our CRT position was 107.86.

In the CRT transactions we take the credit risks of recent Fannie and Freddie underwriting in return for an uncapped floating rate coupon. We continue to believe that housing trends and strong mortgage underwriting will give a robust underpinning to the credit quality of the CRT bonds.

As of December 31, ARMOUR owned 122.8 million of non-agency NPL/RPL securities, nonperforming and reperforming securities. We did not add any of this asset class to our portfolio in the fourth quarter, as spreads tightened in the sector to a level that could not provide us with the levered yield available in other investment opportunities.

At the end of the fourth quarter ARMOUR owned 97.2 million of non-agency legacy MBS. Today we see very few opportunities for investment in the 2008 and prior non-agency MBS asset classes. However, existing assets from that period continue to perform well as they run off.

Like many market participants we continue to hope for a revival in the jumbo securitization market. While we owned more significant amounts in the past, our new-issue jumbo MBS exposure is now only $10.9 million.

Please note that some of our positions have changed since year end, and you can study our asset and liability positions on the Company Update which was released last night, February 15.

Post-election we have entered a new period of volatility. The US economy continues to make steady progress, but headwinds remain in the form of continuing weak economies abroad as well as our domestic challenges with productivity and a strong dollar.

The shape of the new administration's policy remains unclear, with potential positive as well as negative effects on our operations. We have positioned the portfolio and our hedging to reflect this assessment of heightened risks while still allowing us to earn our dividend, which we strongly feel is sustainable in the current environment.

Operator, that concludes our prepared remarks. We'll now take any questions.

Thank you.

(Operator Instructions)

Doug Harter, Credit Suisse.

This is actually Josh on for Doug. I wanted to talk a little bit about the core EPS number. So core income was up quarter over quarter even though average NIM was down, the portfolio seemed to shrink, a little bit less drop income, and I didn't see any meaningful change in CPR. So I was just curious what other dynamics are there that I'm missing that led to the core EPS performance. Thanks.

As you have a smaller portfolio, you also have lower interest expense. We also benefited in the area of some DUS prepayments, which brought some extra fees in that, although not anticipated, we've been getting those regularly over a number of periods.

So on four or five different accounts we just performed a little bit better than internal expectations, and those expectations are the expectations that we relayed to the investment community. So four or five cylinders worked, just a little bit at a time.

Great. Appreciate the color, guys.

Thank you very much for calling in.

(Operator Instructions)

Trevor Cranston, JMP Securities.

A couple questions on the risk transfer investments you guys have been making. I guess first can you say what part of the capital stack you guys have been particularly focused on and generally whether or not you've been more focused on newer issuance deals versus older.

And then the second part of the question is with the magnitude of spread tightening we've seen in that space, which I guess has continued into early 2017, do you still find that bucket to be an attractive place to continue to deploy capital versus other credit investments or the agency side? Thanks.

Well, thanks for calling in. So we own strictly the M2 classes in the Fannie Mae's and in the Freddie the equivalent of that. We bought issues in 2014, 2015 and 2016 issues, some of it in the secondary trading, some of it in the primary issuance market directly from the agencies.

When we first started buying those assets when we were doing the Javelin acquisition was late January or early February a year ago, and spreads, as you may recall, were 625 to 650 off. So we used that opportunity to enter very slowly and over the period of the next two and a half quarters put together the majority of our purchase program.

Now, if you look back at the credit on the group of assets that we own, I believe currently a third of them -- now, these were all unrated when they started -- currently a third of them now have some sort of rating, so they're getting ratings as they improve. But most noteworthy is the fact that the house price depreciation numbers keep going up year after year.

So if you look at any 2014 or 2015 originated paper, you have embedded in that paper 8 to 12 months old, because when they do a deal in 2015 that's 8 to 12 months old, so if you have a 2014 deal that represents stuff from 2013 or even potentially before that. So you have home price appreciation on a lot of these assets of 8% to 12%, even maybe 14%, depending on how old it is.

So the question is what do we think about today? Spreads are very tight, and a lot of the opportunity in that asset class at these tight spreads I think has been -- is wrung out.

So we have not purchased anything recently, and certainly not in 2017, and for us to buy some more I think we'd need to see wider spreads. That being said, the income is very good on that asset class for us, and we at this point do not intend to sell it.

Okay, got it. That's helpful. And on the agency portfolio, there's obviously been increased chatter about the Fed potentially starting to shrink its balance sheet at some time in the not-too-distant future. Can you guys comment on sort of how you're thinking about spread risk and the agency portfolio and how attractive agencies are given that backdrop?

If we'd have had this call two and a half years ago, I would've told you that the dollar value of an 01 on our entire balance sheet for spread risk was $9 million. That is down to closer to $4.7 million today. So we've reduced our exposure to spread risk, and we've done this intentionally.

Our rate DV01, which at one point was $5 million of a basis point move, is now down to $1.1 million. So in two areas we've reduced exposure to risk, meaning spread risk by half, and rate risk by almost two-thirds, as a matter of fact.

We do believe that there will be a point in time which they say we're not reinvesting. It will be well advertised. The market at one point six weeks ago was talking in the second half of 2017.

Us as a group of management does not believe it would happen until 2018, and in that regard even the second half of 2018. We think they'll be very, very slow. We can't envision a point at this hour when we would see them actually selling the assets that they have.

Now, in regard to the next step of risk management, as Scott said earlier, we've taken over a period of time in the fourth quarter, not the middle or the beginning or the end or just a program we were on, to sell some 30 years as they got rich and go into 15 years. For example, yesterday we did the February-March 15-year roll, or maybe that was two days ago, and we're looking at what we believe is a 13% return on that, a levered return based on our leverage numbers. So there are still good opportunities in dollar rolls.

And in turn 30 years have widened out so much on an OAS basis that there's good opportunities there on a levered basis to create the kind of net 10% returns that the market is delivering right now. However, we're not going to go there yet, and we'd want to see some more degradation to that sector before we'd take the risk to enter into it.

So I hope that addresses your question.

Yes. Very helpful. Thank you.

You bet.

(Operator Instructions)

Brock Vandervliet, Nomura Securities International, Inc.

I just had a quick one on the expense base. It looked like expenses were quite a bit lower than what we were looking for. This may tie into the first question on earnings power. Thanks.

I'm going to let Jim answer that question. Jim Mountain is our Chief Financial Officer.

Sure. And I think that if you do a deep dive on expenses you'll see that the biggest savings is in the outside professional, and so that's lawyers and the like. I would say with regard to the lawsuit about the old Javelin transaction, we're getting towards the end of that.

While we're still spending a little bit of money on it, we certainly spent a lot less during the fourth quarter than we had earlier on, and the things that we're spending now are things that were covered by the accrual in the purchase accounting of Javelin. So that's where we've seen some of that savings.

And I would note on a broader scale that there were four or five transactions that took place last January, February, March in our space that you're familiar with, and every single one of those has some litigant trying to stick his or her nose in to grab some money out of somebody's pockets. We're confident in our position, and although it's not conclusive yet, we do believe that around the corner we hope to announce that we've moved on from that issue.

Okay. And as far as total expenses go here, no other callouts that what we see in the fourth quarter is sustainable as a level going forward?

Well, there are always going to be some seasonal fluctuations. While we try to provide for things like annual meeting expenses as much in advance as we can, some of those come in at the time, and so that sort of stuff will gyrate a little bit back from quarter to quarter. But I think as a broad matter your observation that we're on a sustainable expense base is probably pretty close to right.

Okay. Great. Thank you.

Thank you.

Jim Delisle, Wasatch Advisors.

What is the -- in the M2 or the M2-equivalent tranches of CRT that you own, what is the loss spectrum that they absorb?

(Multiple speakers) between 50 basis points to 100 basis points of credit support.

Initial credit support.

Initial credit support, yes.

And as prepayments increase, that credit support can grow, and that's why I believe a third of our portfolios already have rated securities now that were originally unrated securities.

So if I were to look kind of is there a hard and fast number that we could use, rule of thumb, that the attachment point now on average in your portfolio was say, at 80 basis points, or 1.2% of loss, or some number like that?

It's going to be somewhere around that 80 basis points. I think most of our securities have 100 basis points of credit support. If you remember, some of the older securities had -- they weren't actual loss securities. They had a formula for losses. So they were a little different. So it's not apples to apples.

Okay. And what is your average expected maturity on your credit CRTs?

Well, the underlying is usually 30-year securities. Now, it depends on the prepays where the actual average lives are. They're going to be six- to eight-year securities, most likely. Some of them already have the window open where we've already started receiving principle payments on them, which is much earlier than expected, because the speeds are so fast.

However, that'll be driven a little bit by rate factors and the economy. If you have a three and a quarter 10-year note you'll see those prepayments slow down. I suspect the Federal Reserve wouldn't raise rates if they thought that we were going to have a housing disaster. So we might actually benefit in both regards there, whilst the first loss pieces are still growing.

Which I guess is why you treat it with a negative effective duration in your agency/non-agency portfolio update as of January 31, right, Jeff?

Well, sorry, this is Mark. The reason that is that is because of the high dollar price and the nature of the securities. There's a big I/O component to that, and that just causes the duration to be negative.

And you mentioned earlier that you kind of for modeling purposes you use the haircut to kind of determine the application of your equity for leverage for risk purposes. What kind of debt to equity or haircut should we look at the CRT portfolio as having?

Haircuts are in the 20% to 25% range.

So saying that you'd basically for risk modeling you'd be running at, say, three, three and a half times debt to equity.

Well, for -- we look at -- we use the haircuts for leverage return purposes, to see where those returns come out. Risk modeling is a little different in that we look at the portfolio of what we own and run scenarios and stresses on it. So it's a little different model than just how much can we lever it and see what happens.

Right, but I would look at it at three to three and a half, because that's exactly the range we look at. But let me say something a little more broad. We don't look at how our position is levered for agency and non-agency, because you have to have a liquidity pool to support those different asset classes as the prices change on them.

So we look at our total liquidity and how it supports both asset classes. And so, and we've had this discussion with analysts on a one-on-one basis as well as on these conference calls, we look at the total leverage of the portfolio. We will do, as you suggested, an internal of how an asset class would respond to certain leverages, understanding, though, we have to keep a lot of liquidity around to support that.

All right. And this is much more of a philosophical question. You've always been an agency, classic agency REIT. Now increasingly you're moving more towards what we'd call a hybrid REIT. Where would you say you are? How would you define yourself? And how should we think in those terms?

We are a hybrid REIT. Of our haircuts, 45% of the equity in haircuts is in non-agencies right now. So that would definitely define us as a hybrid REIT.

If CRTs were still 450 or 500 off, we would find a way to own more of those and still operate within the rules that the SEC and the IRS have allotted us. So we like that asset class.

There are other asset classes. I know some of our peers have been buying more NPLS and RPLs, and we haven't seen them as attractive.

If you have concerns like we do about some potential spread widening down the road and about some opportunity for rates to go up, increasing vast exposure to the agency asset class right now would not be our goal.

Great, guys. Thank you very much for the candor.

All right. Thank you very much.

And, gentlemen, there are no further questions. I'll turn the call back over to you to continue your presentation or closing remarks.

Thank you very much, everybody, for calling in. If you have any questions, please feel free to call Mark, Jim, Scott or myself. We're at the office. We hope to get back to you in the same day. And with that have a very good afternoon.

Ladies and gentlemen, that does conclude the conference call for today. We thank you for your participation and ask that you please disconnect your line.