Gary Kain - President & CIO
Cheryl Pate - Morgan Stanley
American Capital Agency Corp. (AGNC) Morgan Stanley Financials Conference Call June 12, 2012 1:50 PM ET
Cheryl Pate - Morgan Stanley
Hey, good afternoon. I am Cheryl Pate, the Specialty Finance and Mortgage REIT Analyst here at Morgan Stanley. And it’s my pleasure today to welcome American Capital Agency Corp. AGNC was formed in 2008 as an agency mortgage REIT and has one of the most sophisticated asset selection and hedging strategies in our view. The company has been able to sustain high teen dividend yields and ROEs driven by lower prepayment exposure and AGNC’s focus on preserving book value.
Here to present for AGNC today is Gary Kain, Chief Investment Officer. Prior to joining AGNC, Gary served in several investment management roles; Freddie Mac from 2001 through 2009, including most recently Senior Vice President of Investments and Capital Markets and Senior Vice President of Mortgage Investments and Structuring of Freddie Mac.
So with that, I would like to introduce Gary. Thank you.
Thank you and thanks everybody for your interest in AGNC. Quickly, for those that aren’t familiar with the company, as you can see on slide three AGNC is an agency mortgage REIT as such we only invest in securities backed by either Freddie Mac, Fannie Mae or Ginnie Mae and so since all three of those entities have some sort of backing from the US government credit risk in the portfolio is minimal.
But I want to be clear, while credit risk is minimal, there clearly are you know there are the risks we do manage in the portfolio are prepayment risk and interest rate risk in addition to as a levered investor just liquidity risk as well. And there are host of other factors, but those three kind of stand-out as the key things to keep in mind.
Just as a high level, if you look on the right side of the slide, what you will see is an agency went public essentially four years ago, May of 2008 at a price of $20 per share and then since then we have paid $20.11 in cash dividends and that’s prior to the dividend declaration that we announced last night. So if you bought the stock four years ago in the IPO, you’ve now received more dividends, more cash back and the purchase price.
But importantly, like dividends are really important; they are obviously the big driver of the space. But the other thing to keep in mind is that and we will talk more about this book value is also very relevant and despite, in addition to paying the dividends AGNC has been able to grow book value from $20 per share at the IPO to $29.06 and it has also grown its market capitalization from $300 million at the IPO to over $9 billion where it stands currently. So the larger market cap has greatly improved the liquidity of the stock and it has added significant efficiencies to the operations as well.
Now if you look on slide four what you see is kind of what we view are the most important kind of -- what’s the most important return metric for AGNC and really for mortgage-REITs in general. And this kind of measure, economic returns which we show on the top of the slide shows the combination of dividends and change in NAV or book value.
And again, we want to stress that it is the combination of these two that are really important when it comes to producing total returns to shareholders and the reason is that if we pay you a dividend, but in that process our book value is going down or NAV is going down, clearly that dividend is coming at a cost and all in value creation is going to be a lot smaller than if you are getting paid dividend and you’re seeing appreciation with respect to book value.
So if you look at the slide what you can see is the year-over-year, since 2009 AGNC has produced very attractive mark-to-market returns. Again, our book value is a mark-to-market number and the AGNC mortgage market is relatively liquid. These are obscure assets that are difficult to mark. And we compare those to our peers, the other agency REITs and if you look across the last three years and so far in Q1 in 2012, what you see is significant out performance on the part of AGNC versus its peer group both in terms of dividends paid and in terms of book value growth.
And many people wonder that, okay, they can understand that in sort of in markets where our products are more esoteric or very different, I think to some people this kind of differential performance in a space such as agency mortgages that are all guaranteed essentially by the government, it’s surprising.
And our mindset with respect to that is again, going back to what I said earlier, is while they are all government guaranteed mortgages, what makes a mortgage different? What makes a mortgage different is that it can prepay and so managing prepayment risk and how prepayment risk is correlated with interest rates is extremely important. That can have massive impacts on returns and you can see this on the next slide. This is a very simple illustration in a sense why we obsess so much about kind of prepayment performance on the underlying mortgage assets.
If you look at the, you know if you look at the 10 CPR column, you can see the asset yield at a slower prepayment speed. So 10 CPR assumes -- is a measure that assumes 10% of the borrowers in a given pool refinance on an annual basis. On the far right where you see 40 CPRs, obviously 40% of the borrowers refinance on an annual basis. And since these securities in this case are 30 year, 4% fixed rate mortgage, since they are all currently at premiums, what you can see is that you don't want faster speeds and that really erodes your yield.
But what's more kind of relevant is when you look down and you assume kind of a cost of funds and includes both borrowing, you know, repo borrowing on the security and your hedges which are interest-rate swaps, when you assume the cost of funds, look at what happens to the net margin at different prepayment speeds.
And then apply in the last line kind of let's turn this into a hypothetical REIT with A times leverage and here we are looking at gross ROE; and what you can see is that if you can get very slow prepayments with 10 CPR, you can have a gross ROE in the low 20s. Whereas if that prepayment speed increases then that gross ROE goes to the higher single digits or you know 8.27% and its even worse at you know 40 CPR. And so when you look at this slide, I think it makes it a lot clearer how you can get those performance differences that we saw on the prior slide.
Now the next question to kind of along these same lines, is okay, prepayments clearly matter to all in returns. I could see why that’s critical, but do these things really prepay that differently and/or are they all pretty close together? And the answer is, they perform very differently and you can kind of see that in this graph. And here what we’re looking at are 30 year 4.5% coupon mortgages, guaranteed by Fannie Mae. They were originated; all of these were originated in 2010 under the same underwriting guidelines.
The topline is TBA or generic mortgage securities and as you can see both in late 2010 and then in late 2011, when interest rates hit local lows, instant prepayments picked up dramatically on the more generic mortgages. So you see prepayments in the 25 to 30 type CPR area over I’ll say the last six months. But look at the other two lines.
Now these are also 4.5% coupons originated by Fannie Mae during the same time period, but the difference is that the grace lines are low in balance mortgages. So these are mortgages with average loan sizes between lets say $85,000 and $110,000. And why is it that smaller loans prepay so much slower and are not as responsive to interest rates? Well, it’s kind of straight forward in that, if you think about fixed cost to refinance, if you look at small loan and you are going to refinance into a lower rate and save $50 a month let's say in your monthly payment, it takes a long time to make up for $1,000 or $1,500 worth of fixed cost to refinance. And so borrowers with smaller loans need a much greater incentive or they have to have a much lower rate relative to their existing rate to make it worthwhile to refinance.
But in today's environment there's even a kind of bigger advantage to owning securities backed by smaller loans and what that is, is that these loans or the protection embedded in these loans actually is amplified. The more overall refinancing activity there is on the market and why is that, well, let's take a step back and think about it not from the borrower’s perspective, but from the lender’s perspective.
So let's think about Wells Fargo or JPMorgan or Citi who is operating in an environment like the one we are entering now where rates have come back down, where they are operating near capacity; maybe particular branch can process a 100 loans or a particular loan officer can process a 100 loans. So they get paid a percentage of the loan amount and its how people are compensated in the business both the firm and individuals; so if you are paid 2% on a $100,000 loan, that's $2,000. If you are paid 2% on a $500,000 loan, that's $10,000. The work is the same to process both of those mortgages and in one case you get paid five times as much.
And if you are constrained by capacity which loans are going to get pushed through the system quickly and which loans are not. And what's interesting again, is that if we go into an environment with a QE3 we are ready at again new record lows in mortgage rates. In environment like what we are entering into this kind of protection becomes even more important.
Now let's go to the line below it and those are securities that have already, or loans that have already gone through the either a HARP 1, the first HARP program or HARP 2.0 the program that was kind of, the program that was revamped at the end of last year.
And what's interesting about these securities is that these are higher LTV borrowers that took out loans maybe in 2007. They are now underwater; maybe they have a 100 LTV or a 110 LTV. They are allowed if they are current on their loan, if they are employed then and if their loan was owned by Freddie or Fannie then they are allowed to refinance that loan one time into a lower rate.
But it’s a one time option. So once they go through that program, their loan origination date is past the cutoff and they can't go through that program again. And so if you think about this borrower, its very logical that prepayment speeds are negligible and as you could see on the chart have been consistently kind of around 5 CPR.
So why is that the case, well, now they've gone through this program, they can't go through again, they have a 100 LTV. They can't take out any other loan. They really don't have refinancing options and the question we get is well, they are high LTVs don't you worry about them all defaulting as you could see they are actually not, but think about the borrower.
These were GSE borrowers who have still even though they have been underwater, been current on their loan for let's say five years, they are still employed and now their monthly payment just went down quite a bit. So it's actually pretty logical that you are not seeing a lot of default activity and in the case of owning a Fannie or Freddie security that translates to the loans being pulled out of the pour prepayment to us and that’s why they have been so well behaved.
But take this back and think about these prepayments versus that chart we saw on the other page which is different types of securities have very, very different prepayment characteristics and those different prepayment characteristics translates into very different ROE potential.
Now let’s go to another sector and this is kind of the other side of HARP 2.0. We talked about the loans that the securities had it gone through the HARP program. These are when the GSEs and FHFA revamp the HARP program at the end of 2011, we’ve talked about this a number of time since then and even before that on our third quarter 2011 earnings call and what we told investors at that point was less than 5% of our portfolio was backed by loans that were eligible to go through HARP 2.0 and there is a cut off date made 2009 that if the loan was originated before that then it’s eligible to go through the program.
Well less than 5% of our portfolio was in that camp and what we said to people was we don’t think, we think HARP 2.0 is going to be real and it’s not an area that we want to have exposure to over the six months. And because there was a lot of days in required with respect to the HARP program, speeds didn’t pick up immediately and that led a lot of people to believe that HARP 2.0 was going to be another dud like plenty of other programs that have not lived up to expectations.
But let's look at these feeds on kind of three vintages, 2007, originations of 30-year fixed rate Fannie Mae securities, the blue line being 30-year 5% than 5.5% and then 6% from the bottom.
So even when rates hit their lows and November and December and during 2011, speeds before this program kicked in, were in the say around the mid-25s. Okay, not too big of a deal and a lot of people would talk about the fact that any of these loans from 2007, they’re underwater. If they could have gone through a hard program, if they had any intentions of refinancing, they would have.
Well, now let's look at where we are fast forward to May, and the last month’s prepayment fees and now the 5s and 5.5s are well in to the 40s and I think now if you talk to any mortgage strategist or researchers, most expect that line to continue up for at least another couple more months.
You know, if we get a QE3 or if we stay in a very low rate environment, which is very real, it would be not be surprising to see those in the 50s. So what we want, the point here is that agency mortgages are very different, in different interest rate environments or different kind of environment with respect to GSEs. Different types of mortgages can perform very well and in other times, even if you bought something a year ago that you thought was right for that time, it may not be going forward and what you own can have a really kind of big impact on your returns.
And this is an example of something that was very safe to own both in 2010 and throughout most of 2011, but it's no longer in that category. And again, the way AGNC deals with those kind of situations, we had a fair amount of this, what's called season paper in the mortgage market in 2010 and 2011, now its less than 5% of our portfolio and it was you know and it's not something we have done after the speeds, it's something we did six or you know nine months before you know these things are evident.
So turning to slide eight, what you see here I'm not going to go through this in detail, this just gives an overview of our portfolio at the end of March. I think the key thing is its largely fixed rate with roughly half of the portfolio being a 30 year fixed. There is also some 20 year fixed and then 15 year fixed makes up 35ish percent of the portfolio.
But importantly, we give you and have for a number of quarters right now, the breakdown within the two big sectors, 15 year and 30 year. You can see the percentages that are backed by lower loan balance mortgages or the securities that have gone through the HARP program, you can see those prepayment, actual prepayment speeds after as of April. The other thing on top which is really important and think back to the slide we went over at the beginning, just prepayment speeds on agencies portfolio.
And so if you look over the last seven or eight months, we are balancing around, around 10 CPR. Okay, I mean a couple of months a little higher you know a few more months a little lower and you know that's critical and that doesn't happen accidentally, it happens while being you know very diligent about making sure you both own the best performing mortgages, but probably as importantly that you avoid the worst performing mortgages.
And so moving on to the next slide, I think this is kind of the timely discussion today and we actually went through this same slide and that’s not changed from on our earnings call, which was a little over a month ago and what's interesting at the time, we said that we thought the most likely scenario was scenario one and scenario one was one where there was no QE3, where we kind of growth modeled along, the economy was recovering enough to keep the Fed on hold.
We probably saw interest rates in that environment going at maybe around 2.5% on the tenure but the yield curve steepens. This would have been a Goldilock scenario for the REIT space, why? The yield curve the steeper better for spreads; prepayments are slower because interest rates are higher. That’s all good. Returns on new purchases are going to be attractive in that environment and generally speaking, I think the entire space does very well and exactly what type of mortgage you have, not really going to be that important.
I am going to skip to the third scenario for a second and the third scenario is one where there is a quick upward spike in interest rates and what we said here and what we still believe here is this is a low probability right now. For obvious reasons in terms of global rate environment, central bank, communication, where inflation is, where unemployment is, but look we can’t say that a shock can happen and it can happen quickly.
And so, the reality is, while this is a very low probability, it’s on this page because it’s generally a very bad outcome for the space and for a levered mortgage portfolio because prices would drop, they would drop materially, book value will be under pressure and what's going to be critical here is how you hedge the portfolio and I am not going to go into that in today's discussion but if you look back at our earnings presentation, we talked a lot about our swap portfolio, our purchases are out of the money options and a lot of the tools we use to provide some, you know insurance and to reduce some of book value to clients that would otherwise occur in that scenario.
But now let's focus on the middle column which we said was kind of a medium likelihood again a month or month and a half ago but right now I think in most people’s minds the likelihood of this is going up quite a bit. We are already kind of heading down the path of lower interest rates, you know a weaker economy that is starting to feel like the strength we were seeing in the first quarter was a head thick.
We are obviously seeing weakness in Europe and the odds of a QE3 are clearly going up. I think some street firms have projections as like Morgan Stanley up near 70% to 80%, others are maybe on the low end more like 50-50, we are probably in between there but the reality is that this scenario is no a Goldilock scenario for the space. Why is it not a Goldilock scenario? Well, in an environment where the Fed comes in and buys a significant amount of agency mortgages they are going to do a couple of things.
They are going to drive the mortgage rate lower and yes there's protection, some protection due to capacity constraints in the system and other things but those won't last forever. But it's an environment where prepayment speeds across the industry are going to go up. The other thing is that new purchases as you are trying to purchase, you know if you are as you are getting prepayments or if you are, as you are reinvesting capital in that environment, you are going to be competing with a massive bid from the Fed and your ROEs on your new purchases are going to be worse because of that. You're also going to be in a flatter yield curve environment. When you put those things together, none of those are positives.
But there is other perspective on QE3, and it's very dependent on what types of mortgages you have. And some investors walk in to one-on-one meetings and they say, I'm here because I think this is a great place to be, this is my QE3 play. Okay, and you look at them and you explain the issues around prepayments. But on the other hand, actually a lot of what they're saying is absolutely accurate.
We're a levered holder of agency mortgages. The fed is going to come in and drive the prices of what -- again, it's very dependent on what you own and I'll come back to that -- but drive the prices of a number of mortgage securities up a lot. And so you can, if you're very well positioned, turn what otherwise would be a very difficult scenario into very attractive total returns, especially given what's likely to be available in that kind of scenario.
But let's come back to the benefits, right. So the benefits are -- are all mortgage prices going to go up a lot? No. Are they going to all move together? Not in our opinion. What we would expect to see is the Fed be active in purchasing the lowest coupon fixed rate mortgage securities, both in 30-year and in 15-year. And why do they -- that's all they are currently buying. That's all they've bought in the past.
They don't buy ARMs. They don't tend to buy higher coupons or middle coupons because they're trying to move the mortgage rate lower. That's what's going to stimulate the economy, and that's also where all the supply and all the liquidity is. So that's where they're going to concentrate their efforts. So if you have the lowest coupon fixed rate, you'll benefit from prices going up and you'll get book value gains.
The other thing if you have slower -- if you have prepayment protected or securities that won't speed up like the HARP securities or the low loan balance, if those stories hold up, which we absolutely believe they would, then you can expect to see the prices of those go up dramatically and also you don't get the negatives. You're not going to see faster prepayments. You're -- or significantly faster prepayments.
And you're not going to have to reinvest as quickly into lower yielding, low ROE investments. So how are you positioned in the QE in this scenario, if it occurs, is going to be critical to your total returns. And so then, if you go to the last page and then I'll wrap this up, what we told you, and again, this is an updated slide as of today. This is the slide we used on the 331 portfolio and talked about in early May.
What we stressed was our thought process around asset selection was maintain a prepayment protected portfolio. Our generic securities where going to be in the lowest coupon fixed rate mortgages, which could be sold to the Fed at rich levels if a QE3 occurs. We were also going to kind of reduce our exposure to the highest premium mortgages where the lower loan balancing and HARP securities that were trading at very high premiums to generic mortgages.
We were reducing the coupon on those to reduce some of the risk that those premiums could kind of be at risk under some other scenarios. And then again, avoid HARP 2.0 for all the reasons we've talked about. And then lastly, I just want to touch on leverage.
Which is the other thing we said on the call was that while we closed the quarter at 8.4 times leverage, which was on the higher end of where we've operated over the last two or three years, we said we have brought that down because valuations and mortgages within -- in April had been very strong and we felt that the risk return was much more two-sided. So we've brought that down some.
But one thing that we stressed was we were not going to run -- we were very reluctant to be at low leverage, going into in a scenario where there was a reasonable probability of a QE3. Why? Because when you're underweight or when you have low leverage, that means that you're kind of saying, I'm going to buy later.
Yes, you could potentially maintain that for a long period of time, but you're still, especially if you have faster securities, you're going to be doing a lot of reinvesting. And you're going to be competing with a multiple hundred billion dollar program from the Fed. That's the last time -- the last thing we want to do is wait till the Fed starts buying, to then kind of increase our purchases of mortgages.
We'd rather, if anything, reduce our purchases going forward if the Fed drives mortgage prices to kind of extreme levels. So for that reason, we weren't comfortable in running with kind of low leverage in that kind of environment.
So with that, let me you know stop and open up the floor to questions.
Cheryl Pate - Morgan Stanley
I think in the interest of time we probably have time for maybe one question. Any from the floor? I guess just from my perspective, one issue. We didn't really touch upon, in the presentation, on the financing side. You recently did a preferred offering. How are you thinking about the duration, if you will, of the financing side? And are you looking to extend the repo going forward as well and actually the hedging strategy as well?
Those are very good questions. And so, I think they cover a couple different areas. I'll start with the repo side. I think what you'll see if you look back on AGNC's presentation, we give you our weighted average, repo maturities and average terms of repo. We have been pretty aggressive about trying to extend our repo and reduce, what we call, roll-over risk. While we feel the repo market is stable and solid, it's not limitless and it doesn't grow on trees.
And in general, having longer repo, three months instead of one month and six months instead of three months and moving out, is definitely advisable and it's worth paying five or so basis points more to do so. So that's one thing. With respect to kind of preferred equity versus common, we do view the preferred issuance that we did, which was relatively small at the end of March, as helping in a sense provide longer-term funding.
But we view it as also kind of making sure, as any frequent issuer in the equity markets, you really want to know what all your options are. And in general, we do believe that if you're in a position to issue common equity at a premium to book, if you're going to go to the equity markets, you're generally better served doing most of your equity issuance in common space. But it's important at times and environments change, and you want to know what your options are.
So I think what we would say first and foremost is we think that the returns issuing a preferred, a perpetual preferred that's callable in five years at an 8% -- with an 8% coupon, when you're producing mid-teens returns, is obviously accretive to common shareholders and the economics makes sense. But they also give you a benchmark by which you can kind of watch and where you can't operate against down the road if conditions change.
And then lastly, with respect to the hedging piece, I think it's really important. And again, we go into this in a lot of detail in the earnings presentation. And I'd encourage investors concerned about hedging to go through that. One of the things that comes up a lot, you have a fixed rate portfolio of decent amount of 30 year. There's a lot of duration there. We agree, okay. But there are very simple instruments that have been around forever to hedge duration.
What's harder to hedge is change of duration and how it adjusts in different interest rate environments. And that's absolutely what's -- why it's so important to have reasonably predictable prepayments on your underlying assets.
We went over how we do that. But once you do that and if you can have reasonably predictable prepayments, then duration doesn't change as much as interest rates change, and that makes it more reasonable to be able to manage the interest rate risk embedded in the portfolio. But in addition to that, what we talked a lot about was our use of swaptions. And what a swaption is and the way we were using them is we were buying out-of-the-money put options essentially on interest rates.
And at the end of the first quarter, we had $10 billion in swaptions. And they provide significant disaster insurance in the case of an up 200-type basis point move in rates. Because look, those moves happen -- if they happen, those moves happen because no one expects them, and that's why they're fast and that's when they occur. So we're not telling investors we're not going to figure out four hours or four days or four weeks before one of those things is going to occur.
We might but the odds are against it. And for that reason, we are going to continuously make sure that both via kind of the positioning going in but also via active rebalancing, that our portfolio is positioned to kind of perform reasonably in those kind of scenarios.
Cheryl Pate - Morgan Stanley
Please join me in thanking Gary.
Thank you, guys.