Gary Kain - President and CIO
American Capital Agency Corp. (AGNC) JPMorgan SMid Cap Conference December 1, 2011 2:15 PM ET
Thank you everybody for coming. Very quickly I am introducing Gary Kain from American Capital Agency REIT, Gary go ahead.
Thank you guys and first of all I would like to thank JPMorgan for the opportunity to present at this conference. But I also would like to thank all of you for your interest in AGNC, you know there is really a lot we could focus on given the market between Europe and all the volatility we are seeing in the financial markets and I guess the overall concerns related to both the U.S. economy but more so the global economy.
But really what I want to focus on today is related to the low interest-rate environment that we are in and it's really the prepayment picture because in the absence of any real change to this environment which we don't see happening in the near term, prepayments are really going to dominate performance in the REIT space and it's really going to be a key driver of our returns.
So with that I will try to jump right in because we have a lot to cover today. So, first of all the current market landscape is one where again interest rates are at basically record lows, mortgage rates are at record lows. Borrowers can take out of a 30 year fixed-rate mortgage around 4%; they can take out 15 year mortgages around 3% are very low 3%, ARM mortgages are around 3% as well maybe a little below. The other thing that's really important to keep in mind is that most of the borrowers who took out loans over the last two to three years, two years since 2009 are good credit borrowers that qualified under stricter underwriting and so guess what they can refinance.
They will qualify, vast majority of them will qualify for new loans and what we're seeing is our prepayment speeds and we will look at this in a second prepared to look at this the second on new loans through credit mortgages are fast and they probably will continue to be so.
Now the enhancements to the HARP program that FHFA laid out a few weeks ago. They first kind of highlighted a month ago but then laid out in the middle of November. Definitely increased the risk around season’s mortgages and how they may perform because there is more incentive for the originators to refinance that population as the risk around reps and warranties, some of the fees are lower. There is a number of things that were done via that process that increased kind of the prepayment risk on young season’s vintage of mortgages.
So when you put those two pieces – those two bullet points together in a sense, what you have is the newer mortgages exposed because of they are good credit, the older, more seasoned mortgages largely exposed because of the insurance rates to be to the HARP program, so there is greater exposure there.
So the key point there is that prepayments are a clear and present danger for the space. Now when you layer on this the last bullet point relating to the Fed and the risk were the potential for Q3 and mortgages, now different people think the odds of a Q3 very but the Fed has kind of told you that if there is a Q3 it's going to be center of mortgages.
And so if that happens it's going to drive mortgage rates even lower than they would otherwise be and it will give a lot of attention to mortgage rates and to the fact that borrowers could be refinancing so all of these discussions we have, we are going to have around the prepayment picture will get even more important in an environment if the Fed does go down the path of a QE3 on the mortgage side.
So with that lets look at what's going on, on the prepayment front right now and these are our newer mortgages. These two graphs are both 15 year and 30 year mortgages that were originated in 2010 so on the left side we have 15 year, 4% coupon Fannie Mae mortgages and as you can see and this is the most recent data that’s been released. There is another prepayment release next week but this includes what was released in early November.
You can see that the blue line which are kind of generic or TVA 15 year for us got up to 35 CPRs. So these aren’t, these are good credit 15 year borrowers tend to be very good credit, high FICO scores, low LTVs, people that can afford a higher payment.
So they easily qualify and they tend to react to changes in interest rates pretty quickly. On the other hand if you notice a mean model which is apprised of loans with balances between $85,000 and a $110,000 so call it an average of near a $100,000 the loan size.
Those smaller loans repaid much slower, why? Two reasons, one is because the borrower, the fixed cost to refinance a loan whether it's an appraisal, whether it's courier fees, settlement fees, if you have a small loan those dollars you need a much bigger change in interest rate to make it worthwhile to pay those fixed costs.
There is another really important factor that goes way beyond that and that’s just simple economics for the originators and the loan officers that are involved in the process. Everyone knows that mortgage bankers are basically working at capacity; there is tons of refinanced business going on, because rates are at record lows.
If they can process a 100 loans are they going to process a 100 small loans when they get paid a percentage of the loan or they are going to process a 100 larger loans? Well the short answer is they are going to deal with the larger loans because they will make a lot more money and so first off the borrowers is not incentive to refinance if they have a smaller loan and second of all the originator isn’t incentive to deal with them while they're at capacity and so for those reasons you see this you know a 14 CPR on the same coupon, same year of origination, Fannie 15 year four versus the 32 or 35 that you are seeing for more generic mortgages.
Now take a look on the right side, at the 30 year mortgages, on first look at the blue line which last month paid at 28 CPR this is again TVA or generic 30 year 4.5, the orange line that came in at 9 CPR is smaller loan sizes in 30 year mortgages and again much slower, much better prepayment profile for the same reason.
Since more loans are not going to you to refinance as fast and we have seen this in the data for 20 years in mortgages. The best performer are these is the green line that came in at 5 CPR these are HARP securities.
So we just talked about the HARP program been expanded, well these are loans that are ready went through the HARP program. So they are high LTV loans that don't have another opportunity to refinance except the HARP program but once you've gone through the HARP program you can't go through it again.
So these loans are essentially locked out from prepayments and now that those rules have been solidified with the FHFA’s decisions as to how to change the HARP program, we are very optimistic about the performance of these securities but as other securities are paying considerably faster the HARP loans continued last month at 5 CPR so that’s just the actual number.
And then the lastly on this page kind of goes to the other side and that’s the topline in the redline which is conforming jumbo mortgages. So these are the largest loans between $420,000 and up to $725,000 and these larger loans have the exact opposite kind of affect that the smaller loans have, fixed cost are minimal, brokers have tremendous incentive to refinance the largest loans and that's why you are seeing even after slowing down a little bit these jumbo conforming loans paid at 50 basically 49 CPR last month and the month before that they were in the 60s.
So loan size is really important and what you should take away from this is what the title says prepayment risk is real and this is all pre-HARP and everything but it's not uniform and there are places to hide and why is that so important, well if we look at slide five it really demonstrates that from the perspective of yield and ROE.
So focus on the left side the 30 or 4.5s adding 10 and this is as of price of TVA 4.5s at the end of September, so the end of the third quarter. The yield at 10 CPR was 3.48%, we assume the cost of funds you know and then if you look at the last line we show you what the ROE is assuming for a fictitious portfolio assuming eight times leverage and if you can get a 10 CPR that ROE is 24.5%.
If you get a 30 or 40 CPR the yield drops because of the faster amortization and look what happens to your ROE, a 40 CPR your ROE even with eight times leverage is 4% and if you go back to that start that we just went over there are securities that can get out near 40 or above and there are securities that are, that aren’t likely to stay maybe below 10.
So prepayments are just critical when it comes to returns in this business for a levered mortgage investor that’s really the key takeaway from this slide. Now if we go to slide six, we can look at this similar thing that from the perspective of book value, all right.
So we have seen the difference in prepayment speeds well the market doesn't ignore that and so if you look back to March when rates were higher and people didn't care about the prepayments, the highest priced per year 4.5 security was TVA that's the green line. A little below that maybe 5.30 seconds below that were the HARP loans but they were still even though they are prepaying better now they were actually cheaper because they were less liquid and then another tend to slower than that were the jumbo securities because even then there were concerns if they might not prepaid very well if the interest rate environment changed.
But they were all pretty close together; now look at what happened over the last seven months as the interest rates have fallen and as the market has seen the different prepayment characteristics of these underlying modes. The HARP loans which have performed extremely well now trade at a 2.0 premium to TVA, so they were below now they are more than 2 points above so they are up at the top line they are 107.5 as of few days ago.
With the TVAs at 105.5 and look at what's happened to those jumbo conforming the really large sizes loan size, well they are two points lower than even generic mortgages because people have seen the prepayments and realize you know how difficult that is or how much lower returns will be on those mortgages.
So pretty big pricing differences, so if you think of it in premium terms the 107.5 are the good securities, the premium of that is more than doubled the premium for jumbos at 103.5. So again this is from the perspective of managing your book value, owing the right securities is critical because one of the things that’s helped AGNC and we will look at this over time later in the presentation is it by having securities that have held up well in price.
It allows us to hedge more and it means we can actually also perform well if interest rates go up to. So that would go one other thing that's important to look at and this is probably the least understood advantage of slower prepaid mortgages and it’s the impact on liquidity and thus our ability to lever, and I think everyone understands why I haircuts on a reasonable borrowings is important to how much you can lever higher haircuts, or bad lower haircuts or good.
Well generally haircuts are a little below 5% so for this analysis we have just for simplicity assumed haircuts on borrowings are at 5% but there is another thing that sort of acts like a haircut and reduces our available cash to meet margin calls and that is the intra-month prepayment that we see and the reason for that is that Fannie Mae and Freddie Mac release the prepayment speeds on the 4th business day of the month.
So we will just call that the 5th, the at that point the next day if the size of the security because the pay down went from a 100 to 98, Okay 2% paid down in a month that’s a 24 CPR then the REPO lender calls and says the value of your security, the amount of your security went down by 2%, I need you to post incremental margin. So you essentially have to front the prepayment amount until you are remitted the cash flow on the 25th of the month.
So for two to three weeks in the month you essentially have to make up for any prepayments, so that in a sense acts as an additional margin that reduces on your ability to lever and let's look at how significant that can be in an environment like this. Let’s start by focusing on the right side of the page for a second and let's focus at eight times leverage. This is the far right bar, the haircut again is 5% so an eight times leverage 40% of your equity is going to be on accounted for by the margin on your REPO borrowings.
But then the that prepayment that you have to basically front the margin colon until you receive the cash from Freddie or Fannie is the only 27% so that brings your all in kind of an amount of your equity that you are going to set aside for these two factors to 67% in this example.
Now if you took your leverage down to six times with a 40% CPR now you're at around 50% a year equity with 30% coming from the haircut and 20% coming from the prepayment. Keep those in mind but let's go to the left side of page and look at 10 CPR with some of the securities we have looked at is a very possible outcome.
And let’s go to the eight times leverage, then you have 40% for the haircut because it's no difference but now you have a very small prepayment so you're all in amount of equity okay that you need to handle prepayments and haircuts is only 47% at on a 10 CPR again in eight times leverage.
Compare that now to again the six times leveraged at a 40 CPR, you have more available resources to meet margin calls with eight times leverage if you can count on a 10% CPR then you do at six times leverage if you have to worry about 40% CPRs.
I think that’s a conclusion that most people you know wouldn’t expect which is in addition to the benefits we saw on book value and the benefits to returns slower prepayments allow you to very comfortably use higher leverage and again without compromising your available resources to meet you know kind of even extreme margin calls due to price movements.
So I think that’s a key point to keep in mind that I don't think it's a lot of discussion in the space. So now let's – we talked about the importance of prepayments, we have talked – we have seen some of the differences between how different securities prepay, now let's move on to kind of how is agency positioned in light of this and obviously we are going to work off of our Q3 information on, but at the end of and this is page nine. So our portfolio was roughly 50% 15 year and roughly 30%, 30 fixed with a little bit in 20 year a few CMOs and then roughly 8% in hybrid ARMs but look at the bottom of the page. This is are the prepayment speeds for AGNC’s portfolio actually are actual CPRs declined in Q3 versus Q2. So Q2 CPR was nine, Q3 was eight and what we told you was that our CPR for October one released in early October that is was nine so it picked up a little bit.
So why are these CPRs remaining contained with given the backdrop that we have run over before because we basically have lower loan balance and HARP loans making off the vast majority of our portfolio and as we saw those continue to perform very well. So in the 15 year sector as you can see on page 10, 88% of our 15 year were lower loan balance and the footnotes shows either the average for the loan size was around a 104 for that bucket, a $104,000.
We also had 5% of her 15 year HARPs so between the two the HARP program or the lower loan balance looking at 93% of the largest position. Now let’s go to the 30 year in the table below that, 55% of our 30 year holdings with the HARP securities and I remember the HARP securities are paying extremely slow and we again expect that to continue given that program still does not allow you to re-HARP.
The other main chunk is lower loan balance which makes up another 31% of the portfolio so if you put those two together, 86% of the third-year portfolios is a lower loan balance for HARP and then on the other side, going to back to the policy risk on the changes to the HARP program that could impact the season mortgages.
Season mortgages originated prior to the day HARP cutoff date where the last you could use the HARP program for loans was June, 2009 less than 5% well below 5% of our portfolio is in that category. So we really have negligible exposure to anything that can change on the prepayment front due to the new HARP program.
So I think that’s another thing that investors should take some comfort in is that there a lot of different opinions as to how effective these changes are going to be. We really won't know the real answer from the prior the second or third quarter of next year because of the implementation timeline but I think AGNC investors can rest assured that it's not going to be material to them again given the less than 5% of the portfolio is even in that bucket that where it could possibly to them.
So with that let me go to slide 12, what I want to point is it sort of relates to what we just went over. This is not, the agency mortgages business is not a commodity. There are very different types of securities that can perform very differently, how you hedge those securities is another key distinguishing factor and if you look AGNCs returns versus its peer group over the past almost three years, you can see some pretty materially different results and what we focus the most on is the top which is what we call economic return, why? Because economic return as we have defined it is the combination of the dividend we pay plus the change in the book value which is just total market of the portfolio so it's the markets view, it's the mortgage markets view of our value add right, we are going to pay cash and then we look at what our remaining position is worth so to speak.
And so it takes out like always from accounting and so forth and on that measure when you compare AGNC over the past three years to its peers we feel really good about those results and again paying a dividend on a high dividend is really important and investor certainly care about that. They shouldn't be very happy with the dividend that comes out of book value and so it is the combination of the dividend and the book value that in the end is important.
And you also see and importantly you see those results also reflected in the stock price and that's obviously the total return on the stock price which includes dividends and share price appreciation and you see kind of a very similar story on that front as well.
So with that let’s turn to page 13 and I will conclude by kind of saying looking ahead, yes, a flat year curve with faster repayments generally is going to put some downward pressure on margins. However, we feel that we can produce very attractive returns even in this environment, yes, even with the operation twist and even with prepayment speed that changes to the HARP program and faster prepayment speeds because we have taken actions within our portfolio to really mitigate some of the impacts of those.
So we feel like we can continue to produce good results, we think our funding costs are going to remain contained in low for multiple years. I mean the Fed is telling you that they are on hold for at least a couple of years and even so we have hedged out essentially we have swaps covering 70% of our REPO borrowings so even if it's turns out that interest rates start to go up sooner than that we have a lot of protection against them.
We had $0.85 of undistributed taxable income at the end of Q3 which is also something we have to pay out in dividends at some points in the future and again I just want to stress that the big challenge as we saw is going to be over the next six months, and the prepayment landscape and we feel extremely well-positioned, it's going to help us with respect to book value, it's going to help us with respect to ROE and yields and it's going to help us with respect to liquidity.
So with that I would like to stop now and open it up for questions.
We have used it in the past but generally speaking our undistributed taxable income on a dollar basis has gone up every quarter really since 2009. So we haven’t tended or haven’t had to use it I think, there was one quarter where it may not have 15 go up but we would use it if our taxable income you know was low for a quarter or two and we felt like the returns in the over the longer term were still there and we wanted to maintain a particular dividend.
I do want to be clear though, we are not going to in a sense do whatever it takes to maintain the $1.40 dividend for X amount of time, I mean our mind set is the mortgage market provides us with attractive return opportunities. We have been able to pay a $1.40, generate book value growth and generate excess taxable income so that’s kind of the past.
The future is that we will evaluate our ability to continue to pay a dividend that does not come out of a book value for an extended period of time and I think that’s really it's the combination of being able to pay a dividend not have a drain book value coupled with taxable income and undistributed income that gives you the whole picture.
You are not able to get levered on the intra-month prepaid speed.
No I mean there has been a number of attempts over the years to be able to write REPO agreements in a sense to deal with that and there a couple I would say baby steps to a solution but the short answer is no that intra-month prepayment absolutely affects your ability to lever. And if you remember back to the GSE buyouts in 2010 there were some kind of concerns in the industry it's not – the GSE buyouts created onetime very, very fast prepayment speeds and it definitely creates stress for a number of REITs in the space because of this payment delay. So the short answer is you cannot, you have to have spare cash for that.
All right well if there are no more questions I really appreciate everyone’s interest in AGNC and thanks again.