AGNC Investment Corp (AGNCO) 2012 Q1 法說會逐字稿

Good day, and welcome to the American Capital Agency shareholders call.

(Operator Instructions).

Please note this event is being recorded.

I would now like to turn the conference over to Pete Deoudes, Director, Equity Capital Markets.

Please go ahead.

Thank you, Amy.

Thank you, everyone, for joining American Capital Agency's first-quarter 2012 earnings call.

Before we begin, I'd like to review the Safe Harbor statements.

This conference call and corresponding slide presentation contain statements that, to the extent that they are not recitations of historical fact, 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 those forecasts due to the impact of many factors beyond the control of AGNC.

All forward-looking statements included in this presentation are made only as of the date of this presentation and are subject to change without notice.

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 AGNC's periodic reports filed with the Securities and Exchange Commission.

Copies are available on the SEC's website at www.SEC.gov.

We disclaim any obligation to update our forward-looking statements, unless required by law.

An archive of this presentation will be available on our website, and the telephone recording can be accessed through May 17 by dialing 877-344-7529 and the conference ID number is 10013452.

To view the Q1 slide presentation, turn to our website at AGNC.com and click on the Q1 2012 earnings presentation link in the upper-right corner.

Select the webcast option for both slides and audio, or click on the link in the conference call section to view the streaming slide presentation during the call.

If you have any trouble with the webcast during the presentation, please hit F5 to refresh.

Participants on today's call include Malon Wilkus, Chairman and Chief Executive Officer; Sam Flax, Director, Executive Vice President, and Secretary; John Erickson, Chief Financial Officer and Executive Vice President; Gary Kain, President and Chief Investment Officer; Chris Kuehl, Senior Vice President, Mortgage Investments; Peter Federico, Senior Vice President and Chief Risk Officer; Bernie Bell, Vice President and Controller; and Jason Campbell, Senior Vice President and Head of Asset and Liability Management.

With that, I'll turn the call over to Gary.

Thanks, Pete, and thanks to all of you for joining us today.

Let me start by saying we are extremely pleased with AGNC's performance during Q1, and as we'll discuss on today's call, our asset selection strategies continue to create substantial value for our shareholders.

Furthermore, we remain confident about the future and we're going to conclude this presentation with a detailed summary of how we are positioning the Company in relation to a variety of scenarios that could unfold as 2012 progresses.

Now the big story for the quarter was the continued strong performance of lower loan balance and ARM securities.

In addition to providing stable prepayments, the market valuations of these assets continue to improve as other investors recognize the necessity of having slow and predictable prepayments in the current market environment.

With that, let's review the highlights for the first quarter on slide three.

Our net comprehensive income was $2.44 per share, which was comprised of GAAP income, net income of $2.66 per share and to a loss of $0.22 in other comprehensive income.

Now, net spread income, which many of you refer to as core income, was up materially to $1.42 per share during the quarter.

Now this number does include a one-time premium amortization catch-up of $0.12 related to lower prepayment projections.

Now excluding this catch-up [AM], our net spread income was still very strong at $1.30 per share.

Now the significant improvement in this measure versus prior quarter was driven by the combination of higher average leverage, a reduced cost of funds, and lower prepayment projections.

Now with respect to that last point, our lifetime prepayment estimate dropped to 9% CPR due to the combination of higher interest rates during the quarter; changes in the composition of the portfolio, which included a 24 basis-point drop in the average coupon; and enhancements to our prepayment model provided by BlackRock Solutions.

These changes in lifetime prepayment estimates now bring our projected speeds much closer to our actual Q1 CPR of 10.

And as I mentioned before, BlackRock does update their model periodically as new information becomes available.

That was the case in the first quarter, and we reflected the model change in our results.

Now taxable net income, which is not impacted by the changes in prepayment projections or in the unrealized gains and losses on derivatives, was also very strong at $2.03 per share.

Our undistributed taxable income more than doubled from $180 million to $384 million.

As such, despite the increase in our share count, undistributed taxable income on a per-share basis increased substantially to $1.28 per share at the end of Q1, our highest level ever.

A key driver of the strength in taxable income, and thus UTI, were sales of some higher-coupon HARP and lower loan balance securities.

The book value rose by $1.35, or almost 5%, to $29.06 per share from $27.71.

And while it's still early in the current quarter, we have seen this trend continue, which has given book value a nice tailwind so far for Q2.

The combination of book value growth, coupled with our dividend of $1.25 per share, led to total shareholder value creation, or economic returns during Q1, of $2.60 per share, or 38% on an annualized basis.

Now turning to slide four, our mortgage portfolio increased to around $81 billion by the end of the quarter.

Leverage during the quarter averaged 8.2 times, while quarter-end leverage came in at 8.4 times.

Our net interest spread increased 13 basis points to 207 basis points as of March 31, from 194 basis points at the end of December.

And lastly, while we have raised a significant amount of equity during the quarter, we remain disciplined with respect to the timing and deployment of accretive equity raises.

To this end, the performance of AGNC over the last few quarters and over the last several years should demonstrate to investors that equity can be raised in a manner that is supportive of shareholder value creation.

Now I would turn the call over to Chris so he can discuss what happened in the markets during Q1 and the changes to the portfolio.

Thank you, Gary.

Let me start by reviewing the market data on slide five.

In the top left panels, you can see that the curve steepened as rates sold off during the quarter with five-year swap rates up five basis points and 10 year swap rates up 26 basis points.

30-year passthroughs were largely unchanged in price and 15s outperformed 30s on both a price and a spread basis by about a quarter-point.

Let's turn to slide six to review the composition of the investment portfolio.

The portfolio grew to just under $81 billion during the quarter as we raised a significant amount of equity and increased leverage, given attractive asset valuations.

Our purchases were concentrated into lower coupon, generic mortgages, as well as lower coupon passthroughs backed by lower loan balance and HARP loans.

Our average prepayment speeds continue to be slow, despite record low mortgage rates.

I do want to address the four CPR increase in actual prepayment speeds over the February/March time period.

The increase is largely attributable to a couple new positions that were purchased fully anticipating having higher speeds.

The more significant contributor to the CPR increase is a position in generic 15-year passthroughs that we believe to offer attractive risk-adjusted returns relative to other very short duration assets, despite the faster speeds.

Now let's focus on slide seven for more detail on how our markets performed during the quarter and how we're currently positioned.

As Gary mentioned in his opening remarks, the real story during the first quarter was the tremendous outperformance of higher coupon loan balance and HARP securities.

In the chart on the right, you can see that pay-ups, or, in other words, the market price premium, for some of these strategies outperformed generic passthroughs on a price basis by another half of a point.

It's important to point out that this occurred during a period when interest rates increased, and given that these cash flows do have longer durations than generic passthroughs, this is exceptional performance and a realization of just how cheap these loan attributes were a couple quarters ago.

As you know from our disclosure in prior quarters, the agency was heavily overweight pools backed by both loans with lower loan balances and loans originated through the HARP program.

We still strongly believe that these types of securities should form the foundation of a well-constructed portfolio, but are also cognizant that the risk-adjusted return advantages of these products has, by definition, diminished somewhat relative to other MBS, as valuations have improved.

As such, we sold a material amount of higher-coupon HARP and lower loan balance pools during the quarter, approximately $6 billion, in an effort to reduce the magnitude of our still-overweight position.

Some of the sales were executed at over three points above TBA levels, many of which were replaced with lower coupons and lower payoffs and prices.

So in conclusion, we remain confident that despite reducing some exposure to the highest-coupon, highest-payup pools, the portfolio was still very well positioned with respect to prepayment risk.

With that, I'd like to turn the call over to Peter Federico, our Chief Risk Officer.

Thanks, Chris.

Today, I will review our financing and hedging activities.

I will start with our financing summary on slide eight.

Our repo costs decreased slightly to 37 basis points at the end of the quarter from 40 basis points the previous quarter.

As we discussed on our last call, we expected repo rates to fall during the first quarter as year-end balance sheet pressures subsided.

Repo rates did, in fact, fall slightly.

Offsetting the benefit of somewhat lower rates was our continued focus on lengthening the term of our liabilities.

During the quarter, we increased the weighted average original maturity of our repo funding to 104 days from 90 days the previous quarter.

This maturity extension increased our funding costs by about a basis point.

Turning to slide nine, I will review our hedging activity.

Consistent with the asset growth we experienced during the quarter, our pay-fixed swap portfolio increased to $38 billion at quarter-end from $30 billion last quarter.

During the quarter, we entered into approximately $8 billion of new pay-fixed swaps.

These swaps had a weighted average pay-fixed rate of 1.47% and weighted average term of 6.3 years.

As I have mentioned before, our swaption portfolio plays an important role in our goal of protecting book value.

We enter into swaptions opportunistically as market conditions and portfolio composition dictate.

During the first quarter, we significantly increased the size of our swaption portfolio.

In total, we purchased $8 billion of put swaptions at a cost of $65 million.

These purchase -- with these purchases, our swaption portfolio totaled $10.5 billion at quarter-end.

Given the significant increase in our swaption portfolio, we thought it would be helpful to review the benefit we see in these hedges.

On slide 10, we show the expected performance of our swaption portfolio over a wide range of interest rate changes.

The benefit of our swaption portfolio can be viewed both in terms of its market value, as well as its duration profile.

The graph on the top right shows the estimated market value of our swaption portfolio.

As of March 31, the portfolio had a market value of approximately $78 million.

If interest rates remain relatively constant, or fall, the swaption portfolio will expire with little or no value.

To put this in perspective, losing the entire $78 million would equate to a loss of around $0.26 per share.

Conversely, given the significant positive convexity of these instruments, the market value of the swaption portfolio will increase in a very nonlinear way as interest rates increase.

At the extreme, if interest rates were to immediately increase 300 basis points the market value of the swaption portfolio will increase to approximately $1.5 billion, or about $5.00 per share.

As such, our swaption portfolio is a critical component of our overall book value hedging strategy, and it's specifically designed to provide some insurance against a large and rapid increase in rates.

The graph on the bottom of the page shows the duration profile of the swaption portfolio.

As you can see from the dark blue line, the duration of the swaption portfolio varies considerably as interest rates change, from a base duration today of around 1.5 years to a peak duration of about five years.

In a rising rate scenario, the duration increase associated with our swaption portfolio can be viewed as a partial hedge of the asset duration extension we will experience.

To better illustrate this point, the gray line on the graph shows the duration profile of the swaption portfolio expressed in the terms of the amount of asset duration hedged.

At its peak, the swaption portfolio will hedge about 0.8 years of our asset duration.

Said another way, if interest rates increase significantly and our asset duration extends from about 3.5 years to about six years, a third of that extension would be offset by our swaption portfolio.

This makes the remaining extension considerably more manageable.

Turning to slide 11, our duration gap was positive 0.4 years.

The increase in our duration gap from about flat last quarter was due to portfolio activity, interest-rate changes, and the implementation of the enhanced prepayment model provided by BlackRock Solutions that Gary mentioned earlier.

This model update had the effect of lengthening our estimated asset durations by about a half a year.

Before turning the call back to Gary, I would like to review slide 12, which is intended to help investors better understand the interaction between prepayment speeds on our assets and our ability to prudently deploy leverage.

Prepayment speeds are an important driver of leverage capacity because of the GSE payment delay.

During this period from factor release to principal remittance, we as repo borrowers are required to post additional collateral to cover the principal paid paydown.

As a result, slow prepaying securities require less incremental collateral each month and fast prepaying securities require more incremental collateral.

Leverage capacity is impacted by many factors, but three important factors highlighted on slide 12 are the repo haircut, asset prepayment speeds, and a cushion to absorb adverse price movements on assets relative to hedges.

For illustrative purposes, the graph shows how NAV or equity would be allocated to these three variables across different leverage and prepayment speed scenarios.

To explain it in greater detail, I will focus on the circled bar in the middle of the graph.

That bar shows the amount of NAV needed to support the repo haircut, monthly prepayments, and adverse price shocks.

The calculations assume eight times leverage and assets that prepay at 12 CPR.

Starting at the bottom, the blue section represents the amount of NAV required to support the repo haircut, which in this example is assumed to be 5%.

The next section, in green, shows the NAV required to support monthly prepayments, which is 1% a month, given the 12% CPR assumption.

Finally, the gray section on the top shows the NAV required to support an adverse price move on assets relative to hedges of 2.5 points.

Taken together, these three variables consume about 70% of this hypothetical portfolio's NAV.

In comparison, the stacked bar to the immediate right shows the NAV allocation if asset prepayment speeds increase to 36%, holding the other variables constant.

In this scenario, the total NAV allocation increases to about 90%.

The graph shows similar calculations assuming six times leverage and 10 times leverage.

For example, a hypothetical portfolio with 10 times leverage and assets that prepay at 12 CPR has a total NAV allocation of just under 90%.

Such an allocation is very similar to a portfolio leveraged eight times, but with assets that prepay at 36 CPR.

This comparison highlights the importance of asset prepayment speeds with respect to leverage capacity.

And with that, I will turn the call back over to Gary.

Thanks, Peter.

Before taking your questions, I want to highlight three scenarios that we are focused on as we strategize on how to position the portfolio for the current environment.

So turning to slide 14, the first scenario is the one that the market currently assesses the highest probability to.

It is one where moderate economic growth continues and inflation remains more or less consistent with the Fed's forecast.

In this scenario, we would not anticipate a QE III from the Fed and would generally expect a somewhat steeper curve with the 10-year rate increasing somewhat to maybe 2.5%.

Prepayments on most types of mortgages would likely slow from current levels.

Now with respect to the impacts of the market, we would anticipate that the combination of the somewhat lower mortgage prices, a steeper yield curve, slower organic refinancing activity, and manageable interest-rate volatility would be very supportive of mortgage spreads and thus returns.

Some might even describe this as today's version of a Goldilocks scenario, and in this scenario we would likely expect book value changes to be relatively modest.

The next scenario seems somewhat less likely, but is clearly a very real possibility and something that deserves reasonable consideration from a positioning perspective.

Under this scenario, the improvement in economic numbers that we saw during the first quarter turns out to be another head fake, with economic growth or the employment picture weakening and/or the inflation outlook dropping below the Fed's target.

This scenario could be, in fact, priced in very quickly in response to a number of catalysts, including further problems in Europe or just weaker U.S.

economic data.

If this occurred, we would likely see the Fed implement a QE III, which would likely include significant purchases of agency MBS.

The outcome of this scenario could be more challenging as the Fed's bid would be expected to drive the prices of the lowest coupon fixed-rate MBS to very rich levels.

This, in turn, would cause prepayments to accelerate materially on many types of mortgages, adversely impacting returns on existing holdings.

It would also likely force investors to pay even higher prices for securities such as lower loan balance or HARP mortgages where prepayments would still be relatively stable.

As such, especially given the flatter yield curve, returns on future purchases would be considerably less attractive.

So clearly, this would not qualify as a Goldilocks scenario and us being appropriately positioned in advance will be critical to performance.

But with the right assets and with adequate leverage, book value gains could be significant.

As such, we could actually produce very strong total returns in this scenario.

Again, it will all be about positioning.

The last scenario is one where economic growth or inflation expectations pick up dramatically.

Here we could see intermediate- or longer-term treasuries increase over 100 basis points in a very short period of time.

We view the odds of this occurring as very low, given both the global and domestic economic picture.

The likelihood of this scenario is further limited by a Fed that has been very transparent about its intentions with respect to interest rates and its view of the headwinds facing the economy.

That being said, we cannot just dismiss this out of hand because it does present the greatest risk to a levered portfolio as mortgage prices will drop significantly.

Having appropriate hedges in place will be absolutely critical to attempting to mitigate potentially large book value declines in the unlikely event that this actually happens.

Now, slide 15 is intended to summarize some of the things that were discussed earlier related to how we've positioned AGNC against this backdrop and to provide some insight into our thought process in getting there.

So in short, prepayment protection and lower coupons are critical to returns in the current environment and to reasonable performance in a QE III scenario.

However, as Chris mentioned, we have taken material steps to attempt to lock in some of the book value outperformance that we've achieved over the last few quarters.

But I want to reiterate an earlier point.

It would be irresponsible for us not to maintain a core position of prepayment protected mortgages as predictable prepayment behavior is just plain fundamental to managing a mortgage portfolio.

As Peter said, we also increased the duration of our hedges to improve our performance in the moderate-growth scenario and took advantage of lower option prices to buy a material amount of protection against the risk of a large spike in interest rates, like in scenario three.

And lastly, with respect to leverage, our bias in Q1 was toward increasing leverage as mortgage valuations were attractive and we didn't want to be underinvested.

However, as the prices of mortgages have appreciated significantly over the last few months, we now believe a more moderate stance toward leverage is warranted.

But we are still currently biased against running at leverage levels below, say, seven times as the risk of having to compete with the Fed in a QE III scenario for overpriced mortgages is just too great.

Moreover, we do not expect to be able to buy mortgages cheaper net of hedges, of course, in the base case moderate-growth scenario.

So in conclusion, we feel our portfolio is well positioned to produce attractive returns over a range of market outcomes.

We also hope that by summarizing our thought process in this manner, it helps give investors some transparency into management's approach to the current environment.

But remember, as our performance over the past few years has demonstrated against the backdrop of a QE I, a QE II, Operation Twist, the GSE buyouts of delinquent loans, multiple HARP program enhancements, and several European debt flareups, just to mention a few of the things that have gone on, it is going to be the ability to change direction based on evolving market conditions and then to implement appropriate strategies around asset selection and hedging that's going to be the real driver of market-leading performance.

So with that, let me stop and turn the call over to the operator for questions.

(Operator Instructions).

Jason Arnold, RBC Capital Markets.

Hi, guys.

Great results this quarter.

You did very well with the low balance in HARP-specified pools trade, but it didn't sound like you took some profits here.

So I was just curious if you could share with us your thoughts on -- broader thoughts on value right here between the two segments, and then perhaps comment on the incremental spreads on new investments you're seeing right here.

You know, pool -- the HARP and lower loan balance pools have performed very well, and by definition they're not as attractive as they were a couple quarters ago, but overall we still very much like the risk-adjusted returns of these strategies, or at least most of them, and they're going to continue to be a core position for us for a number of reasons.

Peter reviewed with you on slide 12 the importance of slower prepayments with respect to repo and margin calls, and you know, it's critical to have a portion of your position that you can count on to not drive increases in repo margin calls due to prepayments in a stress scenario.

And these positions are also extremely important from a convexity risk management perspective.

Again, knowing that you can count on these positions to perform in a rally allows you to hedge more effectively.

And so, we're not managing to just today's prepay and interest-rate environment.

We are -- it's important that we perform across a range of different environments, and these positions will help us do that.

And so, they're going to continue to be a core position for us, and we still find the risk-adjusted returns as being attractive.

The positions that we sold during the quarter were largely the very high coupons backed by lower loan balance and HARP loans.

Just one other thing, as slide seven points out, we are, however, cutting our kind of aggregate exposure to those, back down to what we would say are more kind of normal levels.

So our bias over the last few months has been, not that -- we've still been buying some, we've been selling some, but -- depending on coupon and strategy, but our bias has been to be more balanced, given today's pricing.

Okay, So for like dollar for dollar, you're putting probably more into the lower coupon side, and then perhaps kind of balancing that out with other areas of interest.

Is that a fair statement?

Yes, it is, but I do want to stress that it's not all or nothing.

I mean, we still look at -- we don't buy, except in kind of one-off cases, what we would call TVA, so even if we're not buying something like a HARP or a low loan balance pool, we're certainly thinking about the amount of third-party origination.

We're looking at the -- plenty of other factors.

So I don't want you to -- but to be frank, those things don't matter.

I mean, they're important, but they don't matter nearly as much as some of the stories we've talked to you about.

But there's -- I just don't want to give you the impression that it's kind of all or nothing.

Even if you're not involved in something that has kind of really strong prepayment protection, you certainly, in everything you do, you absolutely have to avoid the worst and we feel like we can do that across the board.

Okay, and then, just one quick follow-up.

Thanks for the color on the rate hedges.

I know the plain-vanilla swaps aren't ideal, given convexity and prepayment optionality.

But I was curious if you could comment on what really led to the uptick in the use of swaptions, you know, from a view -- a macro view or rate view standpoint.

And then, perhaps you could also comment on the use of the short TVAs and treasuries, and specifically what component of the rate risk kind of spectrum you're hedging out with those?

This is Peter.

You know, on our last call, I mentioned the fact that we're opportunistic in our purchases and that we were seeing the cost of these options decline once we got through year-end.

So in December and early in the first quarter, we saw implied volatility in the marketplace come down.

The combination of the lower cost from an option perspective, as well as the strike of the option, you know, with rates coming down, the combination of those two things really led us to view these much more attractive in the first quarter than they have been in the past.

Subsequently, later in the quarter, the prices went up a little bit and now they've come back down to close to 12-month lows again, so they continue to look attractive to us.

With regard to the treasury hedges, I'm glad you pointed that out.

We do use a combination of hedges, and in this quarter, we used treasuries more extensively than we have in the past with about $5.5 billion at quarter-end.

And again, it's just opportunistic hedging.

We use treasuries, we use swaps, we use swaptions.

So, we just look at them on a relative value perspective, what do we think is going to happen to swaps spreads, and try to just position the portfolio as best we can from a hedging perspective.

Chris may want to add a comment on the TVAs.

Again, the TVAs give us, again, exposure to different parts of the curve and partial durations, and at times are very effective hedges and allow us to source call protection through some of these specified pool strategies at times where we don't necessarily find the mortgage swap faces attractive.

Dean Choksi, UBS.

Good morning, gentlemen.

Can you talk about your expectations for HARP II prepayments?

Sure, this is Gary.

On HARP II, we absolutely think HARP II is relevant.

You can take comments from any of the large originators or FHFA.

You can look at the flow of the over 125 LTV new mortgage securities into the marketplace.

We're definitely seeing the impacts that HARP 2.0 will have impacts, and we've seen faster -- somewhat faster speeds on the higher coupons.

We think this is still very early and, interestingly, as you can tell from our portfolio, we are still avoiding HARP 2.0 risk.

And let me give you the main reason.

It's kind of interesting.

Again, as we've talked to you about, we can't just manage for one scenario.

We've got to manage over a range of interest-rate scenarios and, for that matter, prepayment scenarios.

And the thing that usually people like about higher coupon season mortgages, and they were a very sizable part of our portfolio back in 2010, when interest rates go up 100 basis points, 50, 100 basis points, they tend to outperform other mortgages, and so they're shorter duration and they add a lot of value in that scenario.

Right now, given HARP 2.0, we actually think they would perform very poorly in that scenario, and the reason is is that right now originators seem to be dedicating maybe 30% of their capacity to HARP 2.0.

If rates went up 75 basis points, all the other refinancing activity that they're doing that's using 70% of their capacity would go away, which would lead them to dedicating all of their capacity, in our minds, or the vast majority, to HARP 2.0.

Plus the coupons are high enough that they'd still be very much refinanceable.

So it's really -- that's probably the biggest concern we have with the universe that's exposed to HARP 2.0.

So I hope that helps.

I appreciate the color, as well as the incremental disclosure in the slides.

Can you also talk about how you deploy the capital around the capital raise a couple of months ago?

Sure.

I mean you can kind of see it from the portfolio.

What you can infer is we put less of it into lower loan balance and HARP securities, but we still bought some.

We put more into 30-year than 15-year mortgages.

But on that front, we also were well-positioned for the capital raise and had acquired some of what we did buy in the specified mortgages earlier.

That is one of the reasons why we use TBA hedges, is to kind of stockpile good assets at good prices.

So when you put all of that together, we were able to deploy the capital rates much quicker this time versus the quarter before where the capital raise was very much opportunistic with respect to a very specific opportunity.

To be perfectly honest, we were not able to forecast it as well, so the deployment time, while I think it was a great time back in October, it took a little longer.

(multiple speakers) helps.

Great, thank you.

Mark DeVries, Barclays.

Yes, thanks.

My questions revolve around the outlook for mortgage spreads.

First, it seems like Freddy and to a lesser extent Fannie have been accelerating the runoff in their routine portfolios.

What, if any, implications has this had for agency spreads?

Very little.

You are absolutely right.

We have seen that same trend as well.

But remember the treasury basically just stopped their mortgage sales program which was around $10 billion a month.

So the incremental slowdown in purchases or net sales from the GSEs right now is not a market moving event.

We don't think that is likely to impact spreads much at all.

It's not that big picture of a factor in the market, especially against the backdrop of fed purchases and reinvestment of cash from paydowns and a lack of net supply into the mortgage market.

So we are pretty constructive still on mortgage spreads across two of the -- most of the environments that we see unfolding.

That being said, they've had a very good run over the last few months and so we are very glad that we had reasonable leverage and had purchased value-added securities before because they've had a good run.

But we honestly don't expect any material cheapening unless you get a real outlier type scenario.

Okay.

What are your thoughts on bank demand for mortgages at these kind of low absolute yields?

Are you seeing that dry up at all or are they still pretty big buyers?

No they've still been pretty big buyers.

Could it dry up in some scenarios or slow down?

Yes but the short answer is banks have a ton of cash to put to work.

I mean their liquidity levels are higher than they have been in ages.

So it's hard to forecast, but the short answer is we don't expect bank demand to really drop off that much.

Okay, thanks.

Arren Cyganovich, Evercore.

Thanks.

With your purchases of more I guess, I don't know, you would call them straight generic 15- or 30-year mortgages after you recycled out of your HARP and low loan balances, I would think that that would increase your prepayment risk associated with the portfolio, yet the model had it going down pretty substantially quarter-over-quarter.

Can you talk about the dynamics of that with your portfolio?

Sure.

First off, again, just if you look at the numbers and the chart on Page 7 as an example, we haven't really gotten out of the specified or the HARP and the loan balance securities.

We have reduced our weighting to them is probably the best way to describe it.

But very importantly, the average coupon in our portfolio dropped 25 basis points, right.

So again, a lot of that repositioning may have been selling a 5% coupon or a 4.5% HARP security at a $1.11 price and then buying a HARP security back by 30 or 3.5% at a much lower dollar price and a much lower payout where the pre-payment speeds would be expected to be slower.

So that is one component but don't ignore the drop in the coupon.

The other thing to be cognizant of when anyone projects prepayments is interest rates, right?

So the 10-year swap rate was up almost 25 basis points during the quarter.

So that's material as well.

Then realistically -- and I think this has come up on other calls -- you know, there were some parts of the market in a sense, in particular HARP loans or whatever but -- where our estimates were good and they were very reasonable.

However, there was an argument, as you got more data, that maybe you were a little fast.

So I think that factored into Black Rock's, their general process for taking in new data and then making logical revisions on an ongoing basis to their model.

So you have all of those factors.

So we are very comfortable with our assumptions against that backdrop.

But I do just want to complete that by saying, as Chris mentioned in the prepared remarks, there are some targeted positions where we're comfortable with faster pre-payments.

The bulk of what we have added to the portfolio is in low enough coupons where, in the absence of a major change in interest rates or mortgage prices, we don't expect to see faster prepayments there.

Okay, that's helpful.

Then in looking at your scenarios between scenario one and three in terms of the 10-year assumption of an increase, we can see some pretty rapid moves in the 10-year.

I think it is down over 100 basis points from where it was only a year ago.

What kind of an increase are you talking about in scenario one versus scenario three?

In scenario one, we're talking about a modest increase.

I think I would say -- I think what I said in the prepared remarks was maybe the 10-year goes to around 2.5%.

So we're not talking about a major shock.

It's kind of where we -- we were near that back in March.

So, that is more of a gradual kind of relatively small increase.

When you get to scenario three, we are talking about more than 100 basis points.

Obviously, there is some room in connecting the dots between all these scenarios.

In no way, shape or form is this designed to give you every possible outcome.

Obviously, we can't do that and there are infinite possibilities there.

But I think the intent of one is, one, where interest rates gradually increase some but it is not a big move.

The intent of the illustration in three is one where interest rates shoot up and the move is large, and large being more than 100 basis points.

That is kind of the intent of the scenarios from an illustrative point of view.

Okay, thanks.

[Scott Smith], private investor.

Just a quick question, maybe it's more educational.

On Slide 4, you talk about a 9% average projected portfolio CPR license.

As I look at some of the other slides, I see March and April have been running more around 12%.

How does one really think about that?

I guess, to a layman, you would think that maybe you would use the higher CPR rate, given March and April.

But how do I think about the 9%, and how is that used, that projected rate, used in your financials?

Look.

That's a great question.

What you should think about is there is a couple of things that are important to keep in mind.

So if you compare that, first off, we are not trying to, with our lifetime projection, the 9%, okay, it's, again, a lifetime projection for the portfolio.

It is going to be -- and the reality is it's going to be subject to change on a quarterly basis based on market conditions, interest rates being the most important, but other -- and composition of the portfolio.

But it's, again, a lifetime speed.

It's not intended to project the next month or two in terms of prepayment speeds.

So you should expect that those are not going to line up actually more often than not in terms of the one-month speeds versus a lifetime projection.

The lifetime projection also considers things, importantly the forward curve, the forward interest rates.

So there is a lot of factors that go into a time pattern of prepayments without getting too technical, but that is a key thing to keep in mind.

The important piece of the projection is it's -- that we give you -- all we can do is give you the aggregate numbers.

We give you somewhat of a breakdown based -- on Page 6 -- based on some of the larger product types within the portfolio.

But those are really the key differences between why a 9% and a 12% won't necessarily -- aren't supposed to equal each other all the time.

Okay, great.

Thanks for that.

Just a follow-on question, albeit small, I guess I was scratching my head as to why you issued preferred stock.

I think, given the costs -- and I thought I read about a conversion feature -- versus the availability of common stock issuances to you?

Sure, no, look, a very reasonable question.

Our reason for tapping the preferred market was one of -- at the highest level, it is one of flexibility.

What you want to know -- any issuer wants to know what their options are and wants to develop liquidity in as many vehicles as possible for accessing funding or equity.

And so, from our perspective, having access and developing kind of price points in a market is an important issue in and of itself.

Now, that being said, we feel that the execution also is very reasonable in that, yes, we are paying an 8% coupon that is callable in five years or it is [patchable] (inaudible) otherwise callable at our option, but our ROEs are well above 8%.

So on those investments, common shareholders are getting incremental returns from that issuance and it makes sense.

Another way you could look at this is you are paying common shareholders a dividend much higher than that.

You're getting equity at a much lower cost of equity than common.

Big picture, that isn't -- there are fundamental differences in the two profile so it is not as simple as the latter example but I think big picture this is not a wave of the future for AGNC at this point.

It's more a matter of keeping our options open, understanding the layout and what you can do in different markets.

Okay, great, thanks.

If I may, one final question.

Could you just comment on your exposure in the repo market to, say, Europe and how one should think about the news we have read about Europe and the potential outlook there and how that relates to your repo portfolio?

Then I will take it off-line.

Yes, we give you that breakdown at least by geography on Slide 27 that shows that about 29% of our repo funding is coming from the order book, but with nine counterparties.

The point is we are very cautious about all of our counterparties.

We try to diversify our exposure.

We also give you on that page a measure not by name of counterparty but by measure of the equity we have at risk with each of our counterparties so you can get a sense on how small on a per counter party basis our actual equity at risk is.

Perfect.

So at this point you are fairly comparable with the outlook for the European banks that are participating in your repo program?

Yes, in particular, the type of counterparties we have we're very comfortable with.

Okay, thanks so much guys.

Great quarter and (technical difficulty).

Thank you.

We appreciate it.

Joel Houck, Wells Fargo.

Gary, I am wondering if you could maybe -- I guess more of a philosophical question talking about the relative size of AGNC now, in terms of looking forward, the value add with the kind of specified pool strategy as opposed to TBA.

Obviously, as you point out, it has come down here the last couple of quarters.

In scenario one, in a moderately rising rate environment, what are your thoughts about how TBAs would perform versus spec pool and how, if any, would you reposition the agency to perhaps get ready for that environment?

Well, look, obviously as you know from the fact that we put out the scenarios -- I'm sorry, your question was more specific to scenario one?

It's more specific to scenario one but it also dovetails into just kind of as you get bigger, you know, $80 billion portfolio, how much more ability do you have to even kind of play in this spec pool market as opposed to the more generic (multiple speakers)?

Let me address that one first.

Look, those are both very good questions.

Look, with respect to the size issue, we could have easily maintained the higher percentage of spec pools that we had had before, if that was our desire.

We went out of our way, so to speak, to sell -- Chris gave you the numbers -- $6 billion in HARP and loan balance security in Q1.

If we didn't do that, then we would have had, irrespective of even purchasing anymore which we could have done, that would have had a pretty decent impact on the weightings first off.

So what I would say is that is a pretty big market and there's something on the order of 20%, 25% of total production, and that will increase with HARP 2.0, that comes in through those products.

So there is a decent amount out there.

We'll be -- size will not, in our minds, inhibit or is not likely to inhibit our involvement in those sectors right now.

But to your point about scenario one, right, that is exactly what we have talked about and why we have chosen to sort of both reduce and reposition the coupons, because the reality is -- and if interest rates go up 50 basis points and prepayments slow across the board, then the price premiums for the specified pools will come down.

At some point, they could overreact like they did before.

In other words, they could become too cheap, in which case that would be a good time to add them.

So we are absolutely focused on trying to balance between being able to handle the current environment, having a core position that is predictable and that can be hedged, coupled with the fact that we don't want to give back all the book value gains and outperformance we have had versus more generic mortgages in the scenario which is probably the -- which could end up being the one that unfolds.

So I think that is the balancing act.

What I want to stress is we think about it.

We don't only think about; it we view it as our responsibility to take action.

Okay, good.

I don't know, Peter, can you give us -- or is it in the queue -- the strikes for the swaptions, the strike levels that you put on?

On the ones -- well we give you the total for our swaption portfolio at 2.7%.

So that is down a little bit.

I don't have the exact number of the strike on the $8 billion that we purchased during the quarter.

It was lower than that --

Okay.

Peter Federico -- but it is in that neighborhood.

Okay.

Thank you.

Because the underlying on most of those swaptions were all mostly seven- and 10-year underlying.

Okay, thanks, Peter.

Sure.

Chris Donat, Sandler O'Neill.

Just one question on the -- with the increase in the size of the portfolio.

I'm wondering a little bit about the seasoning and how that might affect CPRs.

I would just guess, as you add relatively new MBS, there would be some pressure a few quarters out as far as CPR goes for higher prepayments.

But I'm not sure I'm understanding the [credit limit].

I certainly know that you've got your prediction for a 9% average portfolio CPR that factors in a bunch of different attributes.

But anyway, I would just like to understand how seasoning might affect CPR.

Look, it's a great question and we've talked about this on other calls.

It is one of the reasons we don't like to use actual CPRs, right, because if we were buying relatively generic or (inaudible) we have some of those that are one-month-old or three months old, then clearly they are going to prepay very slow today.

But they're going to pick up a year from now unless interest rates go up or whatever.

So our models obviously completely capture that and that is another reason why lifetime CPRs are very different from monthly CPRs.

But keep in mind what we don't have, except in a couple of positions, are lower coupon -- I mean are middle or higher coupon generic mortgages that are new.

So in the case of our generic mortgages, they are very low coupons and so they are new.

So yes, they will speed up but that increase will be moderate, whereas if you had higher coupon generic mortgages that were a few months old that you had purchased maybe before interest rates fell, then they may have been slow over the last few months but they are going to be very fast or could be in the 30s% a few months later.

We really don't have much of that.

So I think, if you looked at Page 6, you can kind of see the breakouts across strategies, but those are some things to keep in mind on that front.

Okay, thanks very much on that one.

Then more of a policy question, as Peter went through the leverage discussion at different CPRs, the underlying mechanics of the MBS market with the pre-payment release date early in the month and then principal and interest payments received later in the month, is there any talk among all of the potential reform of Washington about changing the mechanics of the MBS cash flow world?

Yes, there are.

There are discussions.

FHFA has been public about discussions about designing new securities and there are discussions about combining Freddie and Fannie securities and seeing what the right solution is.

I mean, look, we don't expect "relief" on this issue in the next year for sure, and so we're certainly operating from the perspective that this is the landscape.

But we'll certainly work with all of those entities to try to find longer run solutions.

But again, one thing to really keep in mind is if -- and I think what you should take away from the slide is that there is a lot of extra, excess capacity if you can manage pre-payments to reasonable levels given today's leverage levels in the industry.

So it's really something that, if it's addressed -- it doesn't need to be addressed quickly.

It's not inhibiting an investor with a reasonable portfolio to lever where it's appropriate.

Okay, thanks for that context.

Thank you.

Bose George, KBW.

Hi.

Actually, this is Ryan O'Steen on for Bose.

Just to follow-up on the leverage comments, so you've mentioned on one hand a well-hedged portfolio and slow prepayments can support greater leverage.

Presumably, you are pretty comfortable with your hedging and your prepayments have certainly been slow.

On the other hand, you mentioned a more moderate stance seems reasonable.

Could you just help reconcile that?

Sure.

We -- I was really trying to get to -- and it's a good question.

We disclosed it.

As of March 31, our leverage was 8.4 times.

Again, mortgages -- one, we would feel comfortable levering a good -- a strong mortgage portfolio to the point -- a well-hedged portfolio and one with predictable pre-payments.

We feel our portfolio qualifies in both of those comments at a higher leverage than 8.4.

That doesn't mean that we're going to go there, so to speak.

The reason why we have sort of dampened our excitement or our view on leverage is just completely related to the price discussions.

Mortgages have done very well again in April.

Book values probably in the space are up.

We feel that -- and that is true both of generic mortgages and as we've discussed extensively on the payouts and the value attributed to specified mortgages.

So when you put those two together, the risk-return equation is less compelling than it was three months ago.

So, yes, our portfolio can handle higher leverage and we'd be very comfortable operating that way and did in Q1.

But right now, given the pretty significant book value gains that you have achieved over the past several months, you know, we feel that a more moderate stance is warranted, but we're not comfortable, as I said earlier, at low leverage right now.

There is too much risk that then later you are going to be taking leverage up at a much worse time.

And so we are not comfortable there either, and so I'd say that is why we're trying to give you guys reasonable amount of insight into our thought process today but recognizing that things could change.

Great.

Thank you.

Thanks, guys.

Look, I really appreciate everyone's interest in AGNC, and look forward to speaking to you again.

The conference is now concluded.

Thank you for your attending today's event.

You may now disconnect.