AGNC Investment Corp (AGNCL) 2014 Q4 法說會逐字稿

Good morning and welcome to the American Capital Agency fourth-quarter 2014 shareholder call.

(Operator Instructions).

Please note this event is being recorded.

I would now like to turn the conference call over to Ms. Katie Wisecarver in Investor Relations.

Ms. Wisecarver, the floor is yours, ma'am.

Thank you, Mike, and thank you all for joining American Capital Agency's fourth-quarter 2014 earnings call.

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

This conference call and corresponding slide presentation contains statements that to the extent 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 forecast, 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 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 February 17 by dialing 877-344-7529 or 412-317-0088 and the conference ID number is 10058856.

To view the slide presentation, turn to our website, AGNC.com, and click on the Q4 2014 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.

Participants on the call include Malon Wilkus, Chair and Chief Executive Officer; Sam Flax, Director, Executive Vice President, and Secretary; John Erickson, Director, 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; and Bernie Bell, Vice President and Controller.

With that, I will turn the call over to Gary Kain.

Thanks, Katie.

Good morning to everyone on the call and thank you for your continued interest in AGNC.

2014 was an important year for us and we are very pleased with AGNC's performance, both during Q4 and for the full year.

It is really hard to believe that a little over a year ago, the yield on the 10-year treasury was just over 3% and the overwhelming consensus among market participants and economists was that interest rates were headed higher as the economy improved and the Fed removed accommodation.

In addition, the conventional wisdom was that agency MBS spreads would have to widen as the Fed gradually reduced and ultimately concluded its third round of asset purchases.

Against this backdrop, many investors and analysts were bearish on the mortgage REIT space.

We stressed for investors in early February 2014 on our Q4 2013 earnings call that we disagreed with those views.

We highlighted that the technical backdrop for interest rates and agency MBS was actually very strong, given the combination of lack of mortgage origination, very defensive investor positioning, low inflation, and compelling valuations for US fixed income relative to global alternatives.

We also announced that we bought back our own stock and that we had even purchased shares in other mortgage REITs.

In hindsight, it is now clear that the conventional wisdom and the consensus outlook were wrong.

Interest rates fell materially and agency mortgages performed very well.

Against this backdrop, AGNC produced a strong economic return of 18.5%, despite running below average risk.

In addition, AGNC's total stock return was even stronger, at 27% for 2014 as our price-to-book ratio improved somewhat, albeit from very depressed levels.

In the end, it was a very solid year for the Company.

Now with respect to the fourth quarter, we saw an acceleration of the same themes that played out earlier in 2014.

Long-term rates rallied sharply and the yield curve flattened significantly.

This bull-flattening rally, coupled with an increase in interest rate volatility, should have been a relatively tough scenario for agency MBS.

Fortunately, 30-year mortgages performed well, as the favorable supply-and-demand equation outweighed deteriorating fundamentals.

This performance was a key driver of our strong economic return of 3.4% for the quarter.

We also took a number of actions to reposition our assets and hedges, given the rally in interest rates and the associated increase in prepayment uncertainty.

At a very high level, these included reducing our exposure to generic higher coupon 30-year MBS, increasing our 15-year position, adding more specified and new production pools, and adjusting the composition of our hedge book.

To be clear, we didn't make these changes because we anticipated the 50 basis-point rally we experienced in January.

That said, we pride ourselves on actively managing our portfolio, respecting market signals, and avoiding making big outright bets on interest rates.

This discipline led us to react to the changing interest rate environment as it was unfolding.

These actions turned out to be absolutely critical to our performance, both in Q4 and, more importantly, year to date in 2015.

Chris and Peter will go into some detail on these changes and why we feel the portfolio is well positioned for the current environment.

With that as the introduction, let's turn to slide 4 and I will quickly touch on a few highlights for the quarter.

Comprehensive earnings totaled $0.86 per share.

Our total net spread income improved to $0.92 per share when we include dollar roll income and exclude catch-up amortization.

Our monthly dividends totaled $0.66 per share for the quarter.

Book value increased $0.20 to $25.74 for the quarter and economic return was very solid at 3.4%.

As you can see on slide 5, our at-risk leverage closed the quarter at 6.9 times.

Moving to slide 6, we give you some information on highlights for the full year.

The key takeaway is that AGNC produced an economic return of 18.5%.

Importantly, we achieved this result while running below average risk in the form of lower aggregate leverage and average or even below average interest rate and spread risk.

At this point, I will turn the call over to Chris to discuss the changes in the market and our portfolio during the quarter.

Thanks, Gary.

Turning to slide 7, I will start with a brief review of what happened in the markets during the fourth quarter.

As you can see in the top two panels, both treasury and swap rate curves continued to flatten during the quarter, with five- and 10-year swap rates lower by 18 basis points and 36 basis points, respectively.

Despite the flattening rally in rates and the increase in volatility across many financial markets, agency mortgages performed very well during the fourth quarter, with 30-year MBS significantly outperforming 15s.

Additionally, within 30-year MBS, higher coupons performed much better than we would have expected, given the decline in rates, a flatter yield curve, and the associated increase in uncertainty surrounding prepayments.

For example, 30-year 4s increased in price by more than 1.25 points, while the five-year treasury note rallied by approximately 5/8 of 1 point.

Let's turn to slide 8 to review our investment portfolio composition.

Quarter over quarter, the investment portfolio increased slightly to $71.5 billion; however, its composition was rebalanced significantly in response to the changing interest rate and prepayment landscape.

Given the combination of the outperformance of higher-coupon MBS and the growing prepayment uncertainty, we sold a significant amount of TBA 30 or 4% mortgages in the latter half of the quarter and anticipated closing the quarter with somewhat lower leverage.

However, late in the fourth quarter, 15-year MBS cheapened significantly versus 30s, as well as versus rate hedges.

We viewed this as an opportunity and added approximately $5 billion in 15-year mortgages during the last three weeks of the year.

The 15-year sector generally holds up much better in volatile interest rate environments, as prepayment behavior is more predictable and extension is much more limited.

Our 30-year holdings peaked at 65% at the end of the third quarter and declined to 58% at the end of the fourth quarter.

Conversely, our percentage of 15-year holdings increased to 37%, up from 30% at the end of the third quarter.

Now turning to slide 9, we show the results of the actions we took within our 30-year holdings by comparing a breakdown of the portfolio as of the end of the third and fourth quarters.

Given the sharp rally in rates, the coupons most exposed to faster prepayment speeds are TBA and generic 30-year 4s and higher.

As you can see from the lower left part of this table, the 4% coupon was by far our largest holding as of September 30, totaling $20.7 billion, and it was even higher earlier in the fourth quarter.

But as of year-end, this position had declined to only $12.6 billion, with some of that decline migrating to lower coupon 30-year MBS.

Importantly, the decline in our 4% holdings came entirely from TBAs.

As such, generic 4% and higher coupons accounted for only around 8% of the total mortgage portfolio as of December 31.

To be clear, TBA 30-year 3.5s do have some exposure to faster prepayment speeds at current rate levels, but unlike the 4% coupon, the Fed is still purchasing a large amount of 3.5s, and so we are much more confident that the coupon will perform relatively well as the Fed continues to absorb the fastest paying securities.

In addition, new production 3.5% pools are readily available at negligible pay-ups and should perform fine from a prepayment perspective in the absence of a further rally.

I'd like to quickly turn your attention to slide 28 in the appendix, where we provide a further breakout of our specified pool holdings.

In an effort to further transparency in light of the new prepayment landscape, we expanded our disclosure with respect to specified pools.

Given the strength in housing over the last several years, we expect that certain LTV buckets within the HARP category will behave differently in today's interest rate environment.

For example, many of the loans that were originated under the HARP program two to three years ago with loan-to-value ratios of 80% to 100% at the time of origination have accumulated enough house price appreciation to allow these borrowers to qualify for a regular refinance.

Therefore, prepayment speeds on these pools will be less protected than they had been in the past.

The category labeled higher-quality specified pools are backed by loans with original loan balances less than or equal to $150,000, as well as loans originated through the HARP program with original LTVs greater than or equal to 100% at the time of origination.

The other specified pool category includes loans originated through the HARP program, but with original loan-to-value ratios between 80% and 100%.

This category also includes loans backed by loans with various other favorable prepayment characteristics, such as lower FICO score borrowers; investor loans; loans with original loan balances greater than $150,000, but less than or equal to $175,000; and pools backed by loans from certain geographies.

We still expect reasonably good prepayment performance from this category, but there is more uncertainty with respect to these strategies.

Lastly, on the bottom of slide 28, you can see that the vast majority of our specified pool holdings are in the higher-quality specified pool category.

Now turning back to slide 10, we provide a breakout of our 15-year MBS holdings as of the end of the fourth quarter.

As you can see in the table, over 60% of AGNC's holdings of 15-year MBS are concentrated in either 2.5% or 3% coupons.

Within the 3.5% and 4% coupon buckets, which are the most exposed from a prepayment perspective, AGNC has negligible exposure to generic pools or TBAs.

Our holdings in these coupons are almost entirely backed by seasoned lower loan balance loans, which have consistently prepaid slowly in high prepayment environments.

In addition, these pools should be more burned out at this point and their loan balances will have amortized down further as well.

Generic or TBA 15-year 3s do have some prepayment exposure at current rate levels, but these positions represent only approximately 6% of the total portfolio and there are strategies, such as buying either seasoned securities or new production pools, that can be used to lessen this exposure over time.

With that, I will turn the call over to Peter to discuss risk management.

Thanks, Chris.

I will begin with our hedge summary on slide 12.

At year-end, the notional balance of our hedge portfolio covered 76% of our debt and TBA position, unchanged from the prior quarter.

During the quarter, however, we made significant changes to the maturity profile of our hedge portfolio.

To highlight these changes, we added slide 13 to the presentation, which shows the quarter-over-quarter change in our hedge composition by instrument type and maturity.

In the current environment, where interest rate and yield curve volatility is high, hedge composition can be a significant driver of performance.

To illustrate this point, I will briefly walk through our approach to hedging a new production 30-year MBS.

Initially, we would hedge this instrument with a combination of swaps spread out across the yield curve, with a significant share of those swaps being 10 years or longer.

To this point, most models would attribute the bulk of the duration exposure on this instrument to the longer end of the yield curve.

The challenge from a hedging perspective is that the optimal hedge mix changes over time as the interest-rate environment changes.

For example, as interest rates increase, the duration of a new production 30-year MBS gradually extends.

At the same time, the duration profile also shifts, causing the MBS to gain incremental exposure to longer-term interest rates.

In this scenario, it would be very reasonable to rebalance the hedge portfolio by adding more long-term hedges to the mix.

As interest rates fall and the MBS price increases, the opposite happens in that the duration contracts and its underlying exposure to longer-term interest rates decreases.

In this scenario, it would be prudent to change the hedge mix by reducing the share of longer-term hedges and increasing the share of intermediate-term hedges.

The key takeaway from this example is that hedge composition is a dynamic process, especially in volatile interest rate environments.

As you can see from the table, we made some significant changes to our hedge portfolio.

In our swap portfolio, we increased the notional balance by about $3.5 billion, in part due to changes in our asset portfolio.

More importantly, we significantly altered the composition of the portfolio by adding shorter-term swaps with maturities in the three- to five-year range and terminating some of our longer-term swaps with maturities in the seven- to 15-year range.

The result was an almost 20% reduction in the notional amount of our longer-term swaps.

In our treasury portfolio, we undertook an even larger rebalancing.

Specifically, we sold about $5 billion of five- and seven-year treasuries and bought about $5.5 billion of 10- and 30-year treasuries.

Importantly, our 10-year treasury position shifted from a short position of $2.7 billion at the end of the third quarter to a long position of $2.4 billion at the end of the year.

Likewise, our 30-year position shifted from short $440 million to flat at year-end.

On the bottom table, we show the change in our swaption portfolio.

During the quarter, we added close to $2 billion of receiver swaptions, giving us additional down rate protection, as well as a hedge against higher interest-rate volatility.

Taken together, these changes in our swap, treasury, and swaption portfolios, which were executed throughout the fourth quarter, reduced our exposure to falling rates and gave us additional protection against a flattening of the yield curve.

Finally, on slide 14, our duration gap at year-end was long a half a year, considerably smaller than the 1.5-year duration gap that we started the year with.

The change in our duration gap over the year was the result of the significant rally in interest rates, partially offset by our ongoing rebalancing actions.

Since year-end, given the further rally in interest rates, our risk profile has shifted more toward extension risk.

With extension risk now becoming the predominant risk, we will likely operate with a small or even negative duration gap in the near term.

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

Thanks, Peter, and before we open the call to questions, I wanted to put the current interest-rate landscape in some perspective.

First and foremost, this move in rates has been dramatic, especially against the backdrop of the bearish consensus that dominated 2014.

The table on the top left of slide 15 shows the significant changes in treasury rates over the last few years.

For example, the 10-year treasury has now retraced all of the yield increase witnessed in 2013 during the taper tantrum.

On December 31, 2012, the 10-year closed at a yield of 1.75%, before rising to just over 3% at the end of 2013.

Now, as of January 31 of this year, it was 1.68% or 7 basis points lower than it was at the end of 2012.

Even more surprising is the move in the 30-year treasury, which declined almost 70 basis points since the end of 2012 and almost 170 basis points since the end of 2013.

These are incredible moves, especially given that both US growth and employment statistics have actually surprised to the upside.

On the surface, it also seems contrary to communication out of the Fed that implies they would like to begin to normalize rates around the middle of the year.

Our take is that the bond market is basically saying that long-term inflation expectations have structurally changed as a result of technological advances and the globalization of the workforce.

Market pricing also implies that global central banks will be unsuccessful at achieving their inflation targets of 2% for a very long time.

Now let's just focus on what happened this January.

10-year rates fell 50 basis points and the yield curve continued to flatten, but to a lesser extent than what we saw in Q4.

Against this backdrop, agency MBS underperformed their theoretical hedge ratios, but the bulk of this underperformance was concentrated in higher-coupon 30-year mortgages.

For example, 30-year 4% coupon TBAs appreciated in price only 34 basis points, while the five-year treasury rallied over 2 points.

To put this in some perspective, Chris showed you on slide 7 that during Q4, the opposite occurred, with 30-year 4s appreciating over 1-1/4 points, while the five-year increased at a little over 0.5 point.

What does this mean for AGNC performance?

First of all, we are glad we sold a lot of TBA 4s, and while we have not completed our monthly pricing process, my expectation as we sit here today is that our book value is relatively unchanged in January, maybe down around 1%.

When factoring in a January dividend, economic returns for the month should be close to flat.

We attribute this very respectable result to the significant actions the team described earlier, as well as to some additional moves we executed early in January.

Given that we now provide monthly NAV estimates, you will get our completed book value estimate with our dividend declaration in a couple of weeks.

But now let's turn to slide 16 and I will give you some insight into how we may react to these significant market developments.

First and foremost, the prepayment environment is no longer benign, but we are confident we know how to handle this scenario.

The key challenge will be managing pre-pays while not exposing the portfolio to significant extension risk if some of this rate move reverses, which is very possible.

To this point, we will carefully manage our exposure to lower coupon 30-year MBS and to high pay up 30-year specified pools.

For this reason, 15-year MBS will be a significant focus for us again.

Prepayments on many types of 15-year mortgages will be more contained than their 30-year counterparts, especially given the much flatter yield curve now versus prior periods of low rates.

In addition, as Chris mentioned earlier, even lower coupon generic 15-year MBS have relatively little extension risk and much shorter spread durations than those of 30 years.

Simply put, they are much easier to hedge.

With respect to hedging, to reiterate what Peter said earlier, since our asset durations have shortened significantly, look for us to run a very small or even negative duration gap with rates in this general vicinity.

You can also expect us to continue to actively manage our curve exposure relative to our asset composition and the evolving rate environment.

With respect to leverage, our January month-end leverage was slightly lower than it was at year-end.

Looking ahead, if interest rates continue to rally and mortgages cheapen, we have substantial capacity to increase leverage and would welcome that opportunity.

Alternatively, if MBS tighten and risk-adjusted returns deteriorate, we will be willing to take leverage levels even lower.

In summary, there are more moving parts in the portfolio management equation now.

However, we believe our portfolio is well positioned for the current landscape, with sufficient balance to be able to handle future volatility in either direction.

We also believe the current environment plays to our strength, which is active portfolio management, careful asset selection, and disciplined risk management.

With that, let me ask the operator to open up the lines to questions.

(Operator Instructions).

Steve DeLaney, JMP Securities.

Good morning, everyone, and congratulations on a strong close to last year.

Gary, I'd like to start by asking about the lifetime CPR.

It went up 1%, understandably, in the fourth quarter to 9%.

I am just curious timing of when the analysis was done, what the level of the 10-year was at the time you did that analysis, and that will tell us whether that analysis has to be redone.

Is that something that you might adjust monthly, given the big move that we have had, 40 basis points since year-end, or will the next change be done at the end of March?

Thanks.

Thanks, Steve, by the way, for your comments at the beginning and for the question.

First off, almost by definition we use the rates at the end of December to run that analysis and that it affects the financials for the quarter.

We feel that we have to use the interest-rate environment at the end of December for that analysis, and absolutely given, the move that we have had in interest rates, our models will calculate a different CPR as we move forward.

Now in terms of operationally, it is run monthly.

However, you'll obviously be -- we'll report the results on a quarterly basis.

Yes, clearly speed projections -- lifetime speed projections will be faster, given the rally, but that's one of the things that is really important to keep in mind is the composition of the portfolio, because the areas of the market that are most exposed to the increase in prepayments are areas that are pretty -- that make up pretty low weight in the portfolio at this point.

So I think that's one of the reasons that we feel good about how the portfolio is positioned.

Right, understand.

I assumed -- I had assumed, frankly, that it would have been 12/31 rather than going back and adjusting based on a January level.

And then, just my second point that I'd like to talk about, you have been very opportunistic about dollar rolls.

We certainly have heard some talk that they have become less special here in early 2015, but I think we find that some clients are confused.

I actually like the way you report your combined spread and dollar roll, add the two to get $0.92, but I have had some clients say that to the extent that dollar rolls become -- let's say there is no specialness, per se, to dollar rolls versus repo funding, that the majority of that $0.47 in revenue, which is 50% of earnings, would go away.

We obviously reject that, but I was just curious if there is a way to look at the dollar-roll revenue contribution and assign some range of percentage of that contribution that might be due just to the specialness of the financing.

Yes, Steve, we definitely understand those comments and it was really for that reason -- I believe it was at the end of the second-quarter 2014, and I really would encourage investors to take a look at that, we did an analysis in our earnings materials that broke out the difference between or how much you would attribute to specialness versus how much was really just regular kind of, we will call it, core income that you would have had if you took the position on balance sheet.

The result of that analysis, and it wouldn't be much different right now, was that it was about $0.10 per share that was exposed to a complete loss of specialness.

That's really the neighborhood that we are talking about.

I don't have that exact analysis for Q4, but it's not going to be materially different than that.

What investors should keep in mind is that, look, we have the option and we like the option to roll TBA positions when the financing advantages are there, but it's not a requirement, by any stretch of the imagination.

And so, -- and if all specialness disappeared, the impact on our total spread income is very manageable, especially when you think about that in relation to the dividend.

That's very helpful, and we had been -- for what it's worth, we've been using here 20% to 30% of the dollar roll contribution, so that gets us to -- at the low end gets us to that 10%.

Gary, thank you for the comments.

As always, very helpful.

Joel Houck, Wells Fargo.

My question has to do with the more forward looking.

If one were to assume that the US is going to follow other developed countries in terms of where rates are going, and assume that the US 10-year goes below 1 by the end of this year and eventually settles close to the German bund at 30 basis points, how does -- I guess two questions.

One, from a portfolio management and hedging standpoint, how does an agency REIT look?

And number two, what type of returns would be available in that type of environment, assuming that, obviously, you meet all the REIT rules and simply can't go to all cash?

If you had complete foresight, obviously, you would do that in a hedge fund environment, but given the constraints of a REIT, you can't really do that.

Those are two questions assuming that rate environment plays out the next couple of years.

Interesting question, and, look, I think it's a good question because given what we reviewed in the presentation, the idea that we would be around 1.70% in 10-year notes would have been a very far-fetched kind of idea, again, a year ago.

I think your scenario is a realistic possibility.

Again, I wouldn't -- it's nowhere near a base scenario, but it certainly should not and cannot be ruled out.

What I would say is that what's number one and most important to keep in mind is that the models that we run that show our down-rate sensitivity always assume no rebalancing actions.

As you have seen with respect to this move here, okay, that's really not a reasonable assumption.

We actively manage the portfolio, we will respect market signals and changes, and so we feel pretty confident that we can manage book value, but within reasonable bands in a scenario like that.

But what I would stress is if we end up in that scenario, investors in AGNC should actually take a lot of comfort out of one thing.

Right now, there is a significant price to book discount in the mortgage REIT space, and let's face it, a lot of that is associated with the risk that the Fed is going to hike interest rates and the concern that you're not supposed to own a mortgage REIT before a Fed tightening.

I think it's equally as logical to expect that price to book difference to shrink materially or completely evaporate in a world where the assumption is interest rates are going to be low and very low for very long.

And so unlike normally, investors have a unique situation, which is a pretty large buffer for that scenario and, again, we feel like we can manage book value within a very reasonable range in that scenario.

The other thing to keep in mind is that mortgages will widen in that scenario.

There will be some -- there will be some hit to book value despite active management, but then ROE expectations, albeit maybe lower because interest rates are so much lower, are actually going to be very good on a relative basis, given cheapening of mortgages.

And then, the other safety valve embedded in that scenario is just capacity in the mortgage market around refinances.

The actual prepayment experience will significantly outperform what models project in that environment, because there really isn't the capacity to handle refinances nearly as quickly as a model might think.

So, hopefully, that's a pretty complete answer to a very good question.

That is very thorough.

I guess on the other side of that equation, presuming we get there, you don't have to worry about prepayment risk at some point.

I assume you would agree with that.

Then, therefore, it becomes much easier to manage once you get to that environment.

Would that be a fair assessment?

Absolutely, and I think that's something that will present unique opportunities because what I would say, you don't have to worry about down-rate risk to the extent that the amount that you would have to worry about is very minimal in that if the five-year treasury is trading at 20 basis points, you know what your risk is if you are short it, so to speak.

I think there are a lot of -- that is -- again, it's a potential scenario.

It's not the scenario we envision right now, but it is not something that should be taken off the table completely.

All right.

Thanks, Gary.

Douglas Harter, Credit Suisse.

Hi, Gary.

You mentioned that potentially if you saw the market get -- or track cheap enough, you would potentially add leverage and increase.

Where do you see the market on the relative value attractiveness of the mortgage market right now?

Again, the way we think of it is risk-adjusted returns, and so, right now we view the volatility in the mortgage market -- in the interest rate market, in basically almost any market, currencies, commodities, equities, as being pretty high.

Given that even though mortgages have cheapened materially, we don't see them as, we will call it, a screaming buy at this point, we view them as probably about where they should be under the circumstances.

However, we could get more motivated in the mortgage market either due to more cheapening -- further cheapening of the product relative to treasuries and swaps or due to much more comfort level around volatility and interest rates and the likely -- what we can expect in terms of technicals and production and other factors going forward.

So that equation can change not only due to interest rates or absolute cheapening of the basis, but could also improve or, in fairness, deteriorate due to our view of risk management and potential moves in rates.

Got it, and I know you guys talked about this a little bit with some of the repositioning and more in the 15s, but I guess, just generally, last time we were here in rates and they eventually moved higher in 2013 was your one down economic return year.

How do you think about how you are positioned and protecting book value going forward if we -- the consensus is wrong again and we start moving sharply higher in rates?

Look, very, very good question, and it is something that we've absolutely thought about, and hopefully the theme of the discussion and the theme in our portfolio is one of balance.

The reality is, look, it's very straightforward to position the portfolio for a low-rate, high-prepayment environment.

Buy the lowest coupon mortgages that trade actively, Fannie Mae 3% coupons, and then own a fair amount of just very low loan balance specified pools, and you will be well positioned for a very low rate environment.

But to your point, you will be taking on a fair amount of extension risk.

From our perspective, the right approach at this point is one of a balanced portfolio.

I think that's one of the reasons why we are very attracted to the 15-year sector because we can manage extension risk and, at the same time, manage prepayment risk, and that, coupled with running a very small or potentially negative duration gap, is a way to be able to handle both scenarios.

With respect to this environment, it is a little early to assume that we have to stay down here.

On the other hand, it is also dangerous, and I think this is something that other investors -- a lot of people did in 2014, to assume that rates couldn't go lower or that we couldn't stay here.

So I think we are very focused.

Again, we have relatively low leverage, we have a very small duration gap, and we have, in our minds, a very balanced portfolio, and we think that is critical, given all the moving parts here.

Great.

Thank you, Gary.

Mike Widner, KBW.

You have mentioned, I think, just acknowledged, that spreads have widened a little bit here in Q1.

How much would you say spreads have widened?

We hear numbers 10 to 15 basis points, but what's your -- relative to treasuries, what is your sense of that?

I wouldn't disagree with that number.

What I would just say is you got to be very careful between what coupon you're looking at, whether you're looking at a TBA or a specified pool.

The higher coupons have widened significantly more, based on that -- kind of a measure like that, whereas -- but then, they had performed too well, so to speak, in Q4.

Look, I think that's a reasonable place.

The 10 to 15 is not an unreasonable number, but it's very dependent on the measure you are using and the coupon or 15s and 30s and the type of asset you're looking at.

The other thing I do want to stress is that when we hedge and when we think about this, some spread widening is absolutely built in to our thought process, and we stressed for investors on our prior call our view about the directionality of mortgage spreads, the view that if interest rates rose, mortgage spreads were likely to tighten, and if interest rates fell further, mortgage spreads were likely to widen.

So what we would say is while there was definitely underperformance of mortgages in January, a very good percentage of that would have been in our category of expected underperformance, given the scenario.

The implication of -- obviously, what I am wondering and what a lot of people are wondering is how has the book value performed quarter to date.

If I just take -- just going by the slide in your presentation, right, page 33, if I assume a 10 basis-point widening, it implies about 4% down in NAV.

All of what you are saying seems to imply, well, we did better than that kind of implication.

Am I interpreting that fairly, that you were prepared for that kind of spread widening, so maybe that number that's on page 33 is, I don't know, a little too big a drop, or is that not the right interpretation of what you are saying?

No, I think your interpretation here is reasonable.

Look, we did -- we talked -- actually, what I said in my prepared remarks specifically was that we have looked at the performance of our portfolio.

We look at it every day, and our initial view is that there is not a material change in the book value as of January.

Now we clearly haven't completed our processes.

What, we have had one business day since the end of the month, so that's an early read.

Again, it may be down 1% or something in that neighborhood, but it's a -- again, our take is and we will put this out in a couple weeks, and I think one thing I just want to stress for people right now is that when you think about this volatility, I think the monthly disclosure of NAV is a huge advantage for investors in terms of not getting big surprises here.

We intend to continue to do what we began to do toward the end of the quarter, but, yes, we executed a number of rebalancing options -- actions that we talked to you about for Q4.

We have clearly continued mostly along the same kind of themes so far in January, and as I said in the prepared remarks, those have certainly helped the performance of the portfolio.

Just to follow up on that, you are saying that you have made additional -- because we and several -- a bunch of other people on this call obviously go through and try and mark your portfolios to market, but it sounds like what you are saying is you have made additional changes since 12/31.

Obviously, that makes it a little harder for us to do the mark to market, but that's the implication.

That's sort of what you are saying is that directional set of changes continues with the rate environment changing.

Absolutely, I mean -- look, some of those -- a good chunk of the changes to the portfolio were very early -- were almost immediate in January.

Some, obviously, have come throughout the month.

But the other thing to keep in mind, I think one thing that may be missed sometimes is that the pay-ups on specified pools also appreciate and have appreciated pretty materially in January, almost, I would say, to the point where they are fully priced.

They would normally take a lot longer to move in this direction, especially given that the prepayments haven't shown up.

But one thing to keep in mind is that when you just look at the TBAs, remember, the vast majority as we talked about of our exposure in the coupons that have performed the poorest during the month are -- the vast majority of our exposure is in the form of specified pools and that materially helps the result as well.

Last one, if I may.

Since you brought up the topic, if I rewind the clock, what, two years now, there was fairly big premiums for pre-pay protected stuff, and then we watched as the year progressed those prepayment premiums -- the pre-pay protection premiums evaporated and you guys, as well as a number of others, took some pretty significant book value hits over the course of the first half of the year.

How do you think about those prepaid protected securities, given what you just said about the premiums getting to be sizable again?

Excellent question.

Look, and I am glad we got to this topic.

First, let's rewind to 2010, 2011, 2012.

We were very consistent and we were always in prepayment or from the beginning of our exposure to 30- and 15-year mortgages had a very high percentage of prepayment protected mortgages throughout that entire multi-year period.

They did absolutely appreciate, as prepayments over that period were generally very fast, and we achieved tremendous returns on those positions.

We actually -- and so, yes, there was the first half of 2013, we gave some of that back, a small percentage of that back, and there is -- that is a correct interpretation of what happened.

What's different this time, what's interesting is we don't know how long we're going to stay down here.

We don't know -- we are not getting to successive new lows in interest rates, like we did the last time around.

What I would say is today the equation for specified pools is actually much weaker, okay, than it was the last time around.

From our perspective, what you have seen is not a tremendous increase in our holdings of specified pools.

You have seen a decrease in our holdings of generic and TBA assets for exactly that reason.

Look, and I think -- look, while we also learn from history, we're always trying to improve what we are doing, and what I would say is we do actually think it's -- specified pools are not necessarily the perfect solution for this environment yet.

That could change, but, on the other hand, sticking with generic higher coupons is at least equally risky as well.

But it just raises the question of generic lower coupons, and I think to some of what Joel said, at some point you get to a place where the risks are one sided, to a large degree, and everyone likes to buy low, sell high, and to the extent your pre-pay protected stuff is getting high -- I don't know.

I think you have answered most of that as much as you're going to, but it's an interesting environment.

Yes, it definitely is interesting.

Thanks, Mike.

Thank you.

Arren Cyganovich, Evercore ISI.

The dollar roll that you are maintaining, I guess it was around -- a little under $15 billion at the end of the quarter.

Are you expecting to continue to run around that rate?

Then I guess, also, I'm assuming and I think you said they are more of 3% 30 year today.

What is the thought process in protecting price exposure there if rates do jump back up?

Thanks for the question.

Rolls traded well throughout most of the fourth quarter.

We did bring the position down towards the end of the quarter, as you mentioned.

We have talked about the drivers behind our decision that led us to be comfortable running a relatively large roll position in 2014.

Obviously, one was that implied financing rates were very attractive versus on balance sheet repo.

That's still the case in some coupons, less so on higher coupons.

But the other significant consideration was that we were in an extremely benign refi environment, and so, given the sharp move lower in rates over the last two months, prepayment risk is very much back on the table, as we've talked about.

Our roll position was lower as of year-end; it may come down a little bit, but we're still going to have opportunities to roll positions, especially in lower coupons, which are currently in the 30 to 40 basis points through on balance sheet repo area.

Just a couple things I'd like to add.

You also asked about the risk in lower coupons if interest rates go up, and clearly, that risk exists and that's what the hedge book is for.

We still have substantial 10-year hedges.

We are running a very small duration gap and we'll continue to rebalance along those lines.

But it goes to my earlier comments around, again, the flexibility, and what you'll notice in our portfolio has been there has been some increase in the 3% coupon holdings, but it's still the smallest of the three main coupons in 30 years.

We understand that and we feel that in the 3% and 3.5% coupons, there are a lot of -- there is a lot of flexibility.

You can take advantage of roll specialness, but you want to balance that.

Again, I think that is why -- it is the confluence of all this discussion that is why I think 15 years, almost by default, end up being a focus for us.

Thanks.

Then the new money spreads, at least from what I could see over the past few days, seem to be pretty low relative to historical.

What are your thoughts there if rates stay in this level?

Should we expect some portfolio spread compression or how are you looking to combat that?

I think you have two questions in there.

New purchase spreads are very dependent and we don't like to, in a sense, rattle off a spread.

It depends what you buy, how you hedge it, and so forth.

These are moving parts in this environment.

What I would say is, look, mortgages again have definitely cheapened up some, as we would have expected in an environment like this, but things are in flux.

Let's be practical.

There may be some very good purchase opportunities.

If we back up in rates, mortgages are going to probably tighten pretty materially, in which case new money spreads won't be that attractive, but book value should do pretty well.

But on the other piece in terms of is there going to be spread compression in the portfolio, yes, the yield curve is flatter.

That's not a net positive.

Yes, prepayment speeds are biased higher.

That's not a big positive.

The real question is going to end up being the biggest driver of what happens to, we will call it, the total earnings potential is going to be where leverage ends up.

To some degree, that's going to be a hedge against book value because if book value does extremely well from here, we are probably not going to be willing to take leverage up dramatically, and vice versa.

In the scenarios like where mortgage spreads widen, then earnings are going to be a lot better.

So we're -- it's very difficult to give a projection on where -- which of those scenarios is likely to unfold.

That's fair enough.

Then just lastly, what are your current expectations when the Fed would start tapering its reinvestments?

Do you expect that they will occur around the first rate hike, which it seems like mortgage participants are looking maybe in the September level?

No, I think it will be noticeably and I don't -- after the first great hike and, look, I wouldn't say we have any special insight here.

What I would say is more that I think the conventional wisdom or the general mindset is maybe it would start six months into a rate hike.

I am not saying it will, but I think reinvest -- changes in reinvestment policy are a little ways off still.

Obviously, we could get, and the Fed has tried indicate that the first rate hike could occur in the middle of the year.

The market seems to be saying it is going to occur later than that, which also pushes off the reinvestment -- the end of reinvestment.

Again, the Fed has also indicated they would likely taper their end of reinvestment as well, so that pushes it back on a weighted average basis even further.

Okay, thanks a lot.

[John Carmichael], Investor.

Thank you, guys, for taking my call and appreciate the color that you guys have given and congratulations on dealing with a difficult environment.

I thought you did a really good job.

My primary question has to do with given all that you have talked about and the difficult environment where we are and situation that we are in with the difficult purchase environment and spread environment, et cetera, it seems like a perfect environment to revisit share buybacks.

I'm wondering if you could address that.

Look, that's a very good question, and let me just review for you how we think about that equation and we can compare this to last year.

Look, I think management has demonstrated that we understand the benefits of accretion and share buybacks, and we are prepared to execute them when we feel they are compelling.

What I would say is, look, there are a number of different factors that go into the equation.

Fair enough, the largest is the price-to-book discount.

That's the first and foremost thing that's thought about.

But in a sense, and it's true for investors, you are taking a levered -- you are -- in a sense, the Company takes a levered position in agency mortgages.

Embedded in there is some interest rate exposure, obviously, and there is exposure to mortgage basis.

The one thing I would say is if you compare this environment to a little over a year ago, in our minds and we stressed this a lot on the first-quarter call -- I mean, on the fourth-quarter call for 2013, which was the 10-year treasury was at 3%.

In a sense, we felt it was oversold.

The mortgage basis had widened dramatically in 2013.

We felt it was, to use the same term, oversold.

There were a lot of different factors that all came together at the same time.

When you lose one or two of those factors, then the onus on the remaining factors has to be a little bigger.

So I just want to reiterate very much committed to share buybacks, stand ready to execute them, as we have demonstrated the past, but this is the way you have to think about that equation.

Okay, I guess I am not sure that I understood the answer in respect to the current situation where you've got such a big spread to price to book and you've got also compressing spreads.

In other words, it doesn't seem like new purchases would be as attractive.

I think what you are -- I think what you have to consider there -- first off, I'm not in a position to say whether or not we are going to execute or at what price-to-book discount we're going to execute share buybacks.

We certainly stand ready to, but I do want to reiterate that the price to book is a key variable.

The attractiveness of underlying mortgages and wherever terms are is another variable as well, and those are all going to be factored into our decision.

Okay, and on the point that the last caller just made, I realize it's a macro call and nothing that you guys can really influence, but I guess I'm just -- I am confused as to why the Fed wouldn't consider tapering their reinvestments and even consider, to tell you the truth, starting to sell off some of their portfolio, if not just to stabilize spreads?

Because I think there is some benefits to the overall economy and negatives to the overall economy if the spreads to -- I am talking about spread compression between the long- and short-term rates -- get too narrow.

Even as to other countries, you get too much money flowing into the United States, et cetera, et cetera, you would think that the time for the Fed to act would be when loan rates are going down, as opposed to the opposite of then you are adding fuel to the fire if you start tapering when rates are going up.

Look, I understand the question.

What I would just say is, look, I think the Fed, rightfully so, is very focused on transparency.

What they have indicated is their first move is -- their first concern is starting the normalization process on rates, then at which case they will potentially consider adding to the balance sheet.

I think the transparency is a key driver for them and I wouldn't expect them to flip flop on that, and with that, I think we will move on to the next question.

Ken Bruce, Bank of America.

Thanks for sliding me in.

You have been very nimble with respect to changes in agency policy and the like, and I'm interested in your view as to any changes around the guarantee fee, loan level pricing adjustments, and adverse market condition fees that may be taken, what the implications would be for prepayments, whether you think those are reflected in models in any capacity, and just general thoughts around that topic.

Thanks, Ken, and a very good question and timely, given that FHA clearly has reduced their fees.

There is a lot of market chatter around FHFA and the GSEs.

Let me give you my take.

The first is that what Director Watt said is that they are still evaluating it and the earliest we will hear something is toward the end of this quarter.

It's a little ways off.

Second of all, we wouldn't expect any change from them to be that big, maybe on the order of 10 basis points in aggregate G-fee or something with respect to the loan level pricing or what's called in the market the LLPAs.

But, look, my personal expectation is we will see something from them, and I would take it up like four notches and say the reason is is if you just look at where jumbo mortgage rates are versus conforming rates, that gives you an indication that maybe the GSE G-fees aren't necessarily where they should be, either from a policy perspective or even potentially from a risk perspective.

Second of all, and you can interpret this different ways, the GSE risk transfer securities also indicate that the GSEs are making a decent amount of money at the current G-fees.

So my personal expectation is that we very easily could see some reduction there.

I really don't think it's a big picture issue for the market.

I think that something on the order of a 10 basis-point change in G-fees is a relatively small change in the different variables that affect prepayments and origination at this point, but, again, I think it's a little ways off.

We don't really view it as a big-picture issue, but it is something that guides how we think about exposure to different products and it benefits some things and it makes you more concerned about other things, and we use that in our investment decisions.

Great, and just on the same lines, is there any way to gauge what has been viewed by many as banks' unwillingness to participate in the originations market with the higher-risk mortgages, purely as they look at put-back risk and the like?

Is there any way to assess that and how that has, if any, any implications for the underlying MBS?

I don't think you can specifically quantify it.

I think, look, it's quite clear that some -- in particular, the large banks -- are -- I guess the easiest way to describe it is -- are originating quite a bit of mortgage product, but they are working for wider margins and I think a big -- whether you want to talk about it as put-back risk from the GSEs or you want to talk about it as other kind of regulatory uncertainty in terms of some of the fines that have been dealt out in the mortgage market, I actually -- I think that that is one factor that is in the market and it is definitely something that will help in terms of containing the refinance picture going forward.

I don't think that, in the case of the largest banks, they're going to sit there and fight for a couple points of additional market share in this environment.

I think they will be very happy to make wider spreads that clearly compensate them for this type of risk.

And so, I think that is something that will keep prepayments under control.

Great, thank you for your comments.

Appreciate it.

That will conclude our question-and-answer session.

I would now like to turn the conference back over to Mr. Gary Kain for any closing remarks.

Sir?

Thank you very much, and I want to thank everyone for your participation on the call and we look forward to talking to you next quarter as well.

Thanks again.

We thank you, sir, for your time today and to the rest of the management team.

As a reminder, an archive of this presentation will be available on AGNC's website and the telephone recording of this call can be accessed through February 17, 2015, by dialing 877-344-7529, using the conference ID 10058856.

The conference is now concluded.

We thank you all for attending today's presentation.

At this time, you may disconnect your lines.

Thank you and have a great day, everyone.