AGNC Investment Corp (AGNCN) 2014 Q3 法說會逐字稿

Good morning, and welcome to the American Capital agency third-quarter 2014 shareholder call.

All participants will be in listen-only mode.

(Operator Instructions) After today's presentation, there will be an opportunity to ask questions.

(Operator Instructions) Please note this event is being recorded.

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

Please go ahead.

Thank you, Paul.

And thank you all for joining American Capital Agency's third-quarter 2014 earnings call.

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

This conference call and the corresponding slide presentation contain 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 protections 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 factor 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 November 12 by dialing 877-344-7529 or 412-317-0088.

The conference ID number is 10054108.

To view the slide presentation, turn to our website, AGNC.com, and click on the Q3 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 Kewell, 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.

And I want to thank all of you for your interest in AGNC.

Before reviewing our third-quarter highlights, I want to briefly discuss some of the recent changes we've witnessed with respect to the global economic landscape and to interest rates.

At the beginning of the year, the 10-year treasury was at 3%, and there was a clear consensus that rates were only headed higher and global economic growth would continue to improve.

As we sit here today, however, the opposite has clearly occurred.

The 10-year has rallied about 75 basis points, and the yield on the 30-year treasury has declined by almost 100 basis points during 2014.

At the same time, the global economic growth picture has slowed, and inflation expectations both abroad and here in the States have fallen.

This was clearly not what market participants were expecting when we started the year.

Against this backdrop of falling interest rates and a flattening of the yield curve, agency mortgages have performed surprisingly well.

In fact, this strong performance was in sharp contrast to the consensus view at the beginning of the year, which is that spreads would widen as the Fed reduced their mortgage purchases.

QE3 is essentially done at this point, and the mortgage market remains well supported, with many investors still significantly underinvested in the product.

The key takeaway from all of this is that investors should always question the consensus and the conventional wisdom.

As a manager of a levered MBS portfolio, environments like these serve as an important reminder that we have to be prepared for a wide range of possible scenarios and not obsess about one particular outcome.

Moreover, they reinforce the value of an active portfolio management philosophy that stresses the need to reevaluate and adjust our assets and hedges in light of an ever-changing interest rate and mortgage market landscape.

So what does this new environment mean for AGNC?

I would describe the current environment as being well suited to our asset selection and hedging strategies in that the risk profile in today's market is more balanced.

At a high level, the trade-offs between prepayment risk and extension risk are now more evenly weighted.

The same can be said for expectations around interest rates, as the positive momentum in the US has to be weighed versus significant headwinds abroad, a stronger dollar and considerably more benign inflation expectations.

I generally prefer environments like these because relative value opportunities tend to be more frequent.

For example, prepayment expertise is obviously more valuable today than it has been recently.

This is true with respect to evaluating the sectors where market pricing creates opportunities and also knowing what sectors to avoid.

As such, this environment has the potential to be more like 2010 and 2011 than what we have seen over the past couple years.

With that as the introduction, let me quickly touch on a few highlights on slide 4. In the third quarter, MBS spreads gave back a small percentage of the outperformance they enjoyed in Q1 and Q2.

As such, book value was down 2.7%, and our economic earnings were essentially flat for the quarter and still around 15% year to date.

Comprehensive income totaled a loss of $0.07 per share, comprised of $0.54 of net income and a loss of $0.61 in OCI.

Net spread income inclusive of dollar rolls totaled $0.85 per share.

As we discussed last quarter, dollar roll specialness continues to supplement our returns.

But even if this funding advantage disappeared, our total net spread income would still exceed our $0.66 quarterly dividend run rate.

I do want to reiterate one thing that we mentioned at our investor day earlier this month, which is that we currently expect all of our 2014 dividends to be characterized as ordinary income.

While our dividends will likely exceed our taxable income, they will not exceed our taxable earnings and profit and therefore will not include any return of capital for shareholders.

Turning to slide five, our portfolio decreased slightly to just under $70 billion during the quarter.

Our at-risk leverage was down slightly to 6.7 times at the end of the quarter.

With that, let me turn the call over to Chris to discuss how our mortgage portfolio is currently positioned.

Thanks, Gary.

Turning to slide 6, I will start with a brief review of what happened in the markets during the third quarter.

Given the extreme move in rates in early October, I will touch on the fourth quarter to date as well.

As you can see in the top two panels, both treasury and swap rate curves continued to flatten during the third quarter, with five-year swap rates selling off 23 basis points and 10-year swap rates essentially unchanged as of September 30.

Mortgages generally underperformed hedges.

But against the backdrop of the strong performance in the second quarter, the move was relatively small.

So far into the fourth quarter, rates have rallied materially, with five-year and 10-year swap rates lower by approximately 20 basis points.

Mortgages initially lagged the move lower in rates but quickly recovered, given the combination of manageable supply, Fed reinvestment, demand from a range of investors and no sustainable secondary selling.

Let's turn to slide 7 to review our investment portfolio composition.

As Gary mentioned, at-risk leverage was down from 6.9% at the start of the quarter to 6.7% as of September 30, given a slight decrease in our investment portfolio to $69.5 billion.

During the quarter, we continued to reduce our 15-year MBS position, given the combination of a flatter yield curve, limited extension risk in our aggregate portfolio and our overall comfort with interest rate and basis risk exposure of 30-year MBS.

As of September 30, our 15-year position represented 30% of our total portfolio, down from 38% at the start of the quarter and 51% at the beginning of the year.

Our 30-year MBS position increased during the quarter by approximately $4.5 billion.

In the tables at the bottom of slide 7, we added a column showing the percentage of our on-balance sheet pools that are backed by loans with lower loan balances and loans that were originated through the HARP program.

As you can see, this represents more than 70% of our on-balance sheet holdings.

Given the sharp move lower in rates since the end of the third quarter, we do expect that prepayment speeds will pick up on certain segments of the mortgage market.

However, it's important to recognize that higher overall origination costs and borrower burnouts will likely limit the extent of the refi response relative to the response experienced at similar rate levels in prior years.

From a valuation perspective, a fundamental difference today compared with 2012 and early 2013 is that the quality of tradable float in most coupons is considerably better, given Fed ownership of a significant portion of the worst-to-deliver bonds.

In other words, prepayment differences on TBA deliverable bonds versus call-protected pools will be smaller even in a move to lower interest rates.

The corollary to this is that we expect dollar rolls to also hold up better into lower rates than in 2012 and 2013, given that investor expectations of carry erosion will be less severe.

With respect to our TBA position in today's rate environment, we are not concerned about prepayment speeds on 30-year 3%s, 3.5%s, 15-year 2.5%s and 3%s.

Increases in prepayment speeds will likely be concentrated in coupons with a significant incentive to refinance, such as 15-year 3.5%s and 30-year 4.5%s.

TBA 30-year 4%s do have some exposure in a move to lower interest rates.

However, it is critical to maintain a balance between current income and performance in both rising and falling rate environments.

As we have discussed in the past, we manage our portfolio to perform across a range of potential environments.

And while the move in rates thus far into the fourth quarter wasn't our base case, we are well positioned for it.

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

Thanks, Chris.

I will start with our funding activity on slide 8. Our financing position continues to be very strong, with good access to longer-term funding and significant excess capacity.

In fact, our total repo funding capacity at quarter end was about double our outstanding repo position.

This significant excess capacity gives us the flexibility to take delivery of open TBA positions or grow our portfolio opportunistically.

On the next slide, we provide a summary of our hedge positions.

At quarter end, our hedge portfolio totaled just over $48 billion and covered 76% of our debt and TBA position.

Given the selloff in shorter-term swap rates during the quarter, the market value of our swap portfolio increased $94 million.

The fair value of our treasury hedges, on the other hand, fell by $52 million, given the longer maturity of these hedges and the flattening of the yield curve that occurred during the quarter.

When hedging the market value of agency MBS, it is imperative to have hedges across the yield curve.

To this point, given the composition of our portfolio, the market value of our assets is more sensitive to longer-term interest rates than shorter-term interest rates.

Turning to slide 10, we provide a summary of our quarter-end duration gap as well as what our duration gap would be after various interest rate shocks both up and down.

At quarter end, our current duration gap was long, one year, unchanged from the prior quarter.

In addition, we also provide an estimate of how our duration gap changes in response to various interest rate shocks and assuming no rebalancing actions.

We provide this measure of duration risk to help investors better understand our exposure to both prepayment risk and extension risk.

As shown in the table, if interest rates drop by 100 basis points, our duration gap will naturally shrink by almost two years, leaving us with a duration gap of negative 0.9 years.

Conversely, if interest rates increase by 100 basis points, our duration gap will naturally extend by one year, leaving us with a duration gap of two years long.

As you can see, the movement in our duration gap in the down rate scenario is about double the movement in our duration gap in the up rate scenario.

Our decision to operate with a positive duration gap in the current environment is driven in part by the fact that our assets base more contraction risk than extension risk.

Another important factor in our decision to operate with a positive duration gap is our assessment of the correlation between interest rates and mortgage spreads.

In the current environment, we expect mortgage spreads to widen as interest rates fall and tighten as interest rates rise.

This correlation is opposite to what we experienced in 2013.

If mortgage spreads do indeed move in this way, our aggregate exposure to rising rates, inclusive of both interest rate risk and spread risk, will be less than what is implied by our generation gap or NAV sensitivity tables.

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

Thanks Peter.

As a reminder and as we discussed at our investor day on October 1, AGNC is now paying a monthly dividend instead of a quarterly one.

We believe this makes sense for a couple reasons.

First of all, our assets pay interest and principal monthly, so the income is generated on a monthly basis.

Secondly, we typically just invest this cash intra-quarter in short-term funds which pay very little interest because it is not prudent to leverage that cash when you know you have to pay it out in two more months.

In other words, we believe our investors should get the money sooner and be in a position to invest it or use it as they please.

The first monthly dividend of $0.22 was declared on October 16 and will be payable on November 7. In addition, we will be disclosing estimates of our month-end book values beginning in mid-November, which will relate to our October 31 NAV.

The monthly dividends and NAV disclosures are designed to further enhance our transparency and improve the value proposition for our shareholders.

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

(Operator Instructions) Steve DeLaney, JMP Securities.

Gary, before I ask my question I just want to congratulate you and the team on an excellent analyst day October 1. I hope you will do it again next year.

Obviously, rolls continue to be very important.

And I wonder if you could talk a little bit about how these rolls -- this is a very simple question, Gary -- how rolls become more or less special, given an absolute move in rates.

I guess what's running through my mind is that rates move higher and prepaid risk is reduced, and maybe the generics have less risk rather than the spec pools.

Is it too simplistic to think about it that way?

And I guess the conclusion I'm drawing is if we are down 25, 30 basis points from September rates to the current, then I'm sitting here with the assumption that rolls are probably a little less special today than they were in September.

Thanks.

Steve, look, first off, thank you very much for the comments on the investor day.

We obviously took that very seriously, and we do plan to do it again in the future.

With respect to the dollar roll question, and it is -- it's a very good question, but I want to break out two things.

One is what you mentioned is the specialness, and the other piece is the absolute dollar roll level.

And so implied by your question, the absolute dollar roll level typically would decline as prepayment estimates for an underlying coupon pick up.

So if I used 30-year 4%s as an example, if the prepayment estimate for 30-year 4%s goes from 8 CPR to 15 CPR -- and those are just hypothetical numbers -- then the carry-on and on-balance sheet position, right, if you just put that on repo, would drop because you have a faster prepayment speed.

In theory, the dollar roll level should also decline to adjust to faster expectations on the underlying pools.

However, that often is not the case in that what can happen is, as those changes are happening, if things like the availability of float in the underlying coupon changes or if there are short positions that need to roll those positions, then there's no specific need for that theoretical adjustment to take place.

So what I would say is you are absolutely right in assuming that absolute dollar roll levels will have a downward tendency if prepayment estimates -- in higher coupons where prepayment estimates are picking up.

But that doesn't always or sometimes takes a while to factor through.

But with respect to specialness, which is the incremental value at a new prepayment speed, I think you should assume that's much, much less correlated with interest rates.

And actually, in the first move down it's very possible for the specialness to actually increase.

Okay.

That's helpful.

Appreciate it.

And I guess the message there is we always have to watch the technicals as well as the fundamentals in the marketplace.

Gary, I don't know if you are paying a lot of attention to Mel Watt at the FHFA.

I suspect you are because the press kind of does that work for us.

But I have been really surprised by the more aggressive posture that is being -- the regulator is taking with respect to the GSEs at this point in terms of credit risk, 97% LTVs, maybe doing away with loan-level pricing adjustments, et cetera, et cetera, all driven towards affordability.

So this is a pie-in-the-sky question.

But given that Congress has really not passed any legislation with respect to the GSEs, is there any scenario with the Fed backing away?

And from -- as you concluded with taper and maybe eventually with reinvestment, is there any scenario in which you could see the regulator actually opening up the GSEs trading activity and allow them to, once again, play a more active role in the MBS market?

Thanks for the time.

I appreciate the question.

So first off, I just want to mention that our take on Mel Watt's comments and the move on the part of FHFA is just maybe we will call it a little more balanced in terms of conservatorship objectives and housing market objectives.

We expect the changes that are implied in those announcements to be very much at the margin and not wholesale in nature and not really something that's going to materially impact the mortgage market anytime soon.

Now, with respect to your comment or question around the GSE portfolios, the shrinkage of the GSE portfolios is mandated by the preferred stock agreements.

And we see almost no chance that those would be adjusted.

So while the GSE's guarantee business could adjust over time and will adjust over time, we actually see the GSE's moving much more, in both case -- in both the portfolio case and on the credit risk retention side, toward laying off risk to the markets as a whole.

And I think one of the key things that's not getting enough attention at the GSE is the quantity of credit risk that they are starting to lay off.

And we discussed that at the investor day.

So in terms of risk retention at the GSEs, we feel like they are still going in the other direction.

Got it.

Thanks for the time, Gary.

Douglas Harter, Credit Suisse.

Can you talk about how you view the range of the 10-year and how you are going to position your duration gap around those levels?

Well, first, I want to thank you for not having asked me that last quarter so I could have new information before -- no, I'm just kidding -- before having to answer it.

Look, we feel that one of the advantages of our portfolio, and this is implied by the generation gap sensitivity that Peter reviewed earlier on the call and at the investor day, is that even with 25-, 30-basis-point moves in either direction, our duration gap still remains relatively stable and still remains in a range that we are very comfortable operating in.

And so what I would say is our requirement or need to, as they say in the mortgage market, delta hedge is pretty low even though we are at a somewhat sensitive level of interest rates.

Now, with respect to just our views on interest rates in the near term, look, we respect the market action.

We respect the international factors that are impacting the global economy but also global interest rates.

And it seems difficult to fight that in the short run.

And for that reason, our mindset is that we are in a newer rate range at this point.

That rate range is probably 25 to 30 basis points lower than the 2.60%-ish area we were bouncing around two or three months ago.

And, again, those are big-picture factors and forces, and we certainly respect them.

I don't know, Peter, if you want to add anything.

Well, I would just add that when you look at the way we have set our duration gap this year, as Gary mentioned, at the beginning of the year when rates were 3%, our duration gap was about a year and a half long.

And increases in rates from there would have only put us with a duration gap of around two years, so a very manageable position.

When you look at it in the down rate scenario, given the way we set our duration gap, in the down rate scenario our duration gap gets close to zero, if not negative, when rates pass through, around 2%.

So, for example, the other day when we had the significant rally intraday, our duration gap was probably negative at that point when rates hit their low.

So we feel like our duration gap is well suited for the range that we are -- the wide range of 10-year interest rates that we are currently in.

Got it.

Thank you.

And can you talk about the volatility that you saw intra-quarter of the dollar roll spread?

Hey, Doug.

Thanks.

Yes, so roll-implied financing rates were widened early in the third quarter.

They did better towards the end of the third quarter.

Just to give you a few points, the 30-year 3.5% role is currently trading around negative 50 basis points at 6 CPR.

The 4% role has cheapened up a little bit, but even at 8 CPR it's around negative 30 basis points.

So generally speaking, roles are still trading extremely well relative to on-balance sheet repo financing.

Yes, so , just at a very high level for the year, rolls richened a lot in the second quarter.

Early in the third quarter, they weakened significantly, but from very rich levels.

And really, over the last month or two, they have improved again since the -- from the lows, we'll say, in July and early August.

So again, that's the big-picture trajectory.

Great.

Thank you.

Arren Cyganovich, Evercore.

My question is around the duration.

I guess you say prepayment risk exceeds extension risk, which you can clearly see in your duration table.

But how that relates to the portfolio, if you look at the sensitivity stuff to book value you have later in the slide deck, it shows still more risk to higher rate on book values.

Is the prepayment risk more on an earnings perspective for the portfolio?

Is that what you are talking about there?

I'll start, and either Chris or Peter may want to chime in.

But what you should -- the way you should interpret that comment is contraction risk in our duration gap, the duration of assets or our portfolio, is greater than the risk of extension in the portfolio.

As such, our duration gap extends only a year, whereas it can contract two years or, in that example, 100-basis point moves.

So contraction risk is bigger.

We don't want to give the wrong impression that the current environment creates prepayment challenges only for a couple of coupons, which we have very small positions in.

So we are not worried about the current prepayment environment from the perspective of earnings or the impact on our positions.

But what we are saying is that, to the extent that we get a 25- or 50-basis point move in either direction, the impact on our duration gap or on the overall interest rate sensitivity of the portfolio will be bigger, the change will be bigger in a 50-basis-point decline than it would be in a 50-basis-point increase.

Just to add to that, when you look at the sensitivity tables that we put in the back of our presentation and in my comments at the beginning, part of our decision to operate with a positive duration gap is our belief that mortgage spreads will tighten in a selloff and widen in a rally.

So in a sense, we are using our duration gap to give us some incremental hedge protection against the mortgage basis widening against us in a selloff.

So if you look at our NAV sensitivity table, in a 100-basis-point scenario where we estimate that our loss to our portfolio value will be negative 12%, we don't expect it to be that negative because we expect mortgage spreads to tighten in that scenario.

So really we expect some benefit from the table below.

And what's interesting is the contrast to that and the reason, if you remember.

And we were very explicit on calls in the middle of last year, the corollary last year was that when interest rates rose, mortgage spreads widened.

And so those two factors worked in the same direction, and that's why there was so much volatility in book values during the middle of last year.

And again, given that we are reasonably confident and what we've seen year to date and even quarter to date in this latest move is that mortgage spreads definitely seemed much better supported or tend to tighten when rates go up and tend to widen when rates go down.

And that provides, in a sense, much less interest rate sensitivity than what would be implied by the models alone.

That's helpful.

Actually, if you could just touch on that, why -- what drives that in this environment whenever you see spreads tightening as rates rise?

Is there anything technical that's driving that?

Could you help me better understand that?

Yes, there are a couple components to that.

One is that a reasonable -- first off, there is the expectation around supply and origination and the realities around origination.

As interest rates -- when interest rates were around 2.50% on the 10-year or higher, the amount of new production, the refinancing activity, is very low.

Purchase activity has been subdued for a while now.

And in that kind of rate range, mortgage origination activity is very low.

At the same time, when you are at higher rates, yield-oriented buyers, and this include overseas investors and US depository institutions are probably the two biggest, the demand from these yield-oriented buyers picks up.

So the combination -- you have a combination of both lower supply and more yield-based buying.

And that drives the tighter spreads in higher-rate environments, so to speak, or when rates are headed higher.

And conversely, the same things reverse, at least temporarily.

The yield-oriented buyers typically back away from the mortgage market if there has been a 20-basis-point rally and wait, hopefully, for higher yields before they sometimes adjust their bogies.

At the same time, refinancing activity picks up, and so the supply picture in the mortgage market also increases.

While some of that is temporary, it is enough to certainly impact spreads for some period of time.

But those are the drivers of that directionality that we talked about.

Thanks.

That was very helpful.

Joel Houck, Wells Fargo.

Maybe just to stay on the topic, what is the more normal environment in terms of the relationship between spreads and rates?

Is it the tightening spread environment when rates rise, or is it what we saw last year?

Actually, this is the more normal environment.

Generally speaking, there is a bias for mortgage spreads to widen some when rates are dropping and prepayment risk is increasing, supply is picking up.

On the other hand, I would say this is -- the current environment is certainly more -- those moves are larger than they would typically be.

The environment last year is really consistent with when you are at very, very low levels of rates and mortgage durations are as short as they can be, then you have this issue where extension risk and convexity issues tend to push things in the other direction.

So there's no one -- as we are talking about, these things move around.

But the more normal environment is one where mortgage spreads tend to be pushed a little wider as rates fall.

But clearly, there are times when that's not the case.

Right.

Okay.

No, that's actually very helpful.

Now, does it also follow that in this environment which persists most of the time, is that the volatility of your book value would be less just simply because you are getting offsets one way or the other as opposed to all or nothing, like you saw last year?

Yes, absolutely.

And this is why we have tried throughout the year and even when we, to Peter's point, we were running a 1.5-year duration gap at the end of 2013.

What we have tried to stress and what we tried to stress on the calls last year was given that both spread moves wider and interest rate increases were happening at the same time, it forced us to be incredibly defensive.

We really weren't willing to operate with much of the duration gap.

We had purchased a lot of options.

We were forced -- we had moved to 15 years.

We had to do a lot of things because you couldn't feel comfortable running a duration gap when you knew not only would you lose on that duration gap if interest rates went up, but you would lose on spreads.

You fast-forward that to today, and, to Peter's point, when you lose -- if you some money when rates are going up due to the interest rate component or the duration component, that's an environment where spreads are -- if that is an environment where spreads are actually tightening, then you are actually making money on your spread exposure, which then reduces materially at times the amount of money or the performance -- improves the performance of your aggregate portfolio quite a bit in a rising rate environment.

Conversely, while you might initially think you are going to make a lot of money on a long-duration position, the spread-widening impacts can offset that as well.

And so, to your point -- and this is something that we mentioned earlier on this call; Peter highlighted this.

This is a much more comfortable environment for us because of the fact that aggregate risk to our NAV is lower because of this spread relationship.

And the environment we saw last year was something that forced us to be extremely defensive.

Yes.

And if I recall correctly, some of your best quarters were when rates were rising, presumably because you saw that spread tightening.

But maybe just to switch gears here little bit, and I guess this is one maybe -- if you don't want to answer, that's fine.

But the conventional wisdom here this week obviously is that QE is going to end.

How would you handicap, given everything that's going on globally, the likelihood that the Fed has to come in next year and restart QE?

I don't know if you want to put a probability percent on it or just punt the whole thing.

But I would just be interested to hear your perspective.

Well, I would always prefer to punt, but I won't.

When we look at -- we absolutely expect the Fed to end QE3 I guess it's tomorrow.

And we feel the market expects that; it's completely priced in.

And clearly the markets are -- the bond market is much more forward thinking than sometimes investors give it credit for.

And what I would say is the end of QE3 didn't create the dislocations -- look at 2014 -- that some people expected.

What I would say is what we think is the more -- given the global weakness that we are seeing, given issues with exchange rates and price of oil and so forth, we absolutely expect the inflation picture to be benign and inflation globally but also in the US to remain well below the Fed's target.

Where we see that coming into play is less likely with respect to QE4, but we're not going to rule it out.

But we would put that as a relatively low probability.

But I think what is growing in probability.

And what explains the move in interest rates is that the Fed is potentially on hold for a lot longer than what the market had been pricing in instead of a baseline assumption of the first rate hike being in September of next year.

The probability that it's materially later, not three months but, let's say, a year or two later, has increased.

I think that's what you are seeing in the markets, and that's what we expect.

Again, we are not willing to operate and run our portfolio based on the assumption that the Fed will never raise interest rates and they might put another QE4 in place.

But we also have to operate our portfolio understanding that there's a greater probability that the Fed is now going to be on hold for multiple years.

And the reason is that there are very few periods in history, I don't know if there is one, where you have seen any kind of tightening moves by a central bank when they are significantly missing on their inflation objective.

So that's how we are thinking about the world at this point, if that helps.

Yes.

Thank you very much.

Rick Shane, J.P. Morgan.

My questions have been asked and answered.

Eric Beardsley, Goldman Sachs.

Just back on the topic of spread winding, if you have lower rates and spread tightening when you have higher rates, how much does that color your view of how much duration risk you are willing to take?

Hi, this is Peter.

It's a very significant component of our decision making with regard to our duration gap.

Obviously, we always look at our duration gap in both directions and look at how much it will move.

It all depends on the level of rates and where you are in terms of the convexity profile of the mortgage market.

And obviously that convexity profile changes.

And if we are at a particular point where we face significant amounts of convexity risk, then that too will color our decision.

But the spread component factor is a very significant driver of how we think about our duration gap because, as Gary mentioned earlier, what we don't want to put ourselves in a position is where spreads and rates are moving against us.

We as an investor in mortgage securities face two key risks: interest rate risk and spread risk.

And when we think that they are going in the same direction, we will do everything we can to take our interest rate risk component down such that we are left with the inherent risk of our business, which is mortgage spread risk.

And when we think they are offsetting each other, then we will manage our position accordingly.

So that is the really key driver.

Got it.

Now, is there any risk that if we do have rates move up with significant volatility that you could have spread widening?

Or do you think that's less likely?

It could.

And a lot will depend on the timing.

So for example, if you have a very sharp move in interest rates and it's indicative of something else that's happening, you could have a short-term widening in mortgages as investors try to understand what's happening.

Or, for example, like in 2013, we had some periods in the middle of the year when it appeared like that other managers of other spread products were using mortgages as a hedge against widening of credit spreads and high-yield spreads because of the ability to short mortgages.

So they gave some protection.

So you could have some short-term dislocations.

But I think the longer-term trend in a rising rate environment is the point that Gary made earlier, which is the fact that supply will come down significantly.

Investors are already under-weighted, and so the longer-term fundamental, I think, will win out over time.

One thing I just want to add to that is that, look, to Peter's point, yes, you can have a situation where mortgage spreads widen again in a rising rate environment.

But even in that scenario, I really want to contrast if rates went up quickly today versus what we saw in 2013 -- and there are some huge differences between the environment today and the environment back a year ago.

And I view it as very, very unlikely that we see a repeat.

First and most importantly, the origination volumes last year were to the tune of about $150 billion a month of mortgages that were being produced at the time versus something like $80 billion now and headed lower as you go into the winter, which tends -- those volumes -- what happens, when rates go up borrowers lock the new rates.

There's a huge flush of originations as mortgage bankers who have only half hedged their pipeline sell all of these mortgages into the market.

At the same time, durations were at their absolute lows and they extended significantly, which puts pressure on people to hedge to deal with the larger duration.

They are not in a position to add mortgages.

You throw in the fact that money managers were seeing significant redemptions and bank capital was changing and all of those things were occurring at exactly the same time.

So what I would like to stress is even if we are wrong, and the environment changes and mortgage spreads don't tighten in a backup, we do feel that a repeat of 2013 -- next to none of the conditions that were in place then are in place today.

Got it.

That's helpful.

Thanks.

And then just lastly, you touched on this earlier.

But what has changed in your outlook over the past couple of quarters that led to the shift away from 15 years?

So part of the challenge with 15s, given the move and rates year to date, is that a lot of the realized volatility and expectations for volatility going forward are on the short to intermediate part of the curve, which partially offset some of the benefits of a lower-risk 15-year position.

And as I mentioned earlier, given our overall comfort level with the 30-year mortgage basis, we thought that there was also relatively limited upside in certain 15-year sectors such as higher coupon, TVA-like season 15s, which -- were prepayment speeds are going to start to become an issue.

All right, great.

Thank you.

I want to thank everyone for their interest in AGNC, and I look forward to talking to you again next quarter.

The conference is now concluded.

An archive of this presentation will be available on AGNC's website, and a telephone recording of this call can be accessed through November 12 by dialing 877-344-7529 using the conference ID 10054108.

Thank you for joining today's call.

You may now disconnect.