AGNC Investment Corp (AGNCO) 2020 Q1 法說會逐字稿

Good morning, and welcome to the AGNC Investment Corp.

First Quarter 2020 Shareholder Call.

(Operator Instructions)

After today's presentation, Please note, this event is being recorded.

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

Please go ahead.

Thank you for joining AGNC Investment Corp.'s First Quarter 2020 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.

Participants on the call include Gary Kain, Chief Executive Officer; Bernie Bell, Senior Vice President and Chief Financial Officer; Chris Kuehl, Executive Vice President; Aaron Pas, Senior Vice President; and Peter Federico, President and Chief Operating Officer.

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

Thanks, Katie, and thanks to all of you for your interest in AGNC.

As you know, conditions were extremely challenging in March as the market reacted to the COVID-19 pandemic.

The dislocations witnessed during mid-March were unprecedented in terms of both magnitude and speed, resulting in a significant decline in the valuations of Agency MBS and other fixed-income products.

AGNC's financial performance, like almost all financial companies, was severely impacted by the market volatility, with AGNC posting an economic return for the quarter of negative 20%.

While we were disappointed by this result, we are optimistic that the worst is behind us.

And we believe that we are uniquely positioned to generate strong risk-adjusted returns as we look ahead to the remainder of 2020.

In saying this, we fully recognize the significant uncertainty presented by the pandemic and its associated impact on the U.S. and global economies.

To this point, I intend to dedicate the remainder of my prepared remarks to explaining the rationale behind our optimistic outlook.

Importantly, the bold and decisive actions of the Federal Reserve stabilized the entire fixed-income complex in late March.

As was the case in QE1 and QE3, Agency MBS were again, a critical part of the Fed's actions.

But this program has differed significantly from prior episodes in that the Fed purchased a larger amount of securities over a much more compressed time frame than at any point in history.

In just 1.5 months, the Fed has purchased approximately $575 billion in Agency MBS.

This, coupled with the expectation of ongoing Fed purchases, should provide the necessary support and stability to the sector as the market contends with any future challenges associated with COVID-19.

Against this backdrop, we believe that the financial markets are in the process of transitioning from a focus on liquidity to the next phase where performance will be driven primarily by actual fundamental factors.

Fortunately, our portfolio is comprised almost entirely of Agency MBS, which enjoy the guarantee of timely interest and principal from the GSEs.

As a result, we have very little credit exposure in our portfolio, which is where the bulk of the future uncertainty lies.

In contrast, prepayments, funding and interest rate risk are the fundamental factors that will determine AGNC's ultimate performance.

So with this in mind, let's briefly examine how these factors will be impacted by the current landscape.

First, on the prepayment front, most models tell us that the record low levels of interest rates will increase prepayments substantially.

However, these models are not designed to incorporate the unique circumstances associated with the current crisis.

More specifically, some borrowers will opt to take forbearance on their existing mortgage, while others will have difficulty refinancing due to a job loss, a reduction in compensation, a decline in self-employment income or other adverse events.

Social distancing may also reduce origination capacity and extend closing timelines.

Purchase activity or housing turnover, as it is called in the mortgage industry will likely be impacted to a greater degree as social distancing limits open houses and other showings.

While every scenario is different, there are some similarities between the current environment and the one we witnessed between 2009 and 2012, where the impact of then record low interest rates on prepayments was also materially offset by credit considerations and changes to the mortgage origination landscape.

If we move on from prepayments to funding, the benefits from the current environment are even more straightforward.

The Fed cut the Fed funds target and the overnight repo rate to around 10 basis points and they have offered virtually unlimited liquidity to keep government repo rates near the target.

As such, our MBS repo rates, as Peter will discuss shortly, have generally ranged from single digits on overnights through our in-house broker-dealer to around 30 basis points on 3-month maturities in bilateral repo.

The last element of the fundamental cash flow picture for AGNC is the current asymmetry in our exposure to changes in interest rates.

Normally, hedging a levered position in Agency MBS requires substantial trade-offs as we seek to balance the risk of both significant declines and increases in interest rates on the duration of our assets.

However, if you believe that substantially negative interest rates in the U.S. are unlikely in the near term, which is our opinion, then there is considerably less call risk or downside to being more fully hedged given today's record low swap rates.

So to summarize, the fundamental landscape for Agency MBS is favorable.

Prepayments should remain contained despite low rates as a function of social distancing and credit concerns.

The Agency repo market is trading very well with rates close to 0. And with most relevant interest rates below 50 basis points, we believe there is considerably less downside risk inherent in our hedge portfolio.

For these reasons, it is hard not to be optimistic about the prospects for our business despite the tremendous economic uncertainty that lies ahead.

At this point, I will ask Bernie to review our financial results for the first quarter.

Thank you, Gary.

Turning to Slide 4. We had a total comprehensive loss of $3.61 per share for the first quarter.

Net spread and dollar roll income, excluding catch-up AM, was $0.57 per share, which was unchanged from the fourth quarter, as the decline in our investment portfolio that occurred later in the quarter was offset by lower funding costs.

Tangible net book value decreased 22.9% for the quarter.

At a high level, the decline in our book value was driven by the underperformance of our mortgage assets relative to our hedges, resulting in materially wider spreads during the quarter.

More specifically, about half of the decline was due to lower premiums or pay-up values on our specified pool position, while the remainder of the decline was due to rebalancing costs, incremental losses on our OIS swaps, and our non-Agency assets.

Including dividends, our economic return on tangible common equity was negative 20% for the quarter.

Since quarter end, valuations of TBA MBS have improved modestly, while pay-up values on our specified pool position have recovered a meaningful portion of their Q1 underperformance.

As a result, as of yesterday, our net book value was up approximately 8%.

Turning to Slide 5. Our investment portfolio decreased $15 billion during the quarter to $93 billion as of quarter end.

Our ending leverage was unchanged at 9.4x tangible equity, but has declined commensurate with our increase in book value in April to around 8.5x.

Our liquidity position at quarter end was at precrisis levels with our cash and unencumbered agency assets totaling $3.5 billion.

Importantly, that figure does not include an additional $1.2 billion of capital and excess margin that we held at our broker-dealer subsidiary, or $300 million of unencumbered credit assets.

Our portfolio forecasted CPRs increased to 14.5% from 10.8% during the quarter as a function of lower rates.

Actual prepayments for the quarter averaged 12.2%.

During the first quarter, we also completed $1 billion of accretive equity transactions, including a $575 million, 6.125% fixed to floating rate preferred equity offering and approximately $440 million of common equity issued through at-the-market equity offerings.

Additionally, subsequent to quarter end, we have repurchased approximately $100 million of common stock at substantial discounts to our estimated tangible net book value.

With that, I'll turn the call over to Chris to discuss the agency market.

Thanks, Bernie.

Let's turn to Slide 6. As Gary mentioned, the market dislocations in March were unprecedented.

Agency mortgages widened 100 basis points intra month before TBA MBS recovered 3 quarters of the move as the Fed launched QE4 and purchased $250 billion of Agency mortgages over the course of 2 weeks.

The underperformance of Agency MBS in March was even more dramatic in high-quality specified pools where pay-ups on HLB 3.5s and 4s declined by more than 2 points.

To make matters worse, these pay-ups would have been expected to increase given the 40 basis point decline in interest rates during the month of March.

Money manager redemptions, investor deleveraging balance sheet pressures and a flight to cash led to extreme intra-month moves for all risk assets.

While TBA mortgages recovered a significant amount of the underperformance versus rates by the end of March, specified pool valuations ended the quarter at extremely depressed levels.

When we released our company update on April 8, we thought it was important to inform shareholders that we were able to maintain our high-quality specified pool holdings through the market turmoil in March as we believed valuations would normalize and that these positions would be critical -- would be a critical component of the performance going forward.

Since March 31, specified pools have, in fact, recovered meaningfully.

With pay-ups on HLB 3.5s and 4s, more than 1 point higher, but more importantly, as I will discuss in a few minutes, we believe the cash flow return expectations on these positions are very attractive.

Let's turn to Slide 7. As you can see, the investment portfolio declined to $93 billion as of March 31, the $15 billion net decline was comprised of $4 billion in pay downs and net sales of approximately $25 billion in lower pay-up relatively generic pools with about $14 billion of those sales replaced by TBAs.

As we discussed on prior earnings calls, we had added relatively generic lower coupon MBS in lieu of specs, and these positions were critical to our ability to build liquidity during the middle part of March without having to sell higher quality specs at fire-sale levels.

Turning to Slide 8. We present several illustrative examples to highlight the compelling nature of Agency MBS returns despite unprecedented purchases by the Fed in the second half of March.

The 2 top tables show the potential gross return on equity as of quarter end on 30 year HLB 3.5s and 4s using 9x leverage.

As a reminder, HLB pools are those backed by loans with loan balances less than or equal to $150,000.

And as you can see, potential gross returns, inclusive of hypothetical hedging and funding costs, but prior to convexity costs, were in the mid- to high teens as of March 31.

Importantly, even if prepayments are higher than the base speeds displayed in the table, returns are still attractive.

In the lower 2 tables, we show TBA 30 year 2.5s with and without an assumed funding advantage from dollar roll specialness.

30 year 2.5s are the predominant production coupon, and as such, enjoy the greatest amount of support from Fed purchases.

Roll-implied financing on 2.5s is currently trading flat to slightly through repo, however, we do expect that roll specialness will improve in the coming months and may average 25 basis points through repo, given current origination and Fed purchase dynamics.

As you can see, even without a roll-implied financing advantage, based on our assumptions, potential gross returns on 30 year 2.5s pre convexity costs were solidly in the mid-teens as of March 31.

Since quarter end, valuations have improved.

And so currently, returns are roughly 2% lower than the example shown on this slide.

As we look forward, we're excited about the return environment for Agency MBS, while spreads have tightened materially from the depths of March, they remain wide to historical norms despite QE4.

And while we fully recognize the unprecedented economic challenges presented by the current crisis, our Agency MBS business is somewhat uniquely positioned on a go-forward basis.

I'll now turn the call over to Aaron to discuss the non-Agency sector.

Thanks, Chris.

As Gary mentioned, the credit markets were also very challenging to navigate as they ultimately had to grapple with complete illiquidity across all fixed-income products, and the rapid emergence of serious credit uncertainty associated with pricing in the impacts of a recession.

I'll quickly recap our activity in the quarter and then update you with our outlook on credit.

Please turn to Slide 9. As we discussed on our last quarter's call, we anticipated reducing our CRT position given the further tightening of spreads in January.

Accordingly, we sold close to $200 million of CRT in Q1 prior to the decline in prices in mid-March.

The other changes to the portfolio during the quarter were an increase in residential credit subordinate bonds backed by GSE-eligible collateral and a small addition of CMBS versus a reduction in RPL subs.

With respect to portfolio composition, as you can see from the GSE CRT portfolio pie chart, more than half of our CRT portfolio was issued in 2016 and 2017.

These CRT vintages have performed well since quarter end and have recovered materially from the lows, given their solid credit support and built up HPA.

Additionally, we remain comfortable in large part with our current CMBS holdings.

Almost all our exposure on the conduit side is AA rated or higher.

And while we have some lower-rated CMBS SASB deals, we have no single asset with hotel exposure and minimal retail exposure.

In the second half of March, non-Agency assets were hit with a perfect storm.

Adding to the illiquidity and credit concerns there were material challenges on the financing side.

Increasing haircuts and borrowing rates, along with difficulties in rolling repo with some counterparties combined to produce a brutal a couple of weeks.

The massive intervention from the Fed helped stabilize credit markets.

And once forced selling subsided, non-Agency valuations recovered a meaningful amount of their initial declines.

Against that backdrop, financing has eased a bit, but still remains challenging.

Importantly for us, given the relatively small size of our non-Agency portfolio, we have the option of taking the entire position on balance sheet, if necessary.

Looking ahead, these are some of the things we are thinking about.

On the residential side, while it's impossible to discount the significant drawdown in-house prices, our view is that in the near term, only a shallow decline in house prices is likely.

We came into 2020 with a strong housing market, supported by limited housing inventory and pent-up demand due to strong household formation.

The quick implementation of forbearance programs should both buy time for economic recovery and give servicers time to work on modification programs, should they be needed.

With rapid increases in job losses over the past 6 weeks, it is likely that conventional mortgage forbearance rates will increase from around 5.5%, where we stand today to the 10% to 15% range.

Given this increase, extending the timing of defaults in REO dispositions, far enough into the future should dampen the impact to house prices and is critical to limiting the downside scenarios.

As Gary and Chris addressed earlier, we expect prepayment speeds on Agency's collateral to be slower than many models would project at these rate levels.

This negatively impacts discount price credit securities with slower return of principal and deleveraging.

The GSE's use of CRT to hedge their guarantee business is likely to reduce the refinance programs they would be willing to implement in the future.

If the GSEs were to add new refinance programs that bypass normal underwriting constraints other than the existing HARP like program, they would lose the value of the credit protection they purchased via issuing CRT.

Turning to the commercial front.

The credit curve has steepened out meaningfully as the top of the capital structure in conduit deals has been anchored by the inclusion in the Federal Reserve's TALF program.

Conversely, the bottom of the capital structure is trying to price in assumptions around forbearance rates, material increases in default expectations and the potential for downgrades.

Challenges in the hotel and retail sectors are top of mind, but New York and other big [city] (added by company after the call) office buildings could face significant headwinds as some companies look to diversify the location of their workforce.

These issues will take time to play out.

And in our view, the commercial real estate market faces the greatest uncertainty and biggest range of possible outcomes.

This may lead to interesting opportunities in the coming quarters as we begin to get delinquency data and monitor credit performance.

In light of the tightening in spreads for many structured product securities, along with repo challenges, we believe a patient approach on the credit side is warranted at this time.

With that, I'll turn the call over to Peter to discuss funding and risk Management.

Thanks, Aaron.

I'll start with a review of our financing activity on Slide 10.

Despite considerable market turmoil, the repo market for Agency MBS functioned remarkably well during the quarter.

Given the repo issues last fall, the Fed had already taken significant steps to ensure that the repo rate for Agency and Treasury collateral remain closely tied to the Fed funds target.

On March 20, the Fed began providing an additional $1 trillion of liquidity to the repo market each day.

This dramatic step, coupled with the Fed's ongoing open-market operations, ensured that the repo market for high-quality collateral like Agency MBS remained extremely liquid throughout the crisis.

The Fed also lowered the federal funds rate to the 0 bound with a 50 basis point rate cut on March 3 and a 100 basis point rate cut on March 15.

Our weighted average repo funding cost for the first quarter was 1.8%, down 32 basis points from the prior quarter.

Our repo cost at quarter end fell significantly more to 1.36%.

Today, the repo market has largely repriced to the new Fed funds target.

Overnight repo has averaged about 10 basis points in April.

Importantly, the term market has also repriced with 30-day repo now trading at about 12 basis points and 90-day repo trading at about 22 basis points.

Given these funding levels, I expect our average repo cost in the second quarter to decline to approximately 80 basis points.

Our aggregate cost of funds, which includes the cost of our repo funding and swap hedges, also improved during the quarter, but to a lesser degree, as the improvement in our repo cost was largely offset by a lower receive rate on the floating leg of our swaps.

Importantly, with LIBOR swaps representing only a small part of our portfolio, our aggregate cost of funds will not be negatively impacted as LIBOR converges with other short-term rates.

Looking ahead, I expect our average cost of funds in the second quarter to be about 110 basis points.

Turning to Slide 11.

The notional balance of our hedge portfolio totaled $59 billion at quarter end, down significantly from $99 billion the prior quarter.

In response to the rapidly changing interest rate environment, particularly in March, we repositioned our hedge portfolio by terminating a significant portion of shorter-term swaps and treasury hedges.

As a result, our hedge ratio declined to 70% from 102% the prior quarter.

With short-term rates being close to zero, swap rates repriced dramatically during the quarter.

With shorter-term OIS swaps rallying at about 135 basis points and longer-term swaps rallying about 120 basis points.

The pay rate on new 5- and 10-year OIS swaps is now only about 20 and 40 basis points, respectively.

As such, these swaps provide a unique opportunity to lock in very attractive funding levels for an extended period of time, while also providing protection against the reversal in rates in the event of a sharp recovery.

On Slide 12, we show our duration gap and duration gap sensitivity.

Given the changes that I mentioned to our hedge portfolio, our duration gap remained in a fairly tight band throughout the quarter and ended the quarter at about 0. With interest rates being at such a low absolute level, there is now significantly more extension risk in our portfolio and in the mortgage market as a whole.

You can see this asymmetric risk profile in the table with our duration extending about 2 times more in the up-rate scenario, than it contracts in the down-rate scenario.

Given this asymmetry and the low-cost of longer-term hedges, we will likely favor operating with a flat or even negative duration gap in the current environment.

Such a risk profile is further supported by interest rates being so close to the zero bound.

While negative rates are certainly possible, we believe there will be significant resistance to that as the Fed will likely utilize all of their tools, including especially large-scale purchases and forward guidance before resorting to lowering the Fed funds rate significantly below zero.

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

Thanks, Peter.

And at this point, we'd like to open up the call to questions.

(Operator Instructions) And the first question will come from Rick Shane with JP Morgan.

Two questions.

First, on the non-Agency side.

Look, it's a relatively small part of the portfolio.

As we've seen over the last 6 weeks, it creates a very different risk profile that doesn't necessarily benefit from the same policy interventions and advantages.

I'm curious if going forward, you see the dislocation as an opportunity or as an exit?

Rick, first off, thanks for the comments, and I also want to wish everyone the best in this kind of unique environment to say the least.

But let me get to your question.

We don't view this as an exit opportunity.

But I want to -- I mean, what we have said about the credit portion of our portfolio, is that we do believe that it makes sense to be fluent in the credit sectors of the mortgage market.

Because we do think over the long run, it's naive and limiting to assume that at all points in time, Agency MBS are just clearly going to be the best place to run a levered mortgage position.

So we are committed, and as we have said for a few years now, we're committed to being involved in the credit sectors.

And -- but we will be very patient and very disciplined with respect to that involvement, and we will only increase those positions significantly to the extent that the risk return trade-offs are compelling.

And so let me take that to today's environment and -- so I can kind of put that -- kind of make that a little more specific.

Right now, credit has obviously cheapened up significantly.

But the environment is very uncertain, and there are real credit headwinds to a range of sectors, including CRT or the conventional mortgage market.

The commercial space is probably -- kind of has the most uncertainty, as Aaron mentioned, and we certainly believe that.

But -- so at this point, with the improvement that we've seen in credit from the lows, we don't see credit as being back up the truck cheap by any stretch of the imagination.

But as the fundamental information comes in, as we start seeing delinquencies, as we start seeing some of the stress, then we think there very easily could be some good opportunities there.

It's also very possible that on the Agency side of our business, depending on the magnitude of the Fed's programs, depending on how things evolve over the next 3 to 6 months, we could see a situation like what we saw in QE3 where Agency MBS gets tightened significantly from here.

While that's good for our book value, it will kind of reduce the risk-adjusted returns on the Agency side.

And so there's a very real possibility that 3 to 6 months from now, it will make sense for us to rotate some of our capital away from Agencies into credit.

And having the fluency with credit, having the involvement, but without having it be like handcuffs, is a real positive.

And so we're not saying that's definitely the way we're going to go.

I would say, it's probably less than 50% chance that those conditions will arise, but it's a much more likely scenario now than it certainly was 6 months ago.

So that's kind of what we're looking for.

But no, we're committed.

We feel like we achieved a good balance of, again, being fluent, understanding the products, without it ending up handcuffing us in terms of running our core business.

Look, that makes sense.

I think what I'm hearing is strategically, it's important to keep it as an option.

And tactically, you're not seeing the opportunity quite yet.

One follow-up question.

Given your view on rates and the asymmetry as you sort of develop your hedging strategy going forward, do you think that swaptions will become a larger part of the strategy?

So it's interesting.

The answer there is it might, but not at this point.

It's interesting because the asymmetry allows you really to just pay fixed on, let's say, a 7-year swap that's -- where the pay rate on an OIS swap is, let's say, mid-30s.

So your -- that swap cost you -- you're receiving, let's say, 10 basis points, that costs you 25 basis points in carry.

But if you don't believe in negative rates, how much can you lose on that short position.

Let's say, that swap rate goes to 15, okay?

So you lose 20 basis points.

In most cases, you're going to pay that or more for a premium on an option.

So it's kind of unique that, again, unless you're of the mindset that swap rates can go to negative 50 then options really aren't necessary.

Look, and I want to just mention on negative rates, we're not saying it's absolutely impossible.

We're obviously aware that it's going on in Europe and they have substantially negative rates.

But we do think the Fed will do a host of other things before they go there, increasing their QE would be one thing they'll definitively do.

They'll probably do something around yield curve control.

And there'll be lots of warning signs that will probably give us an exit point on our mortgage position at tighter levels as well.

So we just -- we think in the short run, again, this asymmetry really helps us manage a portfolio, and it's a unique position to be in where the 0 bound really helps.

But let me let Peter add something to that.

Yes.

Rick, let me just add, we sort of already started to implement the strategy that Gary just talked about.

I know you've seen the drop in our swap portfolio, which was about $32 billion.

But we ended up taking off a lot of the much shorter-term swaps, the 1-, 2-year swaps.

But we also added a significant amount of longer-term swaps during the quarter.

In fact, we added about $10 billion worth of swaps between 4 and 10 years.

So we've already begun that rotation that he was describing because those hedges are so much more cost-effective and efficient versus option-based hedges in an environment where interest rate volatility is so high.

The next question comes from Bose George with KBW.

It's Eric on for Bose.

And I hope you guys are all well.

Just how are you guys thinking about the amount of leverage you're taking right now, just given the fact that spreads are already relatively wide and returns could arguably be considered -- and still be considered very strong with less leverage.

I mean how -- why do you guys feel the need to be as levered in this environment as you were at year-end, just given the obvious risks that are still out there and that you also acknowledged in your opening remarks?

Well, look, that's a great question, and I know I'll kind of take it at a higher level, which is -- and I'll answer your question plus, given what we went through in March, has that changed our perspective on leverage big picture.

But first off, one thing, in Bernie's prepared remarks, she mentioned that over the course of April, our leverage has come down not via shrinking the portfolio.

It was mainly due to just the increase in book value.

And in not replacing the full amount of prepayments, so to speak.

So in the end, we are running leverage now, again, that's closer to 8.5.

But I do want to stress that the aggregate leverage number is an important determinant of your liquidity, but it is in no way, shape or form the final answer.

And TBA positions are much, much more efficient for us from a liquidity perspective than on balance sheet pools.

And on balance sheet pools also differ kind of in terms of their impact on your liquidity.

A pool that can pay 30 or 40 CPR is a substantial drain on your liquidity, given the payment delay issue where you have to wait for your P&I receivable until the 25th of the month.

So what I would say is that we are willing to run leverage levels consistent with where we were before, 9.5 to 10, let's say.

But I would say we're putting a greater emphasis on liquidity in this environment because we think that's what shareholders would want from us.

So big picture, I would ask analysts and investors to look a little below -- beyond just the "at risk" leverage.

And understand that there are different components that affect both your liquidity and your ability to handle kind of adverse moves.

And just as a quick example of that, a TBA position through our Bethesda Securities dealer requires a margin that's around 1% to maybe 1.5% in all-in.

Whereas the haircut on an on-balance sheet position in bilateral repo is 5% on average.

And that pool has a pay-down that if it's, let's say, in the 20s in CPR that almost adds an effective 2% more to the haircut.

So that's a massive difference, right, between an effective 7% haircut on a bilateral repo with a pool that's prepaying versus a TBA position.

So what I would say is investors should expect AGNC to be cautious on the liquidity side, but to do it in a smart way that optimizes our risk-adjusted returns.

Got it.

That was helpful.

That, I think, leads pretty well into my next question on funding.

I mean it hasn't been lost on anyone that as bank earnings have come out, the loss reserves they've been booking have been pretty meaningful, which at least on face creates, I think, some risk that they manage their own balance sheets somewhat differently going forward.

So I'm just curious how you guys think about your funding exposure to the banks and their ability to supply funding in general, not just AGNC but other businesses as well.

And as a side to that, I mean, how operationally, how challenging would it be to move even more of your funding let's just say, kind of the bulk of your funding over to Bethesda Securities, if you needed to or wanted to?

Eric, this is Peter.

Hope you're doing well.

Thanks for the question.

First off, I would say that, as I mentioned in my prepared remarks, the repo market for Agency collateral really traded remarkably well throughout the crisis.

And the Fed has done a terrific job of making sure that there is sufficient liquidity in the system, and they're continuing to do so today.

I don't expect that to change.

We have 47 other counterparties in addition to Bethesda Securities.

And those counterparties all performed very well for us during the quarter.

We had no issues with any of the counterparties but having 47 of them, we have our exposure, which, as you point out, is about 45% at the end of the quarter of our funding book, spread out across those counterparties.

So there's really no significant concentration and that allows us to move positions from counterparty to counterparty very quickly because we don't have significant concentrations in any 1 counterparty.

And your other point about Bethesda Securities is a good one from an operational perspective, it is very easy for us to move collateral.

When our bilateral collateral matures, for example, we can very quickly move it to Bethesda Securities.

We have additional capacity, and we could utilize it and we may utilize it to some degree because, as Gary mentioned, it's highly efficient from a margin perspective on the funding side, and it's incredibly efficient on the TBA side, and we would not be able to do that on TBAs, if we didn't have Bethesda Securities.

But we also like the fact that our portfolio is highly diversified.

And so right now, we think the best mix is to have a really large group of strong, individual counterparties.

And I believe that they're going to continue to participate in the market and provide support for us as well as have the ability to move more to Bethesda if we need to.

And I just want to add to that, just keep in mind, right, that the government repo complex didn't go through its shock in March.

It went through it actually last September, right, when there was a huge spike, unexpected spike in overnight repo rates.

And since then, the Fed has been incredibly engaged and has taken ownership of that, of government repo rates.

And to Peter's point, I think a lot of the success of the market in March and the repo market in March, was a function of the fact that Fed had been dealing with it in a large way, okay, buying bills, doing open-market operations, well prior to March to address that.

And then they offered $1 trillion in overnight repo.

I mean the Fed is very, very focused on that issue.

And even before you get to purchasing of treasuries or mortgage-backed securities, so I think that's the last thing that the Fed is going to let have issues.

And having our broker-dealer on the FICC gives us -- it's not direct access.

I want to be clear, but it's closer.

And we've shrunk.

We've definitely let our repo book, given the shrinkage in the portfolio and the move of more to Bethesda, we have lots of spare capacity on the bilateral side.

But thank you for the question.

Next question will be from Doug Harter with Credit Suisse.

As you mentioned, kind of specified pools have performed well in April.

Can you just talk about where there are price wise, pay-up wise, compared to January and February time period?

Yes.

Why don't I let Chris take that one.

Yes.

Thanks, Doug.

I mean relative to the start of the year, the caps on HLB 3.5s and 4s, for example, are slightly higher.

But the thing to keep in mind is that valuations at the start of Q2 were still at very, very depressed levels.

And so while pay-ups have improved over a point or so since then, they're still extremely attractive given that coupon swaps are still very depressed.

They were starting at extremely cheap valuations.

OAS changes year-to-date are still a good 30 to 40 basis points wider.

When you think about just the 125 basis point move in rates, they've actually still underperformed quite a lot, despite the absolute changes in pay-ups.

And so if you look back to the ROE slide that we presented, yields are currently on 3.5s and 4s around 20 basis points or so lower than they were as of 3/31.

And so I'd say, potential ROE is roughly 2% lower.

And so -- but even their returns are still very, very attractive.

Great.

And then again, thinking about -- so then I guess, again, thinking of that slide, where you're looking at the returns, how do you think about balancing out kind of what the returns are on the HLB pools versus kind of the more liquid 2.5s?

And how do you think about what is the right balance to gain the liquidity from kind of the more generic coupons versus the specified pool, which obviously proved to be less liquid in the periods of stress?

It's a great question.

And well, Chris, why don't I start and then you can add.

But what I would say is, first off, look, you don't -- the lesson from the liquidity of March was something we were certainly fully aware of before, which is you don't want your whole portfolio in high pay-up specified pools.

That reduces your liquidity and your options for moving things around to say the least.

But that said, I think, look, both offer competitive returns even given the tightening that we've seen in April.

And they're now relatively balanced.

In the higher coupons, you've got relatively stable cash flows.

There's no origination.

The technical factors are stable in today's environment.

So I don't think you're going to see that much price volatility there.

Your exposure is going to be to kind of realized prepayment speeds.

And are there going to be surprises there?

In the short run, we think the surprises are more likely to be to the downside versus expectations rather than to the upside.

But on the other side of it, there's a lot to be said for lower coupon returns and the potential for roll specialness to improve those.

You have the liquidity benefits where you can run those positions in pure TBA form or in a new production pool that's not going to have a pay down.

So there are benefits there.

And I think the short answer is, we want to achieve a balance of those.

The one other issue you have with low coupons is that there has been and there will be significant production.

So the technicals are how fast is production coming in versus the Fed's bid and any other bids.

So we do expect more price volatility in lower coupons, and we'll look to use that as an entry point.

So that's it from a high level.

I'll let Chris add something to it.

Gary, I think you summed it up pretty well.

I don't have a lot to add other than, I mean, I would say that while specs have improved a fair amount, I mean, the cash flow stability from a spec position is important diversification around our lower coupon position.

We don't want to be overly exposed or reliant to the Fed there.

No, that's a good point.

The next question is from Trevor Cranston with JMP Securities.

A question on the prepaid side.

You talked a lot about the factors that are likely to limit how fast speeds get, at least in the near term.

Can you comment on sort of what you guys bake into your base-case prepay assumptions in terms of how much compression there could be between primary mortgage rates and secondary MBS yields and what you would consider the biggest risks to that spread potentially compressing kind of further than where you currently expect it to?

Yes.

Let me start on that.

I mean I want to say that the modeling, when you look at OAS models or Monte Carlo simulations, prepayment models right now.

This is a stressful environment for them.

Their results are all over the place.

And the thing you addressed how you derive the mortgage rate okay, this primary rate and how you evolve it over time is one of the key variables that differs quite a bit from model to model.

But what I would say is, from our perspective and thinking about prepayments, we do expect the primary rate or primary secondary spreads to compress.

They do in these kind of environments over time.

I think that it's not going to be immediate.

And given how much of the universe is refinanceable today, and given the unique circumstances, we think that compression will be slower than it was in, let's say, other kind of periods of big drops in interest rates.

So we think that will take time.

But yes, that will kind of increase the incentive to refinance.

But on the other hand, as we've said over the next, we'll call it, 3 months, 3 to 6 months, you have these social distancing type headwinds as well as credit headwinds.

I mean we didn't talk a lot.

We mentioned the forbearance issues in our prepared remarks.

But forbearance is a very, very good option for GSE borrowers and generally, there are going to be a lot of people that aren't going to be able to qualify others that don't take forbearance, but won't qualify easily for a new loan.

So practically speaking, I think there are factors at work.

There are a lot of factors that will keep prepayments from surprising to the upside, certainly over the next 3 to 6 months.

But that is predicated, and those are prepayment thoughts and where we think things are going to go are predicated on the fact that the primary secondary spread will compress over time as it normally does.

Again, I think it will take a little longer this time.

Got it.

Okay.

That's helpful.

And then 1 more question just on the credit side.

I think Aaron mentioned it in his remarks that the availability of funding remains somewhat constrained.

I was wondering if you could provide any additional color there in terms of just kind of if you're seeing a significant number of counterparties completely pull away from that market?

And just generally, how much change you've seen in terms of haircuts and where rates are for credit securities?

Yes, sure.

I mean we obviously do finance some non Agencies, as I think everyone knows.

Our positions are relatively small, and it's a very small percentage of kind of our funding considerations.

But I mean, higher haircuts are kind of across the board at this point or almost -- they've been raised.

The rates are much less attractive than they were going into this.

And we have seen some counterparties exit the business altogether.

But to be clear, there are people that are still willing to do the business.

And if anything, over the course of the month, things are a little better now than they were, but you're still looking at rates in the, 2% or more area and haircuts are obviously dependent on the type of security.

And then there are some things that you just can't finance.

And so one of the things that affects our -- that absolutely affects our view on waving in credit, so to speak, are these challenges.

If we were to have sort of a double dip in the economy or some of these fundamental factors were to start looking bad, which is very possible, then we could see more stress again on the financing side.

So we have to build that into our assumptions and ROE.

Look , it's one thing if you're looking at a position from an unlevered cash position, long-only position.

Things are more attractive than when you look at them on a levered position, given the challenges on the funding side.

The next question will be from Kenneth Lee with RBC Capital Markets.

Wondering if you would be able to just give us a sense altogether, how much net interest spreads could potentially improve in the near-term just given the dynamics from the ongoing Fed support of agencies as well as the much lower funding cost that you mentioned in the prepared remarks?

Sure.

Ken, this is Peter.

I hope you're doing well.

As I mentioned, I expect our repo cost to come down around 80 basis points in the second quarter.

I'll sort of give you the building blocks.

And I expect significant improvement from there in the third and fourth quarters, by the way.

As some of our higher cost, longer-term repos mature, it could easily drop to about 50 basis points in the third quarter and down around 40 basis points in the fourth quarter.

The variable for our cost of funds is going to be on our swap costs, which I expect to be net on total liabilities around 35 basis points.

So that would give us the cost of funds of around 110 in the second quarter.

On the asset yield side, obviously, it's going to come down.

But probably in the neighborhood of around 270 in terms of asset yield, might be a reasonable starting spot.

So I would expect our net interest margin in the second quarter to trend toward 150, 160 basis points.

And then from there, it will depend on, obviously, rotations in the asset portfolio into the third and fourth quarters.

And then as well as some of the rebalancing and repositioning that we're going to continue to do on the swap side, which would ultimately, I think, put some downward pressure on the cost of swaps as we move our swap hedges to the longer part of the curve or an intermediate part of the curve, which, as we talked about, doesn't cost very much right now.

So directionally, I think it's heading up into that range.

Okay.

Great.

Very helpful.

And just 1 follow-up, if I may.

In terms of the liquidity position, the $3.5 billion of cash and unencumbered assets -- Is there a way that you could provide us some context or just help us frame how strong this liquidity position is?

Or what's the best way for us as outside observers to determine the relative adequacy of this liquidity position?

Yes, thank you very much for that question.

And you're right, it obviously helps a lot to have some background to that.

So you know where that stands historically.

Let me just give you a starting point.

At the end of the year, our cash and unencumbered at AGNC was $3.6 billion.

And then between non-Agency securities in Bethesda, it was another $1.8 billion.

So it's $5.5 billion right?

And that represented -- excuse me, 52% of our equity at the time, which is a really pretty strong position.

And if you look back historically, you would see that somewhere between 45% and 55%.

So it was at the upper end of where we typically operated, and that gives you a lot of, obviously, capacity to absorb huge shocks.

As Bernie mentioned, the total, the $3.5 billion that we had on March 31, plus the $1.2 billion at Bethesda, which is all cash in Agency MBS and then the $300 million non-Agencies totaled $5 billion, and that was represented -- that was at 54% of our capital.

So we actually had -- at the end of March, we ended the crisis period on a percentage basis with more cash and unencumbered equity.

Yesterday, our position at AGNC was even stronger.

It was $4.2 billion of cash and unencumbered, plus another $1.2 billion at Bethesda Securities and the $300 million.

So we had $5.7 billion of unencumbered and against the capital that we sort of described it being up 8%.

Our percentage of unencumbered had increased close to 60%, which is close to as high as we've ever operated.

So that gives you a sense on how we've prioritized risk and liquidity in this environment, and we put ourselves in a position where we've actually come through the most significant crisis in the history and actually now have a cash and unencumbered position that gives us -- it's even stronger and gives us more flexibility going forward to take advantage of market opportunities.

Let me give you another just quick way to think about it, too.

Which is if we walked in tomorrow and mortgages were across our entire position, were down 2 points, which is obviously a massive move.

I mean we saw worse, obviously, before.

But we would -- that would generate kind of -- or that would be expected to generate margin calls of about $1.8 billion.

Right, Peter?

That's right.

So you're $90 plus billion in mortgages.

So 2 points, we would -- we had margin calls around $1.8 billion.

That would be less than half of our liquidity that we have on hand with zero -- without having to make any adjustment.

So that's another way to think about that in terms of how substantial that cushion is.

Our next question will come from George Bahamondes with Deutsche Bank.

I believe my question may have been answered on the prior question.

It was just to confirm, Peter, that you expect total average borrowing cost to be roughly 1.1% in the second quarter.

I believe you may have just addressed that, though.

Yes, that's correct.

I expect our total cost of funds to be in that range.

And again, the 2 components of the total cost of funds is our repo, which I expect at around 80 basis points, Chris talked about TBA specialness.

I expect that TBA cost to be down around 25 or 35 basis points.

So the combination of that plus the cost of our swap hedges, which at the end of the quarter, if you sort of projected that out going forward, expressed as a percent of our total liabilities would be about 35 basis points.

So total cost of funds of around 1.1%.

But just keep in mind, obviously, we're less than a month into the quarter, and there are lots of moving parts in terms of what we end up doing with swap positions and assets and where dollar rolls go and so forth.

So keep that in mind that there's lots of different things that could move those numbers around.

The next question will be from Matthew Howlett with Nomura.

First of all, congrats on really protecting shareholder interest during the volatility.

Gary, you're always sort of one step ahead of the Fed, last time through the QE program.

So I want to really focus on just some more of those questions first.

Did you address there's been talk of yield curve control.

I don't know if you'd addressed that, but one first, when you talk about Fed policy, do you think there's any risk that they implement something of that nature?

I think there is.

I mean in a sense, it's kind of like they decide where they want the longer end of the intermediate sector and the longer end.

Obviously, they currently peg the very short end.

That's their "job" in setting Fed funds and overnight repo.

And -- but others, such as the Bank of Japan, have chosen to kind of pick a level, let's say, for the 10-year, I'll just hypothetically say they decide they want the 10-year to be around 50 basis points.

And so then they execute whatever that if they were to go that route, they would execute whatever amount of trades they needed to do to keep it in that ZIP code, so to speak.

And so where I think this comes into play, I don't think they're ready to do that right now.

I feel like the they've just unleashed a host of tools, and they're using them extensively.

And I think they -- and obviously, financial markets have stabilized and are kind of performing much, much better than they were.

I don't feel like the Fed feels they have a crisis on their hands at this point.

But let's say things start to unravel a little bit more then these are the kinds of things that the Fed is likely to do.

They're probably likely to up their QE again on the mortgage side and on the treasury side because they've brought their purchases down substantially from where they were a month ago.

So if they felt they were losing ground and the economy was weakening again and liquidity was getting worse, I think they would start to look at what else they could do.

But the first thing they'd probably do is increase QE again, okay?

And probably, first, in terms of the magnitude of what they're doing because they've backed down their 2 main levers, treasuries and MBS.

And then they might look to this yield curve control.

And all of these are things that we think they'll do before they experiment with negative rates, which I think there's really no stomach for right now, either at the Fed or kind of even amongst most financial market participants.

So it's possible.

I don't think it's in the offing in the short run.

I think again, it would be one of the tools they might go to before they went to negative rates.

And when you think about your business model, managing your duration gap, do you feel the Fed's got to get mortgage rates back up to 4%, they're just going to come in and start buying?

And then more would you think they want to bring rates below 3%.

Do you think they have a target in mind that we're going to buy until consumer can see 3% below mortgage rates?

I mean it sounds different in QE, you mentioned that.

So what do you think their target is where they want mortgage rates to be?

I think it's evolving.

I take this in some ways from Chair Powell's press conference yesterday.

But I think they're still in the mindset that their purchases are more designed around liquidity and proper function of the market.

And a month or 2 from now, they're going to start focusing on stimulating the markets.

And when they go to stimulating and you've heard this from the Fed over and over again and from people like Simon Potter, who left the Fed, one of the unique tools that they have that other central banks don't have is the ability to move the mortgage rate.

And right now, it wouldn't matter because of the discussion to an earlier question related to the primary secondary spread.

So they could move mortgage prices higher and it wouldn't affect mortgage rates today because the primary secondary spread would just offset it.

A couple of months from now, that won't be the case, and they'll have more incentive to potentially want to push mortgage rates lower.

So I think that's, in a sense, the next phase of their mindset is how can they stimulate the economy.

Right now, they're still in kind of preserve liquidity, strengthen the financial markets, and they'll go from there.

Next question will be from Brock Vandervliet with UBS.

I was just -- you made a passing reference to your expectation of the forbearance levels rolling from, call it, 6 to 10 to 15.

How do you -- obviously, this is kind of over the end of the horizon at this point, but how do you look at the risk of some of those credits not recovering, delinquency rates pulling up and therefore, the risk of the GSEs is having to buy these out of the pools?

And what effect that could have on pricing?

Yes.

Look, that's a great question.

And I think over the past month, different researchers have written about this and kind of described it as, we'll call it, a significant risk to prepayments.

And that was a reasonable concern early on because the GSEs used to have a policy with respect delinquencies, where after 4 months of delinquencies loans were pulled from pools.

So if they were to have stuck with this boiler-plate practice, than if we got to 10% forbearance and if forbearance was treated like a regular delinquency, you could have very fast prepayments coming relatively quickly from this forbearance equation.

But both the GSEs and FHFA have explicitly, and it was probably about a week ago, put out guidance where they are not going to treat forbearance like a regular delinquency, and they will wait until the forbearance period, which could be up to a year, is over before they start that 4-month clock.

So there are situations if the loan turned out to be modified before that, then there are ways where it could come out of the pool before that.

But I think that's a very low likelihood.

So realistically, delinquencies is getting pulled from these pools is a problem for, let's say, 1.5 years from now?

And then first off, we think and they are obviously very focused on it, we think that a good chunk of those delinquencies will be able to be put on a reasonable repayment plan.

And obviously, the hope is that many of those people will have regained their jobs by then.

And so yes, a percentage of them will probably ultimately default and be pulled out of the pools.

But let's say, you sit there and if you try to say, "Okay, maybe 3% or 4% of these borrowers are pulled out of the pool in 18 months." That's actually a very, very low CPR, so to speak.

And it's not something that we think negatively impacts the performance of our pools.

On the contrary, I think net-net, that's a very good outcome in a way.

I mean again, I don't want people missing payments and so forth.

But from a prepayment perspective, the key change was the GSEs treating this the way they've treated Katrina and other hurricanes where they don't view these forbearances as delinquencies and that they are going to wait a long time to pull these loans out of the pools.

That's critical to the performance of higher coupon securities and specs in particular.

And we were very happy to see FHFA explicitly define that in a press release.

Got it.

Okay.

That's a key distinction.

I guess an analogous question beyond CRT.

You -- obviously, this is credit protection if they forbear and modify, that would threaten the value of CRT, but they've come out right out of the gate with very aggressive forbearance.

How does that fare to the value of the CRT market?

So I mean, look, with CRT, forbearance is good in terms of the fact that we're giving borrowers who would have a short-term interruption to their income and their ability to pay, we're giving them significant time to work through that.

And if they get their job back, then hopefully, there's a plan to get them current.

And I think there will be lots of options.

And as Aaron mentioned in his prepared remarks, just buying a lot of time keeps any pressure off the housing market from REO disposition, and it prevents or likely prevents a downward spiral for house prices.

So that's good for CRT ultimately.

The other thing that's interesting and that Aaron alluded to, just going back to kind of the prepayment question is one of the things -- we've gotten the question at times about, well, you have this favorable view on prepayments, that they're not going to be that fast.

The other thing that CRT does is it really, really handcuffs the GSEs with respect to trying to implement some much faster, streamlined refi program, where they look past income and job situations in terms of deciding whether someone can refi.

But if they do that and they let people refi that have credit issues, then that's a home run for us as a CRT holder because they essentially have this in the money protection that they've bought for 7 years that protects them from the bulk of the losses for these weaker credit borrowers.

And if they just go and refi them and do that uneconomically, they're basically going to waste that protection.

So we certainly do not see that happening.

In a sense for our position, we sort of have a little bit of a built-in hedge for that situation.

Ladies and gentlemen, this concludes our question-and-answer session.

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

Well, look, I'd like to thank everyone for their interest in AGNC.

I hope everyone stays safe through these really difficult times, and we look forward to speaking with you again next quarter.

And thank you, sir.

The conference has now concluded.

Thank you for attending today's presentation.

You may now disconnect.