AGNC Investment Corp. Q2 FY2020 Earnings Call

Good morning, and welcome to the AGNC Investment Corp Second Quarter 2020 Shareholder Call. All participants will be in listen-only mode. After today's presentation, there will be an opportunity to ask questions. Please note that this event is being recorded. I would now like to turn the conference over to Katie Wisecarver of Investor Relations. Please go ahead.

Thank you all for joining AGNC Investment Corp second 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 facts constitute forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. All such forward-looking statements are intended to be subject to the Safe Harbor protection provided by the Reform Act. Actual outcomes and results could differ materially from those forecasts due to the impact of many factors beyond the control of AGNC. All forward-looking statements included in this presentation are made only as of the date of this presentation, and are subject to change without notice. Certain factors that could cause actual results to differ materially from those contained in the forward-looking statements are included in the Risk Factors section of AGNC's periodic reports filed with the Securities and Exchange Commission. Copies are available on the SEC's website at 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. We were very pleased with the performance of our portfolio in the second quarter, with economic return totaling just over 12%, as we recovered a significant portion of our Q1 loss. More importantly, we remain optimistic about the earnings power of the portfolio across a wide range of possible economic scenarios. This favorable earnings backdrop is evident in our net spread and dollar roll income, which increased $0.01 per share to $0.58 in Q2, despite a smaller portfolio and lower average leverage. During the second quarter, market conditions improved materially, as the unprecedented monetary and fiscal support drove a dramatic recovery in equity markets around the world. The S&P 500 recouped almost all of the Q1 losses, while the NASDAQ finished Q2 over 10% higher than where it began the year, a quarterly increase of over 30%. Fixed income credit also performed very well, with most credit spreads recovering close to 70% of the Q1 widening. Interest rates were very stable with the yield on the 10-year treasury ending the quarter within 1 basis point of where it closed on March 31. The short end of the treasury and swap curves performed better in response to growing confidence that the Fed will keep the funds rate near zero for multiple years. Despite very limited interest rate volatility, massive Fed support, and the dramatic recovery in credit-centered product, generic agency MBS performance was mixed, with lower coupons tightening modestly, while higher coupons widened. Specified pools, which underperformed dramatically in March, saw a significant recovery. The outperformance of specifications drove our strong book value and economic return performance for the quarter. Contrary to expectations, the mortgage origination market was less impacted by lockdowns and social distancing. Refinancing volumes remained very robust, and we saw rapid recovery in the home purchase market. The heavier than expected origination volume, which totaled $730 billion in the second quarter, served as a major offset to Fed MBS purchases. As a result, lower coupon agency MBS valuations, while modestly tighter quarter-over-quarter, remain attractive both in absolute and relative terms. This attractiveness is further enhanced by improved dollar roll specialness in lower coupons, a trend we expected to see during the quarter. As Chris will discuss shortly, incremental levered ROE potential on low coupon 30-year TBAs is still in the low to mid-teens, depending on the amount and durability of roll special notes. In contrast, the projected returns on most higher coupon specs have declined to the lower double digits on the back of price increases and faster prepayment expectations. From a big picture perspective, the current investment environment is very different and more favorable for us than the QE3 era. Back in late 2012 and early 2013, the Fed's purchases drove agency MBS spreads to valuations around 50 basis points tighter than today's levels by most measures. While the QE3 tightening was temporarily good for book value, it materially lowered our expected returns on new purchases and set the stage for the significant widening and spread that occurred when the Fed telegraphed the tapering of its purchases. Today, despite Fed purchases of over $850 billion since mid-March, MBS valuations remain attractive, benefit from favorable dollar roll levels, and are easier to hedge, given the zero interest rate now. In summary, we feel good about AGNC's performance in Q2 and remain confident about the earnings potential of the company. Given the lack of credit risk in our agency MBS portfolio and the favorable funding backdrop, we believe AGNC should be able to produce strong returns, regardless of the progression of COVID-19 or broader moves in the global economy. This potential for our portfolio to perform well in either a risk on or risk off scenario is somewhat unique to AGNC. At this point, I will turn the call over to Bernie to review our financial results for the quarter.

Thank you, Gary. Turning to Slide 4, we had total comprehensive income of $1.60 per share for the second quarter. Net spread and dollar roll income, excluding catch-up amortization, was $0.58 per share for the quarter, which as Gary mentioned, was up slightly from Q1, despite a materially smaller average portfolio balance, lower average leverage, and meaningfully faster prepayment production. These earnings headwinds were offset by lower aggregate funding cost, which drove the slight improvement quarter-over-quarter. Tangible net book value increased 9.5% for the quarter, led by a significant rebound in spec pool valuation. Including dividends, our economic return on tangible common equity was 12.2% for the quarter, recovering nearly half of our first quarter economic loss. So far this quarter, we estimate that our tangible net book value is down a couple of percent, given a modest pullback in spec pool pay-up values. Turning to Slide 5, our average portfolio at quarter end totaled $97.7 billion, up $4.7 billion from the end of the first quarter. Our ending leverage was 9.2 times tangible equity, down slightly from 9.4 times as of the end of the first quarter. As I mentioned, given the significant decrease in our average portfolio balance for the quarter, our average leverage was down meaningfully for the second quarter at 8.8x tangible equity, compared to 9.9x in the first quarter. Our liquidity position remained very strong in the second quarter, and is above pre-crisis levels, with cash and unencumbered agency assets totaling $4.5 billion at quarter end. Importantly, that figure excludes both unencumbered CRT and non-agency securities, as well as assets held at our broker-dealer subsidiary, Bethesda Securities. With mortgage rates dropping to historically low levels during the quarter, prepayment speeds increased across the coupon stack. Actual prepayment speeds on our portfolio increased to 19.9% for the quarter, while our forecasted life CPRs increased to 16.6% from 14.5% last quarter. Lastly, during the second quarter, we repurchased $147 million of our common stock at substantial discounts to our tangible net book value for an average repurchase price of $11.99 per share. With that, I'll turn the call over to Chris to discuss the agency market.

Thanks, Bernie. Let's turn to Slide 6. Interest rate volatility was muted in the second quarter, with 10-year Treasury rates ending the quarter 1 basis point lower at 64 basis points. The yield curve did steepen with two-year and five-year Treasury yields 10 basis points lower, ending the quarter at 15 basis points and 29 basis points respectively. Agency MBS spreads were generally tighter, but performance was mixed with lower coupon TBAs modestly tighter, while higher coupon TBAs were modestly wider. Specified pools, however, were the best performers, regaining most of the Q1 widening. The unprecedented support from the Fed, with purchases heavily concentrated in production coupon MBS, drove the outperformance in lower coupons, even with gross supplies significantly larger than expected. As you can see in the lower left table on Page 6, higher coupon TBA 3.5s and fours declined in price during the quarter, as prepayment speeds generally surprised to the upside, as the widely expected COVID-related headwinds to housing and refinance activity did not materialize. Faster prepayment speeds on more generic cohorts and the weakness in higher coupon TBA led to a strong outperformance of higher coupon specified pools in the second quarter. Let's turn to Slide 7. You can see in the top left chart the investment portfolio increased by a little over $4.5 billion as of June 30. Given the outperformance of specified pools, most of which occurred in April, we continued to trim positions in both higher quality and generic higher coupon MBS, versus adding production coupons. During the quarter, we reduced holdings in 3% coupons and above by approximately $12 billion, versus adding $16 billion in 2.5s and twos. As I mentioned on the call last quarter, we expected dollar roll financing to improve, as the combination of very strong origination volumes and large Fed purchases, which continue to clear out the worst bonds in the float, create an ideal backdrop for dollar rolls. Roll financing on lower coupon TBA is currently trading around 20 basis points to 80 basis points through repo, depending on the coupon. With this degree of specialness, the potential contribution to returns is material. As a hypothetical example, with gross returns on lower coupon 30-year MBS funded with repo around 12%, 25 basis points of dollar roll specialness has the potential to add more than 2% in incremental return. Realistically, roll specialness could be even higher as it is today, but there's no guarantee that it will persist. We remain optimistic about the investment environment, given relatively wide spreads, attractive carry, and low interest rate volatility. And while the prepayment backdrop is certainly not the tailwind we had hoped for, our diversified portfolio of higher coupons, specified pools, and production coupon TBA has offsetting risk characteristics that position us well in the current environment. I'll now turn the call over to Aaron to discuss the non-agency market.

Thanks, Chris. I'll quickly recap our current positioning and then update you with our outlook on credit. Please turn to Slide 8. After facing unprecedented price action in the first quarter, both equities and credit markets staged a fierce recovery in Q2. As it stands now, we've seen a V-shaped recovery in credit spreads, yet an uncertain backdrop remains on when the economy will be fully reopened and the resulting toll on the consumer and business sector. To put the magnitude of the rally in perspective, from the start of the year to the wides in late March, IG and high-yield CDX spreads were approximately 100 basis points and 600 basis points respectively. Subsequently, they have tightened about 70 basis points and 370 basis points. Our holdings across the portfolio were largely unchanged, with a slight shift in the vintages of our CRT portfolio to more recently issued securities. These are generally more exposed to and benefit from higher than anticipated prepayments, which is a natural fit for our portfolio. On the residential side, as we mentioned last quarter, the quick implementation of forbearance programs would likely result in reduced downside pressure on housing prices in the near term. The GSE subsequently announced the ability to defer up to 12 months of payments, thus giving borrowers a much easier path to return to current status. This was particularly beneficial for creditor's transfer, as it served to reduce potential modification-related losses. This change, along with faster than anticipated pre-payments, stabilization in forbearance requests, and the improved macro backdrop, led to a strong rally in CRT. Additionally, many of these same themes were supportive of credit spreads in other parts of the residential credit markets. On the commercial front, while increases in delinquencies have stabilized for the time being, we believe this sector remains particularly exposed to the stops and starts in the economy. While we are comfortable with our positions from a risk perspective, we generally remain defensive until we get better clarity around the timing of a potential return to normalcy in the economy. Looking forward, with the Fed actions setting the stage, we're likely to be in a regime of relatively tight credit spreads coupled with an elevated fundamental risk environment for some time. This does present many challenges, particularly so for levered investors, since the risk-return equation is a bit skewed. At the same time, with rates expected to sit close to the zero bound for years, it also means picking the right bonds today could generate reasonable returns, as bonds that ultimately have sufficient credit support are likely to see spread tightening over the coming couple of years. With that, I'll turn the call over to Peter to discuss funding and risk management.

Thanks, Aaron. I'll start with our financing summary on Slide 9. The repo market for agency MBS traded very well in the second quarter, and continues to benefit from the Fed's open market operations, as well as a very significant influx of cash into government money market mutual funds. The substantial demand for high-quality collateral, like agency MBS, has led to a meaningful re-pricing across all repo tenors. As a result, our average repo costs fell to 76 basis points in the second quarter, less than half of the 180 basis points we reported in the first quarter. A key development during the quarter was the Fed's communication regarding their intention to keep short-term rates near zero for the foreseeable future. Chairman Powell's statement that the Fed is not even thinking about raising rates makes that very clear. As a result, the repo funding curve flattened significantly. Today, for example, there was only about a five basis point cost differential between overnight repo, which trades at around 15 basis points, and one-year repo, which trades at around 20 basis points through our captive broker dealer. I expect these favorable funding conditions to continue, and as such, I expect our average repo costs to drop to around 40 basis points in the third quarter, as more of our outstanding repos reset at current market rates. Our aggregate cost of funds, which includes the cost of our swap hedges, fell to 88 basis points in the second quarter, down meaningfully from 167 basis points the prior quarter. This improvement more than offset the decrease in our asset yield and drove the notable improvement in our net interest margin, which increased to 168 basis points. Looking ahead, I expect our net interest margin to improve further in the third quarter. Turning to Slide 10, we provide a summary of our hedge portfolio, which totaled $59 billion at quarter end, unchanged from the prior quarter, and covered 66% of our funding liabilities. As we discussed last quarter, with swap rates at such low levels, we continue to view this as an opportunity to lock in very attractive funding for an extended period of time. As such, we continued to adjust the composition of our swap portfolio in the second quarter, unwinding more of our shorter-term swaps, and replacing them with slightly longer-term swaps. As a result, the average maturity of our swap portfolio increased to 5.1 years at quarter end from 4.5 years the prior quarter. On Slide 11, we show our duration gap and duration gap sensitivity. Given the stability of interest rates during the quarter, our duration gap remained roughly unchanged at negative 0.1 years. And with that, I'll turn the call back over to Gary.

Thanks, Peter. And at this point, I will open up the call to questions.

Yes, thank you. We will now begin the question-and-answer session. And the first question comes from Doug Harter with Credit Suisse.

Thanks. Gary, can you tell us how you and the board are thinking about dividends, given kind of the much higher level of earnings this quarter and your commentary about that? Are those conditions persistent?

Yes, I'd be happy to answer that, and thank you for the question. When considering our net spread and dollar roll income performance this quarter, which appears to be trending positively, one might argue that our dividend cut was unnecessary. We made that decision in April, shortly after the crisis peaked. However, I think it's important to note that while it may have seemed unnecessary, we are unsure if it isn't the best approach in terms of delivering returns for shareholders. Our dividend yield is just over 10%, which remains attractive, and maintaining a dividend below our net spread and dollar roll income allows for a beneficial effect on our book value over time. This is advantageous for both us and our shareholders. Additionally, we have a share repurchase program that we are prepared to use as another form of returning capital to shareholders when appropriate. Moving forward, we will keep evaluating this situation, mindful of the significant gap between our current net income, which is likely to improve, and the established dividend. We are also aware that there are benefits to our current approach, and the board will be closely monitoring this matter.

And then just to follow up, those points all make sense, but I guess how much flexibility do you have from a taxable income standpoint, and kind of how different are kind of taxable and kind of the core flows drop right now just kind of in that dividend discussion?

There is a significant difference between our taxable income and our core earnings. Currently, our taxable income is very low and is expected to remain low, which does not impact our ability to pay dividends. We have considerable flexibility regarding dividends. The low taxable income can be attributed to several factors, including how dollar roll income is treated. Additionally, due to our activities in the first and second quarters, particularly in repositioning our swap portfolio and adjusting duration, the primary reason for the discrepancy between our GAAP net spread, dollar roll income, and taxable income is related to terminated swaps. These swaps are amortized for taxable purposes but accounted for as upfront costs in regular income. Bernie or Peter, would you like to add anything to that?

Gary, that last point is a key one. And obviously, we still have some rebalancing to do there. So, our projection for that taxable income is that that's not going to be a constraint for some period of time.

Great, thank you, guys.

Thank you. And the next question comes from Bose George with KBW.

Hey, good morning. Actually, can you go over the drivers of specialness in the lower coupon MBS? And also, when we think about your return, should we sort of think about a base case return on the spread income kind of in the low double digits and then the specialness being sort of what flexes it up and down from there?

Yes, that's a good way to think about it. I'll address the second part of your question first. When considering a new production 30-year 2.5, if we believe the spread is between 110 and 120, that's an appropriate starting point based on prepayment assumptions. However, the first part of your question is very important. To understand what drives specialness and creates conditions where dollar rolls are special, we should note that while many people point to the Fed as a major factor, the primary piece is actually significant origination volume. We are currently witnessing unprecedented origination volumes, with the last couple of months seeing around $250 billion per month in gross issuance. The origination market functions by selling those forward, sometimes for one, two, or even three months, which lowers the price of out month TBAs. This is the first step in creating specialness. Additionally, substantial Fed purchases continue to remove existing production, accounting for about 40% of total origination. The Fed’s purchases specifically target the fastest prepaying and least desirable pools within the float, which positively impacts the pricing of dollar rolls, leaving better bonds available for other investors. Together, these components create what could be described as the ideal conditions for roll specialness. We noted this in our April call, although the roll specialness hadn't fully materialized then due to balance sheet challenges and other factors. It began to take shape in May and June and continues to do so. We are optimistic about the sustainability of roll specialness, which may fluctuate in magnitude but feels sustainable, especially in the lowest coupons for both 30-year and 15-year bonds in the near to intermediate term, contributing positively to ROEs. Therefore, while we start with our calculations based on repo-funded positions, we believe TBAs will outperform due to the roll specialness.

Okay, great. Thanks, Gary. That was very helpful. And then actually, just on your prepayment expectation, can you just talk about what you're thinking in terms of the primary mortgage rates, like what that does versus a benchmark over time?

Certainly. The primary mortgage rate has been decreasing and is likely to continue to decrease slightly. This is something that both we and the market recognize. Additionally, expectations regarding interest rates play a role. When we assess the situation, we take into account the forward curve, which indicates a potential upward trend in mortgage rates over time. In the short term, we expect mortgage rates to be somewhat lower since the primary rate influences prepayments in general. However, there are other factors to consider, such as the qualifications for mortgages, particularly in an environment with elevated unemployment and stricter underwriting standards compared to three to six months ago. These elements can counterbalance the trends. It's also important to note that in refinancing waves, we often see peaks that can be either narrow or spread out. With the decline in mortgage rates, we may experience some widening, but there is also a factor of burnout where many have already had excellent opportunities to refinance. If someone missed a significant opportunity previously, a slightly better chance later may not motivate them to act. That's a detailed response, but I aimed to address several aspects related to this topic.

Okay, great. Thanks. That was helpful.

Thank you. And the next question comes from George Bahamondes with Deutsche Bank.

Hi, good morning, Gary. You alluded to a favorable backdrop for AGNC given a very accommodative Fed, best entries anchored near zero, hedging dynamics are attractive and Fed purchases are likely to continue for some time. As you think about downside risk to book value and dividend sustainability and kind of near to medium term, what would you say are the biggest kind of things that are on your radar today?

Sure. I'll start by saying that we're not overly worried about dividend sustainability at this moment. The main concern regarding book value in the near term is likely to be faster prepayments and some additional pressure, which we've already observed recently with spec payoffs and higher coupons. It's possible we'll see more of that, but it's only a couple of percent here or there. There is also a risk to book value from lower coupons, particularly during times of high origination volumes that may exceed the Fed's support. If there isn't significant bank buying during those times, we could see periods where lower coupons widen somewhat. However, overall, without major interest rate fluctuations and with the Fed's support, we don't expect significant volatility in book value, either positively or negatively. While some may hope for a scenario where book value significantly increases, we believe that's unlikely. We anticipate a more stable environment for book value, with favorable earnings potential. Our portfolio currently includes a mix of lower coupons that benefit from the Fed's actions and have minimal prepayment concerns in the near term, along with a higher coupon specified portfolio that we're actively optimizing for current conditions. This mix helps reduce the overall volatility of our book value. I hope that clarifies things.

Right, that's helpful.

Thank you.

Thank you. And your next question comes from Trevor Cranston with JMP Securities.

Thanks, good morning. Following up on the question regarding the prepay outlook and your views on the primary mortgage rate, could you discuss your expectations for prepay speeds? For instance, if the primary mortgage rate were to reach 2.5%, recognizing there might be a flatter peak as speeds increase, how sensitive do you think those speeds would be to a 50 basis point decrease in mortgage rates? Thank you.

I'll begin now, and Chris can also provide input. The short answer is that it greatly depends on the type of security, the coupon rate, how long the loan has been outstanding, and various other characteristics. For instance, I anticipate that specified pools with 4s and 4.5s, which make up a significant portion of our holdings, won’t be significantly affected. While it's possible that prepayment speeds could increase, there’s a strong incentive for refinancing today, and we have observed a healthy rise in those speeds. However, even with a stronger refinancing incentive and potential increases, we do not expect dramatic changes, and those increases should taper off relatively quickly. The specified pools in the 3% to 3.5% coupon range are likely to react the most to declines in rates. In terms of TBAs, most seasoned TBAs in the 3s through 4s will also see an uptick, but they are already performing quickly. The concern lies with lower coupons, such as 2.5s and 3s, which may start behaving like 3s, requiring careful monitoring. It will take time, as the Fed will continue to absorb the worst-performing seasoned pools. Therefore, there’s likely to be at least six to nine months of stability in the newest originations in those coupons, but caution is needed with lower coupon TBAs, like 2.5s, as they could quickly enter a prepay window. Nonetheless, this should be manageable given the Fed's purchase program and their efforts to stabilize the float. An additional approach to manage this would be to adjust your coupon strategy. Chris, do you want to add anything?

Yes, Gary, you addressed that quite well. The important thing to remember is that we're already operating at or close to capacity for lenders, with about 90% of the mortgage market influenced by at least a 50 basis point rate incentive. Therefore, even if rates remain unchanged, it will take time for the primary-secondary spread to decrease, as we are nearly at capacity. To directly answer your question, 2.5s would be the most sensitive to a no-cost 2.5% mortgage rate, but we are still some distance from that at this point.

Okay, that's helpful. Appreciate the comments. Thank you.

Thank you. And our last question comes from line of Charlie with JP Morgan.

Good morning, everybody. Thanks for taking the questions today. You noted in the prepared remarks about $730 billion in origination volumes during the quarter, which helped offset those Fed purchases. I'm wondering what your outlook is like for supply through year-end. And I'm wondering if there was sort of a catch-up effect as the lockdowns lifted and things return to normal or if that pace is sustainable given where rates are.

What I would say is that we didn't observe the expected impact from the lockdowns and social distancing in the second quarter regarding origination volumes, refinancing, and the housing market. Interestingly, what happened in the mortgage market mirrored trends seen more broadly in retail delivery. The business adapted instead of shutting down; it transitioned online and relied on technology, which we noted in the mortgage sector as well. Therefore, I would emphasize that we experienced minimal disruption, which was certainly a concern. Moving forward, we believe that production levels will stay high for at least the next couple of months before likely tapering off as people become accustomed to the current interest rates. There is a bit of a catch-up on the demand side, as we expect production to be better absorbed in the latter half of the year, and we're starting to see that slightly in July compared to earlier months like March and April, despite significant Fed purchases. The prepayments are unlikely to accelerate and are already at elevated levels. Additionally, there's a need to catch up on reinvestment of paydowns, including at the Fed, and many other investors have experienced high prepayment months without full reinvestment. Overall, the combination of high production levels, continued Fed support, and the need for reinvestment helps explain the limited volatility in lower coupon prices and MBS spreads as we look ahead. This context also contributes to the continued special nature of dollar rolls.

Got it. Thanks very much for the color, appreciate it.

I'd like to thank everyone for their interest in AGNC, their participation in this call. Please stay safe, and we look forward to talking to you next quarter.

Thank you. That concludes today's teleconference. Thanks for calling in for today's presentation. You may now disconnect your lines.