Good day, and welcome to the third quarter 2021 Annaly Capital Management earnings conference call. All participants will be in a listen-only mode. Should you need assistance, please signal a conference specialist by pressing the star key followed by zero. After today's presentation, there will be an opportunity to ask questions. To ask a question, you may press star then one on your touch-tone phone. To withdraw your question, please press star then two. Please note this event is being recorded. I would now like to turn the conference over to Mr. Sean Kensil. Please go ahead.
Good morning, and welcome to the third quarter 2021 earnings call for Annaly Capital Management. Any forward-looking statements made during today's call are subject to certain risks and uncertainties, including with respect to COVID-19 impacts, which are outlined in the Risk Factors section in our most recent annual and quarterly SEC filings. Actual events and results may differ materially from these forward-looking statements. We encourage you to read the disclaimer in our earnings release in addition to our quarterly and annual filings. Additionally, the content of this conference call may contain time-sensitive information that is accurate only as of the date hereof. We do not undertake and specifically disclaim any obligation to update or revise this information. During this call, we may present both GAAP and non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in our earnings release.
As a reminder, Annaly routinely posts important information for investors on the company's website, www.annaly.com. Content referenced in today's call can be found in our third quarter 2021 investor presentation and third quarter 2021 financial supplement, both found under the Presentation section of our website. Annaly intends to use our webpage as a means of disclosing material and nonpublic information for complying with the company's disclosure obligations under Regulation FD and to post and update investor presentations and similar materials on a regular basis. Annaly encourages investors, analysts, the media, and other interested parties to monitor the company's website in addition to following Annaly's press releases, SEC filings, public conference calls, presentations, webcasts, and other information it posts from time to time on its website. Please also note this event is being recorded.
Participants on this morning's call include David Finkelstein, Chief Executive Officer and Chief Investment Officer, Serena Wolfe, Chief Financial Officer, Ilker Ertas, Head of Securitized Products, Tim Coffey, Chief Credit Officer, and Mike Fania, Head of Residential Credit. With that, I'll turn the call over to David.
Thank you, Sean. Good morning, everyone, and thanks for joining us for our third quarter earnings call. Today, I'll provide an overview of the broader market environment, including our thoughts on the Federal Reserve's reduction in asset purchases, briefly touch on our performance during the quarter, and highlight some of our recent achievements and positioning across our businesses. Ilker will provide more detailed commentary on our agency and residential credit portfolios, and Serena will discuss our financial results. As Sean noted, our other business heads are also here this morning to provide additional context during Q&A. Now first, with respect to the macro landscape, the COVID Delta wave and related factors have led to a moderation in the economic recovery. Labor market gains have slowed relative to the strong pace at the beginning of the summer.
Production bottlenecks and global supply chain disruptions have caused delays that raised prices on products in high demand. While inflation has been boosted by higher goods and energy prices, record home price appreciation has started to filter into inflation's shelter component, suggesting that price pressures may persist for longer than previously anticipated. Meanwhile, interest rates experienced meaningful intra-quarter volatility given the shifting narrative on the economic recovery and inflation. Early in the quarter, rates rallied as markets sought direction on the magnitude of the impact of the Delta variant. Later in the quarter, as a relative reduction in COVID case counts led to a return to economic optimism, rates sold off and the curve steepened to end the quarter. Now, the greater near-term focus in the macro landscape, however, is the imminent reduction in the Fed's pace of asset purchases.
Following the September FOMC meeting, we now have a clearer picture as to what the taper will likely look like. Fed is expected to reduce Treasury and agency MBS purchases by roughly $10 billion and $5 billion per month, respectively, beginning as early as November. This pace would result in the Fed halting balance sheet expansion in the summer of 2022, though we expect reinvestment of portfolio runoff to persist well beyond the end of the taper, consistent with the QE3 experience. Most notably, the Fed's transparent communications have helped to limit the market impact to both interest rates and agency MBS spreads ahead of the official taper announcement. While the Fed has attempted to decouple the taper and eventual rate hikes, elevated inflation readings and more hawkish central bank messaging globally have accelerated investors' expectations of a rate hike.
With markets currently pricing as many as two hikes in 2022, we remain vigilant in managing Annaly's duration exposure, both in the front end and the long end, as Ilker will expand later on. Now, reflecting briefly on the agency MBS supply and demand outlook in light of the taper announcement, private market participants will need to absorb an increased amount of mortgages in 2022. Elevated net issuance remains an uncertainty in the supply-demand outlook, but we currently estimate supply to the private market next year will be similar to levels seen in recent years pre-pandemic. Several factors are supportive of more limited widening in spreads before buyers emerge. Banks are flush with deposits and see little loan demand, suggesting strong appetite for securities is likely to continue, particularly at potential wider spread levels.
Money managers remain underweight mortgages but will likely increase their relative allocation if mortgage spreads become more attractive. Ample liquidity, best seen in the $1.6 trillion pledged to the Fed's reverse repo facility at quarter end, and readily available financing remain the main factors underlying the current accommodative financial conditions. The repo market remains highly liquid and has allowed us to decrease our cost of funds to another record low. In line with efforts earlier in the year, we continue to broaden our financing through an increased use of credit facilities by funding high -quality credit securities for longer terms. These actions help to enhance Annaly's liquidity profile at extremely attractive spread and haircut levels, as Serena will discuss in more detail. Now turning to Annaly's performance in the third quarter.
Our portfolio generated a positive economic return of 2.9%, reflecting a $0.02 gain on book value and earnings available for distribution of $0.28. We achieved these returns amid continued conservative portfolio positioning, maintaining economic leverage at the low end of our historical range and unchanged quarter-over-quarter at 5.8x. Our liquidity remains at our highest levels, with total unencumbered assets of $9.8 billion at quarter end. We continue to see relatively tight spreads and as a result are comfortable with our more cautious approach to managing the portfolio. Now that being said, should spreads become more attractive, our nimble positioning leaves us prepared to take a more offensive posture and increase leverage should it be justified.
Now over the quarter, mortgages performed in line with hedges in this environment as the sector benefited from clarity surrounding the upcoming taper and healthy demand from banks. We increased our agency portfolio by nearly $3 billion in Q3 as we invested a portion of the proceeds from our previously announced commercial real estate sale. Additions to our agency portfolio, which Ilker will cover in more detail, were primarily a placeholder as agency MBS remained fully valued, while credit sectors offer attractive pockets of opportunity, but deployment of capital is more episodic. With respect to capital allocation to end the third quarter, 30% of our capital was allocated to credit, up slightly from 29% in the prior quarter, in line with our view on the relative value equation vis-à-vis agency and our deliberate portfolio positioning ahead of the taper.
Our Residential Credit business represents the majority of our credit allocation at 21% and had another strong quarter as the group continues to successfully execute on their strategy. With assets of $4.3 billion, the Residential Credit Group is now larger than it was pre-COVID, and assets are up over 70% since year-end 2020. We maintain an optimistic outlook on the business given persistent, robust housing market fundamentals and long-term tailwinds driving the need for private capital in the market. The Onslow Bay-based securitization platform remains very active, completing nearly $2 billion of securitizations since the start of the third quarter and nearly $3 billion of securitizations year to date. We expect to maintain our securitization footprint, and we are continually adding to our partnerships to drive new sources of securitization collateral.
Additionally, our correspondent network is adding new partners and Annaly's large capital base and market expertise uniquely position Onslow Bay as an aggregator within the industry. Also, as it relates to OBX, I wanted to briefly touch on the impact of the recent suspension of FHFA's cap on second homes and investor properties. While the decision will have an impact on the delivery of agency-eligible loans to our platform, it does not temper our bullish outlook for the sector. Annaly was an active issuer of agency-eligible investor loans before the cap was put in place, and notably, we issued three securitizations in 2019 and 2020, backed by $1.15 billion of collateral prior to the PSP amendment limiting delivery of investor loans to the GSEs in March of this year.
As the FHFA has reversed the introduction of the caps, we expect Annaly and private markets more broadly to be able to compete with GSE LLPA-adjusted pricing again, assuming securitization execution remains attractive. Now regarding progress on our MSR business, we grew our holdings by more than 40% on the quarter, with the portfolio representing $575 million in market value and 4% of dedicated capital. As we scale the business, we have solidified our position as a reliable partner in the MSR sector, supplementing our bulk transactions with acquisitions through flow relationships. We've increased our MSR holdings by $470 million in 2021, and we anticipate responsibly growing the portfolio through our unique position as a non-competitive partner to originators that need liquidity and capital.
Turning to our middle market lending business, we closed our inaugural private closed-end fund subsequent to quarter end with just north of $370 million of capital. The fund is approximately 2 times larger than the median size of first-time direct lending and private debt funds. It includes a mix of both U.S. and European investors comprised of public and corporate pensions, insurance companies, and asset managers. The fund has supported nearly $450 million of middle -market loan investments to date with an attractive risk-adjusted return profile. Annaly is co-investing 50-50 alongside each fund investment, bringing a strong alignment of interests.
We believe the fund serves as a testament to the track record and expertise of our dedicated middle market lending team, allows for enhanced capital allocation flexibility to further scale the strategy, and provides recurring fee revenue to the REIT. Including the fund, the middle market lending strategy managed $2.3 billion in funded assets at quarter end. Lastly, as the largest mortgage REIT with the capability to invest across all aspects of a mortgage loan, our strategic initiatives over the past year, including investing in MSR and balance sheet and expanding our residential credit business, have prepared us to be a leading source of capital in residential housing finance.
Ultimately, the strength of Annaly's diversified model is enabled by our size and scale, and we remain confident in our ability to generate stable returns throughout various market environments and across economic cycles as we have done historically. Now with that, I'll turn it over to Ilker to provide a more detailed lens into our agency and residential credit portfolio activity and outlook.
Thank you, David. Against the economic and interest rate backdrop David described agency MBS performed broadly in line with hedges, but performance was mixed across the coupon stack. Specifically, lower coupon MBS widened modestly due to continued record supply levels, Fed's taper moving closer, and elevated prepayment speeds. At the same time, higher coupons outperformed hedges into the retracement of higher interest rates that materialized later in the quarter and emerging signs of prepayment burnout. Our portfolio performance showed the benefits of the barbell strategy we have discussed in several of our recent earnings calls. In third quarter, the outperformance of higher coupons more than offset the widening experience in the lower production coupons. We believe the approach is also efficient in protecting the portfolio from any potential taper-induced widening, as nearly 60% of our agency portfolio consists of non-Fed-supported coupons. That is 3% and higher.
In addition, our specified pool portfolio is more than 45 months seasoned on average, which provides a strong source of durable earnings with minimal duration and convexity risk. Turning to our portfolio activity, we tactically increased our agency portfolio by nearly $3 billion during the quarter, predominantly in lower coupon TBAs to opportunistically redeploy a portion of the proceeds from our previously announced commercial real estate sale. Our purchases consisted of primarily lower coupon TBAs, which exhibited spread widening earlier in the quarter. We have favored TBAs over pools in lower coupons due to attractive implied financing rates in the dollar roll market, which remain quite special in the context of negative 30-40 basis points financing and enhanced levered returns. We expect these favorable conditions to persist well into 2022 as the Fed remains a net buyer of MBS throughout the taper.
In addition, the sector offers strongest liquidity should we choose to redeploy the capital into other opportunities. Regarding prepayment speeds, recent prints have mirrored the divergent performance across the coupon stack. Lower coupons have been very reactive in what remains a historically low mortgage rate environment in which mortgage originators have ample capacity, while higher coupons have exhibited moderate burnout since peak prepayment speeds in March. At current mortgage rates, roughly 31% of mortgage universe have greater than 50 basis points of refinancing incentive, which is down significantly from the 72% at the beginning of the year. The restriking of the universe, along with slower speeds and higher coupons, has improved the prepayment outlook going forward. Additionally, the historically strong housing market has led to elevated cash-out refinancing activity, which should help mitigate expansion risk for mortgage portfolios should interest rates continue to rise.
Our portfolio prepaid 13% slower quarter-over-quarter, and our outlook is for a further 10%-15% reduction in the fourth quarter due to higher rates, less reactive borrowers, and slower housing seasonals. In our hedge portfolio, we added duration hedges at lower rates throughout the quarter, positioning the portfolio for modestly higher yields. We added Treasury futures shorts across the curve and increased our longest long-end swap position by exercising $3 billion in-the-money swap options. We also took advantage of relatively lower levels of implied volatility to replace our exercised swap options with higher strike swap options at longer expiries. This proactive approach has already proven worthwhile given the rise in interest rates in September and October. The interest rate outlook remains cloudy due to uncertainties over inflation, the Fed's response, and market positioning. We will continue to minimize our interest rate exposure in fourth quarter.
It was an active quarter for our Mortgage Servicing Rights business. We grew our portfolio through $200 million in bulk purchases and began transacting through flow arrangements, which we see as a good growth opportunity going forward. Additionally, in the fourth quarter, we expect to sell the vast majority of our legacy MSR portfolio for roughly $85 million in estimated proceeds. The sale of higher -coupon seasoned MSR will provide additional capacity to grow our allocation to newer production, low -coupon MSR, which will serve as more effective hedge to the MBS basis.
Additionally, the planned transaction is with an operational partner that ascribes a higher value to customer acquisition, which will drive strong execution for both parties. We believe this transaction highlights Annaly's unique position in the MSR market as a capital partner for the originator community that does not compete for their customers. Moving to our Residential Credit business.
We continue to increase our allocation to the sector as measured by both assets under management and capital deployed. The residential portfolio ended the quarter at $4.3 billion of market value and $2.9 billion of dedicated capital, representing 21% of the firm's capital. The growth of the portfolio was predominantly through our acquisition of residential whole loans, as we purchased $1.4 billion throughout the third quarter. Our Q3 acquisitions were across both expanded prime non-QM markets and agency -eligible investor loans. Our securities portfolio was up modestly, approximately $125 million, as we saw diminished opportunities in third-party securities relative to prior quarters. As David mentioned, our OBX securitization platform was active in Q3 with $1.1 billion of securitizations across three separate deals. Also to note, post-quarter end, we priced two additional transactions, representing another $800 million of issuance.
Life -to-date levered returns of our whole loan strategy remains in low -to-mid double digits, utilizing minimal recourse leverage. Our GAAP residential whole loan portfolio ended Q3 at $5.8 billion, 70% of which is currently term financed with non-mark-to-market securitizations. In summary, MBS technicals remain constructive as trends in prepays continue to improve, and recent price action highlights the amount of support for MBS into higher yields. However, spreads have potential to drift wider as the taper commences and supply remains elevated. We believe any widening is likely to be orderly, but we expect to maintain a conservative leverage profile and proactively manage our basis and interest rate exposure. With that, I will now turn the call over to Serena to discuss our financial results.
Thank you, Ilker, and good morning, everyone. This morning I'll provide brief financial highlights for the quarter ended September 30, 2021. Consistent with prior quarters, while our earnings release discloses GAAP and non-GAAP earning metrics, my comments will focus on our non-GAAP EAD and related key performance metrics, which exclude PAA. At the risk of repeating myself, I would say that the general themes for this quarter's earnings are consistent with recent quarters. That is, this quarter we continued to generate strong results from the portfolio, benefiting from the continued low interest rates in the funding market, marking another quarter of record low cost of funds and sustained specialness in dollar roll financing. To set the stage with some summary information, our book value per share was $8.39 for Q3, and we generated earnings available for distribution per share of $0.28.
Book value increased primarily due to GAAP net income of $522 million or $0.34 per share, partially offset by the common dividend declaration of $320 million or $0.22 per share, and other comprehensive loss of $142 million or $0.10 per share, on higher rates. The negative impact to our book value from our Agency MBS valuations was more than offset by the gains on our swaps, resulting from higher hedge rates and higher mark-to-market valuations on our MSR and resi credit portfolios, which together contributed approximately $0.06 per share to book value during the quarter. Combining our book value performance for the $0.22 common dividend we declared during Q3, our quarterly economic and tangible economic returns were both 2.9%.
As we take a closer look at the GAAP results, the valuation drivers we mentioned above benefited GAAP results as we generated GAAP net income for Q3 of $522 million or $0.34 per common share, up from GAAP net loss of $295 million or $0.23 per common share in the prior quarter. Expanding further on those summary comments, specifically GAAP net income increased due to net realized and unrealized gains on the swaps portfolio in the third quarter of $130 million, compared to losses of $224 million in the second quarter. Lower net losses on other derivatives and financial instruments in the third quarter of $45 million, compared to $358 million in the second quarter.
Net unrealized gains on instruments measured at fair value through earnings in the third quarter of $91 million, compared to $4 million in Q2. As David mentioned during our second quarter earnings call, we anticipated that EAD would moderate slightly. This is reflected in a $0.02 reduction in EAD compared to the second quarter. The most significant factors that impacted EAD quarter-over-quarter included lower interest income predominantly related to the runoff of higher -yielding assets and the reduction in investment balances, which was offset by higher TBA dollar roll as the desk positioned our TBA portfolio based on the relative attractiveness compared to pools. EAD benefited from lower expenses on the net interest component on swaps from the termination of $28 billion gross notional swaps as the swaps portfolio was repositioned to reduce exposure to LIBOR.
Finally, lower interest expense of $50 million in comparison to $61 million in the prior quarter, due to lower average repo rates and balances. It should also be noted that the sale of our commercial real estate business allowed us to shift capital allocation to a higher percentage in resi credit, where we saw higher levels of EAD on whole loan and NPL, RPL purchases throughout the quarter. Now turning to our financing. Early in 2021, we communicated that we were forecasting lower repo rates for an extended period. As such, we have begun to opportunistically target extended terms, that is, 6-12 months for our repo book. This has resulted in higher weighted average days to maturity for our book during 2021 in comparison to recent years.
In doing this, we believe that we have appropriately managed the risk of our liabilities while capturing the lows of the interest rate market. Additionally, given our ample liquidity in the prior quarters, we elected to fund certain credit assets with equity, further contributing to a lower cost of funds. These strategies resulted in the third quarter, marking nine consecutive quarters of reduced cost of funds for the company. Our weighted average days to maturity for Q3 was 75 days, slightly less than the prior quarter at 88 days. This reduction in days is due to the timing of rolling repo extended earlier in the year and not a function of a change in strategy by the company. As David discussed, the market is pricing in rate hikes to begin in the latter half of 2022.
Therefore, we have seen steepening in the repo curve as of late. While longer -term repo does come at a higher cost today, our over $30 billion in shorter -dated 0- to 3-year payer swaps has been of considerable benefit to hedging this eventuality. Given the strong liquidity in the repo market that will likely persist beyond initial rate hikes, we have focused our effort on hedging short-term rates as opposed to repo spreads versus policy rates. Additionally, although our overall repo balances have been reduced since the beginning of 2021, we have tried to maintain steady balances within our broker dealer. This has given us the opportunity to take advantage of attractive overnight funding conditions amid excess reserves and steady decline in TGA balances.
As in prior quarters, we continue to see strong demand for credit assets on the part of repo lenders, and we have opportunistically begun to lever credit assets at very competitive terms, both rates and haircuts. Our average weighted days for credit assets are approximately 100, and we continue to target longer duration funding to lock in those competitive spreads and haircuts at all-time tights, consistent with our prudent and conservative approach to maximize liquidity. Given the growth in our resi credit businesses, we continue to add new warehouse facilities and amend existing facilities to meet the business's needs, with a further $300 million facility put in place during the quarter and increased capacity for our resi credit partnership with a sovereign wealth fund.
To provide additional color regarding our reduced interest expense for the quarter, our overall cost of funds decreased 17 basis points quarter-over-quarter from 83 basis points to 66 basis points. Our average repo rate for the quarter was 15 basis points compared to 18 basis points in the prior quarter. We ended September with a repo rate of 15 basis points, down from 32 basis points at the end of December 2020. With LIBOR reform looming, we took the opportunity during the third quarter to reposition our swap portfolio and reduce our exposure to LIBOR, resulting in the aforementioned termination of $28 billion of swap notional and reduced interest component of swaps for the quarter and future periods.
The portfolio generated 204 basis points of NIM, down 5 basis points from the record NIM level of 209 basis points in Q2, driven by the lower interest income, partially offset by lower interest expense and improved TBA dollar roll income, up 3 basis points. Average yields decreased 13 basis points from 2.76% to 2.63%, mainly because of the change in the composition of assets towards lower -yielding assets in the quarter, both agency and resi credit. Moving now to our operating expenses. Efficiency ratios for the quarter decreased in comparison to Q2's ratio of 1.55%.
As expected, we saw a reduction in our OpEx to equity ratios during Q3 as we realized the benefits of the reduced compensation and other expenses from the disposition of our ACREG business. We saw a reduction due to the timing of certain fee payments and professional fees, along with the true-up of prior period accruals in the second quarter. Our OpEx -to-equity ratios were 1.28% and 1.4% for the quarter and year to date, respectively, with the full -year expenses expected to be at the lower end, if not below, the revised range of 1.45%-1.60% provided in the first quarter.
To wrap things up, Annaly ended the quarter with an excellent liquidity profile with $9.8 billion of unencumbered assets, up from the prior quarter's $9.6 billion, including cash and unencumbered Agency MBS of $5.9 billion. The composition of our unencumbered assets changed slightly this quarter, with an increase in Agency due to lower on -balance sheet leverage and an increase in unencumbered assets due to MSR growth, partially offset by reductions due to investments sold during the quarter and the levering of certain non-agency securities that I discussed earlier. That concludes our prepared remarks. Operator, we can now open it up to Q&A.
Thank you. We will now begin the question and answer session. To ask a question, you may press star then one on your touch-tone phone. If you're using a speakerphone, please pick up your handset before pressing the keys. If at any time your question has been addressed and you would like to withdraw your question, please press star then two. At this time, we'll pause momentarily to assemble our roster. The first question will come from Steve DeLaney with JMP Securities. Please go ahead.
Thanks. Good morning, everyone. Given that you have 13 analysts following the company, I'm just gonna ask one question to lead off.
David, curious, your quarterly dividend, the $0.22, has been stable for the past six quarters, going back to Q2 2020 after COVID. What does management and the board want to see that would give support to increasing that quarterly payout? Thanks.
Good morning, Steve, and it's good to hear from you.
Yes, sir.
First and foremost, just to back up in terms of how we set the dividend. Obviously, our board provides strong guidance on that front. In conjunction with the board, we look at the overall dividend yield, what the earnings outlook looks like, and we want to achieve a competitive yield. Currently, our yield on book value is 10.5%, 10.25% on the actual stock price this morning. In terms of what we want to look for to increase the yield or increase the dividend, we would need to see, obviously, wider asset spreads, a better investment opportunity, more durable earnings. We feel good about where earnings are this quarter, and we feel good about covering the dividend into 2022, certainly.
We do recognize that we have outearned the dividend consistently since the second quarter of 2021. We're comfortable with that, and we're very comfortable with where the dividend yield is today. It's certainly competitive with the rest of the market, and we're content right here.
Yeah. Well, I would certainly agree that a 10% dividend yield is too high, and then therein comes the catch-22 of what comes first. We certainly look at 10% and say there's certainly upside in the stock as that dividend comes down; the credit-oriented names are more in the 8%-9%, and you're certainly adding a good bit of that. Thank you for the comments.
Yeah, thanks, Steve.
Okay.
The next question will come from George Bose with KBW. Please go ahead.
Hey, guys. Good morning. This is Bose. Actually, I wanted to ask about the $1.5 billion ready portfolio. Can you just give us some color on that? What are the returns gonna be and how much capital is that going to use in the end once that's securitized?
Sure, Bose. Good morning. I'll start off and then hand it over to Mike. I would separate both loans from securities. We have had an emphasis on growing our loan portfolio and securitizing. The securitized portfolio has grown, but I will say from a return standpoint, the loan to securitization market is more attractive in terms of what we can generate relative to where our securities are currently priced. Mike, you can jump in here.
Sure. Thanks, Bose. I think we ended the quarter $3.2 billion in securities on an economic basis and then $1.1 billion in terms of whole loans. I think when you think about the conversion of whole loans to securities on balance sheet, we're retaining anywhere between, I'll say , 7% and 8%. We'll say $75-$85 million is what ultimately will be created from that $1.1 billion of whole loans. But to note, we settled $2.8 billion of whole loans year to date, and we ended the quarter with a $1.2 billion pipeline. We remain positioned to continue to go to the securitization markets to the extent that they're open.
I would think about it; probably 7%-8% of the whole loan portfolio is ultimately converted to securities on balance sheet.
Okay, great. The target ROE on that on a levered basis?
Yeah. On securitization, we're in the low double digits right now, Bose.
Okay, great. Thanks. Then on the MSR, investment, how large could that get as a percentage of your capital? In terms of the yields on the MSR that you're buying, can you just give us some color?
Sure. As we've talked about both the last two quarters, we would like to get our MSR portfolio up to 10% of capital, but we've also stressed that we're gonna be patient in doing so. We're not gonna chase returns in the sector. We were happy with the growth that we achieved in the third quarter. Again, it may take time because the market is competitive, but we're finding opportunities. Ilker can talk a little bit about returns.
Yeah. The purchases that we were buying with the hedge benefit will be very low doubles, but at the current pricing it will be high singles. That's why, like, we reduced our purchases. We reduced our purchase pace recently.
Yeah. No. We can also distinguish bulk purchases versus flow, where with flow purchases, you can get into the double digits currently.
Okay, great. Thanks.
Thanks, Bose.
The next question will come from Rich Shane with J.P. Morgan. Please go ahead.
Thanks, everybody, for taking my questions this morning. I just wanted to talk a little bit about the disparity between actual prepayment speeds and the long-term CPR assumption's fairly wide. I'm curious how we should think about the convergence of that over time and the implications in terms of your reported numbers.
Sure. One of the main reason is the steepness of the curve. Long-term speeds are at the forwards, and as the curve steepens, like, mortgage rate goes up and forward speeds goes down. That's one of the reason. Second reason is the burnout. As the portfolio burns out, speeds will slow down. But the curve is the bigger reason. If the curve flattens, you will see those two numbers approach to each other.
It does, but at the same time, ultimately they have to converge at some point. How do we think about that from a reporting perspective, what we should anticipate and really what the timeframe on that convergence should be? I mean, we're wider than 10 points at this point, and that just seems unsustainable. Trying to think about the timeline and the accounting implications.
Yeah. Rick, I'd say there is a consistent catch-up that is adjusted on a quarterly basis, number one. Number two, let's just look back. I realize the disparity between actuals and long-term projection is quite wide now. If you go back a few years, when rates were higher, you did have portfolio speeds that were in the low double digits, very consistent with long-term projections. , to the extent rates do normalize, then we're gonna ultimately approach those longer -term CPRs if we follow the forwards, as Ilker suggested, and potentially even through those levels, and we'll average out. It's all dependent on where actual rates do go out the horizon.
Okay. No, look, I'm not questioning the long-term prepayment speeds, particularly given the burnout that we're all anticipating. I'm just trying to think about what it means if you are persistently above the long-term assumption, because obviously what's left has a very low speed, but there's not much of it remaining. As you replace or reinvest, you could have this situation where you sort of have this persistent gap between the two.
Yeah, you can look at it this way, though. As long as prepayment models were accurate, and if forwards does not realize and curve stays here, yes, short-term speeds will be higher than the long-term speeds. Through that, then you will get the roll -down benefit on your hedges. If the forwards though does not materialize, then your hedge costs will never increase as much as the forwards realize. When you look at the spread of the mortgages to the hedges, you are taking the forward roll down into account. One way is saying that, okay, forwards realized, then your hedge cost goes higher, or other way is saying that forwards are not realized, then your hedge costs will be lower.
As long as what we are modeling or what market is modeling is consistent with where the mortgage rate and where the prepayments are, that those two effects will offset each other.
Okay. Got it. Thank you. I've taken a lot of time. I apologize. Thank you, guys.
Thanks, Rich.
The next question will come from Eric Hagen with BTIG. Please go ahead.
Hey, good morning, guys. So we know that overnight repo is very plentiful in the treasury market, like you noted. Can you share where on the term structure repo liquidity is currently most abundant for agency mortgages and how you expect that could evolve as the Fed begins to taper and more collateral supply gets put onto the market?
Sure, sure. Good morning. That's a good question. I'll just give you a quick run of where bilateral repo is right now. For a month, it's about 12 basis points, three months, 14; 17 roughly for six months, and then in the mid-20s for a year. Obviously we have seen a steepening in the slope of that curve. Now let's look at repo and break it down for clarification. There are two components to repo cost. There's the short rate component or the Fed funds component, if you will. And then there's the liquidity component, which is the spread of bilateral repo over that short rate. And the way we look at the current environment, what we're most concerned about is the rate component.
We've seen hikes get priced in, acceleration of hikes get priced into the market as of late. Obviously that's something we're very focused on. Now, however, when you look at our hedge profile, we're very well covered from that standpoint. , $31 billion in zero to three-year swaps and 80% hedge ratio at quarter end, actually a little bit higher than that now, around 84%. So the rate component and the risk associated with short rates going up, we're well covered from. Now, the liquidity component, given our days are about 75, roughly right now, we are not as concerned about that over the very near to intermediate term because of the liquidity in the system.
A Fed balance sheet of about $8.5 trillion, $1.6 trillion in the reverse repo facility at quarter end, a standing repo facility at the Fed in case there's any destabilization in repo market. We're much more focused on hedging the rate component, which we have, and we're confident that liquidity in the near term is going to be ample. We're not as inclined to pay as much of a premium for term repo; we're actively looking to find good value in, say, six-month to one-year repo.
Got it. That's helpful. It seems like one of the benefits of moving lower in coupon in the agency portfolio is that the premium risk on your capital structure gets reduced or tempered a little bit because you're buying lower -priced securities. I mean, would you agree with that? Do you think that changes the way you think about your leverage and the way you guys manage risk in other areas of the portfolio too?
Yeah. Just to talk a little bit about premium risk, I'd say it's always a concern of ours, but if you back up to the beginning of the year, it's less of a concern now than it was then, because for example, we had $11 billion higher balance in fixed -rate securities in the portfolio, largely premiums, obviously. Rates were much lower at the time, so the risk of refi was greater. Then, perhaps most importantly, we did experience a bit of a repricing of higher coupons in the second quarter. As a consequence of that, premium coupons are priced to very fast speed. We're always concerned about the premium in the portfolio.
Right now, I think it's 27% of our equity balance, which is lower than it's been over the past year, certainly. Does it influence our leverage? We do take a holistic approach, and what most influences our leverage is the attractiveness of the overall agency market and also our capital allocation, as we've talked about in the past. But generally speaking, we are shifting down in coupon into TBAs because the carry is better, and TBAs or lower coupons did cheapen a bit relative to higher coupons, both in the third quarter and on a relative basis to higher coupons a touch in the fourth quarter thus far.
Again, we take a holistic look at leverage and a specific amount of premium obviously has us concerned about refi risk. It's generally a bigger picture than that.
Got it. Thank you, David.
You bet, Eric.
Again, if you have a question, please press star, then one. Our next question will come from Kenneth Lee with RBC Capital Markets. Please go ahead.
Hi. Thanks for taking my question. I'm wondering, on a broader level, could you talk about what are you looking specifically for before you start taking a more aggressive approach to leverage and portfolio positioning? Is it just a matter of spread widening in certain assets, or are there any macro considerations? Thanks.
Sure. Good morning, Ken. Look, this also goes back into the first question with respect to the dividend. When we look at the market and the landscape right now, we are approaching the Fed tapering and obviously, rate hikes a little further on the horizon is something that we're obviously very focused on. Is now the right time to raise leverage? We don't think so. Asset spreads are relatively tight. We're able to generate what we think to be strong earnings, but we would need to see wider spreads, particularly in agency, before we did raise leverage. I'll say that given our liquidity profile, as Serena discussed, and we are at the low end of the range.
We're at the lowest level of leverage that we've been in over five years. We do have ample opportunity should that eventuality materialize. If the market ended the quarter where it is today and the portfolio looks as it looks today, I'd say our leverage would be, , here to slightly lower even. Should spreads widen, we will have the capacity to do so, but we do need to see more abundant spreads to take leverage up. What that is dependent exactly on what the state of the world is at the time. Generally speaking, we're not inclined to increase leverage right at the moment.
Gotcha. Very helpful. One follow-up, if I may. The private closed-end middle markets lending fund, could you talk a little bit more about what you see are opportunities in that area over time with the potential for perhaps additional funds? Thanks.
Sure. Look, I think when we look at the middle market business, we have said repeatedly that the returns in that sector are very complementary to the agency portfolio, given the low correlation. The team has built a great franchise and has very extensive relationships with private equity that we think create a lot of barriers to entry. As a consequence, they were able to successfully raise outside capital, and the catalyst was looking through the lens of the Annaly portfolio. We like the business, and we very much like the assets, but given the fact that it's corporate lending and we're a REIT, there are limits to its potential growth.
Outside capital certainly solves the issue of enabling that business and that portfolio to further scale. It also reduces some of the concentration risk of individual positions. We're very happy with the accomplishment of raising the outside capital. To the extent there's more opportunities, we'll see. Right now we feel like we have capacity on the Annaly balance sheet to add to the portfolio. We're happy with how it's performed. Tim, if you want to add anything to it, by all means.
I think David summed it up beautifully.
Great.
Great. Very helpful. Thanks again.
You bet. Thank you, Ken.
This concludes our question and answer session. I would like to turn the conference back over to Mr. David Finkelstein for any closing remarks. Please go ahead.
Thanks, Chuck, and thank you guys for joining us today. Have a good holiday season, and we'll talk to you at the beginning of the year.
The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.