RBC Capital Markets Financial Institutions Conference. My name is Kenneth Lee. I am the Senior Equity Analyst with the firm covering the mortgage REIT sector, and welcome to our industry panel, The Outlook in Mortgage Finance. Now, I'm very pleased to be joined by my panelists today. To my immediate right, David Finkelstein, CEO and Co-CIO of Annaly Capital Management. Dave joined Annaly in 2013, and prior to that, he served as a primary strategist and policy advisor for the MBS purchase program at the Federal Reserve Bank of New York. To his right, we have Bill Greenberg, President and CEO of Two Harbors Investment Corporation. Bill previously served as a CIO at Two Harbors and joined the company in 2012. Welcome, both of you.
Thanks for having us, Ken.
Thank you.
Before we dive in, why don't we just give a few minutes for each of you to give a brief overview of your companies?
Sure, I'll start. Thank you, Ken, and thank you to RBC for having us here today. It's always a pleasure to do this conference, given the strong attendance and great guests that you have. I'm David Finkelstein, as Ken mentioned. I'm the CEO and Co-CIO of Annaly Capital Management. We're a mortgage REIT. We are the largest and one of the oldest of the mortgage REITs in the sector. We've been a public company since 1997. We have approximately $12.5 billion of capital, and that capital funds a roughly $80 billion balance sheet. That balance sheet is allocated across three sectors, all within housing finance.
The largest of our businesses is Agency MBS, which is approximately a $70 billion portfolio of largely pass-throughs, MBS agency pass-throughs that are levered and hedged, whereby we use swaps, futures, and swaptions to hedge the vast majority of the interest rate exposure. We run about a leverage level on the agency portfolio around seven, seven and a half times. The next largest business is our residential credit business, which some of you may know as Onslow Bay Financial. That business is largely a residential whole loan to securitization business. We do own third-party securities across resi finance, but the core of the business is acquiring whole loans through our correspondent channel, balance sheeting those loans, and then ultimately securitizing and holding the subordinate securities on balance sheet. Our third business is mortgage servicing rights. That has been a business that has been growing considerably for us.
We've been in the MSR sector for about nine years, but we brought it on. We used to own a servicer, and then we brought the business on balance sheet in 2021 and started acquiring MSR on balance sheet. That's about a $3.3 billion market value portfolio of about $200 billion in principal balance of agency MSR that is diversified and relatively low note rate. That's the portfolio. Together, these three portfolios create a diversified capital allocation that we think has considerable synergies across the businesses, operational as well as strategic synergies that gives us a lens in all aspects of housing finance, is not correlated with one another, and gives us relatively smooth book value and earnings and an earnings stream for the investor.
Okay, Bill.
Thanks. I'm Bill Greenberg. I'm the President and CEO of Two Harbors. We're a little bit smaller than Annaly. We have about $2.2 billion of total stockholder equity, a balance sheet of around, you know, $12 billion or $13 billion. We started in 2009. I came in 2012. We have two lines of business, not three, like Dave does. We have an agency mortgage-backed securities portfolio as well as an MSR portfolio. One of the differences between our companies is the relative size of our MSR. We have roughly more than 60% of our capital is allocated to the hedged MSR strategy, and the other 38% or something like that is allocated to a hedged RMBS strategy. We have often talked about our portfolios as being paired in certain ways.
The MSR is providing many interesting risk offsets to the agency MBS position, both in terms of interest rate risk as well as mortgage spread risk. Another one, one of the differentiators between us and Annaly is we service our own MSR in-house. We acquired a mortgage servicer in 2023 called RoundPoint Mortgage Servicing. That has given us some economies of scale and some incremental earnings that we have from that. Of course, having an operational entity itself has given us the opportunity to be able to do more things in the mortgage finance space. We just started a direct-to-consumer lending operation to provide recapture for our businesses. We are offering our borrowers second liens. Of course, it allows us to do other things in the future as well.
I think together, that, again, that provides us with something that two lines that are complementary to one another and provides a benefit to be able to, as I like to say, provide returns in our portfolio that are less volatile than portfolios without MSR.
Okay, great. Before we begin, let me just read a quick disclaimer. Any forward-looking statements made during today's event are subject to certain risks and uncertainties, which are outlined in the risk factors section in Annaly and Two Harbors' most recent SEC filings. Actual events and results may differ materially from these forward-looking statements. Both companies encourage you to read the forward-looking statements disclaimer in their quarterly and annual filings. Additionally, the comments made during this call may contain time-sensitive information that is accurate only as of today's date. The companies do not undertake or specifically disclaim any obligation to update or revise this information. The companies may discuss both GAAP and non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in each company's most recent filings.
Okay, with that out of the way, we're going to keep the discussion relatively interactive, and we will periodically open it up to the floor for any questions from the audience. In the meantime, I'm going to start off with a few questions of my own. Let's start off at a broad level here, and I'll direct this one to you, David, first. Both of you guys mentioned this briefly, but given your business models, mortgage REITs as levered to bond portfolios, the macro backdrop is always going to be an important consideration. I believe that prior to the most recent election, there was a thinking that we can get a little bit more clarity around rates, but there's been a lot going on. What's your latest outlook on the macro environment, rates, and Fed policy?
Look, it's changing by the minute. Let's back up a little bit and talk about where we were when the Fed first started cutting rates in September. Coming out of last summer, we started to have some deterioration in the labor market, and the Fed was obviously concerned about it. As a consequence, they cut 50 basis points in September. At that time, heading into that meeting, there were approximately 10 rate cuts priced into the market from September through 2025. A pretty meaningful easing cycle was factored in, and the Fed was aggressive on that first move. Fast forward a little bit, we saw some stability in the labor market. We saw economic growth in Q4 start to perform relatively well. We ended up growing at 2.5% on the quarter, 2.8% on the full year. Momentum started to rematerialize in the economy.
As we progress, we get to the election, all of a sudden it's a red sweep. You start to think about the fiscal uncertainty and spending and deficits. As a consequence, rates started to really materially increase. Term premium re-entered the market, and we did have a little bit of volatility in rates market. Simultaneously, inflation picked up again in the fourth quarter after pretty calm Q2 and Q3. As a consequence, by the time we got to the December meeting, the Fed, after having cut four times, basically went from the direction is clear in September to policy pause in December. Markets no longer priced six cuts for 2025 into the market. They priced effectively two cuts. The range of outcomes at that time period became very narrow. You know, there was concern about inflation.
Growth was strong, and the bar to hike was relatively high. The market effectively just settled on two cuts throughout 2025. Now, fast forward a little bit, the euphoria surrounding the election dissipated. We're starting to hear a lot of talk about tariffs and immigration. These protectionist and nationalist themes started to erode both consumer and business confidence. Spending in January really lacked on the part of the consumer, a very conservative posture. Consumer confidence and business confidence deteriorated, and manufacturing started to soften. We started to see what we think to be was a meaningful shift in the data. Where we sit today is that we've had a rally back across the curve effectively. You know, the 10-year note, it's bouncing around today, but it's in the context of 4.20% from a high of 4.80% just two months ago effectively.
The market has been shifting quite a bit, and the range of outcomes associated with the economy and markets has widened pretty considerably. You know, our view is that things are deteriorating. The Fed will begin cutting again, likely in June, and we will get three cuts, which is currently priced in the market this year, which generally speaking for agency investors is a good thing. You know, the carry associated with MBS is re-entering the equation, and it brings banks back in and demand for agency. That is a healthy thing. Volatility is typically not very healthy. That is something we're concerned about. What it necessitates is that you maintain a more conservative risk profile associated with both your duration rate exposure as well as your leverage.
Entering into the year, we had the lowest level of leverage that we had since 2014 at five and a half turns. We are very, very measured with respect to our rate exposure. We feel like that's the way to play the market in the current environment. You know, on one hand, it feels like things are deteriorating a little bit. On the other hand, you could see a scenario where trade issues get resolved at some point, not without a lot of damage being done. Nevertheless, things could normalize and we could get back on this growth track with, you know, tax cuts and investment and the like. It's very uncertain. We believe the approach is relatively cautious. Another area as it relates to the macro economy that impacts both of our businesses is the housing market.
Housing has been obviously very strong over the past number of years. Even last year with persistently high mortgage rates, HPA was still 2% - 3%. Consistent, kind of growing at a pace of personal income growth. However, there has been disparity regionally. For example, areas that ran up a lot in 2020 through 2023 started to come off a little bit, mainly in the Southeast and Texas, for example. There are pockets of uncertainty around the housing market. Generally, consumer balance sheets are reasonably strong, unemployment is still low, and we feel pretty good about housing. You do have to be a little bit more cautious as it relates to evaluating the housing market and protect credit a little bit better. Overall, I think we will be okay in terms of how the market will evolve.
You generally have to take a more conservative approach in the current environment. The good news is for both Bill and our business is that returns in the agency market as well as MSR for Bill and MSR and resi for us are still quite ample. You can earn really good returns without taking a lot of risk. That will persist for the time being, we think. We are perfectly comfortable taking a cautious approach because the balance sheet is doing a lot of work to get us the returns we need.
I might add a couple of thoughts to that, Dave. One, I think the history you gave of Fed expectations and so forth was a good one. I think you characterized the whole thing as just the last two years, three years has been interest rate volatility caused by markets changing expectations of what the Fed response function is and what it's going to do. I think while there have been periods of uncertain responses of what the Fed's going to do and more narrow ones, I think now we may be back to more uncertainty, not as wide as it was, but there's still a lot of uncertainty there. I think this administration has shown that always new ways to surprise us in terms of what's expected versus what's not expected.
When I think about interest rate volatility and how that impacts our businesses, I do not see the case for falling interest rate volatility really in 2025 very much. Dave mentioned that mortgage spreads could tighten if interest rates go down. That is likely to bring tighter mortgage spreads and more participants in. I agree with that. I do not think we need mortgage spreads to tighten in neither of our businesses. I think the spreads that are offered are perfectly fine. We are also operating at what I would call moderate leverage. I am hoping for spreads to not tighten, right? Because that just means lower forward earnings potentials as well.
Where mortgage spreads are today, they're certainly historically wide on a nominal basis, but on an option-adjusted basis, they're more in line with historical norms, which is really an indication that volatility is higher than it has been historically. As I said, I think that's fine. I don't think it's going to go down, but the spreads are wide and in support of both of our businesses. I think all of that seems just fine.
Great. Actually, why do we not just stay on the topic for spreads, the mortgage spreads? Maybe we just broaden out a little bit more in terms of what is the outlook there? More importantly, how do you think about supply-demand dynamics over the near term? Do you see a pickup in demand for MBS from other market participants such as banks, asset managers? What could this mean for spreads? I will address this to you.
I think Dave touched on it a little bit. Falling rates as a steepening yield curve improves the carry of MBS, which brings in other market participants, which is generally good for spreads. Prepayments, I think, are expected. As we're talking about, I think for the bulk of the agency universe, unless we're talking about a very large interest rate rally, which I do not think we are, prepayments should remain well-contained and well-supported. Of course, on the cuspy stuff, I think speeds will be fast. I think generally, if this plays out, you'll see more demand on the MBS side. It should be manageable. I think that will drive spreads marginally tighter over the near term.
Great. Why don't we just pivot a little bit towards mortgage servicing rights or MSRs for both of you? Each of your companies have different MSR strategies. Could you talk about the attractiveness of the MSR asset in the current rate environment and how are you positioned should rates decline? I'll start off with you, Dave.
Sure. First of all, I think both Bill and our portfolios are positioned quite well given how low the note rate is on each of our portfolios. Our note rate, average note rate on our portfolio is about 320. We are really over 300-350 basis points out of the money effectively. It takes a real significant move in the market to introduce prepayment risk with most MSR portfolios, just given the universe. The average note rate of the agency universe is around 4.25%. We feel really good about the portfolio, and we feel good about the sector from that standpoint. Now, the vast majority of our portfolio was constructed by providing liquidity to the non-bank mortgage community as they have sold MSR that they have had on balance sheet.
Largely, that was created in 2020 and 2021 when there was a massive refinance boom. Rates sold off beginning in 2022. The originator community business was slowing down. They wanted liquidity on their balance sheets, and they wanted to monetize their MSR because it had appreciated with higher rates. We were there as a capital provider to provide that liquidity. We grew the portfolio much quicker than we anticipated in the years 2022, 2023, and 2024. It slowed down a little bit in 2024. We constructed our portfolio based largely on the 2020 and 2021 origination, which is the majority of the universe out there. It is a very good convexity profile. Delinquencies are very low, so it is very inexpensive to service. We really like the asset class.
Now, a lot of that MSR that was created during that time period has gone from operating platforms, the large originators and medium and small, to folks like us and Bill and other participants and banks. It is diminishing in terms of the availability of MSR. Last year, there was about $750 billion in principal balance that changed hands. This year, we expect about half of that. To the extent that spreads remain attractive, we will certainly grow the portfolio. It is about 20% of our capital allocation currently, and we do not really lever it to any great degree. We would like to get our portfolio to 20% of capital with about a turn of leverage. We effectively could nearly double the size of our holdings of MSR.
We have to be very, very discerning and make sure that we're buying the asset at a spread that is complementary to the rest of the portfolio from the standpoint of generating strong earnings for the company. That is how we feel about it. We think that the market is shifting a little bit from what was a bulk market to more of a flow market. We are transitioning a lot of our activities to take a lot more flow out of the market. We do not have the advantage of an originator like Bill, but we have a lot of connectivity to the origination community that enables us to be there on a flow basis and buy MSR as it is created.
We were, I think, in 2019, 2020, 2021, I think 2022, I think we were the largest flow participants, flow buyer of MSR in the market during those years. As rates rose, that channel dried up, of course, and we participate in the bulk market in general. Just to expand a little bit on what Dave said, there is $750 billion or so of MSR UPB that changed hands last year. There is about $700 billion the year before and about $600 billion the year before that. Those were record years, the most we have ever seen. Prior to that, if you go back another 10 years, the average annual volume was $200 billion a year. The average quarterly volume was about $50 billion a year. I think, Dave, you might be a little bit more optimistic than I am about supply for this year.
I'm probably going to go right back to the long-term mean of $200 billion or $250 billion. I still think that's an ample opportunity for both of us to enjoy some of the things that we see there. One of the things that, because we've been in the MSR business a little bit longer, we didn't necessarily choose this out-of-the-money portfolio. We had been buying servicing all along when rates were rising. Yes, we did buy some low-back stuff as rates did rise, but we also had some at-the-money stuff, which became very deep out of the money at the time. Our average WACC is about 3.50, so a little bit higher than Dave's, but still miles away from the refinance window. I will sometimes call this product 300 basis points out of the money.
We've not seen this before in the world with mortgages this far away from the refinance window. I'll sometimes call this a new asset class. It's not that it's so attractive from the nominal yields that are available because it's not a 20% return or something like that. On a risk-adjusted basis, I think it is unlike anything we've seen before in the structured finance markets. I think it's super attractive. The thing that makes it attractive in particular is the prepayment speeds, the turnover speeds that have been realized over the last year, two years, whatever, have been a lot slower than conventional wisdom would have said. I think the textbooks would tell you that you should expect speeds from turnover, from death, divorce, moving, so forth, should be 6% per year. We're coming in in the 3% - 4% range.
There is a lot of cash flow being thrown off by these assets. It is not very convex given how far out of the money it is. If rates fall 50 basis points, 100 basis points, our portfolios will still be 200 basis points out of the money. In the old days, that would be about as deep out of the money as you could hope to get. We are still 100 basis points away from that. It is easy to hedge. It is easy to extract the value. They are very high-quality loans. Lock-in seems to be still very strong. I think this is going to persist for a long time.
Just to add, in terms of the history, the non-bank mortgage community, the transition to the vast majority of origination going from banks to non-banks has introduced a need for participants like us, capital partners that can provide liquidity and have the balance sheet to take the MSR asset. Now, originators, they're operating companies. They're great at that. We're capital participants. We're really good at providing liquidity from that standpoint. When you consider the profitability, the origination model, the net profit is inside of 100 basis points. To really be cash flow positive.
Which is bound in the MSRs. I think you're about to say that, right?
Yeah. Sell the MSR. Effectively, you got to sell the MSR. That is what our role is in the market. While bulk volumes will slow down, there is continuously going to be a need for folks like us to be liquidity providers.
Great. That is a great segue to my next question. I will address this one to you, Bill. Stay on the topic of MSRs. This is briefly touched upon too. Both of your firms periodically evaluate bulk MSR packages. What is your outlook around the potential supply of MSR bulk packages for this year?
200-250.
A little 300-400 is what we'd say. Yeah, which is $5 billion in market value.
Yeah.
Based on maybe a little bit more depending on pricing. There is enough. We will be okay.
Just current coupon with a little WACC.
That's great.
Yeah.
We'll briefly pause here and see if there are any questions out from the audience before we go along. Right there.
Ask around. Bill, how do you think about for MSRs, how do you define?
Let me just repeat the question before you answer it just for the. The question was about MSRs and how you think about hedging it.
I'll start. First of all, one thing which I think is maybe underappreciated when hedging MSR is that the primary driver of MSR prices and values is the mortgage rate. The mortgage rates for people involved in securities markets is also sometimes called the current coupon mortgage. Those will be the price of the mortgage-backed securities whose price is par. Today, that's Fannie 5s or 5.5s or whatever it is. Even though we have a portfolio which has a gross WACC of 3.5%, which sounds like mortgages which are collateral for Fannie 2.5s, the theoretical hedge for the MSR is not Fannie 2.5s. It's Fannie 5.5s because that's the thing that changes the price, changes the mortgage rate, changes the prepay sensitivity of the asset.
Now, to your real question, you might say that because we're 300 basis points out of the money, we shouldn't have to hedge if rates fall 50 basis points. How is the sensitivity of our portfolio going to change? How are prepayments actually going to change? There are two answers to that. One is, while you wouldn't think there's very large sensitivity to that, you would think, maybe I don't need to hedge it at all. It's a continuum, and one day you are going to have to hedge it. I always find it better to, let's stick it in the models. Let's have it be a continuous function so that we can accumulate the hedge as rates continue to fall so that we can have the right hedge when rates fall 100 basis points or 200 basis points.
While the theoretical hedge is low for 300 basis points out of the money, it's not zero. The other is what I think is even sort of more interesting component of thing is that if you look at the prepayment speeds of Fannie 2.5s and Fannie 3s and Fannie 3.5s, there's a prepay curve there. Fannie 2.5s are, I don't know what the numbers are, Dave, three CPR. And 3s are 3.5, and 4s are four CPR. Even there, even though they're deep out of the money and prepay and refinances are not kicking in or becoming meaningful, the turnover part of the thing is increasing even for small rate moves for deep out of the money portfolios. Those speed changes impact the value, the prices of the MSR, which should be hedged as well.
I'll add a couple of points on that. I agree with everything Bill said. As the market has sold off and we own this deep out of the money MSR, we initially think of MSR as a duration hedge for agency MBS. However, the current portfolios that Bill and I both own have two really important hedges in addition to a little bit of a duration hedge. Number one is the fact that it's a turnover hedge for our low- coupon MBS that we hold on balance sheet. For example, if we own UMBS 3s in the portfolio that have a dollar price of 80 or whatever, our concern is that those do not accrete to par. They do not pay down at a fast enough pace, and they are stuck down there, and they are a longer bond, and their yield is ultimately lower.
The MSR actually hedges some of that. If turnover stays very low on those, at least we're getting good carry on the MSR as well to offset that. Another hedge that is really critical in the current environment is that the MSR portfolio entails holding balances of the borrowers' taxes and insurance payments. We take those balances and we invest it in short-term instruments, SOFR effectively. Banks will pay us SOFR plus something to hold those balances. That happens to be a hedge for short-term interest rates. If the Fed doesn't cut, short rates stay higher for longer, we're getting that float income or that higher float income. These are a couple of other benefits from a hedging standpoint that MSR entails that are very difficult, or at least turnover is an incredibly difficult component of a mortgage to hedge.
That's really one of the only hedges you can think of for a discount bond. You can hedge short rates, obviously, but this is a very eloquent way to do it that gives you positive carry.
I always think it's interesting the floating rate components are not unlike when you pay fixed on swap on hedging in the MBS portfolio in that you receive the floating rate and hedges the short-term rate from that perspective as well. Along those lines, a pure IO that has a 3.5% gross WACC when it's 300 basis points out of the money, you might think even has positive duration, that the speeds won't go down very much, notwithstanding what I was saying. The floating rate components, a floating rate IO, in fact, when rates rise, even if the speeds don't slow, which they do, that also goes up. The floating rate components can be thought of, I like to call it a super floater IO. That provides a lot of extra duration to the MSR asset in and of itself.
Okay. Great. Any other questions from the audience before we proceed? Back there.
All that in general, what are the chief components? What's the dream potential as a result of the GSE from conservatorship? What does this really mean?
I'll let you start with that one.
I'll repeat the question. Question was about GSEs and potential emergence from conservatorship implications for housing finance.
First of all, let's.
How much time do you have?
Yeah, let's talk about the issue more broadly. Obviously, there's been a lot of chatter in the market about the potential exit from conservatorship of the GSEs, which stands to reason given in the last Trump administration towards the end, that was an objective that the administration wanted to achieve, which they obviously didn't. It stands to reason that now we're talking about it again. It is important to note that a lot of the chatter is coming from the common equity holders and junior preferred holders who really are the only ones who benefit from the release of the GSEs from conservatorship. Now, thinking about the likelihood of it, our view is that over the near term, next couple of years, we think it's relatively low in terms of the likelihood for a few reasons.
Number one is that the administration, Treasury Secretary, others, and including the nominee for FHFA director have said that it's not a priority, that there is concern over the mortgage rate, which has to be entered into the equation, and that suggests you don't want volatility associated with MBS pricing. Number three, and this has been stated by not just the Treasury Secretary, but also members of Congress on the Republican side have stated that taxpayer has to get made whole for the GSEs and for the conservatorship. From that standpoint, it seems like the bar is relatively high for what we would characterize to be a hasty release of the GSEs. The second point I'll note is that the math associated with releasing the GSEs is incredibly daunting.
The capital rule that the PSPAs or preferred stock purchase agreements prescribe the GSEs to have before they can be released is about $330 billion that the two, Fannie and Freddie, need to have to be in a position to be durable outside of conservatorship. They're half of that right now. They need to continue to accumulate capital through retained earnings to get to a point where from a safety and soundness perspective, they could be released. That threshold could be reduced.
If you cared about those things.
If you care about those things. Given the administration's perceived tolerance for disruption, they may not. However, I choose to think that politically, the bar is high. Another point to note as it relates to the math is that the GSEs, in addition to getting to that capital level, they owe Treasury $334 billion. Now, that's the amount that was used to bail them out, $190-odd billion, and then another $140-odd billion that the GSEs began to retain from earnings that should have gone to Treasury when they effectively turned off the sweep in 2018 or thereabouts. I do not think politically it is palatable for the administration and Treasury to just forgive that $330 billion. That is $1,000 for every man, woman, and child in the country. There could be some political costs associated with that.
The third point I'll note is that this release of the GSEs is predicated on raising capital in public markets. Now, based on the current capital requirements, if you look at the ROEs of the GSEs, it's mid to upper single digits. Now, if you go to market trying to raise capital with those ROEs, those are like utility-type returns. These are monoline cyclical businesses where investors are going to want to get paid. I do not think anybody in this room is going to jump toward an IPO with those types of returns.
On the largest IPO in history, that would be required.
Exactly. Exactly. Now, like Bill alluded to, you can't underestimate the tolerance for disruption. As a consequence, we have to be very prepared for that eventuality. However, some of the protections in place in the law when they were brought into conservatorship protect legacy bondholders. One of those factors is that bondholders cannot be made worse off through the exit of conservatorship, which means that there is a $254 billion lifeline at Treasury which would go to protect legacy MBS bondholders. If that is there and the government ultimately has to provide support and they get to a level where they have to make a decision, you are more likely to continue to support the GSEs to protect your initial investment. We feel like, from the standpoint of our portfolio, we are in pretty good shape if there was even a hasty release of the GSEs.
Another silver lining in all of this is that at a minimum, we expect that the footprint of the GSEs will be reduced with the new administration, much like it was in the last Trump administration. Just to give context to that, roughly 20% of the loans that are guaranteed by the GSEs are what are called non-core. We are talking about investor property, second homes, higher loan balance loans, and cash-out refis and things like that. To the extent that they put caps on those loans like they did in the first Trump administration or de-emphasize them, price them out of the market, you have a couple of things going on. Number one, it leads to less agency supply, which is good for our portfolios from a technical standpoint. Number two, for our residential credit business, it enables us to compete for those loans.
We actually think that the adjustments to the GSEs that will be made over the near to intermediate term are actually quite beneficial both for our core agency business as well as for our residential credit business. Hopefully that answers your question. We're eyes wide open. Anything can happen, but we feel like we're taking a measured approach and we'll be okay.
I might add just a few more words to that. I agree with your assessment, Dave, about the short-term prospects. In fact, there was, I thought, a very excellent report from one of the primary dealers about a survey that said 11% of market participants thought that you could have privatization in the next two years and 35% over the next two and then 55% in the next administration or never. I thought that was good. Dave highlighted well, I thought, some of the numbers behind the financial numbers of what's required in order to do that. The other risks just associated to the overall housing finance system, I think, are significant. We all thought this was really hard to do four years ago. We thought it was really hard to do 10 years ago, and I think that's still true.
If privatization comes with any changes in what do you call it? Risk ratios. What are they called?
Yeah, bank capital.
Bank capital rules or LCR rules or something like that. The banks own $1.8 trillion, I want to say, of agency MBS. There is $1.3 trillion held overseas. If you change something in the privatization that breaks all of that, that could have a very detrimental effect on the overall financial system, really. That same survey that I referred to said that up to 75% of banks and insurance companies are not sure whether they will need to sell MBS, whether they will be allowed to based on their investment guidelines and rules if the GSE debt is downgraded or if they change the rules in some way. Again, the impact of that is unknowable. I read something again. That same report said that it is unclear whether the Fed could buy MBS if they are no longer an agency of the government, however that has been defined and so forth.
That would tie the hands of the Fed for the next downturn in terms of being able to buy MBS and provide support.
I can assure you they would not buy MBS through QE if they were not in conservatorship or there was not a guarantee.
Right.
So.
They could hold the existing portfolio, but they would not.
Besides all the things that you said about the protections in place and the largest IPO in history and all of that, I think the system is just so interconnected and there's so much risk. I know no one wants to hear this necessarily, but it's mostly working. The mortgage market is working reasonably well. There's lots of unintended consequences and downsides to bringing this thing out. I think there is an element, I think that's understood by the officials who are thinking about this, that if you break it, you own it. It's unclear what the real upside is. I think the likelihood is low in the near term, but all you got to do is read the newspaper to know that anything can happen.
Yeah. One more point. If the administration is looking for a pay-for for tax reform, for example, sweeping the capital from the GSEs, which they could do of $150 billion, and then taking the retained earnings on an ongoing basis over the next decade, that's $400 billion potentially to offset your tax bill.
Okay. That's great. Any other questions before we proceed from the audience? Okay. Let's proceed then. Next one is for you, Dave. For Annaly, and you briefly alluded to this in your prepared remarks, the company derives some benefit from the Onslow Bay platform, specifically the ability to manufacture proprietary assets. And this kind of aspect may be underappreciated by investors. Can you talk more about the platform and what's your outlook on the potential contribution?
Sure. Sure. Just by way of background, our residential credit business, Onslow Bay Financial, has been involved in the non-QM loan sector since 2016. We were one of the original purchasers or acquirers of non-QM loans after the rule was put in place. We developed a bit of a first-mover advantage in the sector. Now, backing up, during that time period, we were actually members of the Federal Home Loan Bank system, and we were able to finance those assets at quite inexpensive levels. It was a very, very accretive trade for us at the time. However, we did start to begin a securitization effort in 2018 as the business started to grow. We were one of the first securitizers of non-QM loans and became quite significant in the sector as the residential credit securitization market redeveloped post-financial crisis.
In 2021, we launched a correspondent channel. Effectively, what that entails is we face currently about 260 different mortgage originators throughout the country that we send daily pricing to. They take that pricing and they lock loans with their loan officers throughout the day. We guarantee them certainty of execution effectively. That business has really grown over the past couple of years as the non-QM market has taken off. Last year, we closed about $13 billion in whole loans on the year, $12 billion roughly of which originated or came through our correspondent channel. What we have done over the past number of years is we have provided white-glove service to these small originators. We have provided them liquidity at times of market volatility such that they rely on us.
As a consequence, they pay a bit of a premium to be able to do business with us. There's an attachment that enables us to extract what we believe to be better returns than other channels.