Two Harbors Investment Corp. Q4 FY2021 Earnings Call

Good morning. My name is Peter, and I will be your conference facilitator. At this time, I would like to welcome everyone to Two Harbors Fourth Quarter 2021 Financial Results Conference Call. As a reminder, this call is being recorded. I would now like to turn the call over to Paulenier Sims.

Good morning, everyone, and welcome to our call to discuss Two Harbors Fourth Quarter 2021 Financial Results. With me on the call this morning are Bill Greenberg, our President, Chief Executive Officer and Chief Investment Officer; and Mary Riskey, our Chief Financial Officer. The earnings press release and presentation associated with today's call have been filed with the SEC and are available on the SEC's website as well as the Investor Relations page of our website at twoharborsinvestment.com. In our earnings release and presentation, we have provided a reconciliation of GAAP to non-GAAP financial measures, and we urge you to review this information in conjunction with today's call. As a reminder, our comments today will include forward-looking statements, which are subject to risks and uncertainties that may cause our results to differ materially from expectations. These are described on Page 2 of the presentation and in our Form 10-K and subsequent reports filed with the SEC. Except as may be required by law, Two Harbors does not update forward-looking statements and disclaims any obligation to do so. I will now turn the call over to Bill.

Thank you, Paulenier. Good morning, everyone, and welcome to our fourth quarter earnings call. Please turn to Slide 3. At quarter end, book value was $5.87 per share, representing a negative 5.6% total economic quarterly return. This underperformance can be attributed in approximately two equal parts. Firstly, to spread widening and higher coupon RMBS and IO securities due to slower-than-expected declines in prepayment speeds and higher volatility. And secondly, to a 20 basis-point widening in the spread of the MSR asset, which we see as being within the typical period-over-period variability of MSR spreads and which arise from using the simple average of three independent broker marks. The overall environment in 2021 was challenging, as the market wrestled with continued waves of coronavirus infections and uncertainty around the path of economic growth and monetary policy. For the year, we delivered total dividends of $0.68 per share, equivalent to an average dividend yield of 10.0%. With the starting book value of $7.63 at the end of 2020, the total economic return on book value equates to minus 14.2%. In many ways, the underperformance in 2021 was a continuation of the market stress of 2020. While the Fed came to the rescue in 2020, buying unlimited amounts of RMBS and stabilizing the market, the Fed's continuing monetary accommodation also had two other effects. First, with the Fed having bought almost $3 trillion of RMBS during 2020, spreads collapsed to all-time tight levels, which reduced the tailwind of carry meaningfully. Second, the Fed's asset purchases also reduced mortgage rates to all-time lows, igniting a prepay wave that hasn't been seen since 2003, and creating a very strong headwind for portfolios such as ours with premium and interest-only cash flows. Nevertheless, we took proactive actions during the year, which benefited the company. We issued and repurchased convertible notes last February, redeemed our Series D and E preferred shares, and we raised common stock through two public offerings. On the asset side, we grew our MSR portfolio, acquiring over $88 billion of UPB at very attractive levels. And in expectation of spread widening, we reduced leverage, increasing the amount of dry powder to be deployed in a more attractive investing environment. Indeed, over the past few weeks, it has become clear that the Fed has shifted to a much more hawkish stance and mortgage spreads have reacted by widening about 20 basis points during the month of January. Despite the rapid and large spread movements, we think there is more to go and expect more volatility in the near term as the timing and pace of rate hikes, balance sheet normalization, and other supply-demand dynamics are sorted out. It is in precisely these kinds of environments that our Agency plus MSR construction should benefit. And indeed, our best estimate of book value through the end of January is approximately positive 2%. Looking ahead, we see a path towards normalization in RMBS spread levels. The refi wave appears to be over and the prospect of higher rates and slower prepayments points to a more constructive environment for both RMBS and MSR. We are well positioned to capitalize on the improved investment opportunities as they present themselves in the coming year. Please turn to Slide 4. During the fourth quarter, economic data, Federal Reserve commentary, and market expectations converged on an accelerated timeline to higher rates and the removal of monetary accommodation. As we see in Figure 1, the 10-year rate fluctuated between 1.4% and 1.6% during the quarter before breaking out in January to the 1.9% range where it sat pre-pandemic. The prospect of more rate hikes pushed short-term rates higher, flattening the yield curve by 53 basis points during the quarter. The market has continued to reprice and expectations have become more front-loaded with almost 5.5 hikes now expected in 2022, with a total of 7 by the end of 2023. The Federal Reserve has announced it will complete its monthly purchases of treasury and Agency RMBS by March, as seen in Figure 2. More recently, while the Fed has said that interest rates will be their primary tool in implementing monetary policy, they do recognize that the balance sheet is bigger than it needs to be and have expressed a desire for the balance sheet to return to a treasury-only portfolio. Chairman Powell has also said that the pace of balance sheet normalization could be faster than it had been in past episodes. The timing and pace of any runoff or outright sales should also affect mortgage spreads. We have consistently said that mortgages at historically tight spreads would need to widen eventually. In our view, the question was not if they would widen, but how quickly and by how much. With the Fed removing its accommodation and banks likely having less appetite for securities, as a function of lower reserves and increased loan demand, some mortgage analysts are projecting that the private sector may need to absorb up to $700 billion of mortgages in 2022. Current coupon spreads remained near historical tights throughout the fourth quarter, although, as I mentioned, have swiftly widened out by roughly 20 basis points since year-end, as shown in Figure 3. However, spreads still remain approximately 15 basis points tighter than other periods where the Fed had not been buying mortgages, let alone periods where the Fed is net shrinking. Indeed, relative to the last period of quantitative tightening in 2017 through 2019, spreads were at current levels or tighter only 8% of the time; 92% of the time they were wider than they are now. In the shaded area in Figure 3, we show how wide mortgage spreads might be able to go. We don't expect spreads to widen and stay at levels much beyond historical averages, but it's instructive to note that during the last period of quantitative tightening, mortgage spreads overshot the long-term average by an additional 20 basis points. Please turn to Slide 5. Since the beginning of the pandemic, the average coupon of the mortgage universe has declined by approximately 75 basis points from 4.25% to 3.5%. As a result of this restriking of the mortgage universe, and the upward trajectory in mortgage rates, the percentage of mortgages in the Fannie, Freddie universe with at least 50 basis points of refinance incentive has fallen to around 30% at year-end 2021, compared with roughly 80% at year-end 2020. With the jump in mortgage rates in 2022 to north of 3.5%, that percentage is projected to fall further to roughly 11%. With such a small percentage of the market eligible to refinance, we are looking at an environment of sharply lower prepayment speeds. How much slower is shown in Figure 2. The dark gray bars indicate the realized prepayment speeds for our current specified pool coupons over the course of 2021. Yes, these speeds were much faster than what was expected at the beginning of the year as a result of high home price appreciation and GSE policies to make refinancing easier. But higher rates will necessarily lead to slower speeds and we show our projections for those TBA and specified coupons in the light gray bars, assuming current interest rates stay unchanged throughout the year. With prepayment rates projected to come down sharply and quickly, the economic carry and cash flows of our paired MSR strategy are expected to dramatically improve. In Figure 3, we show the components of economic carry of our MSR portfolio for 2021 and the projected amounts for our current portfolio over the next year, assuming rates are unchanged, and we do not replace or add any MSR. In 2021, the value loss from prepayments, labeled runoff in the chart, overwhelmed the MSR cash flows despite contributions from our recapture agreements, resulting in total net economic carry of close to negative $200 million. Our static projections for 2022 changed this picture dramatically. Runoff is forecast to drop by over 50%, and we project the net economic carry of our MSR to be close to positive $200 million, nearly the inverse of last year. Of course, these projections are model-based and reflect a single point in time with a number of assumptions. But the idea is that the fast prepay headwind that we experienced in 2021 has shifted and our base-case expectation is for a much more favorable environment in 2022. Now I will turn it over to Mary to discuss our financial results in more detail.

Thank you, Bill, and good morning, everyone. Please turn to Slide 6 to review our financial results for the fourth quarter. Comprehensive income was negative $128.6 million, representing an annualized return on average common equity of negative 24.7%. Our book value was $5.87 per share compared to $6.40 at September 30, including the $0.17 common dividend, resulted in a quarterly economic return of negative 5.6%. As Bill mentioned earlier, the result mostly reflects spread widening in high coupon RMBS, IO securities and MSR. Moving on to Slide 7. Earnings available for distribution was $0.22 per share compared to $0.24 for the third quarter and compared to our fourth quarter dividend of $0.17. Overall, the benefit of our investment in MSR served to partially offset reduced RMBS balances. Net interest income declined by $1.3 million with lower RMBS balances translating into lower interest income and repo funding costs. Interest expense was further reduced by a decline in MSR borrowing costs. Gain and other derivatives was $39.5 million compared to $46.3 million in the prior quarter. The result was driven by a reduction in TBA exposure, which was largely offset by roll specialness, and a lower notional position in U.S. Treasury futures. Turning to MSR. Servicing income rose by more than $4 million to almost $61 million as average UPB balances continued to grow and prepayment speeds slowed. As Bill noted earlier, we expect to settle on $39 billion UPB of MSR in the first quarter and expect to see the economics of that growth come through over the next couple of quarters. Lastly, operating expenses fell by $3 million to $31.4 million due to compensation accrual adjustments, offset by the rescission of the CARES Act employee retention credit that we benefited from in the first three quarters of the year. In the table on the lower right, we show our portfolio yield. Our realized net spread in the quarter rose by 44 basis points to 2.99%. This net increase was attributable to the combination of higher yields on recently purchased MSR, and a greater proportion of MSR in the portfolio. Net spread as of December 31, which reflects our estimate for the near term, increased further to 3.1%, but will depend on our portfolio mix and the pacing impact of interest rate hikes. Please turn to Slide 8. Our liquidity position continues to be strong with unrestricted cash of $1.2 billion at quarter end. Funding in the repo market remains liquid and well supported, and we retained balances with 19 different repo counterparties. While funding costs for Agency RMBS through the repo markets rose late in the quarter on an absolute basis, as the market priced in rising short-term rates and Fed hikes, the spread to OIS and SOFR remains low, currently right around 10 basis points for both three-month and six-month maturities, as shown in the chart in the upper right. We maintained access to diverse funding sources for MSR and our unfunded and committed MSR asset financing capacity stands at $313 million. Turning to leverage. In conjunction with the reduction in our RMBS position, economic debt-to-equity declined to 4.7x at quarter end compared to 6.1x at the end of September. Average economic debt to equity in the fourth quarter was 5.1x, compared to the third quarter average of 6.0x. As a final note, we have included information on REIT taxable income and the tax characterization of our dividend distribution on Appendix Slide 28. For additional information regarding the distributions and the tax treatment, please reference the dividend information found in the Investor Relations section of our website. I will now turn the call back to Bill for our portfolio update.

Thank you, Mary. Please turn to Slide 9. As spreads fluctuated during the fourth quarter at tight levels, we became increasingly concerned about the potential for a more aggressive Fed, and in early December, we reduced our overall mortgage exposure to be more defensive, reducing our portfolio leverage to 4.7x. As we did this, we also utilized option structures to replicate the convexity and volatility characteristics of RMBS, which preserved virtually all of the carry, but without the spread risk. Within our specified pool position, we rotated out of our seasoned 3% pools into new production, investor property, loan balance and geography stories. We also added some very low pay-up 2.5% bank-serviced collateral, which we think will outperform TBA as the deliverable worsens. In high coupons, we were not active, and we let that position continue to run off without replacements. Please turn to Slide 10. Figure 1 shows our story and coupon distribution of specified pools and includes the bank-serviced and new production investor stories that I just mentioned. Figure 2 shows the quarterly total return performance by coupon on TBA contracts in our specified pool holdings. Here, we can see that every coupon with the exception of TBA underperformed their interest rate hedges along the curve and higher coupons suffered more. Finally, Figure 3 compares by coupon TBA speeds to those of our specified portfolio. Overall, prepayment speeds in our specified pools declined 8% to 27.7% CPR and remain much slower than TBA. To the extent that this quarter's underperformance and higher coupons was due to the market being disappointed by the pace of deceleration in prepayment speeds, we are optimistic that that may be short-lived, as prepayments should begin to slow materially in the coming months. Indeed, the February prepayment report released last Friday showed speeds declining by 21% in aggregate and the market is expecting an additional 15% decline next month as well. Please turn to Slide 11, where we'll discuss our MSR portfolio. Our flow and bulk additions almost exactly offset our runoff and market pricing was indifferent to the large curve flattening experienced during the quarter as the long end of the yield curve was essentially unchanged. The percentage of our MSR portfolio which is in forbearance has declined to below 1% by loan count. More than 80% of our borrowers are exiting forbearance in a contractually current state through payment deferral or loan modification. About half of the remaining 20% have prepaid, while the other half continue to work through loss mitigation solutions, including modification. Because forbearances are now very small relative to our portfolio, we will not continue to report it going forward. Our 60-plus day delinquencies are still elevated from pre-pandemic levels by about 100 basis points and sitting now at 1.4%, and we will continue to monitor this closely. In Figure 2, we show the trend of our settled UPB for our flow and bulk channels over the last five quarters. While flow channel settlements have continued to decline, we were able to capitalize on the elevated activity in the bulk market and committed to purchase another $22 billion of UPB. In total, we expect to settle on $39 billion of UPB in the first quarter. In Figure 3, we compare our servicing prepayment speeds versus TBA. Overall, prepay speeds at our MSR portfolio declined considerably by 17% to 22.1% CPR. This is a much greater slowdown than we experienced in our specified pool position and is a result of having more exposure in the lower coupons, which are more reactive to changes in interest rates. Please turn to Slide 12. One of the main differentiating features of Two Harbors is the extensive and robust platform we have created to acquire MSR on a flow basis from multiple counterparties. Two Harbors has been a consistent presence in the flow market through many different environments, and this is shown in Figure 1. The same 10 buyers have accounted for 83% of flow market share in the last five years. During those years, Two Harbors has had an average market share of 25%, and we have been the top buyer of flow servicing in four of those years and #2 in the fifth. During 2020 and 2021, our flow acquisition volumes essentially offset our portfolio runoff as seen in Figure 2. While we are not an originator or a correspondent lender, our flow channel platform acts similarly in many of its effects, one such effect being the ability to source more product in a lower rates and high refi environment. Lastly, in Figure 3, we show the volume of transactions in the bulk MSR market as well as our participation in that market. Since 2014, the amount of MSR coming to market in bulk transactions has been between $150 billion and $250 billion of UPB and a typical success ratio for us during those years ranges from 5% to 10%. For reasons we have discussed previously, a record high amount of roughly $450 billion of UPB came to market in 2021. And 2022 is shaping up to be another record year, as we have seen roughly $125 billion of UPB come to market in January alone. This slide is meant to give a better sense of the supply dynamics in the MSR market, and in our ability to source the MSR that is a core part of our investment strategy. We are very proud of the platform we have built and the reliability to our portfolio balances that it provides. Please turn to Slide 13. As we have discussed previously, the MSR asset acts in many ways like a short position in RMBS, as shown by the gray bars in the top left chart. The blue bars represent RMBS hedges and the amount is chosen to be equal and offsetting by construction. After allocating the RMBS positions necessary to hedge the MSR, we are left with additional RMBS that we hedge with rates as seen in the middle chart. As you can see, we were net short close to $2.5 billion in the 2.5 coupon and had small positive positions in the higher coupons. The short 2.5 position resulted mainly from delevering the portfolio as we believe the 2.5 coupon looks rich, and had the most to lose given the reduction of Fed support. In the second row, we show book value exposure to changes in current coupon mortgage spread. The chart on the bottom right shows a total of 0.8% exposure for a 25 basis-point spread widening in the current coupon mortgage spread. This estimate assumes a parallel shock across the coupon stack and is not relevant for the book value sensitivity to spreads in higher coupons moving in isolation or on MSR. Besides the presence of MSR in our portfolio, this number also reflects the fact that we have very low leverage. Please turn to Slide 14, where we show our interest rate and curve exposures. We are carefully monitoring and hedging the shape of the entire yield curve in general, and we are focused, of course, on the very front end of the curve in particular. With inflation fears in the foreground, the obvious concern is a shock upward in the front end in a continued bear flattening move. In Figure 3, we show that in this scenario, our portfolio has low sensitivity, with only a minus 0.3% impact to book value. Taken together, these three charts demonstrate our commitment to hedging the full yield curve exposure and our desire to have little book value volatility, no matter the rate environment. Finally, I'd like to discuss our outlook for Two Harbors and our return expectations for new investments on Slide 15. The most notable change from the third quarter is the sharp move in current coupon pools hedged with interest rates. The rapid widening of spreads earlier this year has pushed returns from the low to mid-single digits to the very high single digits. These assets are getting close to our target returns, and we believe that they will be an attractive use of capital early this year. In high coupon pools, although we saw some widening in the fourth quarter, they have outperformed somewhat in January. And so combined with a much flatter yield curve, we see static spreads roughly unchanged in the high single digits. TBA roll specialness remains elevated and thus the duration hedges carry, which is not particularly impacted by the recent spread widening, continues to look attractive on the surface. However, despite the specialness, we believe the supply/demand imbalances will weigh on actual realized returns in the near term, and so we are taking a wait-and-see approach to TBA investments. The paired construction of MSR and Agency RMBS continues to be the most attractive way to deploy new capital with returns in the low to mid-teens. Although the MSR market is competitive, we continue to find pockets of opportunity through our market-leading position in flow and our disciplined pricing of bulk packages. This environment with prepayment speeds beginning to slow, and current coupon mortgage spreads widening, is one for which our portfolio strategy was designed. As a result, we are very constructive and optimistic about the forward outlook for Two Harbors and our paired agency plus MSR portfolio construction. Thank you very much for joining us today, and we'll now be happy to take any questions you might have.

Our first question is from Trevor Cranston with JMP Securities.

You talked a lot about the widening of current coupons in January and performance somewhat of higher coupons, which kind of reversed what happened in the fourth quarter, I guess. Can you talk about going forward as taper progresses and the Fed embarks on monetary tightening, how you expect the performance of the coupon stack to evolve over the course of the year? And if you could also provide any commentary on sort of what risk scenarios you guys think about that would potentially cause high coupon underperformance going forward?

Yes, sure. Thanks, Trevor. As we've sort of been saying for a while and shown by the chart on Slide 4 of the presentation, we think spreads on current coupons should continue to widen towards long-term averages. Given the environment we're in, with the Fed removing accommodation and discussing balance sheet tightening, the possibility of an overshoot of long-term averages in spreads is possible. We are positioned accordingly for that. As rates continue to rise, what is known as the current coupon migrates up as well. So looking at the current coupon chart like the one we show is talking about not a single coupon in the coupon stack; it changes over time. Currently, right now, the high coupons called 3.5 and above haven't been the current coupon for a long time. As rates rise and the current coupon shifts up, you should start to see more supply come in those coupons as well. But refinancing speeds across the entire stack, not just the current coupons, will start to fall as the entire refinance universe powers down. I'm not sure if that answered your question.

Yes. Yes, it does. That's helpful. And then when you were talking about the attribution to the book value performance in 4Q, I think you mentioned about half was underperformance of high coupons and then the other component was widening of MSR. I think there was a comment that it was partially related to just the pricing variance based on how MSR marks are provided. Could you clarify that a little bit or just elaborate on if the MSR widening was just sort of that pricing variance or if there was anything fundamental related to expected prepay speeds or anything else going on?

Well, I'll start with that, and maybe Mary can jump in also. Our fair value marking process is to take the average of three independent broker marks, and they have their own information and estimates of our portfolio, and we think the best way to manage that is to take the simple average of those. Within that process, we saw a 20 basis-point widening of that average level.

Got it. Okay. And then the last question for me. I think when you were talking about the refinance ability of the market, I think you gave a number something like 11% after the move in rates in January. Do you have the number offhand for how much of your MSR portfolio remains in the money for refinancing or if that's generally kind of close to what the market level is overall?

Yes. I think that's one of the charts, actually. You see it tracks pretty closely on Page 5 of the presentation. We show this percentage in light blue as the Fannie, Freddie universe and the dark blue line in the chart on the upper right is our servicing portfolio. So it's tracking pretty closely with the universe. There was a little bit of divergence in 2020, but it's largely converged now to the universe.

Our next question is from Doug Harter with Crédit Suisse Securities.

Bill, you talked about— you've talked for a while that the construct of the portfolio with Agency MBS and MSR is meant to dampen volatility in book value. I would say, for 2021, that didn't work. Can you just talk about lessons learned and whether there's any changes to portfolio construction or hedging that you might do to accomplish that lower book value volatility?

Yes, sure. Thanks for the question, Doug. 2021 can be viewed in two ways. One way is that the portfolio construction and the sensitivities that we show, as we say on Page 13, really reflect parallel changes in spreads across the stack. What we saw in 2021 with the Fed continuing to buy lots of mortgages was that current coupon mortgages didn't really widen at all during 2021. They stayed at very tight levels even through the end of the fourth quarter. What we did see was high coupon spread widening during the year. As rates fell and the Fed started buying more low coupons, they weren't making any more high coupons. It was a static pool which you couldn't easily add to. Prepayment speeds were generally faster than the market expected, and as a result, those coupons suffered in actual realized cash flow performance as well as forward-looking spread performance on the high coupons alone. The portfolio was not able to benefit fully from the offset of the MSR, which benefits from low coupon spread widening. That's what we're seeing today in the early part of the first quarter: a very rapid 20 basis-point move in January and further stress in February. In this environment, the MSR does provide an offset. We're switching to a turnover environment out of a refi environment, and I am more confident in the ability of the MSR to provide the offset it was designed to do in this environment.

Our next question is from Eric Hagen with BTIG.

So maybe two questions on my end. Before any offsets from MBS and hedges by itself, can you say what the total return has been year-to-date for the MSR portfolio? And then separately, what was the unlevered yield for the $39 billion that you bought so far this quarter?

I don't think I have the returns broken out on just the MSR of the portfolio without hedges. If I understood your question correctly, that's not something we regularly compute or monitor. We look at the portfolio on a hedged basis and it's not particularly informative to me what the return of the individual components are. In terms of the static yields of the portfolios we've been buying, we see, in general, static spreads in the very high single digits on an unlevered unhedged basis.

Okay. Got you. And then maybe you can shed some more light on the decision to rotate back into specified pools and really just how you expect the premium there to hold up if the expectation is for wider spreads like you guys noted in your prepared remarks?

Yes. So we've talked about this in the past. With the Fed tapering and removing their purchases, we expect role specialness to decline and that should cause TBAs to underperform specified pools. Investing in a static pool, which doesn't change over time, is different from the cheapest-to-deliver dynamic in TBA. As the deliverable worsens, specified pools should benefit relative to TBA. Over time, as roll specialness declines, we expect to rotate more into specified pools, and this was the beginning of that shift; we started by buying very low pay-up securities which should not be particularly at risk as spreads widen.

Got you. And then one point of clarity. The book value color quarter-to-date, does that include an accrued dividend? Or is it net of an accrued dividend?

Well, as you know, the dividend is—

I was just going to say that it would not include a dividend as those aren't set until the Board declares them in March.

Our next question is from Bose George with KBW.

This is actually Mike Smyth on for Bose. Just a couple of questions on the Fed balance sheet. I'm just wondering if you can talk through your expectations for runoff and how that may compare to what we saw last time? And then maybe if you can just talk through where you think the incremental demand comes from as the Fed moves away?

Thanks for the question. I don't have any better insight into what the Fed is going to do than anyone else. They've said that after liftoff they will start to consider normalization and that it could be faster than other periods. They've said they want to get to a treasury-only portfolio. I take them at their word. I think inflation fears have been internalized by the Fed and I think they will act accordingly. If anything, in this episode compared to the last one, there's probably a higher risk that the Fed acts more quickly and more aggressively to reduce their mortgage balances than they did in the past. That obviously puts higher risk on mortgage spreads.

Great. That's helpful color. And then just one more: leverage declined quite a bit again. I'm just wondering if you can discuss what you're looking for to go on offense and take that up, and maybe the timeline for doing so, given that your targeted returns on assets have improved modestly quarter-over-quarter?

Sure. The overarching environment here is one where the Fed is stepping away and there's still lots of supply of mortgages and spreads are tighter than they've been in other periods where the Fed has been removing accommodation. We're watching spread levels carefully; it's a day-by-day decision about the environment and relative value. In the past we've said our leverage could go up to 8 or 9x. Remember, leverage for us means something different than it does for portfolios without MSR because we have the MSR offset. We look at the spread exposures we show on Page 13 more carefully than nominal leverage, although we respect that number. Given the amount of MSR in our portfolio today, we think our leverage can go up into the 7s. If spreads overshoot towards the higher end of the range on Slide 5, it could touch the low 8s. But we'll watch the macro and spread environment carefully and make decisions when we think it's attractive to start increasing leverage over time.

Our next question is from Kenneth Lee with RBC Capital Markets.

Just one on the risk positioning and hedging profile. Based on Slide 14, it sounds as if Two Harbors is biased towards protection around a bear flattening scenario. Just want to gauge if that's a correct reading? Or are there a couple of other scenarios that you guys are particularly focused on?

Yes, we're most focused on that because that's where the action has been in the last several weeks. We knew where the market was heading toward Fed liftoff and typically the yield curve flattens substantially when that happens. The market is moving very quickly; for example, we prepared our remarks last night when almost 5.5 hikes were priced in for 2022 and after recent data there's more than 6 priced in now. We try to hedge all along the yield curve and protect against bear flattening as well as bear steepening or bull moves. We try not to run large interest rate risk positions. Our mandate is to provide mortgage exposure with MSR and not to make directional guesses on interest rates.

Got you. Very helpful. And just a follow-up, in terms of MSRs: you mentioned there could be a pickup in potential available acquisitions on the bulk side. Could you talk a bit more about what could be driving the increased availability this year?

Yes. The supply in 2021 and what's driving 2022 is that during the extraordinary refi wave in 2020 many participants chose not to sell their MSR because they felt prices were low at the time and primary-secondary spreads were wide. Back then, price multiples on at-the-money servicing were probably in the three area as opposed to the more typical four to 4.5 area. Participants held on, expecting better prices, and as rates rose the market healed and price multiples improved. Now those participants are looking to sell those balances. What we saw in 2021 wasn't all of that pipeline cleared. 2021 also had large refinance activity, and many participants were still making money running servicing operations. So not all of the backlog cleared and 2022 should see a continuation of sellers monetizing servicing holdings off their balance sheets.

Our next question is from Rick Shane with JPMorgan.

I want to follow up a little bit on the last question. You point out the differential in the market between bank-serviced versus nonbank-originated lenders. Are you starting to see any consolidation in the nonbank channels? Also, one of the things nonbank originators talked about for volume going forward, even as the rate-term refi wave ends, is the cash-out refi opportunity. How are you mitigating that risk within the portfolio? And are you seeing any peering in pricing between bank and nonbank originators related to that adverse selection?

Thanks, Rick. Let me address that. The addition of some bank-serviced specified pool collateral was a relative value trade between TBA and specified pools; we think TBAs will worsen as spreads widen and deliverables deteriorate. Bank-service collateral prepays a little slower than the generic cohort and these were low pay-up pools, so it's not a big risk position. Cash-out refinances are certainly a factor and we saw more of that in 2021; that's already priced into the market at reasonably fast levels. I think the risk of cash-out refi speed is probably to the downside rather than upside going forward—models and expectations were calibrated to very fast speeds, and it's just as likely speeds will be slower than those expectations. Additionally, in our servicing portfolio we have recapture agreements in place with our sub-servicers, and we're able to monetize some of that value as cash-out refis happen.

Got it. Okay. I will say we enjoyed the slide deck. I think every one of them could be a syllabus for a course on mortgage-backed securities with each slide being one daily lecture, and we're going to have to delve into all of them a little bit deeper.

Well, happy to help with that if you need me to. Thanks very much.

Ladies and gentlemen, we have reached the end of the question-and-answer session. And I would like to turn the call back to Bill Greenberg for closing remarks.

Just want to thank everyone for joining us today, and thank you, as always, for your interest in Two Harbors.

This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.