Good day and welcome to the Two Harbors Investment Corporation's First Quarter 2021 Financial Results Conference Call. Today's conference is being recorded. .

At this time, I would like to turn the conference over to Paulenier Sims, Senior Director of Investor Relations. Please go ahead. .

Good morning, everyone, and welcome to our call to discuss Two Harbors First Quarter 2021 Financial Results. With me on the call this morning are Bill Greenberg, our President and CEO; Mary Riskey, our Chief Financial Officer; and Matt Koeppen, our Chief investment officer..

The press release and the financial tables associated with today's call were filed yesterday with the SEC. If you do not have a copy, you may find them on our website or on the SEC's website at sec.gov. .

In our earnings release and slides, we have provided a reconciliation of GAAP to non-GAAP financial measures. We urge you to review this information in conjunction with today's call. I'd like to mention that this call is being webcast, and may be accessed in the Investor Relations section of our website..

I'd also like to remind you that remarks made by management during this conference call and the supporting slides may include forward-looking statements. These statements are based on the current beliefs and expectations of management, and actual results may be materially different because of a variety of risks and other factors.

We caution investors not to rely unduly on forward-looking statements. Except as may be required by law, Two Harbors does not update forward-looking statements and expressly 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 first quarter earnings call. Today, I will go over our quarterly results at a high level and share some thoughts on the market environment, including the meaningful shift in the interest rates during the quarter.

Mary will give more details on our financial results as well as our capital structure and financing activities. And finally, Matt will discuss our portfolio composition, activity, and risk profile as well as our outlook..

Please turn to Slide 3. At quarter end, book value was $7.29 per share, representing a negative 2.2% total economic quarterly return. The first quarter performance was largely driven by roughly flat portfolio returns with a significant increase in MSR values hedging the decrease in RMBS prices. Lower coupons underperformed higher coupons.

Specified pool performance was mixed as our high coupon securities, where we are mostly positioned, outperformed their hedges..

This quarterly performance also included the impact of the actions we took to optimize our liability and capital structure as we execute on our long-term strategy.

During the quarter, with spreads tightening beyond historic levels and volatility increasing, we reduced our RMBS position somewhat, primarily in 2% and 2.5% coupons, given the unattractive risk reward proposition..

Given these conditions, we would expect to maintain a low exposure to mortgage spread risk.

Ultimately, we believe the rapid pace of economic recovery will lead to a moderation of fed support and spread normalization, at which time we expect to increase leverage and deploy excess cash into investments at more attractive levels than what are available today..

As we mentioned last quarter, the MSR market is healthy, with pricing and activity at pre-crisis levels. We continue to benefit from the strong relationships with our originator partners, settling on $21.3 billion UPB in our flow-sale program, and adding another $1.1 billion UPB through bulk purchases.

I would reemphasize our view that MSR provides strategic value, both as a hedge to mortgage spread risk and as an asset that, on its own, delivers a low double-digit levered return. And this is where we expect to continue to deploy capital while keeping spread risk low..

In this environment where the short-term direction of mortgage spreads remain highly uncertain, we are confident that the agency plus MSR strategy can continue to deliver attractive risk-adjusted returns over time. So with more than a single quarter in mind, we declared a common stock dividend this quarter of $0.17 per share.

As always, we will continue to evaluate the dividend within the context of market conditions and return expectations..

On Slide 4, we highlight some of the noteworthy market dynamics and rate movements during the quarter. A substantially improved market outlook for the U.S economy drove interest rates sharply higher, with 10-year swap rates rising from 93 basis points to 178 basis points and 5-year forward 10-year rates climbing all the way to 2.5%.

During this period, the mortgage current coupon rose 66 basis points from 1.38% to 2.04%..

Looking at the bottom left chart, the spread between primary and secondary mortgage rates continue to compress to more historical levels, especially over the last 2 quarters, meaning that the gain on sale profit for originators has also come down to more normalized levels.

And any further increase in interest rates will likely be passed on to borrowers as higher mortgage rates..

The chart on the top right shows the distribution by mortgage rates of Fannie mortgages in the blue bars, which we believe is a good proxy for the agency universe as a whole. At the end of December, roughly 85% of all mortgages had at least 25 basis points of refinance incentive, as shown by the gray circle.

With the rise in interest rates and mortgage rates now hovering around 3.17%, the percentage of mortgages that are refinanceable with at least 25 basis points of incentive has declined to about 64% as shown by the green circle in the charts..

Additionally, those 64% of borrowers have, for some reason, not taken advantage of refinancing when rates have been lower. So their sensitivity to refinancing is not precisely known.

Nevertheless, higher interest rates, higher mortgage rates, tighter primary/secondary spreads, and lower numbers of mortgages being refinanceable will inevitably lead to slower prepayment speeds, but should be beneficial to our agency plus MSR strategy..

Although rates have shifted higher and prepayment expectations have moderated, RMBS valuations continue to be very rich.

We acknowledged that there are powerful technical factors at play, such as continued role specialness in the 2.5% coupon and not unrelatedly, large-scale purchase demand from the federal reserve as well as large banks as those entities added $380 billion and $180 billion of RMBS respectively in the first quarter alone..

Looking at the chart on the bottom right of Slide 4, the option-adjusted spread on an index of agency RMBS currently sits at 6 basis points as of the end of the quarter. Although you cannot see from the chart, current spread levels are the types they have been since 2006. However, we do not expect these favorable technical forces to last forever.

With the virus receding and the economy gaining strength, the federal reserve will one day begin to moderate their asset purchases..

Most market participants expect that the fed taper will occur sometime in early 2022, and Chairman Powell has said that he will be careful to telegraph those intentions to the market well in advance. And so that communication could come this year.

Given the pace of economic rebound, it's our view that the risk to that conventional forecast is to shorter time lines, not longer.

Furthermore, while the fed has been clear that they will wait until they see the results of quickening economic growth, the large banks, within all likelihood, act sooner to divert capital from investment to making loans and so a bank taper could occur even sooner..

As we discussed on our fourth quarter earnings call, data analysis of historical spreads suggest that spreads are typically mean reverting once they reach extreme levels. As a result, we view the risk of spread widening to be significantly higher than the likelihood of spread tightening.

With a long-term mean of 35 basis points over the last 10 years, spreads have room to materially widen..

Without the benefit of role specialness, the fundamentals of owning RMBS are not very attractive today. With spreads being where they are, the hedge-adjusted carry on the asset only leads to mid to high single digit gross returns.

Given these rich valuations and higher price volatility that we've seen recently, the investment thesis of buying a rich asset and expecting that the fed and the banks will keep buying at higher prices does not seem sustainable to us.

Indeed, as Matt will discuss later, the total quarterly performance on the TBA coupons that the fed and banks have been buying was essentially 0, meaning that we did not give up any economic performance by being underweighted those coupons..

We tend to be patient and deliberate in executing our strategy. We are continuing to add MSR at attractive levels, which, when paired with RMBS, keeps our exposure to mortgage spreads low.

One benefit of the agency plus MSR strategy is that it allows us to keep exposure to spreads low when spreads are tight, and to increase exposure when sprints are attractive. This is what we intend to do over time. .

I'll now turn the call over to Mary to discuss the details of our financial results. .

Thank you, Bill, and good morning, everyone. Please turn to Slide 5 to review our financial results for the first quarter.

Comprehensive income was negative $48.5 million, representing an annualized return on average common equity of negative 9.3% and our book value was $7.29 per share compared to $7.63 at December 31, resulting in a total economic return of negative 2.2% Book value decline was mainly a function of flat portfolio performance, net of core operating expenses, with an increase in MSR valuation offset by the impact of lower RMBS prices.

A higher tax provision and the premium related to the convertible debt repurchased contributed to the overall decline. We recognized a higher tax provision driven by the dramatic increase in the mark-to-market of our MSR assets..

As a reminder, a portion of our MSR asset is required to be held in a taxable REIT subsidiary. Although we record tax provisions or tax benefits every quarter, the amount this quarter was more impactful on our results.

Although we focus on limiting our ultimate tax obligations, which have been historically low, GAAP to tax differences can result in current period impact to earnings and book value. .

Moving on to Slide 6. Core earnings was $0.17 per share compared to $0.30 in Q4. Interest income decreased from $72.5 million to $56.1 million in conjunction with lower asset balances resulting from prepayments and asset sales.

Interest expense rose modestly due in part to a higher convertible debt balance, reflecting the portion of debt maturing in January 2022, which is not redeemed and also due to increased usage of revolving credit facilities to fund MSR.

Gain on other derivatives declined from $43.5 million to $18.9 million, driven by lower TBA dollar roll income as we reduced our positions. We also saw a smaller benefit from roll specialness, which contributed $0.02 to core earnings compared to $0.06 in Q4. .

As we discussed last quarter, we are taking actions to optimize our capital and liability structure. And as part of that, are adding funding capacity in the form of revolvers for MSR. As we utilize the capacity in these facilities, we expect interest expense to increase once again in the second quarter.

While it may not be intuitive to draw up MSR facilities, while we are in a strong cash position, it's important to note that these revolving structures take a significant amount of time to set up compared to standard repo facilities. They are multiyear funding and they typically have minimum use requirements.

It is essential to our MSR strategy that we consider our long-term funding needs and long-term expectations for portfolio assets in construction and plan ahead accordingly.

As a reminder, the liability and capital actions taken this quarter, inclusive of the convertible debt issuance, optimization of MSR financing and redemption of preferred stock is expected to deliver an annual net benefit of $0.04 per share beginning in 2022. .

Turning to the table on the lower right. Our portfolio yield in the quarter was relatively flat at 2.25% and our net spread decreased by 11 basis points to 1.65% due to higher cost of funds related to increased convertible debt and MSR financing.

Net spread as of March 31, which reflects our estimate for the near term, is expected to increase driven primarily by a higher yield due to lower prepayment expectations and a higher proportion of MSR in the portfolio. .

Turning to Slide 7. We continue to maintain a strong liquidity and capital position. Our unrestricted cash balance totaled $1.2 billion at quarter end. The weighted average maturity of our agency repo positions increased to 100 days as of March 31, reflecting the flat term structure market and the attractiveness of longer-dated repos.

We increased the unused capacity in our MSR asset financing facilities to $392 million with the closing of an additional revolver in the quarter. And we also have $180 million in committed capacity for servicing advances should forbearances or defaults rise as the forbearance programs and foreclosure moratoriums wind down in the future. .

Finally, turning to leverage. Our economic debt-to-equity at quarter end declined to 6.4x from 6.8x at December 31, and our quarterly average economic debt-to-equity was 6.5x in Q1 compared to 7.5x in the fourth quarter. .

I will now turn the call over to Matt for a market overview and portfolio update. .

Thank you, Mary, and thank you all for joining today. Turning to Slide 8. Let's discuss our quarterly portfolio activity and composition. As Bill noted, volatility picked up in both the rate and mortgage markets with rates moving sharply higher and mortgages experiencing several bouts of spread widening during the quarter.

If you only looked at the quarter-over-quarter performance of current coupon RMBS, you might assume it was a quiet period. However, that was far from true, and I will spend some time discussing the volatility during the quarter shortly.

Ultimately, RMBS performance was driven by very strong and continued demand from both the federal reserve and the banking system. .

As Mary discussed, we continued to decrease balances during the quarter, reflected by lower economic debt-to-equity of 6.4x. This was in part from sales and paydowns in our specified pool portfolio, where valuations in some stories and coupons have become less attractive. We also reduced our overall TBA exposure from $5.5 billion to $5 billion.

Part of our TBA activity was rotating into higher coupons and reducing exposure to the 2% coupon, due to both an unattractive valuation and a significant decrease in roll specialness. .

We continue to source substantial volumes of new MSR assets through our flow program at attractive levels and have largely maintained the size of our portfolio in this fast prepay environment. Additionally, we opportunistically added around $130 million market value of interest-only securities, or IO, during the quarter.

As we highlighted last quarter, IO positions provide portfolio benefits that are similar to MSR when paired with RMBS in reducing mortgage spread exposure. .

Moving to Slide 9. I'd like to take a moment to provide some color on the positioning in current coupon TBA mortgages in today's environment, where in our view, duration-hedged carry no longer provides adequate compensation for elevated risk. Starting with the upper graph, we show the duration-hedged carry, in ticks, for TBA 2s and 2.5s.

Notably, carry on the 2 coupon has decreased significantly from over 6 ticks per month to around 2.5 ticks at the end of March, caused mostly by the decrease in roll specialness. The graph on the lower left shows the monthly price volatility of the same instruments also measured in ticks per month.

During the quarter, we witnessed significantly higher spread volatility than observed in the prior 2 quarters. Monthly volatility in the 2 coupon nearly tripled from around 10 ticks to around 28 ticks. Putting these ideas together, investors can risk 28 ticks per month in a one standard deviation move for the opportunity to earn 2.5 ticks per month. .

The final graph shows this in terms of how many months of carry are at risk. And you can see that for 2s, 11 months of carry are at risk each month, assuming the observed first quarter volatility. This is significantly longer than we typically observe in the mortgage market by a factor of 2.

With current coupon mortgages at or near all-time tight spreads, we believe the value proposition is quite challenging in the near term. .

Please turn to Slide 10 as we discuss our specified pool positioning and prepayments. In the lower left-hand chart, you can see that performance was a mixed bag with lower coupon specified generally underperforming TBA and higher coupons outperforming somewhat.

In general, the performance across the stack wasn't particularly notable quarter-over-quarter, but as discussed, that doesn't speak to the intra-quarter volatility. .

Today, with regards to specified pools, we remain positioned largely in loan balance and geography stories. In the lower right-hand chart, we show a comparison by coupon of observed prepayment speeds from pools delivered into TBA contracts to observe prepayment speeds on our specified portfolio.

The slower prepayment speeds as compared to delivered TBA pools highlights the reason they command a significant price premium over TBA. .

Moving to Slide 11. You can see that our MSR portfolio was valued at $2.1 billion as of March 31 based on $187 billion UPB and with a gross coupon of 3.6%. That translates into a price of about $0.110 or right around 4.2 multiple.

The balances from the end of 2020 are also shown here, and I would highlight the significant increase in value during the quarter. As the yield curve steepened with the rise in long-term rates, the value of the servicing portfolio increased by over 30% and the multiple increased from 3.2 to 4.2.

Multiple expansion is a natural consequence of rising rate environments as forward interest rates and mortgage rates increase and prepayment expectations slow, increasing future cash flow to the servicing strip. .

We settled $21.3 billion UPB of new MSR through our flow program during the quarter, which was enough to offset runoff in the period. Activity in the bulk market continues to pace, and we continue to find valuations to be situational with some packages trading at precrisis yields while some are clearing at wider spreads.

Around $50 billion UPB in bulk transactions came to market during the quarter, which is a fairly average volume, and we committed to purchasing $13 billion UPB year-to-date. .

In the lower right-hand chart, we compare our servicing prepayment speeds in blue versus TBA collateral in gray. Currently, a majority of the underlying loans in our servicing portfolio have some form of seasoning or prepayment protection, which is why our speeds are somewhat slower than the through-the-box speeds. .

Over the next 2 slides, we display our effective coupon positioning and risk profile. In the chart on the top of Slide 12, we show the combined exposures of agency P&I bonds, MSR and IO as of March 31 as compared to our positioning at the end of 2020, as indicated by the diamond bullets. There are 2 main differences.

One is the significant reduction in exposure to the 2% coupon as discussed earlier, while increasing exposure to the 2.5 and 3 coupon in TBA. The second item of interest is that the current coupon equivalents displayed on the middle chart shifted up from the 1.5% and 2% coupons to become mostly concentrated in the 2.5 coupon.

On net, you can see that we are slightly short the current coupons, while maintaining long positions in the 3 coupons and up..

The lower left-hand chart shows our common book value exposure to 25 basis points spread widening or tightening and it indicates that book value would decrease by 4.2% in an instantaneous 25 basis points spread widening. 4.2% represents an increase in exposure from the prior quarter despite our reduction in specified pools and TBA.

It is due to the extension in mortgage durations that comes along with higher rates and steeper curves. We still believe this a low and manageable risk, and it's certainly lower than the book value risk that would accompany an RMBS portfolio without the presence of MSR..

Moving to Slide 13, here we see our interest rate and curve exposure. Both are low and in line with our historical positioning. I would call out again that the agency plus MSR strategy provides significant interest rate offsets, which you can see by comparing the gray bars to the blue bars in both charts..

Finally, I'd like to take a look at our outlook for Two Harbors and our return expectations for new investments on Slide 14. Working from the bottom up, we believe gross returns for specified RMBS paired with swaps are less attractive than they were, and expect returns to be in the range of mid to high single digits depending on coupon and story..

TBA returns in the 2.5 coupon are enhanced by roll specialness, which is likely to continue for the near future, albeit at less attractive levels. New investments in flow MSR, paired with RMBS today, can also drive returns in the high single digits or low teens. And if you assume roll specialness on the RMBS component, can be even higher..

As Bill mentioned, we are focusing our efforts on adding new servicing paired with RMBS. We continue to focus on our strong partnerships with MSR sellers and internal platform, which gives us the ability to source significant volumes. RMBS valuations are rich, and the near term technical of fed and bank demand are formidable.

Nevertheless, it is our expectation that we are near to the end of QE4 than the beginning, and our portfolio is well constructed for eventual future mortgage spread widening..

Our cash position is strong, and we look forward to the normalization of the mortgage spreads and the opportunity to deploy that cash into more attractive spreads. Despite this quarter's challenges, we continue to believe that longer-term returns are attractive for our agency plus MSR portfolio construction. .

And now I'll turn it back to Bill. .

Thanks for that discussion, Matt. We continue to be excited about our unique portfolio construction and the benefits that MSR bring to the paired agency plus MSR strategy. In particular, with agency mortgages having reached decade-low tight spreads, we think that our portfolio, with low exposure to widening mortgage spreads, is especially attractive. .

Thank you very much for joining us today, and we will now be happy to take any questions you might have. .

[Operator Instructions] Our first question today comes from Douglas Harter. .

Just to get some clarity, Matt, on what you were just talking about, continuing to work to pair to pool agency with MSRs.

Does that mean as you continue to grow servicing, you would look to kind of add back to the Agency portfolio or you're kind of waiting for a better opportunity there and just continuing to grow the servicing portfolio today?.

Doug, thanks for that question and thanks for joining today. I think the answer is largely yes, actually. So we are very focused, like you said, on adding to the servicing asset. I think there's reasons to believe that we'll see increased flows here in the coming months, given the shift higher in rates.

I would think there would be more for sale, and we'd see increased bulk sale. And with respect to how we would hedge it, I think at the moment, we would likely hedge it with TBAs, probably TBA 2.5 in particular, given the attractive roll specialness that still exists there.

And that would -- as you know, that paired construction will continue to keep our mortgage spread duration low in aggregate, but put interesting income into the portfolio. .

Okay.

And then just on the margin, as you talked about that the margin today is higher than the quarter average, can you just talk about what are the assumptions that kind of go into the prepays for that quarter end at margin versus kind of what you're currently experiencing on prepays?.

I'm sorry, Doug, I'm not following.

Which margin are you referring to, exactly?.

The net interest margin, that's on Slide 6 -- or the net interest spread, sorry. Where you show the 1.94% spread as of March 31 versus the 1.65% realized. .

Mary, do you want to maybe take that one?.

Sure. I mean the as of yields, as of March 31, which of course is based on a static portfolio, we do expect yields to go up, and that is primarily due to lower prepayment assumptions and a greater proportion of MSR in the portfolio.

Are you asking specifically what the prepayment assumptions are?.

I guess just curious if that's using prepay speeds that you're currently seeing, or that is based on prepay speeds declining further from here, like using a lifetime assumption. Just trying to understand if that higher yield is going to come through in 2Q, or it's going to take longer to achieve. .

Well, first I would -- go ahead, Matt. .

Well, I was just going to say, that does include -- that yield does include expected future prepayment speeds and we expect that, right? So the increase in rate hasn't quite made its way through into prepayment speeds, but it will in Q2.

We expect speeds from April coming through this month to be down 20%, say, 15% in the months following that, all from the rate effect. And that is what is embedded into the yield assumptions that you see in that calculation. .

The next question comes from Rick Shane. .

This is Charlie on for Rick today. First, I was wondering if we could just get an update on where book value has trended through, I guess, the end of April. And then kind of more broadly, had a question on sourcing MSRs going forward. As you mentioned, you've continued to grow that portfolio even in a pretty fast prepayment environment.

As we look ahead and consider the possibility of rising rates and slower prepayments, I'm wondering if the way that you source MSRs is going to change in any way? Or I guess said differently, how should we think about the mix between flow arrangements, which I assume is a big part of kind of replacing that runoff, and also the bulk purchases, as prepayments begin to slow down and that environment changes? It seems like you've kind of already laid the foundation here with the additional capacity, but just curious to think about how that rate of change might evolve over time, and if growth can really accelerate from here.

.

Sure. I'll start with that one. Charlie, thanks for the question, and I'll hand it off to Bill for the second part. But in terms of April, we haven't quite closed the books altogether on April, but at the moment, we are estimating a total return through the end of the month of about down 1%. And I will let Bill comment on MSR sourcing. .

Yes. Thanks, Matt, and thanks for the question, Charlie. Good to have you on the call today. You're 100% right, that a lot of our MSR acquisitions in recent months have come from our flow channels and relationships as a result of fast prepays and fast current production.

As rates rise, we do expect that to decline, but we do also expect the number of, and volume of, bulk packages available in the market to increase to offset that.

This is something that you often see when rates rise and origination profit declines, that many small and mid-sized originators need to or want to sell servicing in order to generate cash to fund their businesses.

And so that's a dynamic that we think is in play, and we think our platform is really well-suited to switch opportunistically between an acquisition channel of flow and bulk as market conditions change, and so we feel pretty good about our ability to be able to continue to source the amounts of servicing that we like. .

We can now move on to Bose George. .

Actually, just a follow-up on the last question. Did you just give the book value update as being down 1% quarter-to-date? And then just when you think about -- and then just a positioning on the book, I guess it's safe to say if spreads don't tighten from here.

Your book value probably has limited downside and the positioning is if spreads widened, then that's where you kind of essentially outperform?.

Yes, I would say, in general, at the moment, as Matt discussed in his prepared remarks, we are currently keeping our mortgage spread low in general, right? And so I would expect that no matter what spreads do here at the moment, we would have low book value volatility as a result of mortgage spreads.

One of the nice things about -- and I think we said this in the prepared remarks, also, one of the nice things about the strategy is for the presence of MSR in the portfolio allows us to flex the amount of mortgage spread risk that we have so that when spreads are tight, like they are today, we can keep our mortgage spreads very low and when spreads become more attractive, which we believe they will in the future one day, we can increase that somewhat and still have low mortgage spread compared to a portfolio without MSR, but greater than we do today in order to take advantage of those more attractive mortgage spreads.

.

Okay. Great.

And then can you just remind us why some of the MSR needs to be held in the TRS? And does that change the return profile?.

I'll let Mary take that one. .

Sure. So the -- we purchased our MSR into our servicing entity, which is a taxable REIT subsidiary. That's the named servicer with GSEs. And then they kind of dictate how much they allow the excess servicing to be sold into the REIT.

So both Freddie and Fannie have required amounts that need to be held in the actual servicing entity, which is the taxable REIT subsidiary. .

We can go to Eric Hagen. .

So you've got the MSR, which you noted is essentially equivalent to being a short lower coupon TBAs, which I think seems to make sense. And then you were also long some lower coupon TBAs at quarter end, which I think you noted aren't really as attractive from a carrier standpoint anymore.

Can you maybe just discuss what your footprint looks like there now? And if you're still along with those coupons, what you're picking up by being long when you're basically short the MSR, so long the MSR short the TBA?.

Sure. Thanks for the question, Eric. So when we were talking about the position rotation, we're mostly talking about the 2% coupon, that was the one that really, in particular, had much less attractive hedged-adjusted carry and so the roll specialness [Technical Difficulty] quite a bit during the quarter.

The 2.5% coupon actually still is rolling pretty special. I think that dynamic could last at that coupon for some time. Recently, its rolls have cleared in the, say, 40 basis points -- negative 40 basis points range or so. So that's 50 to 55 through repo funding, which is still pretty attractive.

And so like we said earlier, I can see us adding to that position against new servicing purchases as we go forward here. .

Got it. And then the swap portfolio, it looks like it has a pretty short duration. I think that also makes sense. Paydowns are still elevated for specified pools. And the MSR duration, I think, is at the longer end of the curve.

But how do you think about beefing up the swap portfolio if rates pick up a little further? Do you think about just delevering a little bit more?.

Well, I think -- I mean, our swap portfolio is pretty small in general, I would say. Some duration hedging on the short side in the front end and we're -- we need to be along a little bit of duration in the longer end.

You can see that on Slide 13 actually of the presentation that most of the rate exposure is handled by the servicing asset, and the swap book is quite small. As you know, we tend to keep a very tight leash on durations and curve. So I think we would expect to see that going forward as well. .

I can add a little bit more color to that, Eric, if I can. As Matt said, on Page 13, the parallel shift duration, as you see in the lower left chart, is largely offset between MSR and RMBS. But if you look at what the curve exposure is, when you pair those 2, it's generally a curve steepener.

And Matt said this, that most of our swap book is receiving in the long end in order to offset that, but the natural position of the paired construction is to have a curve steepener on, and so our hedges are mostly meant to offset that. .

[Operator Instructions] Our next question comes from Trevor Cranston. .

A question on prepays related to the slide you guys showed with the percentage of the market that's refinanceable. And I think you made the point that it's still something like 60%, but there's a lot of borrowers who, for whatever reason, haven't refied so far.

So I was curious to get your thoughts on the program the FHFA announced last week to try and help lower income borrowers have more access to refinancing.

How impactful you think that could be to prepay speeds in the spec pool market? And if you think there is a risk that more programs like that come out over the course of the year?.

Sure. That's a good question, Trevor. Thanks for that one. I'll take a stab at it. I think that's a fairly new announcement came out last week. So I don't think we know or processed all the impacts from it. But as you mentioned, the idea, of course, is that low-income borrowers haven't been able to take advantage of low mortgage rates here.

So a couple of things I could say is that I think early estimates are that it could affect, say, I don't know, 10% to 20% of GSE borrowers. I think we'd expect that impact to be fairly small for lower coupon RMBS, right? Maybe some impact of an additional CPR for some period of time.

It could be more impactful for higher coupons, right, like we can maybe estimate it as being as much as a 5 CPR increase for some time period. But that would require a very high implementation and a very high pull-through rate.

So it remains to be seen how aggressively it's sort of pushed through the originator community and also how it's received by borrowers. .

As there are no further questions, I'd like to hand the call over to Bill for any additional or closing remarks. .

I want to thank everyone for joining us today, and thank you, as always, for your support of Two Harbors. .

Ladies and gentlemen, that concludes today's conference call. We thank you for your participation. You may now disconnect..