Good morning, ladies and gentlemen. Thank you for standing by, and welcome to the Ellington Residential Mortgage REIT 2020 First Quarter Financial Results Conference Call. Today's call is being recorded. [Operator Instructions].It is now my pleasure to turn the floor over to Jason Frank, Deputy General Counsel and Secretary. Sir, you may begin..

Thank you, and welcome to Ellington Residential's First Quarter 2020 Earnings Conference Call. Before we begin, I would like to remind everyone that certain statements made during this conference call may constitute forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995.

Forward-looking statements are not historical in nature. As described under Item 1A of our annual report on Form 10-K filed on March 12, 2020, forward-looking statements are subject to a variety of risks and uncertainties that could cause the company's actual results to differ from its beliefs, expectations, estimates and projections.

Consequently, you should not rely on these forward-looking statements as predictions of future events.

Statements made during this conference call are made as of the date of this call, and the company undertakes no obligation to update or revise any forward-looking statements, whether as a result of new information, future events or otherwise.Joining me on the call today are Larry Penn, Chief Executive Officer of Ellington Residential; Mark Tecotzky, our Co-Chief Investment Officer; and Chris Smernoff, our Chief Financial Officer.As described in our earnings press release, our first quarter earnings conference call presentation is available on our website, earnreit.com.

Our comments this morning will track the presentation. Please note that any references to figures in this presentation are qualified in their entirety by the end notes at the back of the presentation.With that, I will now turn the call over to Larry..

Thanks, Jay, and good morning, everyone. We appreciate your time and interest in Ellington Residential. The month of March will be remembered as one of the most challenging environments ever for leveraged mortgage portfolios.

The COVID-19 pandemic and the associated measures to contain the pandemic led to extreme volatility and severe dislocations in virtually all financial markets. Economic activity plunged as countries around the world implemented social distancing restrictions. Unemployment claims surged, consumer spending plummeted and GDP growth rates turned negative.

In March, equity sold off across the globe, yield spreads on most fixed income assets widened sharply, and a flight to safety drove record low yields on long-term U.S. treasuries. Portions of the yield curve inverted and interest rate volatility surged.On Slide 3, you can see the extraordinary quarter-over-quarter declines in treasury yields.

Repo financing stresses alongside a drop in asset prices severely reduced liquidity and prompted forced selling across virtually all credit-sensitive fixed income asset classes, and the residential mortgage market was not spared.

The selling pressure was severe, even in perceived safe havens like Agency RMBS.By mid-March, between the heightened interest rate volatility and the ongoing flight to the safe haven of U.S. treasuries, yield spreads on Agency RMBS had skyrocketed to levels not seen since the 2008/2009 financial crisis.

In response, the Federal Reserve slashed short-term interest rates nearly to 0, injected liquidity into the repo markets, launched several credit facilities similar to what it had implemented during the financial crisis and stepped in with unprecedented levels of quantitative easing.

All of which provided meaningful support, especially to the more liquid sectors of the market. The U.S. Congress passed 3 rounds of stimulus packages during March, culminating in the $2 trillion CARES Act on March 27, the largest emergency spending bill in history.

These actions were mirrored by central banks and governments around the globe, and the rollout of stimulus programs continued into April.As the Federal Reserve deployed its full crisis playbook, we saw it was in effect almost a full market cycle compressed into just a few weeks. U.S.

equities bounced back sharply from the March 23 lows as what had been a 34% drop in the S&P 500 in less than 5 weeks was immediately followed by an 18% rise in just 3 days.

The Federal Reserve's injections of capital eased liquidity stresses and yield spreads in the sectors targeted by the Federal Reserve's asset purchase programs tightened sharply, particularly in Agency RMBS, which recovered strongly during the last 2 weeks of the month.Take 30-year Fannie Mae 4s, for example.

By some measures, LIBOR option adjusted spreads on Fannie Mae 4s after reaching their widest level since the financial crisis, tightened an astounding 140 basis points between March 19 and March 31.

In fact, given the persistent high levels of interest rate volatility, which factored greatly into the calculation of option adjusted spreads, Fannie Mae 4 LIBOR option adjusted spreads were, by some measures, actually tighter at quarter end than they were on December 31.In the credit space, yield spreads in some sectors, such as investment-grade corporate bonds also tightened significantly following the Fed's actions, while other sectors, including noninvestment-grade CMBS, noticeably lagged.

Many measures of market volatility subsided from their highs but still remain greatly elevated at quarter end.For Ellington Residential, the precipitous decline in interest rates and high levels of interest rate volatility generated net losses on our hedges.

And while our Agency RMBS assets did appreciate in price during the quarter, they significantly underperformed our hedges.

As a result, we experienced a significant net loss for the quarter, as you can see on Slide 4.As we discussed on our last earnings call, we entered the year with an extremely liquid portfolio and strong balance sheet, which positioned us well to weather the volatility, especially compared to many other market participants, who became forced sellers at distressed prices later in March.

As March progressed, with the asset markets and financing markets looking more and more fragile, we proactively reduced the size of our agency portfolio in an orderly and measured way, which bolstered our liquidity and lowered our leverage. We entirely avoided any forced asset sales, which would have exacerbated losses.

The vast majority of the agency assets that we sold in March were sold either earlier in the month before yield spreads hit their wides or later in the month after yield spreads had already recovered strongly, especially after the Fed removed any explicit limits on its asset purchase programs.

As we reported in early April, we met all margin calls during the quarter.A significant portion of the loss that we experienced during the quarter was related to the market's pricing in lower pay-ups and high-quality specified pools.

This pay-up compression was largely attributable to market-wide liquidity problems, exacerbated by quarter-end balance sheet pressures as well as to the implementation of the Federal Reserve's amplified asset purchase program during the quarter, which was generally limited to TBAs as opposed to specified pools with pay-ups.

Given that the Fed's purchase program dominated the Agency RMBS market, fundamental valuation factors were overwhelmed by technical valuation factors.

So even as mortgage rates hit all-time lows, the market decidedly preferred liquidity of TBAs and the lower capital required to hold them as opposed to paying up for the value of prepayment protection in the form of specified pools.Going into April, we thought this pay-ups on specified pools were artificially low and represented excellent value and upside to earnings, and indeed, specified pools outperformed in April.

While losses are always disappointing, I believe, given that leverage mortgage portfolios, we're in the crosshairs of the distress in the financial markets this past quarter, that it is a testament to our portfolio management, risk management and liquidity management capabilities that we were able to limit those losses and preserve book value to the extent that we did.I will now pass it over to Chris to review our financial results for the quarter.

Chris?.

Thank you, Larry, and good morning, everyone. Please turn to Slide 5 for a summary of EARN's financial results. For the quarter ended March 31, 2020, we reported a net loss of $16.7 million or $1.35 per share. Core earnings was $3.4 million or $0.27 per share.

These results compared to net income of $9.7 million or $0.78 per share and core earnings of $2.8 million or $0.23 per share for the fourth quarter of 2019.

Core earnings excludes the catch-up premium amortization adjustment, which was negative $0.7 million in the first quarter compared to negative $2.5 million in the prior quarter.As you can see on Slide 5, the loss in the quarter was primarily driven by net losses on our interest rate hedges as interest rates declined and were highly volatile during the quarter.

A portion of these losses were offset by gains on our specified pools and also by net interest income on our portfolio, which increased significantly quarter-over-quarter with lower borrowing costs and a smaller negative catch-up premium amortization adjustment.TBAs outperformed specified pools during the quarter, as Larry mentioned, which depressed pay-ups on our specified pool portfolio.

Despite the drop in mortgage rates during the quarter, average pay-ups on our specified pools decreased to 1.67% as of March 31 as compared to 2.05% as at December 31.Also on Slide 5, you can see that our net interest margin improved significantly during the quarter, increasing 20 basis points to 1.2%, driven by lower borrowing costs.

The wider NIM led directly to an improved core earnings of $0.27 per share compared to $0.23 per share last quarter. Our small nonagency portfolio had a net mark-to-market loss for the quarter driven by substantial yield spread widening in that sector.Next, please turn to view our balance sheet on Slide 6.

You can see that we increased cash and cash equivalents significantly quarter-over-quarter to $59.7 million from $35.4 million. In addition to cash, we had other unencumbered assets of approximately $12 million as of March 31. At the end of the quarter, our book value per share was $11.34 compared to $12.91 per share from the prior quarter.

Our economic return for the quarter was negative 10%.Next, please turn to Slide 7, which shows a summary of our portfolio holdings. Our total mortgage-backed securities portfolio decreased to $1.05 billion as of March 31 as compared to $1.4 billion as of December 31.

As Larry mentioned, in light of the heightened levels of market volatility and systemic liquidity risk, we proactively reduced the size of our agency portfolio by 25%, thereby bolstering our liquidity and lowering our leverage.At the end of the first quarter, our debt-to-equity ratio adjusted for unsettled purchases and sales was 7.2:1, a decrease from 8.1:1 as of December 31.

Substantially, all of our borrowings continued to be secured by specified pools. We satisfied all margin calls under our financing arrangements during the quarter.Next, please turn to Slide 8 for details on our interest rate hedging portfolio.

During the quarter, our interest rate hedging portfolio consisted primarily of interest rate swaps, short positions in TBAs, U.S. treasury securities and futures. TBA short positions represented 16.8% of our interest rate hedging portfolio at the end of the current quarter, up from 13.6% at the end of the prior quarter.

While TBAs outperformed specified pools during the quarter, they severely underperformed interest rate swaps and U.S. treasury securities. So our results benefited from the portion of our interest rate hedging portfolio that comprised TBA short positions rather than interest rate swaps.Next, turning to Slide 9.

You can see the significant decrease in our net long exposure to RMBS as our net mortgage assets to equity ratio declined to 5.6:1 from 7.6:1. And additionally, on a more technical note, as of January 1, 2020, we applied the new credit loss standard known as CECL.

Because we have always fair valued our portfolio through the income statement, CECL had no impact on our earnings or book or quarter end book value. Finally, during the quarter, we repurchased 136,142 shares at an average price of $7.24 per share.I will now turn the presentation over to Mark..

Thanks, Chris. I've been in the mortgage market for over 30 years, and I have never seen market illiquidity and dysfunction that characterize the second half of March and early April. That said, while every period of market disruption resulting volatility is unique, each crisis shares common features.

For EARN, early recognition of what was happening, what was likely to come next and what the policy response could be, all allowed our team to substantially mitigate the extent of book value decline.The first step in understanding what happened to Agency MBS during March is to look at the markets during the first 2 months of the year, pre-COVID.

Over January and February, we saw a 75 basis point rally in the 10-year note, and that rate rally unleashed huge prepayment fears into the market. Mortgage originations soared, spreads on TBAs widened and pay-ups on specified pools climbed to very high levels with many pay-ups exceeding 4 points. Then the uncertainty of COVID gripped the market.

This was real uncertainty that caught the market by surprise, and this was the kind of frightening uncertainty that historical data and traditional models cannot quantify.

Without facing the predictive power of models, the market quickly repriced to reflect fear and emotion and also repriced to perhaps the most important technical factor, the proliferation of too many market participants with an adequate liquidity and over-leveraged balance sheets.The next step played out as they haven't passed crises.

The stock market cratered; investors redeemed furiously from mutual funds, including fixed income mutual funds; and repo lenders fretted about the safety and security of their loans. The balance sheet was in short supply.

So any bond holdings that required balance sheet leverage had a target on their back as cash was king.To understand the behavior of repo lenders, you have to put yourself in their shoes. Their best case is that their borrower pays off its repo loan on time and at par, and they could make their modest spread.

Their worst fear is that the value of their collateral could drop below their repo loan. So when repo lenders see prices drop, they first pull a lever that they can control, their margin call.

And when they margin call an over-leveraged borrower, that borrower often has to sell immediately, and that borrower is often forced to sell whatever it can regardless of fundamental value. And to make matters worse in this case, the forced selling occurred not only when COVID panic was at its highest but also as quarter end was approaching.

Banks typically reduced their risk going into quarter end as they had little balance sheet to spare, and so prices cratered, including pay-ups for specified pools.It was almost irrelevant that loan balance specified pools had longer durations than TBAs while rates were rallying. Those are model outputs.

Repo lenders margin called their borrowers aggressively as they simultaneously had to deal with both the risk of their repo book and their own internal quarter end balance sheet pressures.

In response to these margin calls and the subsequent forced selling, pay-ups for specified pools collapsed, which led to more margin calls on leverage agency pool portfolios and so on.

With this vicious cycle set in motion, you can see how thoughtful and forceful the response from the Federal Reserve was.The Fed quickly ramped up QE, which we thought was likely, but they also tailored it to the current market.

Not only was the Fed using QE as a transmission mechanism to lead to lower mortgage rates, but they were also using QE for private sector balance sheet relief. Day after day, the Fed bought tens of billions of dollars of MBS for next-day settlement, and they bought a range of coupons, not just the current coupon.

Buying a range of coupons for short settlement quickly gave the market the balance sheet relief it was so desperate for.

When April finally arrived, with balance sheets in better shape and a new quarter starting, the market could then focus more on relative value and with no longer priced by abject fear and desperate actions of the most over-levered investors.So what did EARN do and what does this mean for future returns? First, the fact that agency mortgage QE is firmly established as part of the Fed's crisis management toolkit makes Agency MBS unique among structured products.

There is a limit to how much an agency portfolio will decline in value if you risk manage that portfolio well enough that you don't have to engage in forced selling to meet margin calls, which is the key.

When spreads are tight, the incremental returns to be captured from that one extra turn of leverage are rarely worth it because it weakens your balance sheets in times of crisis. Another protective measure we take is that we generally structure most of our repo with a 3-month term with staggered maturities.

And we don't try to save a few basis points by using too much overnight or 1 month repo.Another important point about our portfolio positioning going into March is that early in the year, we decided that pay-ups were getting more fully priced so we have generally not been adding high pay-up pools.

So after the crisis hit and pay-ups collapsed, whenever we want to proactively lower our leverage, we had plenty of lower pay-up pools to sell, and it didn't cost us much to sell these pools versus TBA.

In other words, we're able to proactively lower our leverage more easily and more cheaply than if we had allocated all our capital to very high pay-up pools. This is where our research effort really helped again.

We concentrated our research over much of the last 6 months to analyze and understand what types of lower pay-up pools could still provide appreciable call protection but at a low cost.The other dynamic that this period of volatility affirmed on Agency MBS is that many losses from spread widening are reversible if you can avoid being a forced seller because you're not taking credit risk.

In contrast, in many credit-sensitive sectors of the fixed income market, such as many sectors of the high-yield corporate bond market, the economic impact of COVID will absolutely create real fundamental credit losses that will not be reversible.Looking forward, we now see tremendous opportunity in the current market and are well positioned to take advantage.

We anticipate that prepayment speeds will surprise in the low side for borrowers who qualified for mortgages as part of the credit box that had been expanding over the past few years up until the crisis in March.

MBS spreads are still attractive, and we think we can continue to benefit from the Fed backstop and controller prepayment risk with modest pay-ups. In addition, with the distress in the credit risk transfer or CRT market, the GSEs are likely very concerned about their ability to utilize that market to off lay their credit risk going forward.

Faced with the possibility of having to hold much more credit risk on their balance sheet, we think that GSEs will continue to shrink the credit box, which should keep speeds relatively low despite where mortgage rates are.We also like adding more TBA to the portfolio mix in the current market.

In the past weeks, the Fed has reduced its daily purchases from a peak of about $35 billion to $40 billion a day, down to $6 billion a day. But if MBS were to widen materially from here, we think the Fed has the ability to ramp up purchases again aggressively.Another positive tailwind now is the availability of great funding.

3-month repo has declined sharply and should see settle in around -- should settle in below 40 basis points, which gives us a generous net interest margin between asset yields and financing costs.

Since last summer, the Fed has been keenly focused on making sure that repo financing costs for treasuries and Agency MBS more closely track the Fed funds rate. So the MBS repo rates should absolutely be in the lower-for-longer regime.Our analytic models have been very useful for making conditional predictions.

We have no crystal ball to tell when people who go back to work or how high unemployment will get. But we think we are able to predict certain smaller outcomes, such as prepayment speeds as a function of mortgage rates, what happens when the credit box shrinks and what happens when the Fed buys.

Predicting answers to these questions have always been and will continue to be our primary focus to help inform our portfolio decisions.

But at the same time, in light of what happened in March, we continue to monitor the big picture, and we are acutely aware of the uniqueness of the current environment.To recap, we see tremendous opportunity going forward. Agency MBS spreads are still wide. They have Fed support.

Financing is plentiful and cheap, and rolls are very attractive, and we think prepayment protection can be found at relatively cheap levels if you know where to look.Now back to Larry..

They protected book value, and they protected our shareholders. This past quarter, that meant protecting our shareholders, not only by avoiding any forced asset sales, but it also meant protecting our shareholders by avoiding any expensive or highly dilutive capital raises.

And through it all, we were still able to end the quarter with a lot of cash on hand that should enable us to play offense going forward.The current investment opportunities look exceptional, given the continued dislocations in most sectors of the residential mortgage market.

Even with pay-ups recovering strongly in April, specified pools are still attractively priced, and financing costs have come down considerably. Net interest margins in many RMBS sectors are the widest that we've seen in years, and I believe that we are in an excellent position to take advantage of these opportunities.

Along those lines, we are currently considering increasing our capital allocation to our non-Agency RMBS portfolio.All that said, while the global government and central bank responses have provided a boost to liquidity and meaningfully improved market performance in the short term, the path forward for the economy generally and the credit markets, in particular, remains unclear.

In light of this uncertainty, our disciplined approach to liquidity management, interest rate hedging and asset selection will continue to be critical.

We will continue to strive to balance defense, that is building in a margin of safety to absorb additional potential market shocks with offense, that is the ability to capture some incredible relative value opportunities.

These principles will continue to guide us moving forward.Before we open the floor to questions, I would like to take this opportunity to thank the numerous members of the entire Ellington team for their hard work over the past weeks despite difficult circumstances.

And for all of those listening on the call today and to all of those in our communities and around the globe impacted, we hope that you and your families stay healthy and safe.With that, we will now open the call to your questions. Operator, please go ahead..

[Operator Instructions]. And your first question is from the line of Doug Harter with Crédit Suisse..

This is actually Josh on for Doug. We saw lower leverage in the quarter, not surprising. Curious how you're thinking about target net mortgage leverage, just given the uncertainty in the market.

And then, I guess, follow-up to that would be, how does that translate into the ROEs you're seeing on incremental capital deployment? You mentioned increase -- possibly increasing allocation to nonagencies. So maybe just the difference in ROEs you're seeing the different buckets that you look at..

Mark?.

Sure. So on the agency side, I think we can get ROEs 10% to 12%. Spreads are not as wide as they were earlier in the month, but there's also a lot less volatility, right, and the volatility comes with an increased hedging expense.

So I mentioned in the prepared remarks, we've spent a lot of time understanding where we can get call protection without exposing the portfolio to very high pay-ups.

And I think that's the sector we'll continue to invest in and where we think we'll see the best relative value.In terms of the ROEs on the nonagency market, right? The nonagency market came under a lot of pressure in March. There were tremendous selling, not only from mortgage REITS, but also from mutual funds.

And so we've seen a substantial drop in those prices above and beyond what could be justified by higher loss expectations, longer time lines, et cetera. So there, I think, we'd apply certainly lower amounts of leverage because there's still tremendous uncertainty, and so you can't rule out the possibility of asset prices declining.

But with modest amounts of leverage there, I think we can get sort of ROEs 10% to 14%..

Your next question is from the line of Mikhail Goberman with JMP Securities..

Congrats on a very good quarter in difficult circumstances. It's kind of strange saying that given a 12% book value drop, but considering what other firms have posted, that's fantastic..

Appreciate that. Thank you..

Yes. Absolutely.

Could you possibly give a book value update thus far through April and what we have in May?.

We're not prepared to do that. Meaning that -- yes, it's just -- yes, if -- yes. We do have those numbers internally, and we did put out, obviously, a lot of information in our early April release. And we will continue to consider whether we think it's appropriate to put out intra-quarter releases, but generally, that has not been our practice..

Okay. Fair enough. And you mentioned that spec pools are still pretty attractive. You also mentioned, I think, that you like TBAs at this point as well.

Could you maybe talk about your appetite for one versus the other as we move forward?.

Sure. So this is Mark. In regards to TBA, we're in with the Fed support that to us is a material change from the market 3 months ago, 6 months ago. So we spoke a lot on earnings calls we did when we reviewed 2019 performance about improving technology in the mortgage space, how we thought that was adding to negative convexity.

And the big increase in WACC -- the big increase in WACC and coupon that you saw for 2019 and how all those things sort of pointed in the same direction, which is essentially the erosion in the convexity of TBA and the erosion in TBA.

The quality of TBA is a cash flow that you can effectively hedge at a low cost with interest rate swaps or treasuries. There was just a lot of negative convexity that got introduced into the TBA market in 2019.

And you didn't have any Fed buying to absorb the cheapest lever, right? So remember, 2019, the Fed was solidly in portfolio reduction mode, and they were letting the portfolio run off by $20 billion a month, right?So now two important things have changed that we see. First is obviously Fed support, right? So the Fed is now buying certain coupons.

And if you look at the implied roll -- the implied financing costs as that you can infer from the roll market, for the coupon that the Fed's aggressively buying, Fannie 2.5s, Fannie 3, 15-year 2.5s, some of the Ginnie coupons, roll levels are very attractive now.

So now the rolls are not priced to the worst quality TBA they were in 2019, now they're priced to the fact that the Fed is hoovering up the worst pools and they're buying in such size that they're creating a little bit of a front month short many months.

So roll levels on what the Fed is buying is attractive to us.And the other thing is that the Fed -- we mentioned in the prepared remarks. The Fed has already reduced their purchases substantially.

So I don't think -- if we hadn't seen the Fed reduce their purchases, we would have thought it's likely as the market stabilize, the Fed will reduce their purchases, and that can cause mortgages to widen. But that has already happened. So the Fed went from buying as much as $40 billion a day, now buying more in the order of $6 billion a day.

So the fact that the Fed has already tapered their mortgages, now it leaves us with the view that the pace of their mortgage buying is more symmetric and likely to be with us for a little while. So it has -- those two things cause us to like certain TBA coupons more relative to discount specified pools than our view of the world in 2019.

So it's really those two things.And they're both significant, and that can change. That's the one thing I think it's important to emphasize is that to get the best risk-adjusted returns on these portfolios, you have to be dynamic and you have to respond to new information. You can have an investment thesis. That looks great on a Monday.

And Tuesday, new information can come in. And you got to take that information in and maybe that should alter your investment thesis. So that's how the world looks to us right now. But we -- every day, we try to come in and challenge the investment thesis and see how newer information can and should cause us to adjust what we're doing..

Thank you, Mark, for that answer. It's very detailed. One more, if I may. I'm looking at Slide 11 of your deck, which shows a negative spread of 3-month LIBOR versus repo of about, I guess, 20, 25 bps at April 1. I believe in your prepared comments, you were saying that 3-month LIBOR should be settling at around 40 basis points.

So what are you currently seeing for repo? And is there -- could you see that spread sort of turn positive not like it did back in 2018, of course, but basically, what are you seeing for that spread going forward?.

Yes. So I think in the prepared remarks, what we said was that we think 3-month repo is going to settle in before -- below 40 basis points. Now if you look at 3-month LIBOR, it's had some extraordinary volatility in the past couple of weeks.

3-month LIBOR had been shockingly high 2 or 3 weeks ago and very high relative to OIS, and now it's starting to come down.

So where I think 3-month LIBOR is going to be versus 3-month repo, I think they're going to be similar, right? So we're not going to get that big tailwind we've had from time to time when 3-month LIBOR was very high relative to 3-month repo. And then the floating leg you're receiving on your swap is more than covering the repo expense.

So I think we're going to be in a period of time where those two numbers are roughly equivalent.

But an equivalence would be better than what we had for a lot of 2019.And the other thing is that just the yield where we can buy assets, if we're really thoughtful about what we select in the agency mortgage market, that relative to the financing costs, and then adding on the additional costs on the pay fix swaps, that's a very healthy net interest margin.

And if we do that in a way where we're not exposing ourselves to tremendous pay-up volatility, the net interest margins are really healthy right now..

Yes. And if I could just add that our interest rate swaps right now, Mark, correct me if I'm wrong, but they're either exclusively or almost exclusively LIBOR-based swaps..

Yes. No, exclusively, I think..

Exactly. So now there are, as I'm sure you know, there are other types of swaps that some other market participants in our sector have been using so-called OIS swaps. And that's something that we're -- we've looked very closely at.

And I would say that it was obviously very good for us that we were exclusively in LIBOR swaps throughout the stress of March, LIBOR got very high. And so we were receiving higher rates on our swaps on the floating leg of our swaps.

But at the same time, that was helping us with the stress, whether it was -- and whether you want to call that the stress in the repo market or even just what was going on with asset prices, right?So that was a tailwind, a small tailwind, obviously, during March.

But it does show that in times of stress, right, it does make sense to have these LIBOR swaps on when the markets are very stable.

So this is, again, it's hard to have a crystal ball, but looking forward, if the markets become very stable, then our financing costs probably will be more correlated or maybe slightly more correlated with these more OIS based swaps, which don't have the type of credit risk -- as much credit risk embedded in them as the LIBOR swap.

So your -- it's a little bit a function of what's going to serve you better in a time of crisis versus what's going to correlate better with your funding costs in normal -- in more stable normal times..

And we have no further questions. This will conclude today's meeting. You may now disconnect..