Ellington Credit Co (EARN) 2016 Q4 法說會逐字稿

Welcome to the Ellington Residential Mortgage REIT fourth-quarter 2016 financial results conference call. Today's call is being recorded.

(Operator Instructions)

It is now my pleasure to turn the floor over to Maria Cozine, Vice President of Investor Relations. You may begin.

Thanks, Paula. Good morning. Before we start, I would like to remind everyone that certain statements made during this conference call may contain 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 10, 2016, 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. I have on the call with me today, Larry Penn, Chief Executive Officer of Ellington Residential; Mark Tecotzky, our Co-Chief Investment Officer, and Lisa Mumford our Chief Financial Officer.

As described in our earnings press release, our fourth-quarter earnings conference call presentation is available on our website, earnreit.com Management's prepared remarks will track the presentation. Please turn to slide 4 to follow along.

As a reminder during this call, we will sometimes refer to Ellington Residential by its New York stock Exchange Ticker, E-A-R-N, or EARN for short. With that, I will now turn the call over to Larry.

Thanks, Maria. It's our pleasure to speak with our shareholders this morning as we release our fourth-quarter results. As always, we appreciate you taking the time to participate on the call today.

The fourth quarter was the most volatile quarter for long-term interest rates since 2009, with a 10-year treasury yield surging 85 basis points over the course of the quarter and the 30-year treasury yield up an even more remarkable 75 basis points. The day after the presidential election, the 10-year yield increased over 20 basis points which was the largest one-day move in a 10-year yield since the taper tantrum back in 2013.

The agency mortgage market changed rapidly throughout the quarter, shifting from an environment in the beginning of the quarter, where investors were worried about a refinancing wave and watching their yields and durations shrink as a result, to an environment at the end of the quarter, where refinancings were coming to a screeching halt and the durations of MBS were now longer than they had planned for.

The 30-year mortgage rate increased 90 basis points over the course of the quarter, leaving the majority of 30-year agency mortgages out of the money for refinancing. It's almost hard to believe that only a few months ago, in the third quarter, we experienced the highest market-wide level of pre-payments since 2013. And then at the end of the year, the MBA Refinance Index hit the lowest level since the financial crisis.

When the dust settled after this massive shift in interest rates, EARN still had solid performance, generating $0.22 in earnings per share on a fully mark-to-market basis. Even after paying out a $0.40 per share of quarterly dividend, our book value per share only declined slightly by 1.1% to $15.52 per share. And our full-year economic return for 2016 was 8.3%.

The positive economic income we generated in the fourth quarter equated to an annualized return on equity of 5.6%, not bad, actually. Let's wait until all the other agency-focused mortgage REITs have reported their fourth-quarter results, and let's see whether any others will be able to say that they had a positive economic return for the quarter.

So far, all the ones we have seen have reported negative economic returns and often they reported substantial negative returns. Hats off to Mark Tecotzky and the rest of his investment team for an absolutely terrific job this past quarter.

We attribute our performance this quarter to our disciplined hedging strategy and model-driven asset selection. We are very selective, right down to the individual pools we choose to buy. And this process of careful construction served us well this quarter as we successfully shielded our portfolio from blows to MBS valuations.

In 2017, we will continue to execute this strategy in what could be an extended period of heightened interest rate volatility. Fortunately, with volatility often comes opportunity. We will follow the same format on the call today as we have in the past.

First, Lisa will run through our financial results. Then, Mark will discuss how the residential mortgage-backed securities market performed over the course of the quarter, how he positioned our portfolio, and what our market outlook is. Finally, I will follow with some additional remarks before opening the floor to questions. Go ahead, Lisa.

Thank you, Larry. Good morning everyone. In the fourth quarter, we had net income of $2 million or $0.22 per share. The main components of our net income were our core earnings, which totaled approximately $4.9 million, or $0.54 per share, net realized and unrealized losses from our securities portfolio, which were $25.1 million or $2.75 per share, and net realized and unrealized gains from our derivatives in the amounts of $22.2 million or $2.43 per share.

By this measure, net realized and unrealized gains from our derivatives excludes the net periodic costs associated with our interest rate swap, since they are included as a component of our core earnings. Our core earnings includes the impact of catch-up premium amortization, which in the fourth quarter, increased our core earnings by $600,000, or $0.07 per share.

Catch-up premium amortization is calculated based on interest rate levels and prepayment projections at the beginning of each quarter. Last quarter, this adjustment was negative at $1.4 million, or $0.16 per share. If we exclude the adjustments in both the fourth and third quarters, we can see that our core earnings declined by $0.01 in the fourth quarter to $0.47 per share. The premium catch-up amortization generally tends to fluctuate from quarter to quarter.

Excluding the adjustment, the largest variance in our quarter-over-quarter earnings was an increase of approximately $400,000, or $0.04 per share in our interest expense. Interest expense includes our cost of repo as well as costs related to our short US Treasury positions.

The increase in our interest expenses was almost entirely offset by a slight increase in our interest income and decreases in our swap hedging costs and operating expenses. In terms of our net interest margin components, the weighted average yield on our aggregate portfolio declined to 2.75% in the fourth quarter from 2.78% in the third quarter, each adjusted to exclude the impact of the catch-up premium amortization adjustment.

While our weighted average agency RMBS yield increased 3 basis points quarter over quarter to 2.69%, our weighted average non-agency RMBS yields decreased over 3 points to 7.2% quarter over quarter because we sold certain high-yielding positions towards the end of the third quarter. This small portfolio has consistently augmented our overall results.

With respect to our cost of funds, our cost of repo increased 10 basis points to 81 basis points during the quarter. But this was partially offset by a 5 basis points decrease to 25 basis points in the periodic costs associated with our interest rate hedges and our short US Treasury sold.

Overall, in the third quarter, our annualized cost of funds increased to 1.06% from 1.01%. As a result, for the fourth quarter, our net interest margins adjusted for the impact of the catch-up premium amortization adjustment decreased to 1.69% from 1.77%. During the fourth quarter, as interest rates rose, prices declined on our agency RMBS and resulted in net realized and unrealized losses.

However, those losses were significantly offset by net realized and unrealized gains from our interest rates hedges. The fact that we generally aim to hedge out most of our interest rate duration risk helped us greatly this quarter as interest rates rose significantly. Consistent with our portfolio management style, we actively traded our agency RMBS portfolio.

Portfolio turnover for the quarter was 16%. We took advantage of trading opportunities that we identified following the large interest rate move. At the end of the fourth quarter, our aggregate MBS portfolio was just over $1.2 billion and our leverage adjusted for unsettled purchases and sales was 8.3 to 1. I would like to now turn the presentation over to Mark.

Thank you, Lisa. The fourth quarter demonstrates why our strategy with EARN is to capture mortgage spread and focus on relative value which doesn't require bets in the direction of rates or the shape of the curve. The aftermath of the US election was an almost 100 basis point sell-off in the 10-year swap rate, and with it, a large increase in mortgage-backed securities.

Just like the taper tantrum, it was another example of how a macroeconomic premise that is taken as a given and has guided market pricing for years can then be challenged and rejected by the market in a matter of days. By the end of the year, volatility spiked and interest rates closed near highs of 2016.

Mortgage investors that had been laser focused on elevated prepayment rates found themselves worried about extension risk in a world suddenly filled with discount securities. Over the quarter, we had to quickly adapt and shed risk in a declining market.

We came into the quarter with a healthy net interest margin of our assets over our liabilities, and that NIM we captured, together with the advantageous portfolio choices we made and the trading gains from continued turnover in our portfolio, were more than enough to offset the additional hedging costs that we incurred even on a fully mark-to-market basis.

We generated a 5.6% annualized economic return for the quarter against a macroeconomic backdrop that was one of the most volatile in years. We believe that this quarter in particular and this past year in general shows that substantial returns can be generated without the need for large interest rate bets or excessive amounts of leverage.

This past quarter also shows the danger of having one's portfolio construction premised on the prediction of the outcome of macro events. Prior to the elections, the polls show that showed that Clinton would win; and furthermore, financial researchers declared that if Trump were to actually win, both interest rates and stock prices would plummet.

So even if you had a better election crystal ball than everyone else and you correctly predicted that Trump was going to win, you probably would have had a portfolio positioning that would end up losing a lot of money. For all these reasons, we try not to tie the Company's fate to unpredictable exogenous events.

We can't rid inflate ourselves completely but we try to do it to a large extent. Then we focus on sources of repeatable income, like predictable net interest margin, long-term relative value opportunities and short-term trading opportunities. The sell-off that finished the year was the largest since the taper tantrum in 2013.

From high to low in the quarter, the 10-year swap rate sold off 110 basis points. On slide 7, you can see that the duration of Fannie 3.5 on the left axis and the level of 10-year interest swaps on the right. The big consequence of the gap in rates and volatility was the dramatic duration extension of mortgages. For example, the duration of Fannie 3.5s more than doubled.

We dynamically adjusted our portfolio during the quarter, both the asset side and the hedging and liability side, to manage the duration risk of the Company. We also had a tremendous benefit this quarter, in having a significant portion of our hedges in TBA mortgages.

As expected, TBA mortgages extended on a percentage basis and duration more than our specified pools, so our TBA mortgage short position dynamically and automatically hedged a large portion of our specified pool portfolio.

Meanwhile, the mortgage basis came under some stress. In other words, Agency MBS underperformed interest rate swaps, which is not surprising for such a large move in rates. While the mortgage basis widened, the magnitude of the widening was much less than similar sell-offs in the past.

This lack of substantial underperformance makes sense to us for a number of reasons. Since the 2008 financial crisis, the ownership basis in the mortgage market has shifted from highly levered convexity hedges like Fannie Mae, Freddie Mac and certain money managers, to the Federal Reserve.

To date, the Fed has bought and held all its mortgage MBS holdings and even purchased MBS to replace natural run off that has taken away a lot of pressure that the MBS market had to endure in previous episodes. We believe that the Fed's large ownership portion of the MBS universe is in itself supportive because it limits the amount that private participants have to sell at times of stress.

The size of REIT holdings is a good example of this. Currently, REITs own approximately $100 billion less MBS than they did in 2013, and do so with a smaller duration gap and less leverage. Much of that $100 billion may have wound up at the Fed.

In the taper tantrum, it was deleveraging and rebalancing that were among the larger culprits causing mortgage widening, but now that loaded gun simply doesn't exist to the same extent in the current market. However, the Fed has signaled it may allow its MBS portfolio to start shrinking later this year, which has the potential to create a tremendous opportunity for EARN. Prepayment speed slowed down over the quarter as rates have been gradually rising since Q2.

Taking a look at slide 8, Fannie 3.5s of 2014 that had been paying 32% CPR in September, paid 25% CPR in December. That decline became much more dramatic since year-end. As borrowers reacted to the current rate environment, the MBA refined closed out the year at the lowest point since the financial crisis. This is a big positive going forward.

Prepayment risk is much more manageable and prepayment protection is much less expensive now. In the most recent prepay report received early February, prepayments dropped 30% from the prior month. We have always embraced the idea that portfolio composition is supposed to be dynamic over cycles.

The ideal portfolio changes with the times and actively managing our holdings is one of the better tools we have in preserving book value. When prepayments change quickly, we believe that our infrastructure and expertise provide us a distinct edging and monetizing pricing and deficiencies as the market tries to adapt to new information.

We were thus able to thrive in the rate value that started the year and the sell-off that ended it. Even with our vast expertise and understanding of mortgage prepayment behavior, we also have to humbly acknowledge the uncertainty of policy risk. Central banks across the globe were very supportive of mortgage spreads in 2016.

We don't know when, how or even if the support is going to be stripped away, but it's something we constantly acknowledge in our portfolio construction. Policy risk can rise from places other than central banks. The new administration showed on day one that housing policy is on the front burner when it rolled back the recently announced reduction in mortgage insurance premiums.

Policy risk can lead to wider yield spreads for agency RMBS, which could be a challenge to book value in the short term but a benefit in the long term if we could put capital to work at more attractive valuations. With our TBA short positions currently keeping our mortgage basis exposure at modest levels, we have ample ability to add exposure given an opportunity.

2016 was a wild year. Wild is not a price action typically associated with strong performance from MBS. Sure enough, on a duration adjusted basis, agency MBS underperformed treasuries by 10 basis points. The total return for agency MBS in 2016 was only 1.6%.

EARN's compounded economic return, on the other hand, was 8.7% in 2016. We think that this return in this year is proof of concept for our strategy, actively trading MBS with a quantitative analysis of both fundamental and technical factors is an investment approach that can generate great returns without betting on a particular macroeconomic view.

We were able to generate positive economic returns when the market quickly rallied 100 bps earlier in the year. We were able to generate positive economic returns when the market quickly sold off 100 basis points. We are modest in our reliance on our macroeconomic views. And instead utilize our model-driven assets selection process with an aim towards generating strong net interest margin in a variety of interest rate regimes.

This is what served us well in 2016 and what we expect will continue to reward our shareholders in 2017. We think 2017 can be a year of tremendous opportunity. Mortgages may reprice substantially wider in response to actual anticipated changes in the size of the Federal Reserve's mortgage holdings.

Should that happen, look for EARN to take advantage by adding to our mortgage portfolio and reducing our TBA shorts. Currently, our net interest margin is at a healthy level and our repo borrowing costs relative to what we get paid on the floating leg of our swaps are at the best levels in years. Now back to Larry.

Thanks, Mark. Since our IPO in 2013, we have been saying that our objective is to deliver attractive dividend yields over market cycles while mitigating risk, especially interest rate risk, all in the form of an agency-focused mortgage REIT.

This past quarter and this past year were both testaments, not only to our ability to deliver returns but also to the validity of the basic premise of our objective. We are different from other mortgage REITs and proudly so. We seem to use TBA short positions much more heavily than other mortgage REITs to hedge our risk. Why wouldn't we?

Being able to use TBA short positions adds so many extra dimensions to our ability to manage the portfolio, and these are instruments where we have strong convictions that we actually have an edge. On slide 8, look again at the prepayment instability of the non-specified TBA pools versus the specified pools. So when we use TBAs as hedges and interest rates spike like they just did, our TBA short positions not only dynamically extend with our specified pool assets, but they extend more than our specified pool assets.

And each TBA coupon has its own unique hedging characteristics with a very particular portfolio effects, not only on interest rate exposure, but also on volatility exposure, prepayment exposure, and mortgage basis exposure. Our significant use of TBA short positions sets us apart from our mortgage REITs peers and helps drive outperformance in especially volatile quarters.

The mortgage market is notoriously inefficient. For a mortgage manager, being able to use TBAs and our hedging portfolio provides so many more options and so many more inefficiencies to exploit. For us at Ellington Residential, the agency mortgage market offers so many ways to make money without taking undue risk. Given that, why would we want to risk so much on who gets elected?

What's happening behind closed doors at the Fed, and whether large-scale government policy changes will lead to higher or lower interest rates? Furthermore, does anyone have the consistent ability to predict where interest rates are going over multiple cycles? I'm skeptical that there are more than a handful of investors who can do that.

Those who can should be placing most of their wages in different markets. On the other hand, deciding whether to hold low balance Fannie 3.5%s as opposed to low FICO gold 4%s, whether to hedge with TBA Fannie 4%s as opposed to TBA Ginnie 3%s, well, yes, that's something we're confident we are pretty good at. We have been doing that kind of thing for over 30 years and over 20 years at Ellington alone.

2016 certainly proved to be a year full of market surprises. As we move further into 2017, Ellington Residential's objective is to be nimble and ready to adapt to a wide range of market scenarios. While we wait for clarity from the new administration on their policies related to government spending, housing and the financial markets, we could see continued periods of high volatility and interest rates.

Additionally, we should start to see the Federal Reserve's footprint in the TBA market contract further, as prepayments continue to slow in response to higher rates and the Fed receives less principal to reinvest through open market purchases. Of even greater significance, we could also see the Fed finally reduce its MBS holdings by discontinuing its current policy of reinvesting the billions of principal payments it receives each month or perhaps even selling down its portfolio outright.

Any of these policy shifts would have significant consequences for the agency MBS market. The Fed bought $387 billion of mortgage bonds just in the past year alone, in order to maintain its holdings at $1.75 trillion. The Fed currently owns a full 30% of the entire agency MBS market.

With the biggest buyers starting to walk away, we could see a significant widening of spreads. We want to be ready for that scenario. An adaptable strategy is especially important in the uncertain markets, as was evidenced by our strong performance this quarter in comparison to our peers. The prospects for the agency RMBS market remain well-suited to our competitive advantages of diligent and opportunistic asset selection coupled with disciplined and dynamic hedging.

Our goal is to generate sustainable income from a strong net interest margin that not only supports a high dividend but provides room for EARN to grow returns over time in a variety of market conditions. With that, our prepared remarks are concluded. I will now turn the call back to the operator for questions. Operator, please go ahead.

(Operator Instructions)

Doug Harter, Credit Suisse.

This is Josh Bolton filling in for Doug. I have a few questions about your hedging portfolio. I appreciate the comments you made in your prepared remarks. Can you talk a little bit about the timing of the hedges that you put on during the quarter? It seems like they may have been put on largely before the big move in rates after the election. But curious to hear your thoughts about timing of the hedge -- the increase in the hedges? Thanks.

Sure. This is Mark. What we do is every day, and really, during the course of the day, intraday, we are monitoring how much interest rate exposure is in the Company that we are leaving shareholders with. And so that process of monitoring in the real-time interest rate risk and trying to mitigate it in real-time, that's been normal course of business for us since we went public.

So the process wasn't substantially different than what we always do, it was just the magnitude of the changes were bigger and the nimbleness that sort of required on the part of the manager I think was different given the magnitude of the moves. If you look on slide 16, that, you can see how the hedging composition changed over time, right?

You want speak to the timing of the TBA proportion?

Yes, we increased the TBA shorts. We also increased the amount of longer duration treasury shorts. And when we increased the amount of TBAs relative to treasuries, it was our view that when you got this surprise result in the election, it was -- introduced significant enough uncertainty that we thought it was permanently going to raise actual implied levels of volatility in the market. Against that backdrop, we thought it was beneficial to move more of our hedge to the TBA shorts.

That is something we are constantly reevaluating, how much of the mortgage basis do we want to be exposed to? We have been under 20%, I think, at times of our TBA shorts as a percentage of our hedging portfolio. We have probably done close to 50% at times. That's something that we feel is well within our mandate to make judgment calls on in terms of where we think the mortgage basis might be vulnerable. And as Mark said, when we think the market may be underestimating the uncertainties in volatility, following or proceeding an event.

Great, thanks for the color.

Jim Young, West Family Investments.

Congratulations on a great job this quarter with maintaining book or book only being down 1%. That is terrific. My question is, how are you thinking about the leverage at this environment? You ticked up from 8.1 to 8.5 times at the end of December, but as you start to think about how the Fed's involvement with the agency mortgage market, are you decreasing leverage in anticipation of an opportunity? Or how much leverage are you willing to put on if the bonds do widen out like you think they could this year? Thank you.

That's a great question, Jim. This is Mark. So what -- in addition to thinking about how much leverage we have, which is really the amount of our borrowings relative to our equity capital, we think also a lot about our net mortgage exposure. And that is a calculation, where we look at the current face of our mortgage holdings and we subtract from that the current face of our TBA shorts and we compare that to our equity capital.

So if mortgages were to underperform swaps or treasuries substantially, we thought the market had some excessive reactions to the announcements by the Fed, I think you would probably see us do is just buyback a lot of those TBA shorts and put them into either interest rate swaps or treasuries, so then you would see a company that had more mortgage exposure, but wouldn't necessarily be an uptick in our leverage.

It obviously depends on how things move and relative value, but I think that's the most likely reaction we would initially have to a substantial underperformance in mortgages. It wouldn't manifest itself as a material increase in leverage, but it would really change slide 16. It would change the amount of those TBA shorts. And Lisa, I think, wants to add something.

Yes, Jim, if you consider unsettled purchases and sales, the leverage actually didn't go up that much, it went from 8.2 at the end of September to 8.3.

If you look at the earnings release, the last bullet point, actually, in the summary on the first page, we encourage people -- obviously, we disclose it both ways, but we encourage people to look at our leverage after unsettled purchases and sales because if we have a agency specified pool, for example, that is being financed on repo, but we sold it with trade date before quarter end.

But it's going to settle on TBA settlement date usually in the following month, that's going to, as a technical matter, keep our leverage high because that repo still exists as of the end of the quarter. But from a risk perspective, that is coming off -- I mean, that borrowing is coming off when we settle that trade a couple weeks later. So, we show it both ways, and as Lisa said, when you adjust for unsettled purchases and sales both ways, it really was just a very modest pick-up.

Okay, thank you. I recognize this is basically impossible to answer but can you give us your best sense to how much -- how many basis points do you think the bonds could widen out this year? I recognize there is a lot of factors in there, but how much of a widening are you concerned about?

This is Mark. This is obviously just speculation on my part. I think a lot of it has to do with what the Fed says and how aggressively they taper. If they are only going to reinvest half their paydowns each month versus 100%, or if they say they're not going to reinvest any paydowns. So I don't know.

If you look at the taper tantrum, I think the range is, I think at the low end, maybe 8, 10 basis points. At the high-end, maybe something like 30-odd basis points. So then mortgages are a little bit wider than historical standards. The question is, if you own them now and you are capturing a little bit better margin than what you normally capture, is that enough to compensate you for that risk?

It's probably enough to compensate you if the widening comes in at the low end, but it's probably not enough to compensate if the widening comes at the high-end. There was an interesting piece I can send you that Bernanke wrote. It came out a couple weeks ago on what he thinks is sort of the proper size of the Fed investment portfolio right now, and it was interesting because it wasn't that much smaller than the Fed's portfolio is right now.

But, I think that the market can do back flips and work itself up into a frenzy without really knowing much. That is sort of our view now that it made sense to have material and substantial amount of TBA shorts, because I guess we are hoping that you'd get some volatility, and we're hoping that we get a really great point to add more mortgage exposures to the portfolio.

And just to add on that, Mark, just back me up on the calculations, right, so spreads widened 10 basis points, then what do you want to call that? 0.5 point?

Yes, yes.

For arguments sake, let's call it 0.5 point. When I talk about last time, some of our -- we do a lot of premium securities and some of those that have lower durations, but for your typical mortgage out there, that could easily be 0.5 point of price. If they are levered 7 to 1, that could be 3.5% on book value.

Now, if you don't have any TBA shorts against you, then that's 3.5% on book value, right? But if you do, then that immunizes you from those shocks, and given that the Fed is -- their footprint is in the TBA market, right, so one could argue that specified pools would move less -- would do better than TBAs. The other thing that I just want to --

And it supports also your -- then your going-forward NIM is higher, too.

Right, exactly. So -- and the roles come under even more pressure. So that helps you in terms of, again if you're short TBAs is part of your strategy. And I guess finally, I would just say that another mitigating factor is that the Fed -- prepayments are lower, I think I mentioned this in my remarks. So their footprint in the market is already less because their only activity right now is really just replacing the prepayment run off which has gone down a lot.

So if you look at the Fed's purchases to replenish their run-off as a percentage of the total volume of trading in the market, most of which, of course, is not just replenishing run-off, right. You've got money managers getting into the market, getting out of the market so most of the volume is obviously not just replacing run-offs. It's a huge market.

So the Fed's presence is definitely not what it used to be in terms of just because their run-off is less. And so I think pulling that away will have less of an effect. Still very significant but less of an effect than it might have in other environments. Looks like we got one more question, it looks like.

Steve DeLaney, JMP Securities.

I just first want to echo Jim Young's comments. Congratulations. Great job on protecting book value in a very volatile quarter. Repo seems to be something we used to always worry about and it seems like it's sort of another thing that is working well for the mortgage REITs these days. It seems like repo has lagged Fed funds and LIBOR as those other benchmark rates have moved higher.

Just curious if you feel that, that benefit of pricing is going to be sustainable over the next year? And as -- maybe looking out longer term, just curious since you guys have two public vehicles and a lot of private money involved in the mortgage market, has Ellington looked at maybe creating its own broker-dealer so that you would be able to do more direct repo rather than through tri-party repo? Thanks for your comments.

Let me just address that last part in terms of the broker-dealer. It's true. We have two public companies and we have private capital as well, but we're still not big enough to justify creating a broker-dealer just for that purpose.

The other thing I would say is that -- and I know that other people deal with these conflicts, but we are also -- and we and many of our clients and investors are loath to introduce a broker-dealer. And the fact that we have multiple clients, who would own that broker-dealer, would it be the management company? That creates one set of conflicts.

Would it be one of our public companies and not the other? That would set another set of conflicts. So, A, even though we're big, we're not big enough, I would say, to justify that broker-dealer. And I think you'll see that the REITs that have that are much larger, and we also are not crazy about the kind of conflicts that, that might present.

That's helpful, Larry. I understand.

Steve, this is Mark. Thank you for your kind words. That has been a really significant event. We sort of track it where is three-month repo compared to three-month LIBOR? Because three-month LIBOR is what you get paid on the floating leg of a swap.

Sure.

And that swing, since when the fears were most heightened about availability of repo to where it is now, it's probably been a 20 to 25 basis point swing. That means you can just look at mortgages spread to the Treasury curve, or something like that, and you wouldn't realize that you're leveraging these, if you're paying on -- if you are doing them as swaps and repo, you've gotten a 20 basis point benefit from where you had been, so it's really significant.

The other thing which we have observed, which also to us is a significant positive, is that some of the largest US banks, which were not aggressively looking to add repo exposure -- I'm trying to remember maybe a year ago or whenever repo felt like it was a little bit -- there was scarcity associated with it. They have been looking to add.

That to us is significant, right, that you have enormous, trillion dollar balance sheets now aggressively involved in the agency repo market. And it's not the incremental capacity is not coming from smaller non-banks.

Interesting.

Although, we have -- we do continue to see increased interest from foreign banks, I would say. So we've been doing more repo with them. They seem to have a great appetite for the product and the non-bank presence continues to grow. And it's interesting. There have been some US banks that decided, and if you do the math, and it's hard for us to do the math obviously from here but if you do the math based upon Dodd-Frank, it does seem like the return on equity for things like Treasury and agency repo are really low.

But, nevertheless, you've had some banks that have gotten out and haven't really gotten back in, but you have other banks that have decided, we are speculating that it's not really a loss. They are not losing money, it's just not a great return on equity for them compared to maybe some other opportunities.

But I think they feel like -- and I agree with this, to be a presence, a big presence, to have a big presence in the agency mortgage market, which is obviously a big market, this is just something you need to be involved in. You get a lot of information if you do it. You obviously service your customers better, so I think that we are certainly heartened to see those banks staying in the business in some cases and stepping up in others.

That's great. And the foreign players that you mentioned had sort of come in. Are they more Asian or Eurozone?

I would say --

Probably Canada (multiple speakers) --

I was going to say Canada and Asia, I think, are the most notable.

Guys, thank you so much for the comment.

Ladies and gentlemen, this concludes today's conference. Thank you for your participation and have a wonderful day.