Great Ajax Corp (AJX) 2017 Q4 法說會逐字稿

Good afternoon, and welcome to the Great Ajax Corp.

Fourth Quarter 2017 Financial Results Conference Call.

(Operator Instructions) Please note, this event is being recorded.

I would now like to turn the conference over to Lawrence Mendelsohn, CEO.

Please go ahead.

Can you take the music away?

(technical difficulty)

Okay.

Good afternoon, and welcome to the Great Ajax Corp.

Fourth Quarter 2017 Financial Results Conference Call.

(Operator Instructions) Please note, this event is being recorded.

I would now like to turn the conference over to Lawrence Mendelsohn, CEO.

Please go ahead.

Thank you, everybody.

Thank you for joining Great Ajax Corp.'s Fourth Quarter and Year-end 2017 Conference Call.

I want to -- everybody take a quick look at Page 2 with the safe harbor disclosure, and we'll be talking about some forward-looking things.

And then we can quickly jump to Page 3. I'll give you a brief introduction and then I'll get right into it.

We had a good fourth quarter of 2017 and a very important strategic quarter in terms of what we were able to accomplish.

Our loan portfolio continued to perform very, very well.

We had a strong cash flow from monthly payments and prepayments.

We had one of our best quarters for net asset value creation also.

We significantly increased our percentage and absolute dollars of fixed rate nonmark-to-market funding and at very good rates.

We are a credit-sensitive and long-term value-focused business as opposed to being an interest rate and mark-to-market risk taker, and that will -- you'll see more examples of that throughout our call.

Additionally, we grew our joint venture partnership materially in Q4, and we expanded our loan acquisition base as well.

And we're already having a busy first quarter of 2018 even though first quarters generally tend to be a little slower for seasonal reasons.

And with that, I'll jump into Page 3 of our presentation.

It's very important to understand our long-standing relationships to be able to buy loans at the prices we buy them and the kinds of loans that we actually want to buy.

We don't want to be an index fund buying loans everywhere.

We want to own loans in certain specific places with certain characteristics and our sourcing network of long-standing relationships.

We did 12 transactions in Q4 2017.

Almost all of those happened in December.

And in fact, one of them happened on the last business day of the year, 10 minutes before wire cutoff.

We use our managers' proprietary analytics.

A large amount of data -- we analyze a large amount of data to determine the loan characteristics that we want and to forecast performance patterns.

These forecasts and this data analysis isn't just for buying loans.

It also helps us drive our loan servicer, Gregory Funding, with servicing strategies and also with expectations of outcomes for them.

It's really important to understand our affiliated servicer and what it does.

We're very close to our servicer.

It's important for us to be close to our servicer, and we'll talk about in subsequent Events slide that our servicer, we feel, is so important that we now own a part of it.

And we have warrants to acquire more of it if we choose to.

And from a leverage side, our average asset level during the quarter was about 2.7x.

I'll discuss more about this topic because strategically, we made some huge strides in Q4 with regard to our financing.

Now we can talk about Q4 specifically.

We bought a lot of loans.

We bought $220 million, purchase price $241 million of UPB, $363 million of collateral value.

So from a purchase price of UPB, 90%; purchase price of property value, 60%.

All the loans we bought were re-performing loans, we did not buy nonperforming loans in Q4.

We also -- inside of this $241 million, we bought a number of pools jointly in a new joint venture with BlackRock that we briefly talked about in Q3 was in the works.

And it closed in December of Q4 and that's a 50-50 JV with them.

And I'll talk more about that when we get to the securitization structures as well.

We called 2 securitizations from our 2015 issuance, and we issued 3 securitizations.

So all the 5 of these combined occurred in a 3-week period in December.

As a result of calling the 2 securitizations, we accelerated the amortization of approximately $900,000 of deferred issuance costs.

On the 3 newly issued securitizations, we dramatically lowered our funding costs.

This also locks in long-term nonmark-to-market fixed-rate financing.

I know everybody sees rates going up, but as you'll see when I describe the 3 securitizations, you'll see that for us, rates are going down.

Our 2017 B securitization was our first ever rated issuance.

For REIT tax reasons, we only issued one class of bonds.

We have a choice of doing a 63% of UPB AAA or a 70% of UPB AA.

We chose to issue the 70% of UPB AA-rated bonds with a 3.16% fixed-rate coupon at par, no step-up.

That's the coupon and the yield forever.

All of the loans on this structure were at least 12 for 12 in terms of consecutive payments.

Our 2017-C securitization was a single class of unrated bonds.

You may remember our previous 2017 transaction, 2017-A and our 2016 deals were all 64%, 65% of UPB advance rates.

In our 2017-C securitization, we did a 70% of UPB advanced rates, a significant increase over our most recent unrated deal.

We also increased the years and the time to the step-up coupon.

Traditionally, we had done 3-year step-ups.

The market standard is 2-year step-ups.

In this transaction, we extended it to 4-year step-ups.

So as a result, we did -- we increased our advance rate by 5%.

We got 3.75% fixed and we took the bond from 3 years to 4 years before step-up.

So think about that 2017-C deal, 70% of UPB, 3.75% fixed, 4-year step-up.

And then we'll talk about our 2017-D deal, which we did a week later.

We issued our 2017-D unrated bond in connection with our joint venture with BlackRock.

We took the advance rate over the course of 1 week from our 2017-C bond of 70% of UPB to our 2017-D bond of 80% of UPB.

And we also, in that week, went from increasing our step-up years from 3 years to 4 years to our 2017-D deal to eliminating the step-up completely.

And we accomplished both of those changes without any change in rate, 3.75% fixed-rate coupon at par.

So it's a 3.75% unrated, 80% of UPB with no step-up.

So dramatically reducing long-term funding cost in each of these 3 deals.

Additionally, one of the things that really made this 2017 deal -- '17-D deals stand out is that the structure was created with a loan purchase pre-funding account equal to 2/3 of the entire structure for purchases of loans in the future.

So all late December purchases in our joint venture, including approximately $60 million on December 29, the last business day of the year, were purchased directly into the pre-issued 2017-D bond structure, which had proceeds sitting in an escrow account to buy loans that hadn't yet been determined at the time of issuance.

The pre-funding structure locked in the 3.75% fixed-rate, debt cost at 80% of UPB for the life of the loans.

However, because the debt issuance preexisted the loans purchased and many of the loans were purchased at the end of December, we actually had to pay 3 weeks of interest at 3.75% on the bonds issued while the money sat in escrow account earning pretty much nothing.

So we had a negative spread for that week period, for those 3 weeks, on the loans purchased at later dates.

So a pretty remarkable 3-week period, calling 2 securitizations and issuing 3 in each issuance at specifically better terms than any issuances we have done previously.

I want to talk a little bit about the fourth bullet point about portfolio interest income and net interest income.

Since loans are only on the books, we talked about, for 23 days for the quarter, so the average loan was on the books for 23 days because so many loans were bought in December.

And we had, as you can see in the last bullet point, $48 million of cash collections in the fourth quarter, which is a phenomenal number.

The combination of loans only being on -- new loans only being on the books for 23 days and so much cash collections, in the month of October and November, we actually, on average, had less loans on the books than we had 9/30, which then dramatically increased in December.

So net interest income is a slightly lower number, one, because loan performance is so good, it extends duration; two, we had fewer loans on the books for the first 2 months of the quarter; and three, by issuing the pre-funding account in our 2017-D deal, we had interest expense before loans were actually purchased into that structure.

So that works through the portfolio interest income and net interest income.

Now when we talk about income attributable to common stockholders of $6.2 million, $0.34 a share, I want to walk you back to what that actually implies.

So if you think about that is net of $900,000 from the accelerated amortization of the 2 called securitizations, that is expense that would have been taken over a much longer period of time, incurred back in 2015.

But because we called those deals and reduced our funding cost for the loans in those 2 deals, plus another $150 million of loans -- or $200 million of loans in the third deal, by more than 1% on all $300 million of debt, it made a great sense to call those.

It also -- we also were able to get execution and timeliness that was terrific given markets.

This quarter, we had a smaller REO impairment than last quarter, about $500,000 versus $1.1 million in the last quarter.

And you'll notice from our income statement, and we'll go through some of the metrics later in this call, about $300,000 of professional fees related to structuring the tax and accounting sides of our rated deal because of complicated REIT rules with regard to multiple cash flow and classes.

It's one of the reasons why we only did one class of bonds.

As a result of that, you'll see the $900,000 of the accelerated amortization, the $500,000 of REO impairment, the $300,000 of professional fees.

Together, that's about $1.7 million or about $0.09 a share, slightly more than $0.09.

So the $0.34 is approximately $0.43 when you straighten that out.

We actually had some capital gains on actual REO line item in the impairment.

So if we were to take out those gains, it would take you to about $0.42 a share.

This excludes the negative carry in our prefunded debt accounts for our 2017-D structure.

That also probably cost us about $0.01, $0.015 a share in the month of December.

So if you were to add back all the unusual pieces and timing of debt structures versus asset purchases, you would be somewhere around $0.42, $0.43 a share versus the $0.34.

On a taxable basis, we are $0.11.

There's 2 reasons for that: one, much fewer foreclosures in the fourth quarter.

Foreclosures, as you may remember, trigger taxable income at the time they occur, not when you actually receive money.

And number two, taxable income from performing loans and other loans just continued to perform and perform and perform.

Taxable income lags GAAP income because taxable income is based on contractual maturity, and GAAP income is calculated to the expected life.

As you might imagine, when someone gets a 30-year mortgage, the expected life of the loan is shorter than contractual maturity.

And as a result, as long as your re-performing loan portfolio keeps growing, taxable income always will lag GAAP income on performing loans.

We already declared a dividend of $0.30 payable March 30.

It's the same dividend even though taxable income was lower.

We expect taxable income will increase throughout the year, and our board is comfortable with the $0.30 dividend.

Book value at $15.45.

We did the add-on convert in August of 2017.

We had a full quarter of the convert on the balance sheet in the fourth quarter.

By having the August convert and the GAAP calculation of the equity allocation from the convert makes book value $15.45 at December 31.

$53 million of cash on the balance sheet at year-end.

We had plenty of cash and available credit, and we had $48 million of collections in Q4, which obviously helps cash.

But if we annualize this $48 million, that's $192 million over the course of 12 months, which is an enormous amount of cash flow.

It's approximately 17% of our basis in loans when you look at it on a percentage basis.

The portfolio overview on Page 5. You can see that re-performing loans are becoming more and more and more quarter over quarter over quarter of percentage -- higher percentage of our portfolio.

At year-end, they were 96.4% of our portfolio, and nonperforming loans were only 3.6% of our portfolio.

From a property value side, you can see that our principal balance versus the underlying property value is pretty low.

And we'll talk a little bit more of that when we talk about the re-performing loans on the next page.

From REO, you'll see our REO portfolio is about 2%.

Almost all of the REO is held for sale and will turn into cash over a relatively short period of time.

At 12/31, we did have 14 rental properties, which are primarily multiunit properties, although there's a couple of single-family.

Re-performing loans.

We continue to buy very low LTV loans with overall re-performing loan purchase price representing 61.9% of the property value and approximately 82% of the UPB.

Keep in mind that the purchase price to property value does not include any home price appreciation since the acquisition, which would make the price to current property value much lower than the price to acquisition property value.

We continue to play offense and defense at the same time.

That's kind of the mantra that we've been saying for a while.

And if you look at kind of the progression of price to property value based on acquisition, in 2015, at year-end, it was 66%.

At year-end 2016, it was 64.8%.

And at year-end 2017, it was 61.9%.

And if you were to factor in home price appreciation, it would be lower than that number.

So we're very comfortable from a collateral perspective with our portfolio.

On the nonperforming side, we haven't bought a nonperforming loan in a long time.

We didn't buy any in 2017.

Purchased nonperforming loans continued to decline in absolute dollars as well as a percentage of our portfolio.

This is on purpose.

Now there is positive to nonperforming loan prices increasing dramatically.

It's put pressure on NPL buyers to sell quickly if loans re-perform because otherwise, re-performing loans actually will hurt their IRR if they hold on to them for too long.

As a result, we've been able to be a liquidity provider to nonperforming loan buyers who need speed, timeliness and certainty in meeting liquidity requirements and to drive their IRR results, enables us to be very picky and to look at larger pools of loans and only buy subsets because we're timely.

If you look at our 12/31/2017 nonperforming loans, the purchase price to property value for that loan -- that piece of our loan portfolio is about 55%, and the purchase price to UPB is about 66%.

Again, it does not represent current market prices.

Current market prices will probably make that property value higher, so purchase price to property value would be lower.

Page 8. 80% of our portfolio is in our targeted markets.

It's very different than anyone else.

California continues to represent almost 30% of our portfolio; Santa Barbara, south to San Diego is approximately 75% of that 30%, so in the low 20% just in Southern California.

One thing I will mention is we added Houston back in as a target market.

Houston had been a target market up until about 1.5 years ago.

We've made 6 visits before and after Hurricane Harvey as well as looked at a number of assets and analyzed a whole bunch of Houston and nearby data.

Based on this, we've added Houston back to the target market, although we're particularly interested in the small balance commercial side in Houston even more so than the single-family analysis.

We are currently looking at Indianapolis.

We're evaluating it as a target market.

We're in the data analysis period of Indianapolis.

And we, in our portfolio, have noticed certain characteristics of loans in Indianapolis have caused dramatic performance and excessive expectations.

And we're spending a little time analyzing and trying to do a forecast of what characteristics we would look for and whether it's Indianapolis driven or some characteristics of the loans in Indianapolis or both that are really driving that performance.

Portfolio migration.

I cannot say enough about this.

This is really kind of the heart of what we do, which is NAV creation.

First, I'll talk about that -- loans over a -- loans that have made more than 12 consecutive payments to us.

So keep in mind, this is just payments to us, $649 million of loans that made 12 consecutive payments to us.

And if you started 7 for 7 to us, that's $869 million have made 7 payments in a row to us.

This has increased $160 million in the last 6 months.

Even more so is any loan that we haven't owned for 7 months can't be 7 for 7 to us.

So this is even taking a relatively smaller subset of the portfolio that's eligible.

And I'll talk a little bit about that more in a minute.

Second, if we owned -- 12 for 12, if we owned a loan for 12 months, if we haven't owned the loan for 12 months, it can't be 12 for 12.

So any loan bought after January 1, 2017, cannot be 12 for 12 in this chart, okay.

So the $649 million that I just mentioned of 12 for 12 excludes any loan bought in the year 2017, okay.

Third, if we take into account payments made to the prior servicer on loans purchased in 2017, this would increase the 12 for 12 payments or better loans significantly.

12 for 12 loans with our LTVs sell at very, very high prices, even higher now than in Q4.

And Q4 prices were higher than in Q3, okay.

In the last few weeks and throughout 2018, we've seen some 12 for 12 that are large pools sell at sub-4% yield.

And in some cases, we've seen them at close to 3.75% yield.

This is significantly higher than what we paid for a 6 for 6 or a 7 for 7 payment loan.

Our loans, once they hit 7 for 7, get to 12 for 12 a little over 90% of the time.

So 7 for 7 for the loans that we buy and with the characteristics of the loans that we want have a very high continuing payment rate once they cross the 7 for 7 point.

So we would expect, if you include payments to previous servicers and then as 2017 carries on, that this percentage of clean pay loans would keep increasing as it already is.

In addition to the increasing cash flow and as I just discussed, the increasing net asset value that this creates, the significant outperformance of our loans also lowers our asset-based cost of funds over time.

We can see this from our December -- our 3 securitizations we did in December 2017, that the loan cash flow velocity enables us to push advance rates up.

So keep in mind that we push them between our 2017-A securitization, our 2017-D securitization from 65% of the UPB to 80% of the UPB.

And we pushed the step-up time from 3 years to 4 years to never.

And cash flow velocity enables us to do this.

So our cost of funds, and you'll see this -- you won't see it as much in Q4 because all the securitization goes through in December, but you'll really see this in changing and decreasing cost of funds at the asset-based level over time.

The 3 securitizations, remember, are only on our balance sheet for a very short time in Q4.

But they'll be much more apparent, as you see, throughout 2018.

And lastly, and this is the NAV topic.

Subject to REIT tax rules, we may sell some of our 12 for 12s.

And we have plenty of loans with better pay history and 12 for 12s also.

Prices have gone up for these 12 for 12 loans and for 18 for 18s and 24 for 24.

And we are currently evaluating that as well and have room under the tax rules to be a seller of some loans.

On Page 10, it's a complicated page.

Our joint venture -- our 2017-D joint venture with BlackRock.

Because it's 50-50, we have to consolidate it.

So we've actually put a middle column in, which deconsolidates it so that it only shows, versus the balance sheet, what we actually own and get out of it, along with all the other metrics in combination.

So there's a couple things I want to mention.

First, under average loans, average loans was almost unchanged in Q4 from the previous quarter.

But the year-end number was -- in absolute dollars was much higher, and that's because so many loans closed in December that the average understates the end of period.

You'll see that average loans is only about $20 million higher, but year-end loans are considerably higher as a result.

If you look at average loan yield, going from 9.6 to 9.2, that's really performance-driven.

The material increase in loan performance continues to extend duration, but it's dramatically increasing cash flow.

We captured far more discount than expected when we bought the loans, but we captured it over a longer period because the loans keep paying.

However, when we compare these yields to where 12 for 12 clean pay loans and larger pools transact to sub-4% yields, it's great for NAV.

So performance, performance, performance, we love it.

If you look at total interest expense, there's a bunch of different factors.

One, we had a full quarter of the convertible add-on from August.

So we had 3 months of it versus 2 months of it in Q4 versus Q3.

But loans, as you saw, from average loans, didn't really increase until December.

And with the pre-funded structure, we paid interest for 3 weeks on a lot of funded bond money that was sitting in an escrow account waiting to buy loans at the end of December.

So we had a negative spread on 2/3 of that issuance in Q4.

Next, asset-level debt cost.

You'll see it came down a little bit, even though everybody else's is probably increasing with LIBOR.

Our asset-level debt came down, but that doesn't really reflect the true heart of it, which is the 3 securitizations, because they were all done in December.

It only reflects a few weeks of that over the course of the quarter as opposed to the entire quarter.

And if we look at the noninterest operating expenses, $5.3 million growing to about $5.7 million.

The difference in that is primarily the increase in professional fees principally related to our rated transaction, our 2017-B transaction, which wouldn't be replicated now that we have that structure all set up and tied out and to be duplicated.

The next page.

We'll talk about what's happened since January 1, and the answer is a lot.

We bought $18.5 million of performing loans, all performing loans in the month of January.

We paid 89.5%.

You'll see the price to property value of 53.8%.

It just keeps getting lower.

We also bought a 32-unit apartment building in inner city Phoenix, for about $100,000 a unit, the first of those in our portfolio.

In February, as usual, in the month of February, didn't buy much, but we're making up for that in the month of March.

When you see pending acquisitions were agreed upon, about $91 million in UPB, high purchase price because the underlying loans have approximately 6% interest rate.

We expect that this will also be added into our 50-50 JV with BlackRock.

We also have $2.7 million of small balance commercial loans expecting to close in the month of March.

We're also -- simultaneous with this, and I don't have information as of 45 minutes ago yet, we're negotiating on 2 other significant pools that would also be part of the 50-50 joint venture with BlackRock that would be early to mid-April transactions.

And hopefully, I'll have more information on that in the next 20 minutes to 24 hours since we've been haggling for about 3 weeks on these transactions.

Very important, January 26, we closed on our acquisition of interest in the servicer, Gregory Funding.

It's a 2-step transaction for regulatory reasons.

We're buying an 8% interest, and we're also getting 3 sets of warrants, totaling another 12%.

The 8% interest and the warrants are split into 2 pieces: a 4.9% interest that is already closed; and the 3.1% interest that closes 121 days later.

So 121 days from January 26, we'll close on the other 3.1%.

It's just subject to the passage of time.

One thing that's important, it's all new capital to the servicer.

There are no selling shareholders in the transaction.

I mentioned earlier in the call, we already declared our $0.30 dividend.

It's payable on March 30 to stockholders of record of March 15, 2018.

We're keeping the dividend constant.

We expect taxable income to increase over the course of 2018 from fourth quarter taxable income, and we felt comfortable keeping the dividend at $0.30.

With that, thanks for listening to me.

And we are happy to answer any questions you might have.

(Operator Instructions) The first question comes from Tim Hayes with B. Riley FBR.

My first one, you had a lot of loans closed at the end of the quarter.

But have they all boarded in and started earning interest at this point?

They have all boarded.

The last pool closed December 29, and so we're earning interest on all of the loans that we bought in Q4, the last of which we closed on the 29th.

So we're earning interest from that day forward.

Got it, okay.

And is your expectation last quarter that you could achieve mid-teens ROE if you were to execute on your securitizations, which you did?

Is it still your expectation, especially now that you've deployed capital and have a bunch of loans of that boarded at the end of last quarter and then now in this quarter so far that you can still achieve those type of returns?

So let me give you an idea of simplest way to look at it.

It's not exact but it's simple.

If we think, for example, our 2017-D structure, which was 80% of UPB at 3.75% fixed forever, okay, the loans that went into that transaction, the $200 million or so of loans were bought at approximately at 90% of UPB dollar price.

So you received senior financing of 80% of 100%, which is 88% of the purchase price.

So you've got 8/9 leverage at 3.75% fixed forever.

So what they -- you bought those loans to a 7% yield gross minus your cost of funds of 3.75%, okay, with 8x leverage, so you lever up the spread by 8x, and then you add the yield, it's a very high number.

Now that's before deal expenses, so the 3.75% all in is probably a little over 4%.

And it's before the servicing costs.

So from that leverage structure, it's pretty easy to get mid-teens, if not higher than that.

If we look at our 2017-B deal, okay, which was our rated securitization, where we did 70% of UPB at 3.16% fixed at par forever, what it shows is that when our loans that we buy in the 80s become 12 for 12, we can get 70% of 100 or approximately 4:1 leverage on those loans because we get -- on our purchase price in the 80s, we get 70 in the mid-80s, okay.

You'll get about 4:1 leverage, maybe 4.5:1 leverage at 3.16% forever.

So when you start saying, okay, we bought those loans when they were 6 for 6, we're probably earning on portfolio Y over 9% on them.

And we put those into a securitization with a 3.16% coupon once they're 12 for 12 and you get 4.5x leverage.

All of a sudden, it becomes a pretty high number.

So the cost of funds and the advance rate in our structure are really important.

That's why I mentioned in the beginning, strategically, this is one of the most important quarters we've ever had because we got our first rated deal done and we put -- we're able to push the bar on our unrated securitizations pretty dramatically from our 2017-A deal of 65% advance rate with a 3-year step up to our '17-C deal of 70% with a 4-year step-up to our '17-D deal of 80% with no step-up ever.

So as result, from the funding side and from the strategic kind of asset-level debt side, it was a terrific and strategically important quarter.

Yes, understood.

That being said, we did 3 -- we called 2 securitizations and issued 3 in the 3-week period.

And I think, it took a couple weeks for people to get their sleep patterns back after that.

Well, I hope you guys are on a better schedule now.

Yes.

And so to your comments, like this is a very active quarter for you guys on the securitization front, obviously, but can you just talk about your appetite or pipeline for future securitizations at this point given kind of the execution you believe you can get today?

Yes.

Our 2016-A securitization was the most expensive one we've ever done, and that becomes eligible for call in the -- right after the spring, so in April.

And our expectation is we will probably call that deal early on in its 2-year -- post its 2-year call period and add in other collateral and do another securitization.

We also expect that we'll do either an additional unrated securitization or an additional rated securitization, my guess, between now and approximately June 30.

The next question comes from Steve Delaney with JMP Securities.

Congrats in all the transactions, Larry.

Great quarter.

I wanted to -- you surprised me a little bit with the apartment building in Phoenix.

Can you let us -- give us some color how much you paid for that?

Will you lever it?

And what type of cash on cash return are you going to be looking for there?

Sure.

We paid $3.4 million for it for 32 units in inner city Phoenix.

It's fully rented and in a location that we think is changing over a 5- or 10-year period.

So think of it as, in New York, say, Bed-Stuy, 10 years ago, and that's the way I would think about it.

We like that particular market on the small multifamily.

For us, 32 units is actually a big property.

So that would probably be as big a Phoenix property as we'd ever buy.

We are looking for other inner-city properties in 5 or 6 markets, Phoenix being one of them, Atlanta being one, parts of Los Angeles, although we've not seen anything in Los Angeles where the prices make sense.

But we like certain markets there.

We've been pretty active in risk financing other acquirers in that market.

The -- I wouldn't say that we're going to dramatically drive the apartment side, but we do think that we bought this to a current cash flow in the 7s, and we'll be able -- and we paid cash but we have the availability to finance it on a credit facility if we choose.

But given that we have $50-something million of cash lying around, there's no reason to finance it yet.

Understood, understood.

And do have a plan for some CapEx that would allow you to roll some rents up going forward?

Like raising them with some upgrade?

Yes.

We do have that, and we're also looking to acquire some other properties in the same neighborhood.

And ideally, we would do that sooner rather than later so that we can actually have some scale in the CapEx itself.

Okay, great.

And just one final one for me.

The Gregory investment, I think we all agree that that's important strategically.

How should we think about any income -- the income recognition going forward on that investment?

Will that be equity method?

Or how will you account forward?

And will there be any cash flow coming on that as well as maybe a GAAP income recognition?

Sure.

So Mary says it will be the equity method, okay.

She's sitting next to me.

She says it will be the equity method.

The servicer does, from time to time, pay some dividends, although there -- I would not expect it to be material to Great Ajax from a cash perspective for future investing.

But the servicer paid 2 dividends in 2017, and I would expect it would likely do similar, that basically are like a 3% or 4% dividend yield.

But the real benefit, the servicer is able to drive its scale as well as its brand.

There's a lot of the -- as you can see from our joint ventures, there's a lot of parties who think the servicer does a really good job compared to other servicers.

And as a result, we think there's material upside to it, and we also think that there's great benefits to Great Ajax getting in early.

We have some pretty sophisticated outside directors who negotiated this.

And when JV -- potential JV partners are calling yourself, trying to get Great Ajax to invest, they quickly figured out that if you're going to make an investment to a servicer, you make the investment before you give it permission to service for others, not after.

So as part of making the investment, the outside directors have permitted Gregory to service for pure third parties so long as it doesn't have a negative impact on what it does for Great Ajax itself.

The next question comes from Scott Valentin with Compass Point.

Just with regard to the ATM, I noticed you guys issued some shares under the ATM and just wondering what your philosophy is there in terms of when you decide to issue shares.

How does the price of issuance affect your thoughts on issuance?

Sure.

So the issuance was done at about $14.60.

Almost all of it was to one buyer.

And the Board of Directors gave us a specific cap where the follow-on convertible that we did in August was about $20 million, and they gave us the authority to do a maximum of up to 1/3 so it would make the debt ratio 3:1 on the follow-on.

And that was the bulk of the logic.

So there was a specific very low cap, and probably wouldn't have done it if it wouldn't have been one reverse inquiry buyer for most of it.

Okay.

That's very helpful.

And then I think you were going to discuss leverage, obviously, with the 80% advance rate in some of the ABS transaction.

Is there room to take leverage up to from where it is today?

Or do you feel comfortable leaving it where it is, around 3, 3x?

Yes, 3 is a good number, although our board is comfortable with a little bit more leverage on the asset base side for good performing loans.

So they wouldn't want to be over-levered with mark-to-market repo in a rising-rate environment where LIBOR is your enemy as opposed to your friend.

But in a fixed rate, nonmark-to-market, match funded structure, our board is comfortable with a bit more leverage on the good performing loans in our portfolio.

That being said, it's not going to be 4x leverage.

It's going to go -- I would anticipate leverage going up a bit on the asset level side, not corporately.

Okay, great.

And another question, just the change in the tax laws, both, I guess, just thinking from a corporate structure perspective as well as maybe the asset class you guys invested in the housing market.

Are you seeing just any changes?

Are you seeing any changes?

Well, the tax law is complicated.

I'm not sure that anybody knows 105% of the rules, including the IRS, at this point yet in terms of how it's going to be interpreted.

Obviously, our board, as a matter of fiduciary responsibility, wants us to look at about -- I hate to say it, at about 24 different scenarios.

We thought it was 8. Then they had 3 scenarios for each of the 8. So it's now 24 scenarios of both REIT and non-REIT for the same scenario to see what the difference in tax rate means to total value creation as a structure.

And that will take a while to figure out, and the board has given us certain target dates to give them information and output.

And that -- so I'm not saying that we don't expect to be a REIT.

What I'm saying is the board feels that they have fiduciary responsibility to at least examine the difference between the new tax law and the REIT structure as a board.

And then think about what all of that information means.

It's a complicated tax law.

The -- and I think that over the next 6 months to a year, there's going to be a lot of interpretation of it that's not necessarily based on the language in the tax law but based on some rules that come out, to opine on it.

And I don't think -- until we see some of that, it will be very hard to make a decision as to what it actually means from a structural perspective.

Okay.

And then just with regard to the state local tax change, any views on high-cost housing markets, particularly in Southern California?

Do you see any impact there on housing at all?

Not so much in the properties we are involved in.

So if you were to take markets -- that 80% of our loans are in 10 markets, basically.

If you were to take those markets and split housing into deciles 1 through 10, 10 being the most expensive and 1 being the cheapest, we are probably -- 90-plus percent of our assets are probably in deciles 4 through 7 or 4 through 7.25, with the biggest fall probably deciles 5, 6, 6.5.

So the -- especially on the property tax side, it's not as big an issue for our portfolio than if, for example, our portfolio was all -- California has relatively low property tax, but if our portfolio was all New Jersey, New York or Illinois where -- Chicago with high-property tax, we'd be a little bit more nervous, especially on the higher end.

But for our property kind of characteristic and decile levels, location by location by location, we don't think it would be particularly impactful.

It's one of the reasons why we stay away from the very high decile properties, deciles 8, 9 and 10, especially 9 and 10.

They're very hard to comp.

A lot of it has to do with a matter of taste, and a lot of it has to do with who the next buyer is, not who the last buyer was.

And as a result, it's hard to figure out.

And we tend to stay away from it because of its lack of predictability and lack of liquidity.

And in many locations, the new tax laws are -- make those much more expensive to own.

(Operator Instructions) Okay.

Seeing no further questions in the queue, this concludes our question-and-answer session.

I would like to turn the conference back over to Larry Mendelsohn for any closing remarks.

Thank you, everybody.

I appreciate all the good questions and listening to me go on and on for the last 50 minutes.

Thanks for joining us on our fourth quarter, year-end 2017 conference call.

And we're seemingly always around, and if you have other questions that come up over the next few days, weeks, feel free to give us a call.

We're happy to help any way we can.

And thanks again and thanks for being on 6:00 Eastern Time.

Okay, thank you.

The conference has now concluded.

Thank you for attending today's presentation.

You may now disconnect.