Great Ajax Corp (AJX) 2018 Q1 法說會逐字稿

Good afternoon, and welcome to the Great Ajax Corp.'s First Quarter 2018 Financial Results Conference Call. (Operator Instructions) Please note, this event is being recorded.

With that, I'd like to turn the conference over to Larry Mendelsohn. Please go ahead.

Thank you very much. Thank you for joining us on our first quarter 2018 quarter end call. I'd like to -- before we get started, have you take a quick look at Page 2, our safe harbor disclosure and a discussion of forward-looking statements with regard to this presentation and any Q&A. And with that, I'll jump right in.

As first quarters go, this was pretty typical of a first quarter in January and February. And then March, it got very busy. Usually, first quarters are 1 of the 2 slower quarters of the year, usually quarter 1 and quarter 3 are slower and quarter 4 and quarter 2 are busier. And January and February followed that trend, and March picked up dramatically as we'll -- even more so than usual as we'll discuss later on in the earnings discussion and the subsequent events. But in March, we agreed to buy a lot of loans and get a lot of due diligence for closings in April and also for closings this week and next week.

With that, I'll jump to Page 3 and give you a quick overview of our business and any changes. One of the most important things is where we get loans, how we find them, our sourcing network is extremely important to our ability to acquire the types of loans we want and at the prices we pay relative to other people. We frequently owe liquidity providers to funds and to banks who have balance sheet issues or ratio management to do, and you'll see from what we've acquired and what's coming through in April, May and June, though we're in May and June now, at the prices and the loan-to-values and the types of loans.

Our manager's proprietary analytics, also very important. We spend a lot of time analyzing data to determine the target loan characteristics we want. We don't want to be an index fund of loans based on where loans exist. We only want to own loans that we actually want to own, and we use a lot of this data analysis to forecast patterns of what's going to happen in loans and what's going to happen in properties in certain markets and as well as demographics and rents and things like that. Not only does the data analysis help drive the loan acquisition strategy, but it's really helpful as well in driving loan servicing strategies and predictions of outcomes loan by loan by loan. Makes us a lot more efficient. It also gives us the ability to buy loans that, based on pattern recognition that we've built, other people don't necessarily see or understand.

Having a captive affiliated servicer is really important. Our servicer and our closest to our servicer, it really helps create outcomes in loans that other people don't get. As of the end of January, we now own an interest in our servicer, and we also have warrants to acquire more of our servicer. And as our JV structures get bigger and bigger, which they've done in April, and we'll also get -- be in June, it also -- we get an extra benefit through increasing servicer valuation as well and the brand of the servicer that the institutional joint venture partners rely on. It's very important, and it's good for the overall value of the servicer. It's good for both our interest in the servicer as well as our warrants in the servicer.

We use moderate non-mark-to-market leverage. You'll see as we go through this that our ending leverage at the end of Q1 is actually lower than the end of Q4 and lower than the average leverage throughout Q1, even though we are pretty busy. But that's something that we'll talk a little bit more about.

If we jump to Page 4 and the highlights. First, in the quarter, we bought $17.5 million of re-performing loans. We only bought re-performing loans. We didn't buy any nonperforming loans in Q1. Our purchase price was approximately 89% of the unpaid principal balance, but it was only 54% of the underlying property value. So very low relative to property value, good average property values. We also bought an apartment building in Phoenix, a 32-unit apartment building for about $100,000 per unit in an area of Phoenix that we believe has a significant upside and in the urban location.

From a revenue interest and -- net interest income, there's a lot to talk about today. First, interest income of about $25.5 million. Net interest income about $13.1 million and $0.41 a share of earnings. Some important factors to talk about when we talk about interest income and net interest income. Net interest income increased some. We had more loans on average for the quarter, but the yields on our loans were down approximately 30 basis points. And this is going to be an odd thing -- they're down 30 basis points due to over-performance. It's an unusual concept and a trade-off. Monthly performance on the loans is significantly better than our expectation, and that causes yield to decrease.

We received significantly more cash flow, but duration extends because far fewer loans than expected are defaulting. So we have the usual problem of too many loans are paying that continues, and you'll see from the collections and the way leverage changed in this quarter that the performance level is even significantly better than even 3 months ago.

We received so much more cash flow, that our Q1 quarter end leverage is lower than our quarter average leverage during Q1 and lower than our Q4 2017 ending leverage. In addition, our cost of funds declined as well, despite an increase in LIBOR. And we'll talk a little bit more as we see another securitization structure that we did in April as to what this over-performance of loans really enables us to do on the financing side.

Also, in these numbers is an REO impairment of approximately $400,000, although that's offset a little bit by about $500,000 of REO sales gains. We talked before on some calls that REO impairment happens first under GAAP, is when you get the REO at foreclosure, you have to make that determination. And sales gains happen later when you sell REOs. So sales gains always have a lag versus taking the impairment.

Our REO portfolio overall continues to have expected future gains on a net basis, net of any existing or future impairments that we're aware off. So -- but remember, gains always come second, and REO impairment always comes first.

Other highlights. Taxable income significantly higher than last quarter, $0.37 a share. Interest income as well as payoffs as well as tax gain from loan modifications. So not GAAP gain from loan modifications, but tax gain from loan modifications. The GAAP income comes from those over time. The book value of $15.53, cash collections of $50.5 million in a quarter. $23 million of that is payoffs. If you were to annualize the collections, it'd be about $200 million or a little more than $200 million, which is approximately 18% of the entire basis of our loans, so it's really a remarkable amount of collected money.

And to give you even kind of more -- a bigger sign of that, over a 3-day period, March 30, the last business day of the quarter as well as April 2 and April 3, the first 2 business days of the second quarter, on those 3 business days, we had combined payouts of approximately 1/2 of 1% of our entire portfolio in 3 business days. So it's looking like Q2 payoffs and Q2 cash flow is following on from the velocity of Q1.

At the end of the first quarter, we had $47.5 million of cash, and we had, on average, $51.5 million of cash throughout the quarter. It's pretty remarkable that our leverage actually went down during the quarter, but our cash amount stayed the same because our collections were so high. And having an average of $51.5 million of cash during the quarter. While it's great for future acquisitions and capital availability, it does hurt your interest income number having that much cash versus having less cash.

Quarter end leverage ratios, asset level, 2.8, corporate level, 3.1. Those are both down from the end of Q4. Leverage declined during the quarter, but our cash balance is unchanged. Having so much cash decreased the income a bit but it also -- all the cash flow decreased our leverage, while at the same time, we still maintain the same amount of cash. It was kind of a unusual set of overlapping items.

January 26, we talked about this as a subsequent event in our call a few months back. January 26, we closed the first step of a 2-step transaction acquiring an 8% interest in our loan servicer. We acquired 4.9% in late January, another 3.1% will close at the end of May. We also acquired warrants on additional 12% of the servicer so we can capitalize on the optionality of value created at the servicer and that the brand that the servicer is creating for itself.

Our portfolio on Page 5, we're still about 96.5% re-performing loans and about 3.5%, nonperforming loans. But that will change a little bit as we'll see later in the presentation in later Q2. On the REO side. REO for us is principally held for sale, and we'll turn it to cash over a relatively short period of time. At 3/31, we did have 15 rentals, primarily multiunit buildings, including the 32-unit multifamily in Phoenix that we purchased in January of this year. But for the most part, REO is held for sale, and we expect it to turn into cash and reinvest it.

On Page 6, when we look at our re-performing loan portfolio, it continues to grow. We continue to buy lower loan-to-value loans within the overall RPL purchase. Our purchase price represents 62% of property value and approximately 83% of principal balance. This price of property value doesn't include any appreciation that happened since acquisitions, so our belief is that our prices actually significantly below the 62% of property value.

We continue to play offense and defense at the same time in the world we live in as well as typically financial markets defense is harder than offense. And we found that a methodology to play defense and offense at the same time, and it works well. And what it allows us to do on re-performing loans is, from a performance level, if a loan doesn't pay, it actually can make our yields go up rather than down. Although we would always like our loans to pay.

If you look at the progression of our re-performing loan portfolio, down in the bottom graph, you'll see March 31, '18, our purchase price of property value is 62%, versus a year ago, it was 65%; versus 2 years ago, it was 66%; versus 3 years ago, it was 68.5%. So our purchase price to REO has come down -- our purchase price to property value has come down significantly, and that's without actually including home price appreciation. That's just straight -- the way we buy loans and the kinds of loans we source.

On the NPL side. Our NPLs have continued to decline in absolute dollars. However, as part of a joint venture, we'll talk about in a few minutes, we expect to acquire some additional NPLs in June of 2018. During the quarter, we had 45 foreclosures, of which 18 of the 45 properties in the foreclosure sales sold directly to third parties for cash. That gets accounted for through loan pool accounting because it never actually turns into a property we own. So it's as if it's a type of loan payoff. But 27 of the properties became REO. And we also sold 27 REOs in this quarter. So on a net basis, we have the same number of REOs that we had the quarter before from that perspective.

Where are our loans? Well, California continues to represent approximately 30% of our overall portfolio, and Southern California is about 75% of the 30% or about 22% of our entire portfolio.

We're seeing very consistent payment and performance patterns for loans in California, especially in California urban centers like San Francisco, Los Angeles, parts of Orange County and San Diego County. And we find that loans with certain characteristics in California are very predictable into what they will do.

Probably the only item that's potentially different in the future is we continue to evaluate Indianapolis. We've been looking at it for a few months now. We started looking at it late in Q4, and we're still on the data gathering, and we've made a number of visits there, feet on the ground to determine whether or not it should become one of our target markets.

On Page 9 of portfolio migration. This -- the numbers here are unbelievable. I don't have another word that I can describe it with. So $955 million UPB of our portfolio is 12 of the last 12 payments or better. That's up approximately $300 million from Q4 of 2017, okay? In addition, $1.06 billion is 7 of 7 or better. And the significance of 7 of 7 is the analysis of data on loans that we buy.

So going back to early on, where we've been buying loans subsequent to now -- or all the way up until now, the loans we buy have specific characteristics that are data-driven. And what we find is that, once the loan hits that cusp point of 7 consecutive payments, that loan, 91% of the time, gets to be 12 of 12. And so when you think about it, we have $955 million plus another 12 of 12 or better, plus another $108 million or $109 million of 7 for 7, which have a 91% chance of becoming 12 or 12 or better.

So our expectation is that we're going to wake up in 6 months and other than loans going in because of payouts, we're going to have close to $1.1 billion, of our portfolio being 12 of 12 or better. And over $600 million being 24, 24 better, really remarkable in terms of performance. That goes back to when we bought these loans, we knew the 7 of 7 cusp, but getting to 7 of 7 is the hardest part. And what we have found is that so many loans have more -- have well over-performed our expectation that we're getting to the point where every loan or the lion's share of loans are just paying every month.

In addition to increasing the cash flow and NAV, this payment pattern, the significant outperformance of these loans because of the payment pattern, lowers our asset-based cost of funds over time. We've seen from the 3 securitizations we did in December of 2017 that our -- the cash flow velocity on our loans enable us -- enables us to push up the advance rates and at lower cost than in previous years.

And on the next slide, you'd be able to see that our total average debt cost declined 20 basis points in Q1 versus Q4, despite the fact that there was a significant increase in LIBOR. So we look at some metrics on Page 10.

And on Page 10, what we've done is, you may recall we did a significant joint venture in late 2017 with BlackRock that was a 50-50 joint venture, which requires us to consolidate it. So what we've done in Q4 would have been the first quarter of consolidation. So what we've done is we have shown Q2 and Q3, and then we're showing the deconsolidated Q4 and the deconsolidated Q1. To tie them back to the consolidated, we actually put in an extra page on the next page to show both consolidated and deconsolidated.

But if you look at the deconsolidated, so you can see a rolling 4-month period, you'll see the average loan yield has gone down from about 9.4 to 8.6 over a year, primarily because of performance, okay? Material increase in loan performance continues to extend duration, but it's dramatically increasing cash flow and that cash flow is over a longer period of time. When comparing yields in the market, though, to -- for larger pools of 12 of 12 loans, and this is great for NAV creation because yields in the market are significantly lower than these yields for 12 to 12.

Also, one thing to keep in mind is that our yield calculations don't permit us to calculate principal above 100 LTV. So if we have loans that are 100 LTV or higher, and they're paying monthly payments of principal, we're not getting income on that in our model until they go below 100 LTV because we can't take income for principal we didn't expect to collect. So a well-performing high LTV loan actually hurts yield versus helps yield. It's a very unusual GAAP pool accounting piece that -- where we comply with, and it's part of the modeling.

If you look at asset level debt cost, it's come down as well. And if you look at total average debt cost on our entire balance sheet, it's also come down from Q4 as well as from Q3. The -- if you look at the noninterest operating expenses, you'll see those are up a small amount. That's primarily driven by 2 things. One is interim servicing fees to the previous servicer for loans in our 2017-D transaction. That joint venture that we closed in the last 2 weeks of December, we had to pay the previous servicer interim servicing fees, and that shows up in January, the -- and there was a large number of loans.

And number two is property valuation expense on loans in our credit facility. So in our credit facilities, plural, at our JPMorgan and Nomura facilities, require updated property values every so often, and all those occurred in 1 quarter.

Quarter ending leverage -- if we go to the bottom. Quarter ending leverage is 2.8x. And at the corporate level, including our convertible bond, is 3.1x. We're running a 10.5% return-on-average equity with a 2.8x leverage, which I think is not typical in the mortgage REIT world, that's for sure.

Page 11 is the reconciliation that we're required to do since we deconsolidated on Page 10, and we reconsolidated on Page 11. So that everybody can see the tie out. And then I'll jump to Page 12 because there's a lot going on. On April 26, we closed another joint venture. Our 2018-A bond structure. It included $160 million of UPB and of which, 80% were senior bonds, are about $128 million, and the balance was a equity certificate. One thing that made this a very unusual structure is that, at the time of closing on April 26, 90% of the structure was set up as a pre-funding account. So only about 10% of the money was spent on the day of closing to buy loans, and the remaining 90% can be used to buy loans until the 10th of June.

The senior bond was done at a 3.85% coupon at par, with no interest rates step-up during the life of the senior bond at all. So similar to our 2017-D, with no step-up over the life and an 80% senior -- the 2017-D, we pushed the envelope by having about 40% pre-funding account. The 2018-A, we push the envelope by having a 90% pre-funding account. You can see, on the left-hand side of this page, the loans that we expect are going to go into that pre-funding account. You'll see the first is what closed on the 26th into the structure.

And then the next 2 batches, the $138 million of RPL and $5.8 million of NPL, are expected to close during the month of May or early June into this structure, totaling about $160 million. If you look at the purchase prices to the collateral value of the RPLs, 61.8 and 57.1. And then at the NPLs, 39.6, you'll really see that we're continuing our pattern of low LTV loans and low purchase price to property value. The big difference, though, is that people and bond investors have become so comfortable with the way we do due diligence and the way we find loans and the way we value loans that they're willing to put -- set up a 90% pre-funding account for loans to be identified to go into the pool.

In addition to that, we have a new joint venture that we expect to close in the second week of June, that'll be our 2018-B structure. And that, as of right now, will have about $124 million of nonperforming loans. The purchase price to collateral on those is about 63% of collateral. The average property value in those nonperforming loans is about $300,000, so the average discount to property value of the purchase price is about $111,000 of, what I'll call, comfort zone or protection in the purchase price.

Separate from the JVs, we have about $18 million of re-performing loans that we expect to close in the next few weeks. Again, if you look at the price to collateral, it's about 60% of the underlying property value, again low price to property value. We continue to play defense and offense at the same time. On April 26, that's the joint venture. We just talked about 2018-A.

We also, as part of our press release today, our board declared a dividend of $0.30, payable to stockholders of record on May 15 on May 30 payday. Taxable income was $0.37 in excess of the $0.30 where our board is certainly comfortable with the $0.30 dividend. And as we see the next few quarters, they'll take a look as to whether that dividend should be improved based on continued payment performance of the loans as well as loan modifications in any fee income we might have.

On Page 13 and 14 are the statements of income and balance sheets. And rather than go through those in detail, I will be happy to open it up for questions.

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

Can you just talk about your kind of the -- how the strategy differs between all the new JVs you've entered into and how you view executing a new JV versus on balance sheet investment? And then just how you plan to allocate investment between all these vehicles?

Sure. All right. Depending on the JV itself and what loans are identified, our JV partners -- some have different yield goals or different performance goals, or they're looking for loans with different pay histories than we might be. And in some cases, we'll take a much smaller share. In some cases, we'll take a larger share of a JV. One of the reasons why it was so important for us to make an investment in the servicer and have warrants, even more importantly than just the investment but have warrants on the servicer, it's because these JVs also can have a material impact on the evaluation of the servicer itself. And because clearly, from the JVs, these institutional investors have see our servicer as a real performance brand in servicing land. And in fact, the servicer has delivered those results to them. So the JVs provide us the ability to see more loans. I mean, in some cases, some of these institutions will see a portfolio that we don't see, and they'll bring it to us so that we can do the analysis for them to help them understand what we think will happen with the loans. So it helps us see loans that we don't see before. It helps us have the servicer, have more access to servicing loans for these institutions, and we and all these JVs get to choose how much equity we wanted to put into each one. It's not preset. For example, in the 2018-A transaction, we have 2 institutional investors, name brands in that transaction, and we only retained the 9.5% interest. And in the December 1, we retained a 50% equity interest. And in the June JV, the 2018-B, our expectation is we'll be about 20% of that transaction. So it kind of gives us a benefit to have -- to decide how much we want to do ourselves. We're the operating partner. We're the ones with the rights to call the structures or to sell loans out of the structures after 2 years, and we're also the ones who do the due diligence and negotiate the prices for everybody. And we also have our affiliated and captive servicer who gets to manage these loans on a daily basis for us. So we get to know a lot about them. It also -- it gives us the ability to have a very cheap financing, with 80% senior bonds, with low coupons and no step-ups for loans that we want to acquire. So low [is a]cost of funds gives us more optionality as to what we want to buy versus less of, and it also increases the value of the servicer that we own 8% and have warrants on another 12% at the same time. So we kind of -- it kind of helps us in 3 different ways. And in REIT land, everybody thinks that you're always going to raise capital. And what we find is that we can create significant multiple value on other pieces here. We own 20% of the manager, so it gets value. We own upside on 20% of the servicer, so it gets value, plus we get to pick and choose when we -- when and how much of our own balance sheet we want to use.

Makes a lot of sense. And your -- on the Subsequent Events page, you showed that you're acquiring some NPLs, and you talked about on the call a little bit. And clearly, it looks like it's going mostly to the JV, but do you see Ajax balance sheet taking on more NPLs? Are you seeing better relative value there? Or is it really just more in line with the strategy that your partners are looking for?

No, these NPLs -- the one thing we like about these NPLs that were different than most of the NPLs we see is: one, these NPLs were pretty far along in the process; and two, they had much higher average property value than we're used to seeing at NPL pools. We're used to seeing, in NPL pools, properties that are worth under about $150,000, and they tend to be similar to what you see in HUD pools. And these pools are much more Fannie Mae and jumbo-type loans. So better property values and further along in the process. So with our purchase price, we're able to buy them at a price where we have about $111,000 in discount. And NPL land is really kind of a fixed-cost resolution business. So you want as much absolute dollar margin as you can have, not percentage margin. So 63% of a $300,000 property is a lot cheaper than 50% on an $80,000 property or $100,000 property. So we like the price to the property value, and we like the -- how far along they are in the process. We can decide how much we want in this pool up to 50%, and my guess is that we'll choose to be 20% or 25%. Just because I think that more loans will ultimately end up going into this pool to the extent that we identify them prior to mid-July because we'll set this up with a pre-funding structure also most likely. And then if you see the nonjoint venture pending acquisitions, those are all RPLs low purchase price relative to property value, with a coupon in the mid-5s and high-percentage California. So those RPLs really fit our existing portfolio almost in -- as if you were just pasting it on top.

Right, right. And then one more for me. You talked about on your last call potentially calling that 2016-A securitization. I'm just wondering if you're still thinking about doing that. And then if you do, just kind of how much cash you think you could potentially pull out and the type of expense savings you could see there?

Sure. If we were to call that, we would pull out, probably, about $30 million to $40 million. And we -- if we were to re-securitize it with some other loans, we could probably lower the direct funding cost on those assets by about 0.75%. So it wouldn't just be -- it wouldn't be 0.75% on the $30 million or $40 million. The $30 million or $40 million would be net cash [we buy]. It would be on the [face] amount at the bonds rate.

The next question comes from Stephen Laws with Raymond James.

Kind of following up a little bit on Tim's questions with regards to the loan pipeline as well as your comments at the beginning of the call on seasonality. Can you maybe talk about the pipeline, what you're seeing? I appreciate the color you gave on the NPLs with regards to absolute dollars as opposed to percentage of property values a second ago. But it looks like, for the quarter, you closed about roughly the amount you talked about in early March as having done in January and February. Some of the events here really point more towards agreements in place to close loans here this month. So can you talk about the pipeline and kind of how we should think about the portfolio building or investments being closed as we move through the remainder of the year?

Sure, sure. Second quarter and fourth quarter tend to be the 2 busy quarters. You usually start seeing the Q2 loans in mid-March. A lot of banks have June 30 regulatory capital dates. And we find in -- a lot of funds like to get some liquidity at June 30 as well for whatever reasons they have. The -- there's others that [need] June 30 under contract dates. So second quarter closings tend to roll into, at the latest, mid-July, maybe by default the 15th or 20th of July. And then it gets slow again August and September for closings. The -- I would expect that this 123 plus 18, so 141 will -- the 18 should close in the next couple of weeks. The 123, our expectation is that will be a second week of June closing. And that we'll probably set up a pre-funding account to buy some other loans that we identified between now and the second week of June, that would close sometime between the second week of June and the 15th or 20th of July. We were very close to adding on about $100 million of re-performing loans and -- but couldn't come to an agreement as of 2 hours ago. So I don't know what will happen with those, but we, at any point in time, we're probably looking at somewhere between 6 and 10 different loan pools that are anywhere between $0.5 million and $100 and some million. So I would anticipate that there'll be a number of pools in addition to these that we'll identify during the month of May that would close in the second half of June or early July.

Great. That's helpful. And one smaller question to the other income line. It looks like that doubled sequentially higher than the run rate. Is this a new level that's appropriate? Or can you talk about what went into that $1.5 million of other income this quarter?

Yes, there's -- let me just pull up the income statement real quick. So apples-to-apples. So one thing that happened in another income, as we have a number of fees paid based on loan modifications, which was a pretty material amount. Let me look as I have some notes written exactly on that. We have some -- about 500,000 of REO gains. But keep in mind, the real estate operating expense has $400,000 of REO impairment. So the net REO gain is $100,000, so you kind of have to net those 2 numbers. But you should see continued other income of REO gains become pretty regular because we're getting to a point now where we're selling a lot of REO each quarter, and the impairment happens before. The other big thing is that, in the last quarter, you had deferred issuance cost of $900,000 from calling a securitization, which we didn't have. So that was something in the last quarter. So you didn't have a negative -- you didn't have $900,000 of expense related to that. The other thing is HAMP fees, so the Home Affordable Modification Program. The -- when you finally get borrowers to have paid 12 consecutive months on a loan mod, you're entitled to some fees. And we were -- you can see from the performance of our portfolio that we would expect to get more fees in the future, also. And we don't -- in our yield models, we don't have any fee income. So that's all just gravy.

The next question comes from Scott Valentin with Compass Point.

Just with regard to 2 things. One, on the overall margin, Larry, point out cost to fund is coming down as you finance high-cost debt. Yields are down because of duration extension, so that means cash [yields] are up, but yields are down. Just wondering if going forward, how to think about you getting excess cash flow. Does that -- more like it's reinvest in the business, obviously, but is there room now to maybe -- is there a potential for a buyback of stock? You're trading below book value, does it make sense to do that, given the JVs now or another avenue of kind of funding [them] as we wait upon the capital markets to grow? Just wondering how you think about the whole capital management strategy in light of the JVs that provide the capital flexibility and the higher cash flow you're seeing off the portfolio.

Sure. It kind of -- first thought would be, from our board's perspective, is to find additional assets that we think are cheap and have good long-term value. I think that a number of our large shareholders tell us that they would prefer us not to really buy back stock because it would decrease the market cap and they would actually like the market cap to be a little bit bigger. And they think if we were a little bit bigger in market cap, that it would trade at a better value. I can't prove whether that's true or not true, but that's -- we clearly heard that from a number of our larger shareholders. To the extent the taxable income keeps -- stays up at these levels, obviously, we'd have to increase the dividend relative to that taxable income. The other thing that our board, given that they think we're going to have significant cash flow coming off the loans and available debt capacity on the structured credit side in the securitized bond market, is they have us actually looking at are there other platforms that they think are on a cheap basis, also, where we could look at those as asset acquisitions as well rather than just individual loans. And so they have a -- I feel bad for Mary and her group because -- Mary Doyle, our CFO, is here because they have her running dozens of scenarios based on different assumptions for different kinds of investments. And so they're working hard. The -- There's a lot of interesting opportunities out there, both on the individual loan side, the property side, the bridge financing side, in commercial, the redevelopment side for providing financing and commercial as well as there's some asset-based platforms that we could look at as well, where we understand the underlying assets. So it's not about -- I don't want you to think we're looking to be an operating company, but we do see that there's places that we can buy assets, not just from our current sources.

Okay, fair enough. And then just -- you brought up the taxable earnings and for a while there, I think for the last 4Q and 3Q, taxable earnings were below GAAP, and now it's kind of flat. Just wondering

(technical difficulty)

is GAAP tax timing, I understand that. I was wondering how you think of taxable EPS going forward? It's one of the concerns that people look at are the optics of having taxable EPS below your dividend. Just again the trends you see in taxable EPS going forward.

Sure. Well, given the amount of cash flow we're having, taxable EPS mathematically has to go up over time. Just from the payoffs, think about every time there's a payoff, the discount becomes taxable EPS, the purchase discount becomes taxable EPS. The -- every time you foreclose on a loan, that foreclosure becomes a taxable event. One of the reasons why our taxable income was going down was because we have less and less NPL, so we have less and less foreclosures. I would expect that, as part of buying $125 million of nonperforming loans, that you'll have a little more early taxable income generation from those throughout the second half of '18 and certainly 2019. What you see from NPL is taxable income tends to peak about 9 to about 21 months after the NPLs themselves are acquired. These NPLs are a little later stage than typical NPLs, so maybe that's 6 to 18 months rather than 9 to 21 months. But time will tell on that. But just the sheer velocity of cash flow coming up our loans, at the end of the day, if you stop growing, when you have no loans left, tax and GAAP have to be the same. So tax has to grow considerably. That being said, if loans pay for 30 years and never miss a payment, taxes to contractual maturity and GAAP is to expected life. If you didn't expect them all to go to 30 years, you would have a lot of taxable income at the end and GAAP income a little earlier. But I don't think all of our loans are going to 30 years.

Okay. And then one final question. On the last call, I think, just having the amount of work that Mary has to do. You guys were reviewing 24 different scenarios under the new tax law, just figure out. Just wondering the progress on that, that process?

Yes, and the good news is the board meeting took those -- the board took those scenarios, and they only gave her about 7 more or 8 more to do. So that's progress. But I think that the board will make a decision, I'm expecting, at the July board meeting. Theoretically, if they were to decide to ever not be a REIT, they would make that decision -- they'd have to make the decision prior to filing the tax return as opposed to on the day they decide. That being said, a lot depends on some of the different scenarios they've asked to run. The -- And the -- having a kind of a defensive ability to the extent the tax law gets dramatically changed a couple years from now, which isn't a 0 probability. And you need to have in place the ability to re-REIT if you ever were to de-REIT. And that's pretty complicated and don't know how easy that would be to figure out between now and July. That's -- when you get into the complicated tax world, it requires spending a lot of time with our tax people, Deloitte as well as mixing these different scenarios as to -- if the tax law changed, what would be the changes that would be really bad, and what would be the changes that would be okay, all right? And just protecting yourself against any of that. So I think different board members have different opinions. I don't think there's a -- at this point, there's 100% unanimity on either side.

(Operator Instructions) And our next question comes from Lazar Nikolic with JPL Advisors.

I'm an investor in Ajax. And is every call on the -- several consecutive conference calls, I don't know several quarters ago, you went for this exercise where you tried to estimate the NAV of the company based on observable trading prices of subordinate bonds that -- without your -- comparable to your net assets. And if I remember correctly, you came up with a value of $18-plus per share. So given this number, I'm a little bit puzzled as to why you had issued shares at $1,400 per share in Q4 2017 to a single buyer. On a call, you kind of made it seem like, ah, it wasn't a big deal. We did it. We didn't have to do it. I mean, this is more than 20% discount to your own NAV estimates, and it's from a shareholder standpoint, it's really troubling if we keep getting diluted at [such a huge amount]. So I just want to hear your thoughts on that.

Yes. Yes, we issued about $5 million at about $14.60 in Q4, almost sold that to one buyer in the reverse inquiry. The board was -- wanted to do it because they look at it as we did an add-on to the convert at the time they did the convert, they wanted to do about 1/3 of the add-on, also in equity, but so long as the combined exercise prices are convert, and the equity issued was approximately $15.35, which was the then book value. So they wanted the mix to be approximately booked of where the convert was issued because the convert was issued at a premium in August, and the $5 million of equity. Absent that, they would not have issued the $5 million.

I see. So going forward, you don't foresee issuing equity anywhere near the current valuations, right?

No. No, if we did, I would be the buyer.

Yes. But I mean, diluting the shareholder pool at large discounts is always a bad idea. Another question I had was...

I do, although -- look, hang on one second. I agree with that, although I would say one thing, is that it does depend what you do with the money. So for example, if someone said -- if we saw an opportunity that we thought we could make 40% a year on for 3 year -- 4 years, we might be willing to issue a little bit of equity to make that acquisition. But if someone said we saw an opportunity that we thought we were going to make 8% for 4 years, we wouldn't issue equity.

Right but even in a situation that you could make 40%...

Invested 40% unlevered. It would -- issuing a little bit of equity would actually be accretive to earnings, not necessarily dilutive. Even if you were diluting book value, it would be accretive to earnings and ultimately, the book. Just like if you issued equity at '18 and you invested it in cash, I would say it would actually be dilutive.

Right. No, I understand that. But I mean, there is [recyclable] capital. Couldn't you sell the subordinate bonds in crystallization dollars per share NAV, and then we deploy that in 40% returns, you could commission a new equity.

Absolutely. Yes, but I 100% agree with that. Yes.

Okay. All right. I hope there won't be future dilutions of this kind of magnitude.

I'm a large holder too, so...

At this time, this will conclude the question-and-answer session for today. [So I'll] turn the conference back over to Larry Mendelsohn for any closing remarks.

Thank you, everybody, for joining us on our conference call for -- to go through first quarter of 2018. Second quarter is already very busy, and we're looking forward to early August when we have our conference call for second quarter ending June 30. And we're around if anybody has particular questions, we're happy to answer.

The conference is now concluded. Thank you for attending today's presentation. You may now disconnect your lines.