Great Ajax Corp (AJX) 2016 Q3 法說會逐字稿

Good day, and welcome to the Great Ajax Corp third quarter 2016 financial results conference call. (Operator Instructions). Please note, this event is being recorded. I would now like to turn the conference over to Mr. Larry Mendelsohn, CEO. Please go ahead.

Thank you very much. Thank you for joining Great Ajax's third quarter 2016 call. I want to introduce Russell Schaub, our President, and Mary Doyle, our CFO, who are also here with me. I also would like to have you take a quick look at page 2, the safe harbor disclosure with regard to this presentation and the call.

I want to give you a quick overview, as we always do, of the business we have here at Great Ajax. We use a high-quality, very active and diverse sourcing network to acquire loans and privately negotiated transactions. Over 90% of what we acquire are private negotiations, nobody else but the seller and us. We've done 160 transactions since our inception. We did 14 transactions in Q3 of 2016.

It's very important to understand how reliable and necessary our sourcing network is to acquire the types of loans that we want to acquire and in the markets that we want to acquire them in and just as importantly the prices we pay relative to others for the loans we buy.

We use our manager's proprietary analytics. We price loans loan by loan by loan. We negotiate with sellers loan by loan by loan. It's very different. We're not auction buyers. We're not trying to be an index fund of loans all over. We spend a lot of time and effort, and we do a lot of data analysis trying to figure out what loan qualities we want to own and recognizing what performance patterns are based on which loan qualities.

A typical deal is $5 million to $20 million, up to about 100 loans, although we have done significantly larger ones, as you'll see. And it's also important to understand that we have an affiliated servicer. They service only for us and entities that we're invested with that integrates the servicing with our analysis and acquisition to determine patterns and the qualities in loans that we want. We manage assets loan by loan by loan. We buy assets loan by loan by loan, really important to understand that.

We tend not to be giant users of leverage. We know there's plenty of mortgage-related REITs out there with significant leverage. We've slowly been increasing our leverage. At quarter end, we were 1.9 times. At June 30, we were 1.6 times. When we did our quick little follow-on equity offering in June, we talked about how, despite the equity, our leverage ratio had actually increased. And it has from 1.6 to 1.9. And we would expect that to continue increasing throughout the fourth quarter as well.

We use -- as part of that moderate leverage, we tend to use securitizations. We've done seven securitizations totaling $825 million of UPB, getting about 3.1 times leverage overall from just the senior bonds.

In August, we did our seventh securitization, Ajax Mortgage Loan Trust 2016-B. It was $82 million of seniors, about 64% of UPB senior bond. We did this one, the first of its kind in a very long time. We put in a prefunding account for acquisitions that we made within 30 days after the issuance of those bonds in early August.

While we paid interest during that period for bonds issued, it locked in fixed rate nonrecourse financing in advance of those acquisitions. And it was good to get that done. We'll talk about the securitization market later as -- in our subsequent events as well.

One thing I want to talk about is that, since early 2015, we purposely focused on buying lower LTV loans with underlying properties in what I'll call housing deciles 4.5 to 7.5. And that's -- we talk about it in deciles rather than in price range because, in a city like New York or in Brooklyn, decile 5 is different price range than a city like Dallas, where decile 5 is a lower number. So, we talk about it in deciles of specific locations.

And also, with absolute minimum dollars of equity thresholds, we've -- it's clear that performance isn't just based on LTV. In fact, it's -- even a higher reliable factor is the absolute dollars of equity related to loans. Our analytics suggest performance repayment are also tied to this absolute dollars of equity.

As you might amount, the larger absolute dollars of equity are also a hedge, a natural hedge, against housing price declines. They're a hedge against slower economic conditions. And they're also a hedge against recidivism in reperforming loans.

Additionally, nearly 50% of all the loans we own are either arms or separate coupons and, as a result, provide some built-in interest rate risk protection. We've all seen the bond market come off recently. And we've seen distortions in LIBOR. The distortion in LIBOR has actually had an interesting phenomena as we've seen an increase in prepayment in LIBOR-based loans as a result of it. We're able to build our portfolio based on all these specifications because of the way we source loans.

I want to get into the results of the third quarter. It was a -- from accomplishing things, it was a great quarter. It was a little aggravating on a couple of items that I'll talk about, but as quarters go, we have nothing but good problems.

We accomplished quite a bit. We're seeing more loans than ever before. We just keep on seeing more and more. And we're even -- we get the luxury of even being choosier than we've ever been able to be.

We acquired $260 million in UPB. And as part of that, we also added $100 million out of that $260 million to our joint venture with DoubleLine Capital.

We invested $216 million in reperformers, entirely reperformers, in Q3, aggregate UPB of $259 million. At the end of the quarter, we had about $950 million UPB of loans. Purchase price in the quarter, 83% of UPB, and low 60% on property values. Our interest income was about $18.7 million, net interest income $11.8 million, and earnings to common shareholders $7.6 million.

I want to talk about a couple of things that were in the quarter, want to mention they're a little frustrating. You'll see from looking at our financials that we show $100,000 of transaction expense. In Q2, we had $400,000 of transaction expense related to a transaction that closed in Q3. And in Q3, we had $400,000 transaction expense that closed on September 30 of Q3, so again, another quarter where we had $400,000 of transaction expense related to an acquisition for which we received no interest income.

That September 30 acquisition was supposed to close the first week of September, but the bank seller had some difficulties getting all their documents together, caused a delay. As a result, we have the transaction expense related to that.

Now, one thing that offset that $400,000 of transaction expense is we did have some transaction expense recovery that we recaptured as part of increasing our DoubleLine JV from some of our acquisitions.

But, as a result, you have transaction costs with no related or corresponding interest income, about $400,000. And you also don't have the interest income from a $70 million acquisition that occurred the last day of the quarter.

Another thing that I want to talk about that's also in this quarter is what I'll call, for lack of better term, REO impairment. We have about 130 different REO on our balance sheet, represents about $20 million in book value. GAAP says that -- I'll give you an example.

Let's say that we have 20 REO that we think we will take a loss on, and we have 110 REO that we think we'll have a gain on. GAAP says that you take the loss on the ones you expect to have a loss on today, and you take the gain when it happens. So, as a result, REO that you have a built-in gain on does not offset REO where you have a built-in loss, and you take the loss as an impairment. We took an impairment of REO.

If you think back to our acquisitions in Q4 of 2014, we acquired about $70 million of nonperforming loans, REO peaks between 18 and 30 months generally after an acquisition like that. We're in approximately month 20 right now. So, REO will be in its peak period for the last two months and for probably another 7 to 10 months. But, again, it's GAAP. GAAP just requires the impairment to be today as opposed to being offset from built-in gain.

We do believe we have built-in gain in the balance of our REO. But, on a quarterly basis, as our REO is built up, it's just the way GAAP works. But, as a result, it's a timing difference, it could be as much as $0.03 or $0.04 a share in this quarter.

Two other things in the quarter that we should talk about. One is it's the first quarter after our equity offering where we had the full amount of shares. So, we had all 2.5 million additional shares from our late June equity offering. And we also had average cash over $50 million during the quarter. Part of that was cash that we had in the beginning of the quarter. Part of it was from the securitization we did in August that did a significant amount of refinancing. And also, transitioning loans over to DoubleLine gave us a significant amount of cash as well, much of which we used to pay down our credit facilities with the ability to reborrow it, some of which we've reborrowed in Q4 already.

Taxable income $0.18. RPLs continued to perform. And the more they perform, it defers taxable income. We've also accelerated our growth rate of our portfolio, which also defers taxable income at a -- to a slightly lower percentage of the total portfolio.

Book value's $14.99. We also raised about $81 million in net secured borrowings through securitization primarily. We increased our DoubleLine JV $78 million of cost, about a little under $100 million of UPB. Most of the proceeds went to pay down one of our repurchase facilities.

We actually paid down one of our repurchase facilities $20 million more than we were required just so that we wouldn't have to pay the interest so that, if you look at cash at quarter end of $23 million, that actually is -- could've been in the 40s if we had decided not to pay down the additional principle on one of our repurchase facilities.

We continue to have cash and credit availability. We continue to have significant joint venture capacity as well. One thing I also want to mention is that collections on loans in REO was approximately $28 million in the quarter, which also is far in excess of accretion as well as it provides cash in excess of financing pay down.

If we look at our portfolio on page 5 of the presentation, you'll see it's now 91% reperforming, only 9% nonperforming and REO combined. The -- as I mentioned before, since early 2015, we've been very focused on lower loan-to-value reperforming loans. And those have continued to reperform.

In fact, on page 6, you can really see it. In the first nine months of 2016, reperforming loans have increased by $200 million net. But, the underlying collateral as of the acquisition date -- so, ignoring home price appreciation, the underlying collateral increased by $319 million versus only $200 million of RPL. So, the net increase in reperforming loans at an accelerating rate has still only been at 63% of property value, continues the theme for buying lower LTV RPLs. It's something we kind of look at as we get to play offense and defense at the same time because of our sourcing network.

On the NPL side, you can see the continued decline as a percentage of our portfolio. NPLs remaining on the balance sheet are at approximately 59% of the underlying property value, again using property values as of the acquisition date, so ignoring any home price appreciation.

On page 8, our target markets are the same as in Q2. A couple things I'll say, California still represents nearly 30% of our overall portfolio. Southern California is 75% of that 30%. We have a number of pending acquisitions in the fourth quarter that will actually increase this California percentage portfolio wise by a few percent overall.

And if -- the other market I want to bring up is our RPL portfolio in New York. Our largest concentration there is in Brooklyn. It's increased a little. However, as a percentage of our overall portfolio, New York's a little smaller than it was in Q2. We acquired $70 million of loans, of reperforming loans, very low LTV, on September 30th. And in that $70 million, no loans were in New York. And that will have the effect of decreasing the percentage in New York in our overall portfolio somewhat significantly.

Page 9 is my favorite slide. It really gives me a feel for kind of the -- what we do here in terms of being very value based. We updated all these numbers to our 9/30 portfolio of loans. And if we think about net asset value as forecasted by the fixed income market again, build these tables.

The other thing I like to look at is what this means for return on equity, that if we took our whole portfolio and we did our latest securitization -- I'll talk a little bit more about that -- 2016-C, which we did 65% of UPB senior bond advanced rate at 4% at par. If we -- and we created two subordinate bonds of 5% of UPB each.

If we did that exact same transaction in our portfolio, we would end up owning for a cash equity basis of $16.5 million, a $237 million face residual. We'd effectively own 25% of the UPB of our portfolio for about $0.069. And that would be 90% reperforming loans that are outperforming our performance expectations. So, when I look back and say building net asset value, these numbers really speak out to me.

And if you look at the bottom, that's our August securitization, which was done at 64% of UPB. You can see the bond market is giving us additional leverage just since August. If we wanted to finance our B1 and B2 bonds -- we've kept them all, but if we wanted to finance them, we can. We've financed to date a small percentage of them.

And in fact, as part of the securitization, you'll see on the next page we paid down about $14 million or $15 million of financing on our B1 bonds that had previously been out there. And so, while the securitization picked up leverage, we actually delevered and created new subordinate bonds that we could finance if we chose to.

Subsequent events, I think the easiest way to say this, it's been really busy here. In October, we acquired -- in four transactions, we acquired just under $76 million in UPB with about a 5.25% coupon of reperforming loans. The price was a little higher. These were mostly cleaner pay loans, of which a large sum, a large percentage were already serviced by our captive servicer Gregory Funding.

As you might imagine, we know a lot about these loans. And in fact, a biggest chunk closed a week ago Monday. And we've had $1 million worth of prepayment in the five days that we've owned those loans.

We acquired a small amount of nonperforming loans just yesterday in fact, small UPB, $1.75 million at a -- purchase price was about 47% of the underlying collateral value.

On the pending side, we have an agreement to buy in seven different transactions about $93 million UPB. And you can see very low LTV, $135 million collateral value. The purchase price is 82% on UPB, but only 56% on collateral value. They're 100% RPLs. We've continued our low-LTV RPL acquisitions. In this case, over 40% of the underlying loans are in the State of California in these pending acquisitions.

We also completed last week our eighth securitization. I talked about our seventh securitization that we did in Q2. We did our eighth securitization, closed it last week. We called our 2014-A and B securitizations, put them into a new trust, and also added in about $12 million of small balance commercial mortgages and issued new bonds. Those bonds are issued at our best execution so far, 65% of UPB, 4% coupon priced at par. It was about 4.5 times oversubscribed.

It's done wonders for our cost of funding. As our leverage has increased the last few quarters, our cost of funds has actually gone down rather than up with the related leverage. This securitization, we ended up -- the refinancing provided about 15 points of free cash, which we used to pay down some of our bond -- our B1 bond refinancing.

The other thing is, as part of calling the 2014-A and B, there's about $600,000 of remaining deferred issuance cost. And that will be taken as expense in Q4. But, the success of the securitization and the reception we got from bond buyers more than offsets the $600,000 of deferred issuance cost that will accelerate, rather than amortizing over the next 12 months.

With that, I'm happy to take any questions anybody might have.

(Operator Instructions). Steve Delaney, JMP Securities.

Thank you. Larry, good to be on my first Great Ajax earnings call. Thanks for taking the question.

Welcome aboard. Glad to have you.

Thank you. So, I wanted to just ask about the DoubleLine joint venture. Obviously, it's a high-profile, wonderful strategic partner to have. But, when you were putting that together, because you just -- you basically just sold them I guess $78 million of loans or sold into Ajax E Master Trust. What was kind of the -- what was the thinking there about sort of running that separate -- I guess I would call it like a separate account alongside the core Great Ajax portfolio. Can you just talk about it and sort of how you saw the synergy in running that pool of loans with 95% of their capital and 5% your capital and how you saw that as a benefit to AJX shareholders?

Sure. There's a number of different perspectives. One is kind of the simplest way to think about it is they've been very involved with us on the bond-buying side for about a year and a half now and in fact have come in for 100% of each of our last five securitizations with orders.

The -- from a loan-buying side, they came and spent a great amount of time here looking at Gregory Funding, the servicer. And they really believe that nobody cares as much as the people whose money it is. And having Gregory here and only servicing for us really gives them a perspective that's different than what they see elsewhere.

As a result of that, they approached us about buying loans together probably late last year. And we finally decided to do it with them in late first quarter on a portfolio also of about $100 million.

They have a little bit different way of looking at loans. They're a little less focused on location and LTV than we are. They're a little less focused on absolute dollar amounts of equity. They like kind of payment patterns a little bit more. And we -- but, we have a common theme that we like to play offense and defense at the same time. So, it's -- if you buy something and you're wrong, you need to make a certain amount of money. And if you buy something and you're right, you need to do better than that. And so, it can band your returns.

They have a little lower threshold than we do for returns, being a big fixed income mutual fund as opposed to a public REIT with return on equity targets. They do everything unlevered. We obviously use some amount of leverage.

So, as a result, we have -- I won't go as far as to say we have an exact set of specs. But, we have kind of like a decision tree with our Board of Directors that we go through determining whether something is suitable solely for Great Ajax or whether they would prefer Great Ajax to be a joint venture partner.

And so, in some of the acquisition we made in Q3, it fit the DoubleLine JV profile a little bit more specifically than it fit just the Great Ajax profile. We Great Ajax would be happy to -- would be willing to own it. But, we think that other things we buy also we might prefer to own in their entirety more.

So, as a result, you saw that the pool of loans we bought on September 30, that was $70 million that we bought that ourselves. And that was specifically negotiated loan by loan with specific target locations and specific loans. That was $70 million that we went loan by loan out of $180 million pool that a bank delivered to us and asked us to meet certain specifications and that they needed.

And so, for when we go loan by loan by loan by loan by loan, it's -- the pool generally meets what we're specifically looking for in target percentages as opposed to being a little bit broader and in more locations. So, DoubleLine's a little less location sensitive than we are. We like to have 80% to 85% of our loans, and if you look at our map, basically 8 or 10 places.

Right, that's very helpful. And then along the same --

The other thing I would say about it is it gives us the ability to see more loans in locations that we wouldn't necessarily be looking for. And as a result, more sellers come to us because of it as well.

Exactly. Even if it's -- so, just being involved in more -- looking at more collateral, talking to more potential sellers has got to be a good thing. And --

That's exactly right. And from a brand perspective, as you might imagine, given the joint venture, they spend a lot of time looking at us operationally, and they like it.

Right, and I think that's a credit to the platform that you have there that investors should take note of. Related and this along the same theme, you had the opportunity to buy some loans in some trusts that were being serviced by Gregory.

Your RPL portfolios, you've generally owned it about 75% of UPB. But, these are obviously very -- cleaner loans in terms of pay history. So, you paid up. Do we get the -- when you look at those, are you still modeling out a similar type of ROE? And do you get there through a lower loss assumption? I'm just trying to understand the economics of that relative to your normal purchase.

So, the answer is yes and yes. These loans have a 5.25% coupon, which is something you don't see in modern America, 5.25% weighted average coupon, especially when they paid for somewhere between 24 and 42 consecutive months.

The -- we know everything about these loans. We could tell you in many cases the name of the children of the borrowers from conversations we've had with them over the years. Gregory has serviced these loans. And they have a great history with them.

In fact, we've owned them now for -- I think today is the eighth day we've owned them. And in the first five days we owned those loans, we had $1 million of prepayment. They're lower LTV, lots of equity, good prepayment perspectives, and in good locations. And we're very familiar with the loans.

So, the other thing is we thought about two other pieces that go with it. One is a substantial number of these loans are probably worth par or over par. And one of the things we thought about is -- and we'll work with Deloitte, our tax advisors, about whether there's a way to carve out a subset and resell a small portion of them.

The other thing is we're -- we've started preliminary discussions with rating agencies about doing a rated transaction, which would be longer-duration financing and with no step-up rates over time and provide more leverage. And that would provide even more ROE than we might've modeled.

Interesting. Well, thank you for the comments, Larry.

Sure.

(Operator Instructions). Kevin Barker, Piper Jaffray.

Good evening. I just wanted to follow up on the acquisitions that you announced at the end of the second quarter. It appears that some of them you didn't close all the UPB. Was that -- was there some runoff in those portfolios before they closed?

So, some of what we -- we actually acquired most of it. Some of it subsequently was moved into the DoubleLine transaction.

Okay. So, it's just a change between the DoubleLine and --

Right. And then in -- there was a transaction that we talked about that was supposed to close beginning of September that we said was about $90 million. That ended up being about $78 million. And it closed on September 30. And when I mean September 30, I mean 10 minutes before wire cutoff on September 30.

Okay. And then when we think about the transaction -- the $400,000 transaction expense that you mentioned, the -- and I'm assuming that's coming through other income, which had a negative $220,000, right? And so, historically, you would run -- sorry, go ahead. I'll let you go, Larry.

So, if you see loan transaction expense, it shows $100,000 on the expense side.

Yes.

So, that includes the $400,000 offset by about $300,000 of recapture from DoubleLine.

Okay. All right. And then okay. And then when we look about -- then the -- then what happened with the other income where you had a negative $220,000?

So, that's the REO I referred to. So, if you think of our REO portfolio, and if you -- we bought $70 million of nonperformers in the fourth quarter of 2014. And REO tends to peak about 18 to 30 months after an acquisition like that. And we are -- basically, 18 months was June. So, we're in that peak REO period. So, we have about 130 REO, call it book value $20 million. And if we said we thought actual market value of the REO was $23 million, but there's 20 that are minus $600,000 combined, GAAP says you take that $600,000 loss today because it's impaired, without regard to the fact that the rest of your portfolio you think you have built-in gain.

Okay.

Just the nature of GAAP. So, when you think about that, that costs probably $0.03, $0.035 a share just because you don't get to have goods offset bads in marking REO to -- and we don't mark to market. So, you impair ones that you think you'll sell at a loss to your basis, but you don't get to offset that with the ones you think you'll have a gain on.

Okay. And then so, when I think about this quarter, you've got the $0.03 per share from the impairment on the REO. You've got $0.03 to $0.04 from the transaction expense in the third quarter that flows --

And no associated income --

Right.

-- from those transaction expenses.

And then your average balance on the portfolio obviously be higher going into the fourth quarter.

That's right.

Right. So, on normalized basis, you're probably -- yes, you're adding back maybe $0.06 to $0.07, depending on how you want to look at it. And then obviously, the average interest income will be higher. How big was that portfolio on September 30th that closed? Sorry.

It was about $78 million. It was supposed to close the first week of September. It was day by day excruciating dealing with the seller.

And then remind us again your limit on the leverage. I believe it's 2.25. Is that right, or are you going up to 2.5?

Well, we don't have an absolute limit. But, our Board has given us some guidance. Our Board has given us some guidance that, given how our portfolio's evolved to be significantly more reperforming, that they're comfortable going to 2.25. And then we'll see it from there. And from our securitization that we just did and from the acquisitions we've already made in this quarter plus pending acquisitions, I would anticipate our leverage would creep up from the 1.9 times.

Okay. And then regarding the REO, do you expect any other impairments in the next few quarters, or do you think that will -- ?

I would think that REO on a quarterly basis will have some amount of impairment. Now, to the extent that we have gains or losses from actual liquidations, that will offset them to some degree. But, in any particular quarter, while we have this peak REO, I would expect that we would receive REO back, some which we expect will have significant gains and some which we think would have some losses. And unfortunately, losses come first from a GAAP perspective.

And I think the other point on REO is, since we're reaching the peak foreclosure process and the REO coming out of the pool, we just have a large glut that's really been coming out over the last two quarters, which is driving the impairments.

That's right. So, this is not a forever phenomenon. In fact, the more our portfolio becomes reperforming, the less of an item it becomes. But, it also -- on the flipside, because we made a significant REO invest -- or I'm sorry, a nonperforming loan investment in one quarter, which was Q4 of 2014, and REO will peak 18 to 30 months from Q4 of 2014, we're in that kind of 12-month peak period. We're about a third of the way through that 12-month peak period.

Okay. And then finally, on the October and the pending acquisitions, the price to UPB is relatively higher than past acquisitions. It seems like a little bit of a trend that you're setting a higher price per UPB.

Higher price per UPB, lower price to property value.

Yes, lower risk value.

And lower price to property value, and more arms and step-rate loans. Now, the other thing you'll see if you look at page 10 is that you'll also see, in the pending acquisitions, of the $92 million, that's a price UPB of 82%, 56% of collateral value, which is a very low collateral value price, and the price UPB similar to what it was last quarter in the -- what we acquired in October, that price is higher at 90% because most of the loans we bought were clean pay for 24 months with a 5.25% coupon.

Okay. And --

So, those were low-LTV, long-term clean pay, high prepay that are -- we have the ability to do different kinds of financing with those as well.

Okay. That's helpful. Thank you.

Jessica Ribner, FBR & Co.

Hey, guys, how are you?

Good. How are you?

Okay. Having phone issues. Most of my questions have been -- that shouldn't surprise you. Most of my questions have been asked and answered. One just kind of remaining. Are you seeing sellers with the same motivation as you've seen over the past few months?

Even more than ever. It is -- I don't know if it's the risk retention rules, bank regulatory capital, fear of more regulation if the Senate changes. I don't know what it is, or some combination of all of the above.

But, we have seen sellers being in a hurry. We had -- our September 30th closing was a bank seller who came with $150 million and asked us to work out a pool where they got a minimum number of proceeds and a specific dollar price. And we could buy the loans we wanted to buy, as long as we met those two criteria.

And then the loans that were -- have pending, that'll probably close the week after Thanksgiving is my guess. There's seven transaction. But, one in particular, which is the biggest, about $65 million or $70 million, that was over $200 million pool where the seller had us go loan by loan and negotiate prices as long as the seller got a minimum, a certain minimum amount of proceeds and a certain specific dollar price.

So -- and in September, it was a bank seller, and in October, it was a fund seller. So, I can't even tell you it's banks instead of funds or funds instead of banks. There's -- I think there's a lot of demand for marks for liquidity for whatever reason. And I think there's a bit of a fear of bank regulation.

And in terms of when the loans close, should we continue to assume that you're only holding your acquisitions on your books for 30 days of the quarter?

I think that's probably reasonable. The one thing I will say is the more ratio management that sellers do, the more closer to the end of the quarter the closings happen.

Okay.

So, for sellers who are looking for specific marks or specific ratios, those tend to close later in the quarter, usually in the last week or two weeks in the quarter. And for sellers who it's just part of their business and it's -- whether it's strategic or it's liquidity or it is a smaller bank being acquired by a bigger bank or a regional bank trying to offload something in advance of an acquisition, whatever, those tend to be spread out more evenly in quarters. But, ratio management sellers tend to be in the last -- closings happen in the last two weeks of the quarter.

So, it might even be better to assume that you're holding your acquired loans for less than a third of each quarter.

It's possible. In this one, we had a number of big acquisitions in -- at the end of July. And then August was a slow month from acquisitions. It's a great month from getting a securitization done in early August. But, what that also did is it put $16 million of cash on our balance sheet on the 10th of August that didn't get invested until September 30.

Okay. So, there was, right, drag there.

Drag, right. So, great execution, but too good for a month.

Fair enough. Right, I see that. The average daily cash balances were higher quarter over quarter.

Yes, it was $50-some million quarter over quarter. And in fact, when we -- we actually -- it would've been higher than that when -- if we hadn't overpaid down debt when we transferred some of the loans to the DoubleLine JV. We actually overpaid our lines down so we wouldn't have as much cash.

Fair enough. All right. Well, thank you so much.

Sure.

(Operator Instructions). Brock Vandervliet, Nomura Securities.

Good evening. Thanks for taking the question.

Sure thing. Thanks, Brock.

If you could just spend a moment on the funding dynamics in terms of the puts and takes between your secured borrowings and the repo, and is there any benefit that may accrue in the fourth quarter with respect to funding cost?

Yes, both from our securitizations and our repo, our funding costs have gone down quarter over quarter. Q3 new funding is substantially cheaper than Q2, both on a repo basis as well as a securitization basis.

Our August securitization was done to yield four and an eighth. And our October securitization was done to yield 4%. And the August securitization was higher advance rate than before. And the October securitization was a higher advance rate than the August securitization.

And in fact, if you look back to the deals, the August securitization called our 2014-A and B deals, those were combined at a 50% advance rate at 4%. And our August -- or our October securitization was 65% advance rate at 4%. So, that was significant.

On the repo side, we funded most of our July acquisitions on the repo facility. We -- when we transferred some loans to DoubleLine, we paid down both the debt portion and the equity portion related to that piece. So, we overpaid down the repo facility. We reborrowed a little bit when we closed in late September. And then in our acquisition in October, we reborrowed about three-quarters of the overpay down as part of that.

So, our leverage ratio has come up since Q2 and Q3. But, our loan portfolio has grown dramatically as well. But, our cost of funds has come down, both on the repo side as well as on the securitization side.

And I guess as a follow up, and you touched on this with respect to earlier questions, the loan yields should continue to come down as you continue to buy what sounds like relatively clean paper. So, do you think -- ?

Two things. One is we generally buy loans that are approximately six or seven consecutive payments. This acquisition of clean paper was price, opportunity, and financing and knowledge induced because we knew the loans for 3.5 years already.

I would not say we're out there trying to compete with MetLife buying the cleanest pay paper out there. Opportunistically, if we see good portfolios at prices that we have an advantage in the acquisition of, we'll buy it. But, we're not out there trying to competitive bid against the premium buyers of absolute clean paper that pay very high prices. We would rather be a seller to them than a buyer against them.

And we may in fact, subject to tax rules, look at reselling a portion of this clean pay to one of those buyers. The -- but, on the -- separately, our regular traditional portfolios that we buy that for -- since early 2015, we made a point of buying very low LTV as a hedge against property values.

Those have had continued performance far in excess of our expectation. It causes a duration extension and lowers yield. But, it also significantly increases cash flow over time from those same loans. And if I'm going to have any problems, having too many loans pay too much is what I'm willing to have.

I guess another way of asking that, is there any reason not to expect the loan yield to continue to drift lower?

No, I would expect that continued high performance of not just the new loans we buy, but even more so of the loans that we've acquired over the two years. Those continue to show significantly better performance, which extends duration and reduces yield, but increases cash flow over the life of the loan, increases money multiple on the acquisitions, also allows them to be financed with significantly higher advance rates at significantly lower cost of funds.

Okay. All right. Thank you.

This concludes our question-and-answer session. I would like to turn the conference back over to Mr. Larry Mendelsohn for any closing remarks.

Thank you very much, everybody, for joining our third quarter 2016 call. We will all be around for a while and for days and days and days, as it's a very busy quarter ahead of us. And happy to talk to you and get together whenever's convenient. And appreciate it, again. And look forward to doing this in three months.

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