Alliance Data Systems Corp (ADS) CEO Edward Heffernan on Q4 2018 Results - Earnings Call Transcript

Alliance Data Systems Corp (ADS) Q4 2018 Earnings Conference Call February 7, 2019 8:30 AM ET

Company Participants

Viktoriia Nakhla - AdvisIRy Partners

Edward Heffernan - President, CEO & Director

Charles Horn - EVP & CFO

Conference Call Participants

Sanjay Sakhrani - KBW

Darrin Peller - Wolfe Research

Ramsey El-Assal - Barclays Bank

Dominick Gabriele - Oppenheimer

Robert Napoli - William Blair & Company

Andrew Jeffrey - SunTrust Robinson Humphrey

Operator

Good morning, and welcome to the Alliance Data Fourth Quarter and Full Year 2018 Earnings Conference Call. [Operator Instructions]. In order to view the company's presentation on their website, please remember to turn off your pop-up blockers on your computer.

It is now my pleasure to introduce your host for today, Ms. Vicky Nakhla of AdvisIRy Partners. Madam, the floor is yours.

Viktoriia Nakhla

Thank you, Operator. By now, you should have received a copy of the company's fourth quarter and full year 2018 earnings release. If you haven't, please call AdvisIRy Partners at 212-750-5800. On the call today, we have Ed Heffernan, President and Chief Executive Officer of Alliance Data; and Charles Horn, Chief Financial Officer of Alliance Data.

Before we begin, I would like to remind you that some of the comments made on today's call and some of the responses to your questions may contain forward-looking statements. These statements are subject to the risks and uncertainties described in the company's earnings release and other filings with the SEC. Alliance Data has no obligation to update the information presented on the call.

Also on today's call, our speakers will reference certain non-GAAP financial measures, which we believe will provide useful information for investors. Reconciliations for those measures to GAAP will be posted on the Investor Relations website at alliancedata.com.

With that, I would like to turn the call over to Ed Heffernan. Ed?

Edward Heffernan

All right. Thanks, Vicky, and good morning, everyone. Joining me today, as always, is Charles Horn, our CFO. He's going to update you first on the full year results. I'll wrap up '18 and then we'll shift over to '19 with our guidance and our thoughts. We'll keep this relatively brief, so that we'll have plenty of time for Q&A. So Charles?

Charles Horn

Thanks, Ed. Pro forma revenue increased 5% to $8.1 billion, adjusted EBITDA net increased 7% to $2.1 billion and core EPS increased 17% to $22.72 for full year 2018. Revenue came in short of our initial guidance. Conversely, our core EPS achieved the midpoint for the year. EPS increased 24% to $17.49 for full year 2018, aided by a lower effective tax rate, approximately 27% in 2017 compared to approximately 21% in 2018. We were active during 2018 transitioning Card Services toward more attractive clients and verticals. We sold $1.2 billion of held-for-sale receivables during 2018 and ended 2018 with $1.95 billion in held-for-sale receivables, with the goal of selling the majority of the balance early in 2019.

From an accounting perspective, held-for-sale receivables are removed from card receivables and the related allowance for loan loss. They also no longer impact delinquency or principal loss rates. In addition, provision for loan loss expense is no longer recognized after the reclassification to held-for-sale but, rather, the market value of these receivables is continuously adjusted through a charge recorded in operating expenses.

The mark-to-market charges on held-for-sale receivables were over $100 million in 2018. Optically, the different classification makes provision expense look low and operating expenses look high. During 2018, we spent $443 million on share repurchases and $125 million on dividends. In addition, we reduced our corporate debt by $342 million during the year, lowering our corporate leverage ratio from 2.7x to 2.3x, well below our covenant of 3.5x. Turning into the segments on the next page. Starting with LoyaltyOne, pro forma revenue increased 4% to $1.4 billion. AIR MILES pro forma revenue decreased 3% versus the prior year, primarily due to lower redemptions. AIR MILES issued were flat for the year, as we saw a pullback in promotional activity by our sponsors in the fourth quarter. BrandLoyalty's revenue increased double digits for the year, but product mix and higher-than-expected expenses kept adjusted EBITDA essentially flat.

Epsilon's revenue decreased 4% to $2.2 billion, while adjusted EBITDA was $475 million, consistent with last year. The decline in revenue was primarily due to double-digit declines in our agency and site-based product offerings. Despite the top line decrease, we were able to maintain adjusted EBITDA due to expense control and a positive shift in product mix. Lastly, Card Services revenue increased 10% to $4.6 billion, while adjusted EBITDA net increased 11% to $1.5 billion.

Average receivables increased 8%, while active card receivables increased to robust 23% for 2018. Net loss rates were 6.1%, the same as the prior year, aided by an improving recovery rate. The reserve rate at year-end was 6.1% on reservable receivables, down slightly as a rate from 2017 due to improving loss trends. The delinquency rate remains elevated due to a slowing growth rate in cards receivables, one the industry we call primarily growth map.

I will now turn it back to Ed.

Edward Heffernan

Great. Thanks, Charles. If you'll turn to the slide entitled 2018 Full Year. A bit of this is repetitive, but just to hit the key bullet points. For the non-card segments, which would be the Epsilon and LoyaltyOne parts of the company, again, with Epsilon about $2.2 billion in revs and $475 million on the EBITDA side. It is a business that has a mix of growing and stable as well as declining products. And primarily on the declining side, the agency and the traditional site-based display offerings are in decline. but the margin itself continues to improve as we're shifting to the more tech-based solutions. On the LoyaltyOne side, again, that consist of two primary businesses: one, the AIR MILES business based out of Canada; as well as our BrandLoyalty business based out of the Netherlands. Combined, we call that LoyaltyOne. They did $1.4 billion in pro forma revenue and a little over $0.25 billion in adjusted EBITDA. BrandLoyalty, which, again, does a lot of the shorter-term grocery loyalty programs across the world, the revenue was up double digits, and AIR MILES revenue was down slightly.

On the EBITDA side, essentially flat with the prior year. Really, the nice step-up in revenue from BrandLoyalty was largely offset by higher expenses. AIR MILES issuance, which is a key metric because it's essentially how we get paid, was effectively flat with prior year, primarily because in the fourth quarter, a number of programs were kicked by our sponsors into the earlier part of '19. So that's more of a timing issue. But that's sort of a wrap-up on the non-card segments.

Let's turn now to the full year discussion of Card Services, which had a good year. The revenue of $4.6 billion in adjusted EBITDA, again, it's a term we use in cards, but it does reflect the cost of both funding the portfolio or cost of funds as well as the provision expense for credit losses. If you put those in there, you'll still net to $1.5 billion, and that's a very solid double-digit growth rate in both revenue and EBITDA.

We had a record number of new client signings. We're running at 2x sort of the already elevated historical run rate that we talked about. And I think, probably, most importantly, about this one is the signings are all in growth verticals where we have a proven track record of winning, such as home décor, children's, beauty, jewelry, e-comm, areas that actually are doing quite well on the retail landscape. If you look at some of the names, such as IKEA, Wyndham, Academy Sports, Floor & Decor, Penn Gaming; pure e-comm players, such as Appliances Connection and Adorama; as well as 3 more that we've signed we haven't announced yet, but very, very good year for signings and also is a good step, as we continue to transition the portfolio into healthier, higher-growth-type brand. So nice progress on that, and those will be spooling up as this year -- as 2019 unfolds and we move into '20 and '21.

Tender share remained strong. A lot of people ask me what tender share means. Tender share is nothing more than the dollars that flow that are being spent at that retailer, what percent are flowing through the card product. And if you look at the clients that we've signed in '15, '16, '17 and '18, we're seeing nice tender share gains of roughly 120 basis points year-over-year. So that's what we want to see.

At the same time, the core vintage, which, again, is the portfolio that has brands from 2014 and prior, we continue to hold very solid tender share, roughly in the mid-30% range. So roughly 1/3 of all sales will flow through our product.

Importantly, also, credit quality was stable and was similar to prior year. And as we look ahead, and you talk to the folks whose job it is to watch over this stuff, you're not hearing any types of red flags in terms of consumer distress. And so we think that this, as we call it the normalized loss rate or our long-term loss rate of roughly 6% right around, seems to be a good number. That's what it was in '17-ish, '18-ish and as we move into '19. So that's a good sign.

We are, as I mentioned, moving towards the more attractive clients and verticals, was largely completed in the fourth quarter. Again, we're signing those types of clients. But at the same time, we are also moving out certain clients that can no longer benefit from our services, and many of them are having difficulties in today's changing retail landscape. So think of it as we're accelerating this pivot into the healthier verticals by moving out over $2 billion of card receivables in these troubled verticals.

Overall, reported average receivable growth of 8%. But what we look at here from a longer-term perspective is how about those active brands that we have, those that we're going to be relying on to keep us going well into the future. And that was up well over 20%, which bodes extremely well as we move into the latter part of '19.

If we can turn now to the credit metrics. Charles, why don't you hit this one?

Charles Horn

Sure. So we're over on Page 8, and I thought we'd talk through some of the key performance indicators for the year. I think it's important to start with normalized average receivables for the simple fact it includes our held-for-sale receivables. And that's a relevant metric because, really, we continue to earn revenue and incur expenses on held-for-sale receivables until sold. So really, in evaluating gross yield, operating expense leveraging, you have to consider the held-for-sale receivables, which were about $1.95 billion at year-end.

Our gross yields were down slightly, about 30 bps, and that's primarily due to vintage mix. What that means is more of our receivables growth are coming from younger vintages, and it takes a younger vintage about 3 to 4 years to really hit a run rate gross yield.

Operating expenses, we actually were good on the leveraging there. They were up about 10 basis points year-over-year, even though the charge was 70 bps, as we talked about before, about 60 bps is due to the mark-to-market adjustment occurring value for the held-for-sale receivables.

Lastly, the delinquency rate, which we talked about a little bit before, continues to be elevated on a year-over-year basis, primarily due to the slowing growth rate in card receivables. That pressure should moderate as card receivables growth reaccelerates. And we're anticipating that will be more in the third quarter back half of 2019.

With that, I'll now turn it back over to Ed.

Edward Heffernan

Okay. So I'm on Slide 9, Card Services, the average receivables growth. This is what we use here to sort of gauge, are we heading in the right direction as we pivot the card business itself. And if you look at the active clients, those are the brands that we're counting on to drive the longer-term growth of the business. It's masked quite a bit by the fact that we did, as we've talked about, discontinue our relationship with a number of brands. And if we turn back to the active clients, you'll see is a very healthy year for growth, up over 20%. That's nice.

What probably is more important is what does that mean for the future. And you'll see to the far right that roughly 1/3 of the growth coming from these newer vintages in the healthier, probably more exciting verticals that we've been talking about, and that's up 200% from the prior year.

So what's it all mean? What's it all mean is that the 2015 to '18 groups of clients that we signed represented roughly $4 billion of the portfolio in '18. And when those things are fully ramped up, because many of them are in these early stages, that's going to be roughly $11 billion to $12 billion. So there's a lot runway there with these signings.

We mentioned the various verticals that we're excited about, names such as IKEA and Wyndham and Academy Sports and Penn Gaming. But also in there are some of the earlier signings, such as the Wayfair and the Ulta Beauty and the Signet Jewelers, Williams-Sonoma, Diamond International, Build.com, Viking Cruises. As you can see, a lot of different verticals from what we've been known for, for the first probably 15, 18 years of the company, which was primarily mall-based specialty apparel. So it's -- we've made a lot of progress. And we're going to see the benefits of that play out as '19 moves into the latter part of the year.

All right. So let's move to Slide 10. I think this captures it succinctly, which is if you were to look at what is the benchmark in the industry in terms of growth rates in -- yes, if you want to put in general-purpose cards, you want to put in co-brand, you want to put in Private Label, most of it is captured in revolving debt. Revolving debt has been growing around 3% or 4% if you go back 7 or 10 years. And against that, we've been growing 18%, and that even includes the divestiture of those noncore brands that we talked about at length earlier in the call. So even with the $2 billion removed, we borrowed roughly 18% growth versus an industry that does 3% to 4%, and we expect that to be one of the attributes of this company going forward as well.

We believe that we play in a sandbox that will allow us to grow at a significant multiple above the industry as a whole. A lot of this has to do with where we play in terms of our clients tend to be much smaller than the very large co-brand card programs that are out there. And if you were to look at our portfolio of $18 billion, that's spread over 160 different brands. So again, these tend to be smaller portfolios, very focused on brand and on high-end quality service, and that's our specialty.

Additionally, to the industry-leading growth rate, our industry-leading return on equity. So our ROEs are 30% and/or more, which is anywhere between 2 to 3x what the industry is at. A lot of people keep sort of scratching their heads saying, "How can you be growing so fast and have ROEs that are 2 or 3x the industry?" And the answer is, again, we play in a sandbox that, we believe, we have unique advantages. And the uniqueness of our advantage is the fact that everything we do is in-house, from the actual network itself, we don't outsource that. The customer care is done in-house. The collections are done in-house. The marketing, the database -- the databases that we build, they're all done in-house. And with that, that allows us to approach the industry in a holistic manner and allows us to have the uniqueness that comes with a close-loop type network that can extract not only who the customer is of the client, but also what she purchased down to the SKU level. And we use that type of information to then go back on a one-to-one personalized basis through the various digital channels as well as some of the more traditional channels to drive that incremental purchase. And it's those incremental purchases that sort of set us above and apart from sort of the more traditional banks and card players in the industry.

And so that's -- we're not for everyone, but we believe that the total addressable market, or the TAM, that we're going after is about $50 billion. And so we've got a long way to go to get there, and that gives us a lot of runway as we look forward. And when we talk about the TAM for our business, we're not talking about these big co-brand programs. We don't really play in that area. We believe that the Private Label offering is really the type of offering that is sought after by the client base that we're after. We'll occasionally do a co-brand as in a combination for some of those clients who already have a PLC, or Private Label card. But otherwise, we're trying to remain very disciplined in the products we offer and the deal terms that we'll except. And that's why we believe that even moving into these newer verticals and healthier in faster-growing verticals, we're keeping the same types of discipline, such that those ROEs can remain at that 30%-plus level.

Okay. Let's move into a little discussion about losses, which, I think, is always -- garners all sorts of different questions of why do you run the portfolio at a 6% loss rate, not a 3%, not an 8%, not a 1%. And the answer is we've tried to find sort of what we believe over the last 20 years to be the right level to run the portfolio, to garner both significant tender share and, therefore, value for the retailer, while at the same time not going too far out on the risk spectrum. And so if you're to look at the typical cycle, I always get asked about the cycle, the typical cycle for our business would be we run at a 6%-or-so loss rate. And a lot of folks thought that, "Oh, my gosh, the last couple of years, losses are creeping up. And what does that mean? Is the portfolio getting worse?" And the answer is no. What basically happened if you go back to the Great Recession was that losses ran quite a bit higher than they typically do. And as a result, you effectively drain the pipe in terms of potential new customers. And as a result, you wound up going from above trend to well below trend. And all that happened over the past several years is we've returned to our long-term trend line. It's nothing more than that.

We think the fact that '17 and '18 were very stable, '19 looks to be stable as well that we're at trend and we want to keep it right around that level going forward. So it's not going to jump up too much, not going to jump down too much. This is where we want to run the programs to optimize what we talked about earlier.

Also Charles touched briefly on delinquency rates, which we're running around 60, 70 basis points above prior year. As we exited 2018, there were some noise in that number. What's going to happen is you're going to see that start stepping down pretty quickly. And so we expect that to drift down as we move throughout 2019 and we lose the noise.

In terms of the outlook for '19 for loss rates, again, we expect it to come in, in that 6-ish-type range for the year. The trends look good. From a seasonality perspective, just sort of a heads up, the way it works in our Private Label business is first quarter tends to be sort of the high watermark. And you'll be in the 6s there, and then you'll start drifting down. So look for about mid-6, maybe a little bit better in Q1, and then we'll drift down. And for the full year, we'll wind up right around back at that 6% level. So that's that. Finish up 2018.

We now move to 2019. To the extent there are no transactions completed, if you were to just look at status quo for the noncard businesses, we would expect to see mid-single-digit growth in revenue and adjusted EBITDA. I don't think that status quo is going to be what we're going to see, and we'll talk about that in a little bit.

So let's go to Card Services. We expect to exit 2019 at roughly $20.5 billion. It's up a little bit from last call where we thought it was around $20 billion. So about $20.5 billion compared to exiting this year at 17 -- this past year at $17.9 billion. Or said differently, we expect that both active as well as reported growth rates will converge at the end of the year as we push through this $2.1 billion discontinued programs. And both metrics will show an exit rate in the mid-teens, which, again, sets us up for a nice run exiting this year and going into next year.

As importantly, the new vintages that I talked about earlier, these are the clients from '15 to '19 that we're excited about. We now expect them by the end of the year to account for over 40% of the receivables. So in terms of pivoting the portfolio, it's going very, very fast, and that's both from the new signings as well as the discontinued programs. It's a little bit of pain that we need to go through in terms of getting to the first part of the year without those programs that have been discontinued. However, the end result is clearly a much healthier and stronger portfolio. '19 also assumes a modest level of growth coming from some relatively small files that are out there that we have our eyes on. And we assume that we'll take a few of those in-house, nothing too significant, however.

We talked about the transition to the home décor and children's, beauty, the jewelry, again, away from that sort of mall-based specialty apparel sector that's having such a difficult time. Importantly, we are keeping very strong discipline in terms of signing for these new vintages. And although, obviously, it takes them a while to ramp up, we are building the models and signing deals, keeping similar ROE characteristics. So we do not expect that to be an issue going forward.

We talked about the stability in credit quality and the loss rates being relatively flat, with delinquency rates year-over-year beginning to come in as we move throughout the year. So the net result is it should be a very strong year for the cards in terms of these critical metrics that we look at: active growth, the convergence of active and reported growth to 15%; exiting the year losses; delinquencies. But it will be dampened, especially in the first half, on the disposition of those card receivables. So that's sort of the price that we've decided is necessary to get us pivoted in the right direction as fast as possible and get this behind us.

All right, Page 13, the 2019 outlook. For the Epsilon divestiture, obviously, people are very curious as to how that's going, and I need to be somewhat brief in my remarks. Obviously, '18, we spent a good deal of the time with the board coming up with our overall strategic review of the various businesses. It was quite clear that Epsilon has many valuable assets, both from a technology side as well as a people side, that seem to have more value or is ascribed more value out in the marketplace than it was within the company itself, I think, primarily due to the size of cards. And so the decision was made to go ahead and move towards a divestiture to capture that value and find the right home for Epsilon. The strategic review took a good chunk of the year. We did, in the fall, hire bankers and lawyers and all sorts of folks. And in November, we moved to a very -- to a formal process.

Where are we today? The process continues to move. We did have a great deal of interest in the businesses. Initial bids have been received. We are currently finishing the selection of the final round. The process is moving smoothly. And I have to say, if consummated, the use of proceeds will be focused on a combo of further debt reduction as well as significant share repurchases. And the only comment I can make on that is it's -- it was good to see, as we move into the final round, interest from both the strategic side as well as the sponsor side. So stay tuned. We're moving into the final throes on that.

All right. We're getting near the end here. In fact, I think I'm actually on my last slide. So if you go to the last slide on 2019 guidance. The initial annual guidance that we're going to throw out there is a little over $8 billion of revs and $22 as the base case in core EPS. We expect, as we talked about, very strong growth in ending card receivables. What that also will mean, however, is you need to build a provision for that big growth. And that will -- is one of the reasons that earnings will be slightly dampened this year.

Additionally, obviously, you had a couple billion of receivables that were generating income for us in '18. Those have been moved out. And so as a result, that's the other reason why you're going to see some dampening of the earnings. Both of those are timing issues, which is good. The first being the provision buildings that the newer vintages are ramping up nicely, and that should pay off as we move into the following year. And then the dispositions that we talked about, again, we are doing that thoughtfully. But we wanted to do it and get it over with. We don't see anything else on the horizon.

First quarter revenue will be down mid-single digits. Core EPS will be down high single digits, primarily due to the dampening effect of the divestitures of those -- the $2 billion-plus of receivables. Normalized average card receivables, again, will be down because of this divestiture. However, what we report from an active perspective should be nice and healthy in terms of growth rate.

Importantly, from a guidance perspective, it's -- you're going to have quite a bit of movement, I believe, on the guidance as we start moving through the year. If you were to assume that there is an Epsilon divestiture coming up in the use of proceeds, the timing of that, how it's used, is it debt, is it equity, et cetera, et cetera, that's going to change things quite dramatically. And so we're trying to sort of give everyone enough to start their models and give enough color with the understanding that you should expect things to be updated and changing, both Charles and I believe, fairly dramatically as the year unfolds.

So from a summary perspective, the divestiture of certain card receivables will create some near-term headwinds. We've talked about that at length, and this was done on purpose to make sure that we can move faster to get this thing back to very healthy, strong, mid-teens-type growth on the file. We are -- the repositioning, as we've talked about, the brands that we sign suggest that we're on the right path. And as a result, strong double-digit growth in active clients throughout the year. The gap that we've talked between our reported receivables and the active receivables, we'll obviously, by definition, now to 0 by the end of the year and should go back to that 15%-plus, as we finish that anniversary.

Credit quality is stable. We've talked about the ROEs in the business. Again, the newer clients, the disciplined approach we're taking would suggest that we expect those ROE -- ROEs to remain at elevated levels, vis-à-vis the industry.

The Epsilon divestiture, again, we think we will find a good home, the right home, while, also, at the same time, unlocking a good deal of value for our shareholders. And at the same time, it does help the narrative in terms of beginning to make the whole story a bit easier to explain and a bit more focused on what we're trying to do. So I'm going to wrap up now just with a thought or 2, which is what we're going through right now in terms of repositioning the card portfolio into -- to reflect where the health and the growth are in today's retail environment is something that needed to be done. And we very aggressively are pursuing that approach. And I think the payoff will be superb, as we move towards the back part of the year, as well as taking a look at other businesses within the company that may not be valued as highly -- as part of Alliance as they could be outside of Alliance as well as potentially complicating our overall story.

To some, it may seem like this is an awful lot of stuff going on at once. It is. But for those of you who've been around, this is not so different from what we did roughly 10 years ago, back in the '08, '09. In the teeth of the Great Recession, we divested three separate businesses that were viewed as noncore, simplified the story. We then did a massive share repo at recession-level pricing. And so those two moves paid off quite handsomely for our shareholders. I view today as sort of act 2 sort of 10 years later. At the end of the day, the critical things as part of the strategic plan are that we're pivoting the portfolio to healthier verticals. We're exiting those brands that we can no longer help and they can no longer benefit from our services.

Additionally, we are divesting businesses that have value -- higher value elsewhere and probably have not gotten the type of attention that they needed within Alliance. And then when it's all said and done, frankly, we'll be taking advantage of these, what I view, as recessionary-level valuations to benefit our longer-term shareholders with proceeds from these divestitures. So a little bumpy, but strategically, it's a sound plan. We believe very strongly that it will pay off, especially for those who recall that it's not so distant, which what we did 10 years ago.

With that, I'll put the pen down and will open it up for questions.

Question-and-Answer Session

Operator

[Operator Instructions]. And your first question comes from the line of Sanjay Sakhrani with KBW.

Sanjay Sakhrani

I guess, I've got one multi-question on card and one on Epsilon. On card, could you just walk us through this noise on the delinquencies and sort of why you're confident it'll improve over the course of the year? And then as far as, like, the provisioning is concerned into 2019, I know it's probably going to be a headwind from the growth, but any cadence associated with that, Charles, would be helpful. And then as far as the profile of the customer of these new partners that are growing -- you're growing with, has that changed in any way, like the demographic profile? Because it used to be that this customer that you'd get would have a secondary income or additional income that they wouldn't have -- that they wouldn't sort of notate and they'd have a very good credit quality profile. Is it sort of consistent with the same in low lines?

Charles Horn

You want to start off from the delinquency?

Edward Heffernan

Sure. I'll take the easier one first. And then I'll give the hard one to Charles. But in terms of the type of customer we're seeing in the newer verticals, no, I don't think the quality is any different. In fact, since we have the right to maintain a certain level of quality within the overall portfolio, Sanjay, we can obviously adjust the knobs as necessary to sort of track to that goal of roughly 6%. So I would say, the only big thing that has shifted -- when you -- when the vast majority, when 90% -- plus are -- of our cardholder base are women, I don't see that changing a lot. But what we are seeing, obviously, as we're moving into the wayfair.coms and in some of the pure e-comm plays is -- as you might expect, is you're going to begin to skew down more into the millennials, and you're beginning to see what the Ulta Beauty and stuff like that, the Gen Zs or the iGens. And so for those types of clients, we're getting them in the door a little bit younger. And so you would adjust your credit lines accordingly.

Charles Horn

And on the delinquency, Sanjay, you know quite well when AR growth decelerates, it pressures delinquency trends. You have less receivables going into your current bucket. And then when you get back to a point where AR growth is accelerating, it gives you some benefits on your delinquency trends. Based upon the profile we see now, I think you'll still see pressure on delinquency rates until you start moving into the growth periods of Q3, Q4. And I'd say, the same will be true of your provisioning. Most of that will be more back-half weighted. Again, as you saw, we expect receivables to be down Q1, Q2, accelerate in Q3, Q4, with quite a bit of growth based on the time, what we see now, in fourth quarter. So I'd expect it to be your largest quarter in terms of provisioning expense.

Edward Heffernan

And I think on the delinquency side, Charles, check me if I'm wrong here, but we should see the year-over-year spread begin to narrow, like, very soon. So I think that's in the release.

Sanjay Sakhrani

Hope so. And then on Epsilon, I know you guys are limited to what you can say. But could you just characterize if you're reasonably pleased with the bids and if you envision a total or partial fail?

Edward Heffernan

Well, I've got counsel sort of hanging over here. But no, I think we've probably said all we can. I mean, clearly, if you're going into -- if you're selecting final bids, final round for bidders, you're not displeased with what you're seeing. So I think I'll probably have to stop there.

Operator

And your next question comes from the line of Darrin Peller with Wolfe Research.

Darrin Peller

Let me just start off. I understand the split and the rationale on the restructuring on the -- from the card side, obviously. Just -- can you give us a little more of an idea when do you expect the held-for-sale portfolios could actually get sold? And then thinking about the freed-out capital there, should we consider that being invested in growth receivables? Should we think about portfolio acquisitions? Is there even anything out there?

Charles Horn

So I'll start with the held-for-sale. You saw there, we sold $1.2 billion during the course of the year. We ended the year at $1.95 billion. A chunk of that relates to one bankrupt retailer that's in liquidation we're working on. I think that one could get done in the first quarter. And then I think a lot of it can be addressed first half of 2020. It's just some of it may take a little bit longer. What was the second one?

Darrin Peller

Just the freed-up capital plans.

Edward Heffernan

Yes. I mean, clearly, if we're expecting the -- both active and reported to hit that 15% growth rate by year-end, clearly, we're deploying that capital back into the business. A lot of it will be, obviously, to fund the continued ramp-up of the '15 to '18 vintages. However, to your point, Darrin, there are a handful of files that are not likely to attract the attention of the big banks, but very nicely in our wheelhouse in terms of size. And I think you can expect to see a combo platter from us in terms of both the organic vintage ramp-ups as well as a handful of these smaller portfolios coming in.

Darrin Peller

Okay. And just one follow-up. When we think of the pro forma entity for the restructuring for Epsilon, I mean, we have a large card business now and we'll have, obviously, the LoyaltyOne business. So when we think about that, I mean, is the LoyaltyOne business still going to operate as a standalone business under the EC Alliance Data umbrella? Does it make sense to keep it? Does it makes sense to -- I just -- I never really saw a ton of synergy between that and the card business, so I'd just be curious of your thoughts there.

Edward Heffernan

Yes. No, it's a fair comment. I think -- I guess, the easiest way to explain it is we had a lot of balls in the air right now, and so we're taking things one step at a time. You've got -- sort of what can we say has been completed in terms of the overall strategic plan. We've completed sort of the necessary divestitures of those brands that we talked about in the card business. And we've set the card business up now, in my opinion, to have a very strong run for several years. We also have determined that there's a lot of value in Epsilon that we can unlock, and so that's sort of number two on the list. Obviously, there are further discussions going on about what else could we do in terms of continuing to focus and clean up the narrative on a go-forward basis. Are there other assets that may or may not fit? And at this point, we're getting around to that. But right now, cards was first; Epsilon, second. And that's sort of what we're focused on right now.

Operator

And your next question comes from Ramsey El-Assal with Barclays.

Ramsey El-Assal

I'm going to try again on Epsilon. Can you share any details on potential timing? Is this something that we should expect prior to the next earnings call, somewhat imminently, in terms of some resolution there on Epsilon?

Edward Heffernan

I guess, we can keep trying to answer the question the same way. But I guess, the best way to think of it is talk to your banker friends. And when someone says they're going into the final round, you can take from that how long it usually takes from final around to when a potential announcement could come out, and we're likely to be on that track. So that's about as much of a nonanswer as I can give you.

Ramsey El-Assal

Okay. And if the end state of the entity here, let's just say that you have Epsilon and that, potentially, LoyaltyOne is to follow. I mean, at some point, does it make sense to change your corporate structure from this ILC structure? I mean, would that not unlock quite a bit of trapped capital that could be put to work? I was just curious about your philosophy of kind of the long-term maintenance of that ILC corporate structure.

Edward Heffernan

That would certainly be a possibility.

Operator

Your next question comes from the line of Dominick Gabriele with Oppenheimer.

Dominick Gabriele

Just as you continue to develop your relationships with the card partners and their needs, many of these partners are seeing more sales online -- on the online channel. Can you talk about how the -- how ADS' capabilities in a post-Epsilon world can continue to help your partners drive sales through the online channel? And also what did Wayfair see in ADS and your capabilities that some new perspective partners in the card business may also gravitate towards?

Edward Heffernan

Sure, and it's good question. I think, obviously, we've had the capabilities for quite some time to do omnichannel-type transactions, whether it's bricks-and-mortar or catalog, online, whatever it is. To us, they all flow into the central bucket. You're absolutely right. We're seeing -- even with our traditional bricks-and-mortar, you're seeing a larger and larger footprint on the online space. You're seeing, at times, our online sales representing 40% of our overall sales volume. So it is an enormous piece of our business today. What we're trying to do on the online side is continue to develop those tools that will make it, if you want to call it, an effortless transaction on behalf of anyone who shops online. We've all been in that position where you've got the screen pops up and then you get all these fields to fill in, and it takes you forever. And so the abandonment rate is huge.

People just get frustrated with that. So what we do is you have your card and you have your number and your ability to get in there and sort of that one button, one-push-type transaction, which is so critical to closing the sale, that type of technology, we have rolled out or in the process of rolling out across all of our clients. Additionally, the ability to understand who is shopping online and offer right then and there a personalized-type discount or earn point or something that we get them to, what we call, a trigger marketing to make that one extra purchase that they normally wouldn't have done is critical. And so to sort of answer your question of what do we do that others don't, it really comes down to, again, understanding who the customer is, whether online or in the store, and proactively offering up a very, very focused offer that could trigger that one incremental purchase that they otherwise wouldn't have made. And so our whole business is driven off of driving incremental sales as opposed to saving money from the client on interchange or something like that. And so the data itself and how we use the data won't change, whether Epsilon is part of the overall ADS or outside of ADS or either be a service arrangement or, frankly, we've built a 500-person division within cards itself over the years of folks, whether data scientists or marketing experts, to help drive that. So that's sort of the special sauce.

Dominick Gabriele

Right. And then just one more, if I can. You just talked about the average balance size that you have currently versus where this average balance size could go when you keep moving away from specialty retail. And that -- how much do you think that piece of the loan growth story in '19 and '20 is attributed to that?

Charles Horn

I don't have that off the top of my head. I can get that for you once we go offline. I just don't have that number.

Operator

Your next question comes from the line of Bob Napoli with William Blair.

Robert Napoli

Ed, maybe I'm not sure if you could answer this or not, but I'll try. With Epsilon, would you -- with the company, would ADS not sell Epsilon if they didn't get offers that were within -- had a valuation that added significant value to shareholders on the -- with the capital deployment you would make? Are you committed to selling this? Or if you got the bids that were insufficient, would you pull back?

Edward Heffernan

I think that there's no question there, it's a floor. And it's got to be above the floor, for sure. And so we wouldn't be moving to final rounds with bidders, unless we felt comfortable we're heading in the right direction.

Robert Napoli

Okay. And then just on -- I think last quarter, you had put out kind of a 2020 receivables number. Just first of all, are the new assets generating the same returns that you've generated historically? Are they a little bit lower? I mean, 30% ROE is wonderful. But as we try to think about earnings growth into 2020, yes, as you're -- you build the portfolio, how should we think about the returns relative to the historical levels that ADS has delivered?

Edward Heffernan

Yes. I mean, I think, look, we're pretty -- there's a lot of deals that we walk away from, obviously. And what we're trying to do is keeping the pricing discipline where it has been in the past. And so far, as long as -- frankly, as long as we stay within, I keep calling it, our sandbox, we seem to be maintaining that level of ROE hurdles that we've had in the past. Now clearly, when you're taking a very mature portfolio that's been around for 15 years, not really growing, it's throwing off a lot of cash and profit. But longer term, obviously, it's not the healthiest. But those are being divested. And you're spinning up the newer brands, which take a little bit of time before they can actually generate those types of returns, right? I mean, it's the nature of these vintages. So -- but steady-state, no, we -- we're keeping the same discipline we've seen before. And frankly, the people who are looking for the special sauce that we're -- that we offer, they're pretty focused on how you're going to drive that incremental sales and prove it to me. And as a result, I think we're -- we should be in good shape on the ROEs.

Robert Napoli

Last question, real quick. On the $22 guidance for next year, and I understand there's a lot of moving pieces there, you're going to change that as we move through the year, but does that include share buybacks?

Charles Horn

It does not.

Operator

Your next question comes from the line of Dan Salmon with BMO Capital Markets.

Unidentified Analyst

This is William on for Dan. Just following up on Epsilon with a more fundamental-oriented question. You called out agency and site-based displays obviously being weak. But I was wondering if you'd go into more detail about areas of strength and which offerings are showing the most promise right now.

Charles Horn

The way I'd break it down is Technology Platform, which you remember is really the older part of Epsilon, has -- was very stable this year after being down the prior year. The 2 big drivers last year really were the auto vertical and the CRM -- Conversant CRM. We did not see the same level of growth this year with those 2 growth drivers as we did the prior year. But we still see building backlogs. We still see a big runway for them to develop. But even with our 2 bigger growth drivers, they didn't really produce as much in '18 as what we thought they could.

Edward Heffernan

Yes. I mean, the big growth drivers going forward would be the CRM engine, the more traditional tech platforms, which are building the big loyalty in CRM databases and then the auto as well.

Operator

Your next question is from the line of Andrew Jeffrey with SunTrust.

Andrew Jeffrey

With regard to the ongoing discussions around Epsilon, and I wonder if you can just comment on sort of how broadly, or from a high level, you're thinking about the pieces of that business that historically have been important to driving the value in card and whether or not -- or how you contemplate the sustainability of your ROE and the value prop, which we, I think, understand so well in the context of selling the entirety of Epsilon perhaps. Just how intertwined are those businesses?

Edward Heffernan

Yes. It's a great question, Andrew and, obviously, something that we give a great deal of thought to. By "unlocking" the value with Epsilon, are you, in fact, going to hurt the offering in cards? And so what we've done is a number of things. First out of the gate, as I mentioned, over the last several years, we've built up a, what I would call, mini Epsilon within cards itself. So you've got a 500-person dedicated division of marketing and data scientists who, you could say, normally would be over at Epsilon. But since cards was so big, we built it in cards. So that won't suffer. It's more on the technology side. So all of our brands in cards use the Epsilon platform for the various reward and loyalty programs. That will continue to be the case. It will be a service level agreement -- service agreement between the 2 entities. Additionally, these brands would use the various digital channels, whether it's Conversant or whether it's the digital e-mail platforms. Again, those would be direct service agreements with Epsilon. And then finally, a lot of the demographics, psychographic data that is housed within Epsilon, again, would be a service agreement between the two entities. So to put it succinctly as I can, which is always difficult, Andrew, it would be the people-based stuff, we think, is all taken care of and housed within cards. The technology and the data that we need from Epsilon, we are carving out very specific contracts and agreements between the two entities. So the goal is not to have anything that disrupts what we offer to the client set.

Andrew Jeffrey

Okay. That's helpful. And then Ed, I just wonder, I think becoming a pure play or closer to a pure-play in value-added Card Services makes a ton of sense for Alliance shareholders. Are you -- but is there a question about where we are in the cycle as you get more of a -- as you become more of a pure-play in a credit business?

Edward Heffernan

Yes. I mean, I think that's a lot of the stuff, right, that has sort of hit valuations of the financial sector and consumer finance companies. And all we can do is, frankly, what worked best for us should a recession approach in the next couple of years, what worked best for us in the Great Recession if -- to do around, was the fact that we actually grew. We basically did not pull in during the Great Recession. In fact, we expanded and actually grew the portfolio and took advantage of the environment to, frankly, get some really nice clients with very good ROEs over the longer term. And my guess is we are positioning the company right now, if you think about it, as we deleverage and take advantage of what we believe are very attractive valuations out there for us. I don't see us doing anything differently. So we're going to play through in terms of right now, times are good. And so we're going to continue to sign and to grow to the extent there's some types of recession on the road. When you're making the returns that we're making, losses from a recession, you're probably talking 1 point, maybe 1.5 points at most, which still leaves a ton of profit flowing from the card business. So we would follow the exact same game plan as before and deploy as much, if not more, capital towards growth if there's a downturn. All right. Okay. Thank you, everyone. Appreciate it, and we look forward to our next call. Bye.

Operator

Thank you again for joining today's call. This does conclude today's conference. You may now disconnect.