Bread Financial Holdings Inc (BFH) 2017 Q2 法說會逐字稿

Good morning, and welcome to the Alliance Data Second Quarter 2000 (sic) [2017] Earnings Conference Call. (Operator Instructions) In order to view the company's presentation on their website, please remember to turn off the pop-up blocker on your computer.

It is now my pleasure to introduce your host, [Ms. Analise Han] of FTI Consulting. [Analise], the floor is yours.

Thank you, operator. By now, you should have received a copy of the company's second quarter 2017 earnings release. If you haven't, please call FTI Consulting at (212) 850-5721.

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 the 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. Reconciliation of those measures to GAAP will be posted on the Investor Relations website at www.alliancedata.com.

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

Great. Thank you, [Analise]. Today, we're going to stick with the same format we did last quarter, which is probably a fairly brief commentary by Charles and myself, and then we'll open it up for questions. So obviously, joining me today is Charles Horn, our CFO, and he's going to walk you through the second quarter results. And then we'll turn our attention to the full year outlook, how we're doing against our goals for the year. And then also, we believe, at this point, we feel pretty good about putting a stake in the ground for 2018.

So with that being said, Charles?

Thanks, Ed. The second quarter came in slightly better than expected, with both revenue and core EPS up 4% from the second quarter of 2016. We saw continued strength at Epsilon and Card Services, and positive momentum at AIR MILES in both miles issued and adjusted EBITDA margin, which returned to the mid-20s range during the quarter. However, BrandLoyalty continues to fight through some timing issues, revenue down about 30% for the second quarter. And we now expect that trend will continue through the third quarter.

Setting the unexpected softness of BrandLoyalty aside, we are delivering on our 3 major objectives for 2017: first, restoring Epsilon to both positive revenue and adjusted EBITDA growth; second, delivering on credit normalization; and third, retooling the AIR MILES model.

During the quarter, we concluded our $500 million share repurchase program. In addition, we augmented our liquidity during the second quarter by replacing and upsizing our corporate credit facility. Its maturity is now 2022, and its capacity is $4.6 billion, up $500 million from the previous facility.

Let's go on to the next slide and talk about LoyaltyOne. It was a soft quarter for LoyaltyOne. If we break down the segment by major business, AIR MILES' revenue decreased 12% to $179 million for the second quarter of 2017, driven by a 13% decline in AIR MILES reward miles redeemed. The decline in miles redeemed was expected given the elevated redemption rates in the second quarter of 2016. The burn rate, which is miles redeemed divided by miles issued, dropped to 76% this year versus 86% last year, and that's the range that we really try to operate this program, 76%.

AIR MILES issued was down less than 1% year-over-year, a nice improvement from the 4% decline in Q1 due to several sponsor-driven promotions in May and June. We continue to see steady improvement in collector engagement, in activation, and our collector retention rate currently sits at 98%. As a result, we expect issuance growth to turn positive in the back half the year.

Our adjusted EBITDA margins at AIR MILES improved during the quarter and are trending to our mid-20s guidance for the year. We continue to look for ways to streamline costs while better serving and adding value to our collectors.

During the quarter, we significantly expanded travel options for our collectors, while also providing more ways to book and earn miles.

BrandLoyalty's revenue decreased by 32% to $101 million for the second quarter of 2017. Several programs that were executed in the second quarter of 2016 are now scheduled for the fourth quarter of 2017. Timing of these short-term promotions are driven by the client, so having program execution shift from one quarter to another is expected.

Adjusted EBITDA margins for BrandLoyalty dropped into the mid-single-digit range for the quarter due to its high fixed cost structure.

Let's move on to Epsilon. Epsilon carried its solid momentum from Q1 into Q2, with revenue increasing 5% to $544 million and adjusted EBITDA increasing 4% to $107 million compared to the second quarter of 2016. Once again, we saw solid top line performance within our key product offerings, coupled with strong expense control. In particular, salary and wage expense increased only 1%, but was offset by onboarding costs associated with 2 new significant clients, moderated EBITDA margin expansion during the quarter.

Breaking down revenue by offering. Our Agency and Auto offerings continue to grow double digits, with Auto benefiting from solid first half wins, such as Hyundai and Volvo. Our digital media offering, known as Conversant CRM, continued its double-digit growth trend, increasing approximately 30% year-over-year. Our data and CJ Affiliate offerings were stable with low single-digit growth.

Importantly, Epsilon's Technology Platform's offering continued to show progress toward a promising turnaround, narrowing to a 3% decrease for the second quarter, sequentially improving again since exiting Q4 2016 with a double-digit decline. The client base is showing stability, and we have seen an increased win rate on the basis of the introduction of cloud-based, more packaged products into the market.

Finally, Conversant's Agency business continues to be a drag on Epsilon's growth rate, about 150 basis points for the quarter. Stabilizing this noncore offering has been a challenge, but frankly, its diminishing size lessens the overall revenue impact moving forward. Putting these trends together, we are looking for revenue growth of 7% or better in the back half of 2017.

Let's go to Slide 6 and talk about Card Services. Card Services had another solid quarter as revenue increased 13% to slightly over $1 billion and adjusted EBITDA net increased 11% to $306 million despite a large provision build. Notably, this was the 22nd consecutive quarter of double-digit revenue growth for Card Services.

Total credit sales increased 6%, with core programs, those greater than 3 years old, delivering low single-digit growth despite softness in the retail sector. A couple of reasons we continue to grow credit sales. First, diversity of clients. Specifically, brands such as Wayfair, Zales and ULTA Beauty have helped to mitigate weakness in the apparel vertical. Second is the migration of sales from bricks and mortar to nonstore channels. Over 30% of our credit sales are nonstore versus the industry average of about 15%. Card receivables grew to just under $16 billion, an increase of 17% over the second quarter of 2016, consistent with our expectation of mid-teens growth for the year. We continued to drive operating leverage as operating expenses, expressed as a percentage of overall receivables, dropped from 9.5% in the second quarter of 2016 to 8.6% in the second quarter of 2017, a 90 basis point improvement.

The net loss rate was 6.2% for the second quarter of 2017, a 10 basis point improvement over the first quarter of 2017. The gross loss rate continues to trend with our delinquency forecast, which suggests lower rates in 2018. Our recovery rate is down this year. It was about 18% in Q2 of 2017 versus 23% in Q2 2016, pressuring the net loss rate, primarily due to lower pricing in the third-party recovery market. Ed will talk later about how we are stepping up our in-house recovery efforts to combat the lower pricing.

I will now turn it over to Ed.

Okay. Thanks, Charles. I'm thinking here we'll also have the wedge that was provided to you and that's just sort of illustrative of how we're doing against what we first actually introduced, I believe it was in October of last year. So if you will turn to Slide 9, which is the second quarter of 2017. Again, just real quickly, reiterating some of the points Charles made. Consolidated, plus 4. Bottom line, plus 4. We're looking for about that range on top line. And that's a little bit better than what we had anticipated on core EPS, where we had guided to more like flat.

Again, in terms of the businesses, Card Services growth continues to be quite strong. Epsilon, which is the second consecutive quarter of what we like to call repeatable growth, so we don't have all the big dips that we've had in the past, so those are the false starts and then disappointments. So it looks like things are going along pretty nicely there.

And then in LoyaltyOne Canada, some good news there. We -- the actual EBITDA margins up there came in, in the mid-20s, which is really where we want to be for the full year. So the model has been retooled. And now we're just getting ready to crank it back up.

Probably the one disappointing area was on the BrandLoyalty side, less so in Q2. We knew Q2 would be soft, but a bit more in terms of when we expect to bring on a number of these programs. We'll talk about that a little bit later.

Again, the 3 goals for this year that we've laid out over the last 6 months would be: one, Epsilon/Conversant, let's make sure we have repeatable growth quarter after quarter, both top line and bottom line. And we feel comfortable at this point that we have seen that -- we're seeing the daylight. And in fact, we were looking to see a little bit of acceleration on top line as we go into the back half. Number two was the big old question of credit normalization. Are losses going to continue to go up or are we finally beginning to see a plateau, and this is just a normalization process? And we're dead on track with the wedge that we put out back in October. So we can check the box there. And then the third was the retooling of the Loyalty program in Canada, and that is coming along pretty nicely.

So let's go ahead and turn to Epsilon and our full year outlook. Again, we talk about the repeatability of the performance. That's sort of the key thing we've been focused on, and the ability to deliver sort of, at least this year, that mid-single-digit sort of top and bottom line, which will hopefully begin to strengthen a bit as we move into '18 and beyond, as we finish the turn in the Technology Platform. But we did have a decent second quarter, both top line and bottom. It's the second consecutive quarter. And frankly, this hasn't occurred since back in 2015. So as we look into Q3 and Q4, we see that trend continuing.

The big sort of work we had to do this year was to confirm that we can retool and repackage the Technology business, which is about 25% of Epsilon's revenue. These are the big platforms that we build both out of the big database platforms and the big loyalty platforms. And if you recall, that got hit pretty hard last year as we realized that our pricing was not competitive and our time-to-market was not competitive. We repackaged all of that. And the revenue, which was down in that segment 13% in Q4 of last year, was cut to 7% in Q1 and 3% in Q2. And we are very nicely on track to get that to flat by year-end. And then we'll be up low single digits as we move into '18. So the turn there is for real. And that's certainly going to benefit us going forward. So that's a good relief.

During the quarter, and really for the year, the major growth drivers, obviously, the big digital media business, the Conversant CRM offering is really on fire and doing extremely well, as are both the Auto businesses and the Agency businesses. In the Auto business, you hear some weakness in car sales and stuff. That's not really where a lot of our focus is. We're more on after the sale is made, a lot of the communications and personalization that go out there to an existing car owner about what's going on and time to bring it in for this or that from a servicing perspective, so we expect that to continue to be strong. Also, we're onboarding a couple of major new brands, which should set us up nicely for '18.

Full year guidance, comfortably on track for what we've been talking about, which is sort of the mid-single digit rev. In EBITDA growth, I'd say the one tidbit of new news would be that we do expect top line to accelerate up to around 7% plus in the back half based on what we're seeing in the -- and how quickly the Tech Platform has turned. So that's probably the new news coming out of Epsilon/Conversant.

I'll finish up on this with Conversant itself. If you were to rip apart the various pieces and look at just the Conversant stuff that we acquired back at the very end of '14, what you would find is that, that business -- those businesses are running in the high single digit, which is exactly what the original acquisition model was based upon, so a little bumpy getting from A to B. We probably went through most of the alphabet before we got to B. But right now, it looks like that acquisition's beginning to pay off nicely, so that's where we are on that one.

Why don't we turn our attention to the next group, which would be LoyaltyOne. And again, this is where you've got a little bit of some messy numbers to sort of sort through. Again, in terms of guidance, no change. For Canada, we guided this year to about $760 million of revs and about $180 million of adjusted EBITDA. That would put our margins right around in the mid-20s. And if you recall, with the resets and the issues that came up last year from a legislative perspective, that knocked our margins down into the teens. I can tell you in Q2 and for the rest of the year, we're already back in the 20s, mid-20s actually, which is very nice. So we've put the changes through. They're working. And so we believe that, that model is once again going to deliver what we expect to on a go-forward basis.

Again, there were the questions out there in terms of was there damage done to the brand with all the noise last year, and the place you look for that would be on both the sponsor side, those are the big names that pay us to issue the miles, and then the collector side, which are the actual consumers who use the program. And as of right now, we feel very comfortable that we don't expect to see any attrition whatsoever on the sponsor side, which is great news. In fact, most of them, the general theme that we're getting from them is enough already, move on and let's get going. So that's good news there.

And then on the consumer side, which is sort of the issuance side, a measure of health is are people using it to drive additional sales, additional issuance of miles. We were down 4% in Q1. That's now swung to, really, minus 1%, almost flat, in Q2. And we're on track to be back sort of our long-term plus 5% run rate by year-end. In fact, we'll break our heads above water in Q3. So again, I think the program, after taking a couple of body blows last year, has come back nicely without any permanent damage. And we were able to retool the model.

Okay, BrandLoyalty. We don't really talk too much about that. It's been such a consistent grower since they joined the company back in '14. And in fact, if you look at the last 3 years, I think the numbers are high teens annual revenue growth and sort of low double digit in terms of EBITDA growth. Very high grower, consistent grower over the last 3 years. It looks like we've got a bit of a timing issue this year. To explain it a little bit, we've got 135 clients in the business over -- across 40 countries. Roughly 200 to 250 programs we run each year. Again, these are the grocers, and these are their sort of quarterly quick hit promotional-type programs to drive sales during the quarter. And they're heavily influenced by major events. Again, this is our international business. And so things like the Rio Olympics in '16 and the Euro Cup in '16 drove a lot of programs. We knew we didn't have that in '17. In '18, the World Cup, so we expect that naturally will bring on a lot of promotional programs. So the question is in '17, what made us think we were going to get that double-digit growth when we didn't have the big events out there helping to drive it? Frankly, we were hoping that the big agreement we signed with Disney would have happened a little bit sooner and we could have gotten some traction out of that. Again, that's the agreement that allows us to provide a lot of these programs with Disney-type merchandise across all of EMEA. And that, we expect, which is coming online in Q4, is going to be a strong driver for '18, along with the World Cup. So you're going to see sort of a very, very strong year in '18, less so in '17.

So what does that all mean? From a visibility perspective, we were looking at the big ramp-up being Q3, Q4 of this year. It shifted to Q4 and Q1 of this year and Q1 of next year. And so you've got a shift of really a quarter. We have strong visibility into Q4. We've got the program signed. We're looking at 25% plus revenue growth in Q4 and 40% EBITDA growth in Q4. And we expect a very strong jump-off in Q1 as a number of these programs ramp up.

So I think that, while disappointing that we didn't get these things wrapped up sooner, the good news is there's no issue from a business model perspective. We just have a timing issue on a business that usually has been very consistent on an annual basis year-to-year. So you get about $0.40 we were planning on in Q3 of '17. That's been shifted into our '18 guidance. So we've increased '18 guidance accordingly.

All right. Card Services, let's go to them. Receivable growth, 15% plus. Again, very strong. Pipeline robust. Tracking to another $2 billion vintage. Again, that means when all the signings have ramped up, they'll add $2 billion of portfolio growth to the business. And where are they coming from, right? This is the question we get all the time. And it -- they're coming from varied sources. Obviously, our focus has been in apparel and soft goods and home furnishings and jewelry. And there's a ton of wood to chop there. And -- but the type of retailer is changing a bit. And so you will see our announcements will be more of a combo platter of traditional hybrid, which is both store and online as well as pure online players. And in terms of the pure online players, these could be startups that weren't even around a couple of years ago. But also, there's quite a few sort of established names, very well-known names, that no longer feel the need to use department stores as their primary platform and are actually striking out on their own and developing pure e-commerce models. And that allows us to step in. And they would be perfect sizes for us as opposed to some of the monster department stores, which the portfolios are just too big for us. So we view it as a pretty good opportunity to continue to grow the business.

In terms of the financials, the yields are stable. We're getting nice operating leverage out of the business. And then probably the second big question over the last 2 years has been around credit losses and, "Oh my gosh, are they normalizing? Are they not normalizing? When do they stop going up?" Because that really drives a good chunk of how the earnings flow through the business itself. And we brought out last October what we call the wedge, which is essentially looking at the best future predictor of losses, which, of course, are delinquencies. So a certain percentage of delinquent accounts after 180 days are written off. And those can be very predictable over at least a 6- to 9-month period. What sort of amazes me is the fact that we put out the chart back in October, and we are dead on with the wedge. So we have looked at Q1, where delinquencies were up 50 basis points year-over-year. Q2 averaged 40 basis points over last year. And now we're entering sort of the fun part, which is Q3, Q4. Q3, you're going to see that thing drop pretty dramatically, and we're going to wind up about 20 basis points over. And then we'll be flat in Q4. And all that means is flat delinquencies means that loss rates will be no longer going up in '18 and will be flat to lower for 2018. And that's when we know that you have the Slingshot that we've talked about so much in earnings, which is you're not having to set aside all those reserves.

All right. In terms of any noise in the business, the principal loss rates, again, we thought that loss rates, gross loss rates right now are tracking up about 50 basis points, which is what we expected. The noise that you're hearing in the marketplace has to do with recoveries. And again, recoveries account for lowering the loss rates by as much as 20%, 25%, so they're important. And what you've had in the market this year is that there's a lot of paper that's being sold to third parties on the market. We participate in that program. We also do a bunch in-house. And what we've seen this year is a very, very soft recovery market. A number of reasons have been given. I don't really know which is the correct one. But -- so we're looking at a situation that we ran into back probably in '09 and '10, where we sort of swung using the external third-party market and decided we're going to do it all in-house and -- because we're going to get a better yield on that. And that's what we're going to do this year. So you're going to move from a model where you're using the third-party market, you get the sale, you book the recovery amount. It's you get the short-term benefit against the quarter. But because of the pricing, it's going to be a lower benefit than in the past versus if we swing it in-house, we hire our own people, we ramp up that process, we're going to get recoveries that are going to be back in the low 20% range, which is really what we want. And so that pushes out the benefit a little bit. But from a cash flow perspective, it's a no-brainer.

We're going to be tinkering with that for the rest of the year, which we're definitely swinging more to the in-house. How much more -- will it be 80%? Will it be 100%? We don't know yet. But we're going to tinker with it. We're going to see how that's going. We're clearly going to be going in that direction until the third-party market firms up, because we know we can get the type of recovery rates that we need by bringing it in-house.

What does it all mean? At the end of the day, from a full year guidance perspective, we certainly still expect mid-teens revenue growth on the top line. And then importantly, the sort of what I would call operating cash flow or what we call adjusted EBITDA net, which is -- it's a little bit of a funny term, but it essentially includes all these provision costs that we talked about, the credit loss provisions. It also includes our cost of funding the portfolio. So it's sort of what the business is throwing off from an operating cash flow perspective. And because it includes the recoveries and how much are in-house and how much is going to be third-party, et cetera, et cetera, we -- the simple thing that folks need to remember is regardless of how much we bring in-house versus sell on the third-party market, that key metric there is going to be growing 10% plus, and that's our goal. So we will tinker with the other stuff, but the net result of all of it is the yields are strong, operating leverage is quite good, delinquencies are dead on track, gross losses are dead on track, recoveries are fluctuating a little bit, but we'll manage it to the point where we'll do 10% plus on our cash flow growth.

Okay. Let's finish up with our '17 outlook. We have -- from a consolidated guidance perspective, we're increasing our revs from $7.7 billion to $7.8 billion, up $100 million, up about 9%. On core EPS, we're going to ding core EPS by the $0.40 of timing issues at BrandLoyalty that we've talked about, and so you'd have $7.8 billion and $18.10 for guidance for this year. And then you'll see the quarterly rollout. We have good visibility on Q4 in terms of the ramp-up of the Slingshot, so mid-teens growth in Q4. It's about a quarter before we normally would throw out '18, but we wanted to give people our initial cut. We've been working quite a bit in terms of where are the various businesses headed and in terms of our comfort level with growth in the Card business and credit quality in Epsilon/Conversant, the timing of BrandLoyalty and the retooling of the Canadian model, there's a bunch of pieces here. But at the end of the day, we feel comfortable at this point of actually putting the stake in the ground a little bit earlier than we normally would. And we're looking at a return to more than double-digit top line. We're looking to grow top line almost $1 billion to $8.7 billion or up 12%. And then on core EPS, we were looking at the mid-teens. You throw in the $0.40 from the BrandLoyalty timing and you're actually getting closer to 20%. We put in $21.50 or 19%. So plus 12%, plus 19% for us seems to be a doable and achievable initial cut at guidance. We haven't factored in things like what are we going to do with all the free cash flow, so we'll figure that out as the year unfolds.

So to sum up and then we'll turn it over for questions. Look, I think we're executing on the big 3 goals that we had this year, which was the Epsilon/Conversant, sort of that repeatable sustainable mid-single-digit top and bottom line. We're actually looking at a little bit of an acceleration in the back half on top line, so we feel good about that in taking that into '18. We're looking for our ability to grow through sort of the macro retail challenges that are out there. We're seeing mid-teens to high-teens portfolio growth. We're looking at opportunities with different types of retailers. So we expect that to continue through '18.

Credit normalization, it's been a long process. It will be almost 2 years since the normalization process happened. But based on the delinquency curves and the wedge, we're dead on with that, which means we're going to have flat to lower losses in '18. And then finally, you had our LoyaltyOne business, where Canada had some trauma last year. We had to retool the model. Are we -- could we keep all the sponsors? Are the collectors going to get reengaged? Everything we're seeing right now is yes. And can we return the model to the mid-20s type EBITDA margin? And we already saw that in Q2. So check the boxes with all 3. And of course, with us, it just wouldn't be Alliance unless we had a bit -- one fly in the ointment. And this time, it's the business that has consistently done strong double-digit top and bottom line growth. We do feel comfortable, however, that's a timing issue. And we think that you're going to have a really significant 2018 out of BrandLoyalty. So overall, we feel good. That's why we're giving our '18 guidance a little bit early. But right now, we think things are heading in the right direction.

So I think we took a little bit longer, or I took a little bit longer than necessary. So let's open it up for questions, please.

(Operator Instructions) Your first question comes from the line of Bob Napoli with William Blair.

I guess the -- looking at the business and the BrandLoyalty business has always been pretty volatile by quarter. And I guess that's continuing a little more. But do you feel, Ed, that the business has become more complicated over time than you'd like it to be? I mean, are all of the pieces, do they fit? I mean, your multiple, obviously, is way below where it's been historically, your valuation, your PE multiple, and maybe some of that is the additional complication of the business. I just would love your thoughts around that.

Sure. It's a fair question. I think a business like BrandLoyalty, which has sort of these shorter-term hits in terms of loyalty programs, I don't think it's gotten more complicated. It's always been, as you correctly pointed out, very choppy by quarter. For the past 3 years, that choppiness sort of has always evened out on an annual basis to give us that double-digit growth. This is really the first year in 4 years that they've been part of Alliance where we're seeing the chop actually pushes it into the first quarter of the following year. So from our perspective, we can see the visibility. We see the pipeline. We see the programs that are getting booked. And we know that with the World Cup coming and with the Disney program signed, it's going to be a very big '18. It's unfortunate that it hit -- had to hit our '17. But no, I don't really think it's gotten more complicated. But from a volatility perspective, it's a great double-digit business growing annually every year. It's got a huge market, but it doesn't have the type of visibility the other businesses have.

Okay. And just my follow-up question would just be on credit. It looks like the delinquencies, as you said, are right in line with what you had laid out a year ago. The conversion -- so are you expecting higher credit losses now in the back half the year as you get that -- stop selling as much and then bring it in-house? So should we expect charge-offs to stay in the 6% range or a little bit above that for the rest of the year as you make that conversion?

Yes, I don't think we really -- can really pin it down. What we're trying to do is basically say, like, the wedge is dead on. So we know we're going to be heading toward flat loss rates or better in '18. The gross loss rates were up right where we put them at 50 basis points for the year. The recovery is going to vary by how quickly we can get the in-house stuff moved up. So the comfort that we can give people is the fact that regardless of what we do in-house versus third party, the overall earnings of that business will be up 10% plus. So we want to give that comfort. And then look, we're not going to be stupid about it and pull the rug completely out of doing third-party sales, but it's -- it really benefits the company if we can be a bit more patient, deliver our 10% plus on the earnings side for that business and get it in-house. So stay tuned.

Your next question comes from the line of Sanjay Sakhrani with KBW.

I guess, Ed, just a question on Epsilon. Obviously, you guys have pretty decent momentum, it seems, as we look out to the intermediate term over the next several quarters. But as we look ahead to next year, are there any other areas that we should be concerned about? Because it seems like there's been a lot of volatility within the various segments of that business. Is there any one that you're particularly concerned about at this point?

Well, the big one, obviously, Sanjay, was the Tech Platform business, which was the -- about 1/4 of the combined Epsilon/Conversant. And that was a source of a fair amount of angst last year because that used to be the workhorse for the entire segment. As you know, it used to pound out 8%, 10% growth. And then we got flat -- caught flat-footed with sort of pricing was too high, delivery time was too long, it was too complex, it had too many bells and whistles. So that's been retooled. At the current rate that we're seeing that business turn, which is a little bit faster than we had anticipated, and the backlog of orders that we're seeing for these more shrink wrapped offerings from tech, I've got to tell you, I think between the Tech Platform getting its head above water as we exit this year, it should do low single-digit growth in '18 versus sort of down 14% last Q4 and about flattish in the third quarter, probably. So that's no longer the big worry. And so if that can stabilize, which it has, and actually grow a little -- Auto, again, which is more on existing autos, those big programs look solid and the big CRM business, those 2 are the -- are really the big drivers of sort of that double-digit growth in those 2 businesses. And then the other businesses are sort of more mid-single-digit, the Data and Agency and stuff like that. So it's a very long-winded answer to say we don't see any trains coming at us right now. And so can we finally move Epsilon/Conversant from sort of that mid-single digit and start creeping it up into the plus 7, plus 8 model that we really wanted when we took on Conversant? And that's sort of where we're -- what we're shooting for.

Okay. And my follow-up is on the Card Services segment. You guys talk about the stronger revenue growth. When I looked at the -- where revenue came in relative to our expectations, those were a little bit weaker. And it seems like it was driven by this uncollectible fees and finance charge billed. Maybe Charles, you could just talk about your charge-off outlook, which is a little bit more noncommittal, and sort of your expectations for the yield given this phenomenon. And then just on the expenses in that Card Services business, I guess, the run rate year-to-date improvement in that business has been quite strong. And obviously, you're expecting a little bit of moderation. But is there anything specific that drove that improvement in the first half?

So let's first start off with the gross yield that Sanjay is referring to. If you recall, we keep an allowance for principal receivables, meaning receivables we don't think we'll collect. We also keep a reserve for billed finance charges that we deem uncollectible. There's a direct correlation between the 2, except that the billed finance charges are of those against revenue, so it's a gross yield impact. With that direct correlation, with loss rates being up in the high watermark in Q2, the net reserve for your billed finance charges will go up, and that gross yield is negatively affected. And then when your loss rates come down, that pressure abates and your gross yields go up. So what we see, it's a short-term issue. It's a direct correlation with the loss rates. As the reserve, basically, the loss rates come down on principal, you'd see that reserve come down and expects gross yields to come back up in Q3, Q4. On the OpEx, we do think we can get 40, 50 basis points of improvement as a percentage of average AR. Some of the improvement this year is just timing, meaning we're shifting some marketing out of Q1, Q2, Q3, Q4. That's why we've done really solidly in Q1 and Q2. We'll give some of that back. But 50 to 60 basis points is a realistic target for 2017.

And I think, Sanjay, also, to make sure, we certainly expect losses to be down in the back half versus the front half. It's just a little unclear as to how much based on what we do with recoveries. But earnings themselves should be up that sort of double digit in that business.

Your next question comes from the line of Darrin Peller with Barclays.

Credit sales, I think came in about 6% growth versus the high single-digit trends I think you guys were expecting. And yet, the organic growth of the receivables side continued to be even stronger than that. Can you talk through the dynamics there in terms of what you're seeing in terms of the overall purchase volume trends and what your expectation are for that embedded in your guidance for throughout this year and next? And then if we should expect that gap organically between the purchase volume growth and the receivables growth rate to persist, are you including also Signet in the portfolio -- or in the revenue growth outlook? I know it's not as a full year impact, but some benefit to '18. And then lastly, on the credit side, the cost -- I'm just curious what bringing recoveries -- the collection efforts in-house do to the cost structure of the business.

Sure. I'll take part of it, then I'll kick it over to Charles. Yes, the relationship between sales and portfolio growth, you will have periods where sales growth is higher than portfolio growth. And you'll have periods when portfolio growth is higher than sales growth. So it's not -- they're not in absolute lockstep. This is -- what you'll probably see as the year progresses is you'll begin to see that narrow a bit. We do expect to see our credit sales walk back up to the low double digits as we move into the back half, based on what we're seeing out of the new -- it's primarily from the new clients spooling up. And so you're going to see that gap narrow as the year progresses.

Yes, what I'd add to it, Darrin, you talked about Signet. Since it's coming in so late in the year, coupled with purchase accounting, it's really not going to drive a lot of lift to revenue or profitability. So it's really more of an '18 event for us. To your other question on the core, the core growth in credit sales is consistent with what we thought for the year, basically in that low single-digit range. So there's really been no change or modification there. On the cost of bringing things back in-house, we have factored that in, in terms of the guidance we're giving you of 50 to 60 basis points OpEx leveraging. So there's really nothing incremental. If Ed decides to bring it 100% back inside, we're good. Or if we keep it somewhere in between, we're good. So don't expect that to be any additional pressure in terms of their profitability for the year since we've already factored that in.

Yes. And to give everyone sort of a sense of why the recovery thing is so important to us in terms of realizing what the true values are on the recovery side, there's sort of this breakeven point of do you farm it out or do you crank up other resources in-house and do it. If we can book somewhere in the 20s, let's say recoveries are between 20% and 25% on losses, then we're relatively indifferent when you do the math between in-house versus pushing it to a third party. But the outside market this year has brought this rate down to more like in the mid-teens. So it's fairly dramatic and it behooves us, because we know in-house, because we're also running in-house programs that we're collecting in the 20s. So this is found money once we bring it in-house. So that's sort of the trade-off.

And just to be clear, I mean, just to close the loop on that, you are including some economic benefit from Signet in '18's guidance, but not additional portfolio acquisitions?

That's correct.

Okay. And just a quick follow-up on the AIR MILES side. Just remind us again, or reiterate the visibility you have. I mean, I understand your sponsors are sticking with you and saying they want to do more. But I guess, from a consumer standpoint, you talk about the brand not being damaged from last year. Just more evidence you can give us that we should expect AIR MILES issuance to rebound. Is it just -- have you really surveyed the population, the client base?

Yes. Look, that's a fair question. And to say the brand hasn't been damaged, I probably misspoke. I think there definitely was damage. I mean, it was terrible press, and clearly, the political environment was such that it was open season on the AIR MILES business from the government's perspective. But from the consumer side, the best metric that we look at is are people engaged. What's the activity rate? And if we've got 2/3 of the entire country of Canada active in our program, we'd say all right, is that number staying stable? Or is it down? And as you started off the year, quite frankly, it was down. And we were, I guess, the activity down, what, Charles, 4%, 5%? Then it was down 2%. The last -- we don't even have to do surveys. You just look at all the active account holders. We're back to flat. So that would suggest -- that's our best predictor of saying hey, it looks like people are over it and are moving on. But I can tell you, I'd rather not go through that one again.

Your next question comes from the line of Ashish Sabadra with Deutsche Bank.

Just a quick clarification. So despite the push-out in BrandLoyalty, you were still able to increase the revenue guidance given the momentum in the rest of the business. But the EPS guidance, that was pre -- was mostly because of the fixed costs in the BrandLoyalty, is that right? That isn't really any other pressure in other parts that business, which is putting any kind of pressure on the earnings?

Yes. We'd say, Ashish, that we're not really seeing any pressure against Epsilon. If you look at it, we've raised the revenue growth, we said we're going to have some costs associated with the onboarding of 2 new clients, which is why we didn't really push any incremental EPS from that. And that's not unusual. If you remember back 2 years ago, we onboarded a big client. You incur all of the costs to get them onboarded before you get the revenue stream coming through. With Card, what we're basically saying is you've got some upside on your OpEx. You've got maybe a little upside on your gross yield. But conversely, you've got a little risk in the recoveries which would mitigate it, so we don't want to pass through any increase there. So then if you look over BrandLoyalty, when we started the year, we really thought BrandLoyalty would do about EUR 115 million in EBITDA. It's tracking closer to more like to EUR 85 million, so about EUR 30 million delta in euro, which equates to the $0.40. And so that's why we're saying it's really an issue where we've got good performance with Card. Epsilon, not ready to pass anything through yet. But as Ed talked about, we've got very good visibility with BL. It's not coming through in Q3 the way we expected. That's the shortfall we're talking about, $0.40, and we're just slipping it into 2018.

Yes. Let me sort of be crystal clear on this one. It's something where we spent a lot of time internally talking about if we wanted to stay with, in our Card business, the third-party sales even at the lower rates, let's face it, you could have come up with that $0.40 to offset the pushout of BrandLoyalty. But we feel, from a business perspective, that's just the wrong way of thinking. And so I know people are not going to be thrilled that there's $0.40 that they're pushing out into next year. We were very thoughtful in what we did. And there was potential of overperformance -- significant overperformance in cards if we kept doing what we were doing. But we do think that using some of that overperformance and pushing out to in-house recoveries is the best thing for the business. It's going to benefit '18 and beyond. And it's real cash flow. But it can't be used to cover off the $0.40 in BrandLoyalty this year. About as plain as I can put it.

That's very, very helpful. And maybe just a follow-up question. So as we think about the impact of the recoveries, how should we think about the provisions in the back half and also allowance for loan losses reserve rates for the back half of the year? Should we see more provisions in the third quarter and then maybe taking -- more [heading] off in the fourth?

What I would say, Ashish, we are still looking for Q2 being the high watermark for net loss rates, so dropping in Q3, Q4. If you look for, the second quarter, about a 6.65% reserve rate. What I would expect, it could drift up a little bit in Q3, but it will end Q4 right around that 6 -- 6.6%, 6.5% reserve rate. So I think we've pretty well got the rates where we expect it to end the year. And now you've just got a couple of quarters for it to flow through.

Your next question comes from the line of Wayne Johnson with Raymond James.

I have 2 questions, one regarding Epsilon, and the other one regarding capital structure. So on the new Tech Platform, Ed, is it fair to say that the coding is complete on this? If you had to give it a percentage of completion, how far is it completed? And when -- if it's not, when do you think it will be?

It's -- yes, it's done. We're selling and we're delivering. So there's no more from that perspective. Now it's just a question of completing the deliveries and booking the revenue. So that's why I think you're going to see us get our head above water before the end of the year. So the work's done.

Okay. I appreciate that. So just a quick follow-up on that. So how many customers have you boarded on it today? Or is that just an -- that's in process, as we speak? I'm just trying to get some more clarification on that.

Yes. I mean, we probably have a fairly robust pipeline. I would say you probably have 12 to 15 that are either have been delivered on or in the process of getting spooled up, but that's a solid number for us.

Okay, that's helpful. And then just on the capital structure side, do you still think it makes sense to keep all 3 divisions, all of your business units together? Or do you foresee some point in the future a reason to split the company up?

That's -- people have asked that along the way. And it's going to be the standard answer, Wayne, of if this model isn't getting traction over a period of time in terms of value, then we're clearly compelled to look at other structures. Right now, frankly, if we are correct and this acceleration and Slingshot happens in '18, frankly, we would be disappointed if people weren't pretty excited about it. And we're getting closer and closer to it. We'll see what people think about it once we start really cranking out the double-digit growth on a consistent quarterly basis. Hopefully, that will get folks excited. There's a school of thought on both sides. I mean, as we look at it now and you look at all the disruption that's going on in the various industries that are out there, and what you're really seeing, right, is you're beginning to see tech beginning to -- or continuing to disrupt verticals that no one had even contemplated a couple of years ago. And if we were to look at, for example, the Card business and the retail space and everything else, if you believe that disruption is occurring there, which we do, we have the type of structure where you have all this technology, which is in the form of all this SKU data and all those unique IDs that we use at Conversant to identify folks and their proclivities online. And so the question is can we have -- can we be the tech-type provider that is the big disruptor as it comes to the card industry itself? And frankly, I think we can. There's so much fragmentation and disruption going on in just the retail space that our ability to bring data and SKU level information, analytics, the digital distribution channels, the omnichannel-type distribution approach, is something that's pretty compelling to these retailers today. And the ones who are breaking off from the big department store platforms, frankly, they're looking at the Tech side of us while they're looking at the Card side of us. So right now, it kind of makes sense. But we need to execute on the Slingshot.

That's really helpful. Would you say that these services at Epsilon and Card Services to some extent as well, would be in more demand because what's happening with Amazon coming into the physical retail world with the potential acquisition of Whole Foods and potentially others to follow?

You kind of faded on me, Wayne. Could you say that again?

Would you say that Epsilon and Card Services could be in more demand to fight off the impact of Amazon getting into the physical retail market?

Yes. I mean, I don't just say Amazon, I mean, all the big tech platform players who are disrupting more traditional verticals. There's no question that when we have discussions with clients or prospects today, let's just take in the Card business, it's less and less about tell me about the Card product. And it's more and more about how does that help me have a differentiated product, how does it help me reach the consumer across all channels, how can you provide sort of that personalization approach to the consumer that these big tech platforms are promising. And so that's why I see sort of this -- when people are down about where retail is going, what we're looking at is actually all these hybrids, brand-new e-commerce players, those splitting off from department stores, for us, it's actually, this type of volatility and disruption is driving a tremendous amount of business for us, and that's just in the retail space. And as we go into other verticals, that's right in Epsilon's sweet spot. So yes, I mean, from what we're seeing, this disruption is probably a good thing for us. We're going to do one more.

And your final question comes from the line of Larry Berlin with First Analysis.

Two quick things, I hope. First of all, on BrandLoyalty, how is the U.S. and then Canada deployments going? And are you seeing strong pipeline building in those 2 countries going for next year?

Yes. In Canada, you've got the benefit of having a 20-year presence, a 25-year presence in the country. So we're in pretty good shape in Canada in terms of the programs that are there, and they're up and running. There's either 4 or 5 brands that we have programs with. In the U.S., it's a question of when does the first domino fall. We had -- we have one grocer that we're doing now. We have a couple of pilots. We have a couple of calls that have come in since the Amazon-Whole Foods thing popped up. So I think, again, we've never really been able to find a way to penetrate the grocer vertical in the U.S. Every other vertical, we got, but grocer has always been tough, I think because the margins are so tight. But there's a new sense of urgency out there, so I would expect a couple of these things to fall. But we're not relying on some monster deal in the U.S. to push us through '18. Our bread and butter has been the 40 countries that we're in, in Europe, Asia, Latin America. Now it's Canada. But I do expect to get some traction in the U.S. Another one or 2 names would be nice. And if one of the big ones comes into the tent, then it gets fun.

Cool. Then on a totally different topic, stock buyback. What are you guys thinking going forwards in this year and next year for resuming the program and so forth?

Yes. I'd tell you, Larry, we completed the $500 million authorization this year. I'm sure we'll do at least another $500 million authorization in '18. I'd say at this point, we've not really decided do we do another buyback authorization for '17? Do we look to pursue a little bit of M&A? Or do we just look to pay down a little of debt? I'd say at this point, we're just being flexible around the remaining free cash flow utilization for '17.

Thanks.

Bye, everyone.

All right. Thank you. Bye-bye.

This concludes today's Alliance Data Second Quarter 2017 Earnings Conference Call. You may now disconnect.