Bread Financial Holdings Inc (BFH) 2016 Q4 法說會逐字稿

Good morning, and welcome to the Alliance Data fourth-quarter and full-year 2016 earnings conference call.

(Operator Instructions)

It is now my pleasure to introduce your host, Mr. Edward Sebor of FTI Consulting. Sir, the floor is yours.

Thank you, operator. By now you should have received a copy of the Company's fourth quarter and full year 2016 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. Thanks, Eddie. Just to -- as we continue to evolve over the years, the commentary we received back is to have management commentary shorter, so that we'll have more time for questions. So we're going to try to keep management commentary to no more than 20 minutes. That being said, I'll turn it over to Charles, who is Charles Horn, our CFO, who is going to talk about our fourth quarter and full year results. And then, I'll wrap up with 2016, and then shift to 2017 outlook. So Charles, go ahead.

Thanks, Ed. Let's flip over to page 4. If we look at the results, it was a little bit of a noisy fourth quarter for LoyaltyOne Alliance Data due to expiry. Not only did we have more AIR MILES redeemed than expected due to heavy media attention to the issue, but also had a new law enacted in Ontario, Canada which forced a one-time charge. Remove the noise though, and it was a terrific 2016.

Let's start with pro forma revenue of $7.4 billion, up 15% from 2015. We estimate that elevated redemption activity in Q4 added about $175 million or 3 points of growth to 2016. Essentially AIR MILES that would have redeemed in 2017, were instead redeemed in 2016 ahead of the expiry, creating a pull-forward of revenue recognition into 2016.

Adjusted for that, revenue was approximately $7.2 billion, up 12% and consistent with the guidance. The core EPS was $16.92, up 12% from 2015, and $0.02 better than guidance. In calculating core EPS, we added back the $242 million related to elimination expiry, but not the full breakage reset.

Let's turn over, and talk about LoyaltyOne. Adjusting the breakage reset and revenue pull-forward out of the numbers, we estimate that revenue and adjusted EBITDA for LoyaltyOne both increased about 4% for 2016. Let's begin with BrandLoyalty, which had a terrific year with high single-digit growth in both revenue and adjusted EBITDA. Expansion efforts in Canada continued to exceed expectations, and the US market could be a growth driver in 2017 based upon early pilot successes.

For AIR MILES, it was a tumultuous year with all of the publicity surrounding the upcoming miles expiration date. AIR MILES issued increased 1% for 2016 below our expectations to about 3%, as sponsor promotions dropped in Q4, as there was simply too much noise in Canada to be effective. We believe the issuance growth will return to 3% range in 2017.

In addition, we reduced the breakage rate for the AIR MILES Dream program from 26% to 20% at the end of 2016. The first 1.5 point reset resulted from higher than expected Q4 redemption activity. This is a normal change in accounting estimate and is reflected in our operating results. However, the second 4.5 point reset is different, in that it was driven by the enactment of a new law, a one-time event, thus we added it back for adjusted EBITDA and core EPS purposes.

Flipping over to Epsilon. It was a disappointing year for Epsilon as revenue increased 1% while adjusted EBITDA decreased 6%, both substantially less than our 5% growth expectations for 2016 Breaking down revenue by [P] offering, let's start with the good. Collectively auto, CRM, affiliate, and data grew revenue double-digits compared to 2015. In addition after a slow start to 2016, due to the pull-back of a large client, agency's growth rate turned positive in the back half of the year.

Let's now shift to the problem childs. Technology platform growth turned negative mid year due to some client attention -- attrition. In attention, in addition, the non core offering, essentially the old value click business was a consistent issue all year, as we endeavor to pivot it towards a more data-driven solution. This offering alone was a 3 point drag to Epsilon's revenue growth.

Turning to adjusted EBITDA. We started off the year in a big hole, down 22% in Q1. By significantly lowering our expense base, we cut that to a minus 9% in Q2, turned to flat in Q3, and then [up] 3% in Q4. We believe our lower expense base will enable us to be more price competitive on certain product offerings going forward.

Let's now flip over to card services on page 7. It was another stellar year for card services with 24% growth in revenue and 14% growth in adjusted EBITDA net, despite over a $200 million build in allowance for loan losses. Credit sales increased 16% for the fourth quarter and 18% for the year. Growth in core credit sales which are our pre-2013 programs was flat for Q4, and up 4% for the year, driven by tender share gains of about 150 basis points.

Growth in the core weakened as the year progressed, as comparable sales growth turned negative at several large apparel-focused clients. Gross yields after being down in Q1 through Q3 due to card holder changes made to the program in 2015 stabilized, and increased slightly in the fourth quarter of 2016. We believe that these changes have fully burned in at this point.

We continued to achieve operating leverage. Operating expenses up 21%, compared to revenue growth of 24% for 2016. As a percentage of average card receivables, operating expenses decreased 50 basis points compared to 2015. Average receivables growth remained robust at 24% for 2016. Growth slowed over the course of 2016 due to softness in core spending as discussed before, resulting in ending card receivables growth of about 20%.

The principal loss rate was 5.1% for 2016, just slightly higher than original guidance primarily due to the slowing in receivables growth rate. This creates what we call the denominator effect. Importantly, losses continue to track in line with our delinquency expectations for 2017, as we show in the wedge. With that, I will turn it over to Ed.

Great. Thanks, Charles. If you'll turn to slide 9, where it says 2016 wrap up, a couple things we want to walk you through. There's a couple of moving pieces here, with the noise from the Canadian business. If you look at revenue, revenue was up to $7.38 billion or up 15% when you exclude the charge. And so, that was a couple hundred million higher than our guidance of $7.2 billion.

When you put in the charge, then your revenue is decreased to the $7.14 billion, up 11%. And so, you start at sort of a $7.4 billion. The charge takes it down to about a $7.2 billion. What we should also say is -- if you can't really look at just the $7.4 billion. That does include a pull-forward of about $175 million that Charles talked about, with what we call the run on the bank in the Q4. We had expected to come in the year about $7.2 billion. It looks like we came in about $7.4 billion, so the real run rate is really about a $7.2 billion, once you factor out the run on the bank.

The net result is, I think if you were to look at sort of what is the right number, I look at the $7.2 billion, up about 12%. Core EPS $16.92, up about 12%. That factors in sort of the effect of both the charge, as well as the pull-forward, and that should give you a good starting point. Double-digit revenue and core EPS growth, despite absorbing a 12 point drag in earnings per share from the increase in the card services loss rate.

In terms of the use of cash, we were fairly busy. We did about $800 million in repurchases, $350 million to take out the final 30% of our European-based BrandLoyalty business, another $500 million to support growth in our card business of roughly $2.7 billion. We did establish a 1% dividend. You wrap it altogether, and we still have leverage below 3 times. So again, it attests to the strong free cash flow of the Company.

Let's go through the businesses, slide 10. LoyaltyOne, we printed revenue up 17% before the reduction in the -- due to the expiry charge, and adjusted EBITDA was about up 4%. Again, the revenue included about $175 million in revenues pulled forward from 2017 into 2016, as people were trying to cash in their points before expiry. And obviously, this was in front of the law that was changed up in Canada. So if you were to look at adjusted numbers again, and take the $175 million pull-forward out, you're probably talking about plus 4%, plus 4% for top and for EBITDA.

BrandLoyalty continues to be a bright spot, with strong growth from existing and new markets. And in North America, Canada, they are firmly established in Canada, and in the US we won our first client, and we expect to have some announcements that are pretty exciting as we go through 2017.

For AIR MILES itself, as Charles mentioned, Ontario's Parliament enacted a new law after -- let's see, it was, we had a five year expiry, and after four years and 11 months, it was determined that that was unacceptable. And as a result, the Parliament passed a law prohibiting us from completing the expiry process. And as a result, resulted in the one-time charge to reflect loss breakage revenue. So overall, if you were to do the math, LoyaltyOne combined did about a [$1.580 billion] in top line, and that's up 17%. If you want to take out the pull-forward from 2017, your starting point would be like $1.4 billion, and that would have been up 4% versus the 2017 that was reported.

All right. Epsilon, as Charles mentioned, revenue up 1%, adjusted EBITDA, down 6%. And again, we have plenty of bright spots, and we've got a couple of spots that need some hard work. The core, as we've defined it, is 92% of total revenue, and effectively all of the profit. If you looked at the core, the revenue was up 4%, which did include absorbing a 5% drag from the technology platform business.

The India office ended the year at a full scale run rate of about a 1,000 associates. So we have now hit the point where our office in India is at critical mass, which means we're beginning to see the flow through in terms of the leverage on the labor side. And so, we have completed part one of two for the turnaround in the technology platform business, which is the expense side. We think the expense side is in good shape, as we head through 2017. And now we got to work on making sure that our products and our time to market are consistent with the competitive landscape.

So what's left over, you have the non core revenue which used to be about 10%, 11% has -- is now only about 8% of revenue, and did cause a 3% drag on total revenue growth. But obviously, it's becoming a smaller and smaller piece of the business.

If we turn to card services on slide 12, obviously a very, very strong year, revenue was up 24%, receivables up 24%, adjusted EBITDA up 14%. I think that the key thing here is, you had those types of growth while absorbing a 60 basis point increase in loss rates. And to put it in perspective, that's about a $165 million hit to EBITDA or $1.85 a share. The net result is, if we grew earnings a share at 12% that included absorbing another 12% from the normalizing loss rate. So we would have been somewhere in the mid 20%s this year without the loss drag.

So again, this gets down to our thesis, that once credit normalizes, this thing is going to be a slingshot in terms of the acceleration in growth. We are nicely on track to have this normalization take place during 2017, with the key driver there being the delinquency shrinkage year-over-year. And as soon as that thing anniversaries, you know that you've hit the normalized part. So that's the thesis, and I think we're tracking nicely to it. We mentioned before, it was going to take two years to get this thing normalized. We're through certainly the first full year, and I would say over the next six months that should pretty much clean it up.

We had a huge year in terms of new signings, $2 billion vintage. The vast majority are start-ups, which again will cause -- you won't have the immediate benefit to the portfolio growth rate, but what you have, as it comes in over a two to three year period as they all spool up. But you've got some names here that are going to really add some juice to the file. So very good names, as a huge year of signings. And then finally, a150 basis point increase in tender share, which essentially means if our retailers, their sales do X, then we're going to do better than that, in terms of sales on our card.

All right. Let's turn to 2017 outlook. If you look at revenue, on our last call we said look, we wanted to do [10%], [10%] on both top and bottom, and what the -- that would have come out to be sort of your 7.8[%], 7.9[%] in terms of growth rate. And then, what we try to do here is, we're saying, look, we came in hotter than that for this year, hotter than the $7.2 billion. We actually came in about a $7.4 billion, but that did include really pulling from 2017 about $200 million of redemption revenue, which carries virtually no margin, but it is revenue. And that's why from a guidance perspective, our $7.2 billion for 2016 came in like a $7.4 billion.

For 2017, if we were guiding to about a $7.9 billion, obviously, we had that surprise sort of run on the bank in Q4 which increased 2016. It's taken it out of 2017, so that brings sort of your run rate for 2017 to about a $7.7 billion. I know that it's a lot of noise, but at the end of the day, you had this sort of pull-forward which is pulling out of 2017 into 2016, and we want to make sure that everyone adjusts their models to reflect that. On our core EPS, it didn't really affect that, so we're sticking with -- we should go from $16.92 to about $18.50, which is about 10% growth rate. And that's pretty much the guidance that we're going to have for the year. We feel good about it.

In terms of growth rates, if you were to sequence these things, you would see that the whole slingshot thesis behind the model is really driven by the fact that as we sort of exit the final normalization part of the credit cycle, and we start rebuilding the model that we need in Canada, you start moving out of sort of the mid singles flattish type first half, and you start ramping up in Q3. You hit full run rate really in Q4, and then you should have a huge jump as we go into 2018.

Again, if you're looking at our history over the last 10 years, revenue has grown consistently at 13%. EBITDA is up 12% a year for the last 10 years, and earnings are up 17% in the last 10 years. So we've been a little bit shy of that 17% in 2016. And in 2017, you would therefore expect in 2018, you're going to be well above that. So that's again the thesis that we have here at the Company.

So if you'll turn to the next slide, on the consolidated slide, we expect all segments to contribute. We expect solid growth across the businesses while absorbing the final credit normalization and revamping the AIR MILES model to be more consistent with the new laws in Canada.

In 2018, what you're going to have, it's sort of just math, is the slingshot where you're going to have stable loss rates, which turns into an accelerated growth rate on your metrics. If you look at the different businesses, LoyaltyOne, we expect BrandLoyalty to continue to have strong performance approximately 10% top and bottom line. We think all areas, Europe, Asia, and the US will be driving that growth, so expect a strong year from BrandLoyalty.

In Canada, again, the -- I think we've sort of have beaten this thing to death, but we'll keep going. The pro forma Canada was up about 24% in 2016. That did include a $175 million pull-forward, so maybe the run rate that you use is more like a [$760 million]. We expect that to be relatively flat in 2017, and we expect the margins which will go from 27% to 24%.

Essentially, we're recapturing about a little over half of the margin lost from the new law. But the way it will work is as we implement the new model in Canada, that will drift in as the year progresses. And as a result by Q3, Q4 you're essentially going to be well above the margin rate in Q1, Q2. So again, feeding into -- we're looking for that sort of momentum pick up and slingshot effect as we go into 2018. AIR MILES issued looks about 3%, and then we expect on a full year basis, the margin to be fully recovered in 2018.

On Epsilon, it's time for Epsilon to step up and complete what needs to be completed, and return to a consistent revenue growth model. We expect the core to run about 5% growth, overall about 4%, so there's about less than 10% of the business that's viewed as non core. EBITDA we also expect that to grow 4%, 5% as we go through 2017.

What gives us comfort as we look at it, well, the big piece we need to make sure gets turned, is the technology platform which is about one-quarter of our total revenue. The step one, which is what we've been doing the last 12 to15 months has been to significantly lower the expense base. That's done now, so you can check the box on that. We finally got that done. We've got the duplicative costs out of the system, so we should have nice leverage there.

And we're right now in the middle of step two, which is we need to standardize the product. And we've got to have a much faster time to the marketplace. That's just the way the market has changed, and that's the goal as we move into and through 2017. What's left over in the non core piece, again was 11% of the business in 2015, 8% in 2016m and 6% in 2017. So again, less and less. It should really be not material in terms of the drag on revenue, no more than 1 point in 2017.

Card services 2017 outlook, again we expect nice growth. Card receivables growth of just about $2.5 billion, which is about 15%. Now you will note that it is less than the 20% growth that we did in 2016. So you do have a deceleration in the growth rate for 20% to 15%. Primarily that is comes back to the fact that our vintage that we signed are primarily start-ups, which means the growth will reaccelerate as we move out of 2017 and into 2018. We have fewer portfolios that we brought on -- in fact, we have none, and we have really all start-ups that should have nice yielding receivables as they grow.

Pipeline is robust. It's a good market to be in. We expect another $2 billion vintage year. We expect gross yields to be stable. We're not seeing pressure on the competitive side. We like where we are, staying in our sand box, and it seems to be working.

We're going to get a little bit of operating leverage as we continue to grow, and then, of course, the all-important question of what's going on with credit. And the credit normalization as we've talked about is right on track. And we talk over and over again about the wedge, again delinquencies being the best predictor of future losses. So as delinquencies begin to anniversary with delinquencies rates of last year, that means losses will do the same within three months to four months.

So in Q1, we expect delinquency rates to be about 50 basis points over last year. Q2 that's going to go down, and then as we move into Q4, it's going to be flat to prior year. What does that all mean? It basically means you've completed the normalization process, and your loss rate is going to follow shortly.

Principal loss rates for this year, we expect in the mid 5% range. We're going to drift to about 6%, 6%-ish in the first half, as again we talked about the denominator effect of the step down in growth in the portfolio; however, we expect our loss rate to be below 5.5% by Q3. So again, it looks like we are tracking nicely. So we expect the first half to be negatively impacted by the slowing growth and the card receivables which is the denominator effect. And overall, even with the higher loss rates, we still expect 8% to10% earnings growth from this business.

If you were to look at guidance, and looked at our sort of our history. At the beginning of 2015, we had guidance of mid 4[%], we came in at 4.5[%]. Beginning of 2016, we gave guidance of approximately 5% and came in at 5.1%. And today, we gave guidance somewhere in the mid 5%, so overall we're pretty good on the guidance side. But to allay any other concerns, our guidance for today for our overall revenue and earnings also includes some squish room, if you want to call it or buffer, should the rate drift up a little bit above where we thought, just to give us the comfort that we need on the guidance side. If it doesn't, then that's great news, and we'll flow that through as the year goes. In 2018, again the whole thesis rests on flat earnings -- I'm sorry, flat losses compared to 2017, which should show your slingshot and accelerate your growth.

All right. Finishing up, if you look at the outlook, closing the wedge so to speak. Essentially again, this is what we track to. This is how we run our business. We know that losses will have fully normalized, and will therefore be flat to prior year, shortly after delinquencies are flat to prior year. So we expect the wedge to close as the year progresses. Everything we're seeing suggests that's the case. And again, if that happens as we expect, then you're going to have a very nice acceleration of earnings as we go into 2018.

Final commentary for LoyaltyOne. Look, in the past year, it was a very difficult year, and we went through a very disruptive event. We're past that. We've got a plan in place for the model. Issuance has returned to normalized levels, and redemptions have come way back down to more normal levels as well. So the game plan is to comply with the new law, while we retool our model. We have good visibility on how we're going to get that margin back, and we expect really the last half of 2017 to reflect the benefits of that retooling.

On Epsilon also, a tough 2016. We have a good plan to return to mid single top and bottom in 2017 which is sustainable. That's the key. The expense issue is now fixed. We're working on faster time to market, and a more competitive tech offering. We talked about the non core knocked 3 points off of growth in 2016. That'll decline to 1 point in 2017, and then be not meaningful really for 2017 onwards.

For cards, credit quality visibility grows each month, and remains consistent with the key predictor delinquencies. Flattening out in the latter part of 2017 mean losses will be flat in 2018, and we'll have that slingshot. We appreciate all of those who have hung in there, as the credit normalization turned 20%-plus EPS model into half of that in 2016 and 2017. Fortunately, we were still able to deliver double-digit in 2016, and we'll do as well in 2017, with the slingshot hitting in 1018.

With our 10 year history of 13% annual revenue growth, 17% earnings growth, we drifted below that in 2016 and 2017. Again, we expect to start making that up, as we move into the latter part of 2017, and then for all of 2018. So you should really begin to see the beginnings of the slingshot in Q3 of this year and certainly in Q4. We think we have all our ducks in a row for this, and we will continue to aggressively take advantage of any opportunities in the market to pick up additional shares in front of the slingshot.

That's it. Let's turn it over to Q&A.

Thank you. Our first question comes from the line of Jason Deleeuw of Piper Jaffrey.

Thanks, and good morning. So first question is on the receivables growth outlook for 2017, the 15%. I understand there's portfolios, new portfolios ramping. But could you also parse out and help us understand the components from the same-store sales? It looks like that slowed. And just the tender share, and then just kind of give us a sense for that, so we can have a little bit visibility with a potential reacceleration into next year?

Yes, sure. The core of those have been with us for three years or more. If you looked at those retailers across-the-board, their growth was actually negative. So you saw very strong consumer spend out there, but it seems to have been on auto and home improvement and healthcare, not so much at our traditional retail clients. Their growth was negative. We picked up about 150 basis points of tender share, which is call it, about 4 points of growth from the core.

So whatever the core does, you add about 4 points, and that will be what we get in terms of growth on our cards. That traditionally has been more like, the retailer would do plus [3] and we would do plus [7] or [8]. So clearly, we did well from a tender share perspective, but it's off of retailer results that were quite soft. That means that the -- a bigger chunk needs to come from the newer signings, which is why the $2 billion type vintages are so important.

Therefore, we would expect your next sort of 10 points to come from these start-ups that we talk about, and that's where you get to about 15% overall growth. If there are a couple of portfolios out there that we decide to bring on the year, that will certainly be additive to that growth. But think of the core as providing 4 or 5 points of growth, and think about the new vintages, the start-ups spooling up as adding about 10 points. And then, if we have portfolios, we can do above that.

All right. Thanks for that. And then, just thinking about the provisioning level when charge-offs historically have normalized or peaked, versus the charge-offs, annualized net charge-offs or -- I'm not necessarily looking for exact numbers, but I mean, directionally, there should be less of a reserve build, as we get into peaking or normalizing charge-offs. Is there any kind of sense that you can help us with, and historically where -- what level has that been on the provisioning side?

Sure, it's a great question, Jason. When we knew the loss rates were picking up, we'd have a spread over the LTM loss rate of about 1 point, or a little over 1 point. Once you get to a normalized loss rate, you could probably see that drop to 70 basis points. So there is some flexibility in the reserving, as we basically got ahead of the increasing loss rates. So as they stabilize, you could see the spread to the LTM loss rate narrow to some degree.

All right. Thank you very much.

Our next question comes from the line of Dan Perlin of RBC.

Thanks, guys. So let's go back to beating the dead horse a bit. So when we think about 2017's guidance, I guess, a couple things for Loyalty, and in AIR MILES in particular. One is, I just want to make sure I understand what's fully being baked in, in terms of your breakage rate assumptions for the year? I want to know -- maybe not exactly, because you don't want to fully telegraph, but what are some of these possible remedies to the program that you're really thinking about? And then thirdly, when you're talking to your sponsors, it's clear that they held back, and it would seem as though the program is in question right now. So I want to know what are they, I guess, worried about, and then, what are you effectively telling them, and what are those early conversations like? Thanks.

Okay, I'll take a stab at it, then Charles can jump in here. What we've essentially done is, we have taken out and cancelled the five year expiry policy per the law that was made at the end of the year. And as a result, all you have left are really inactive accounts that can be expired. So that's consistent with all of the other loyalty programs out there.

In terms of what our assumptions are, again you're looking at a program that historically, we did not make margin or much of a margin when people redeemed for various rewards. We put all our margin into the assumption that X number of people would essentially lose interest in the program, and would not redeem. We need to change that. And the more traditional loyalty programs would have some margin coming from a rewards. And that's essentially what we're going to be doing is, we are going to be trying to create a portfolio of rewards that the collectors will be excited about.

And have various events and specialties, and a lot of marketing and promotion that could in fact get them excited. But drive them also into those rewards that are not only of value to them, but that we get a good expense leverage from, and therefore we can realize margin on those rewards which we never did before. That's essentially what we're doing. It's a very traditional loyalty program, it keeps it right in the lane of what's considered quote/unquote, normal in this business.

Again, we had a fairly unique model. It turned out at the very, very end, it was unappreciated, and so we're retooling it to make it fall right squarely in the middle of a traditional model, in terms of how you generate revenue, and how you generate earnings. Our goal right now is to really make sure that the collectors and the sponsors are excited about it again.

I think that the fact that we did go out there, we did take the hit. We are promoting the program fairly aggressively right now. We do see already some nice pick up on the issuance side, redemptions. There's no more fear of people losing their points, so those levels have dropped off dramatically. There have been obviously a couple of stories in the media that were off-base, and things were taken out of context. Right now, we believe our relationship with our sponsors are solid. Obviously, they would have preferred not to have gone through this with us, so it's our job to make sure that they're comfortable with it going forward.

Okay. So just to be clear, are you saying breakage is at 20%?

That's correct.

Was it -- okay. And then, on this AIR MILES issued number of plus 3% for the year, it declined in the quarter. Historically, when you have a couple quarters of year-over-year declines, and maybe it sounds like you're alluding to the fact that's not going to happen. But you've called out things like air pockets in the past, right, because of how it's ratably recognized? And so, if it dips for a couple quarters, it makes it harder and defers it out a little bit to rebuild it. So I'm just trying to make sure I understand how you get back to a plus 3%, in the face of all this noise, and it sounds like it is still pretty deferred. Clearly, back-end-loaded, but are we expecting it to be negative potentially at all in the first half?

Think of it this way, Dan. We start issuance growth of 1% in 2016. Q4 is usually our heaviest promotional period with our sponsors. With all of the noise in the market, we had promotional activities drop a little bit in Q4, which is what pressured in 2016. We believe, and obviously we're going to support additional promotional activities in 2017, so we think getting back to that 3% range is not going to be an issue for us.

Okay. And let me ask one other question unrelated to the loyalty, more about card services. And this is maybe a bigger picture strategic question, given the shareholder that's in your stock. But remind me why you feel compelled to keep the Company in this consolidated holding company structure, versus spinning or selling the business? I mean, credit has been a fantastic business for you guys, and although it's got a little bit of a slowdown here, it still looks to be enormously appealing asset, maybe even more so in a different type of regulatory environment. So can you just briefly, Ed, explain why you feel so compelled to have to keep this under this one umbrella structure? Thanks, and I'll jump off.

Yes, sure. I mean, it gets back to what the thesis of the Company, the thesis of the Company is, essentially we all do the same thing, which is we have a number of platforms. Some have a credit component to them, some don't, and they're just flat out single loyalty programs. Some are coalition loyalty programs, like in Canada.

At the end of the day, this Company is essentially based entirely on -- we've got some type of platform that is used to collect very valuable information called [SKU] level information, which is then we use to derive insights about the consumer, use that information and that intelligence from the data, to then go out into the marketplace and drive additional sales. Those could be additional sales at the grocer up in Canada as part of the coalition, additional sales at a retailer in the mall in our card services, or additional sales at a financial institution, their credit cards or a hotel in Epsilon. So it's essentially all the same model, we just make money different ways.

If you look at the card services business, their core loyalty engine, that's an Epsilon platform. If you look within the card services business, we've got 500 people that are hard-core heavy duty data and analytical types, think of it as a mini Epsilon. So they also share the same messaging platforms for e-mail. We'll be using the Conversant asset to reach out with our clients, and reach them over the various digital channels across the Company. So there's a lot of commonality.

The net result is, we built this Company to find ways to extract sort of the gold nuggets, the SKU level data that we can get from these platforms, which we think is really the gold standard for understanding the consumer and driving incremental sales. That's why we think it fits together nicely; however, obviously, there are other arguments that would suggest maybe they don't fit together, and look, we look at everything every year, and we'll continue to do so.

That's helpful. Thanks, Ed.

Our next question comes from the line of Ramsey El-Assal of Jefferies.

Hi, guys. Thanks for taking my question. On Epsilon, you mentioned now that you have your cost structure settled at this point, and now sort of phase two is pushing to drive, to reaccelerate growth in some of the problem areas. Can you speak to your confidence level that the headwinds there are surmountable with extra effort, meaning not just related to changes in the demand environment for the products, or sort of environmental things that are more difficult to get at, with just incremental investment?

Yes, it's a great question. I mean, it's -- if you look at the business at Epsilon, you've got sort of three-quarters of the business is chugging along pretty nicely, mid single-digit or higher, and you've got the final quarter, which is what we call our technology platform business, which for years and years did anywhere from 7% to 8% growth, you just chunked it out. These are the massive loyalty platforms that we would build for a Citibank, (inaudible) or a Walgreens or something like that.

And what has happened over the last sort of 24 months is that the marketplace has changed. The biggest question we needed to answered was, is there still a huge amount of demand for these big loyalty platforms out there? The answer that came back is, for sure, now more so than ever, because it's the best way to collect that SKU level data that we talked about.

So the next question is, who are we losing share to, and why are we losing share? And it quickly came back that people, while they loved having these massive platforms with all of the bells and whistles and everything else, it was no longer practical. And so, they wanted something that was more cost effective with bringing in an in-house SaaS based solution and hiring 100 analysts to analyze the data. So we needed to get our cost structure back in line. That's been done, so that was the whole India initiative.

Now we need to go back, and get a product that we can get to the market, not in 12 months, but more like 4 or 5 months. That's really what the big hang up has been, is the fact that we have a lot of Chief Marketing Officers who say, look, I don't want to take this thing in-house, I don't want to hire 100 analysts. But at the same time, you guys were too expensive, and it took forever to get the thing delivered. So we've taken care of the expense. We've got to basically strip off the bells and whistles, have a more modular approach to the technology platform, sort of a plug and play. And that's what we are working on right now.

We believe with those two things solved, you're going to have a very nice turn in the technology platform. There will always be a huge market for CMOs who want a in-house solution, and there will always be a huge market for CMOs who want to be focused on what their Company sells, and they don't want to worry about the actual technology and platform for their loyalty and CRM needs. So we're going to appeal to that market, but we needed to fix two things. We've got one fixed, we're working on the other right now. I think we're going to be in good shape, as the year goes on.

Okay. Switching over to the cards business, the tender share ends in the quarter were a little lower than last quarter, which I guess, the question is, can talk about the kind of trajectory of that [rate]? I mean, to a certain point, you would think that those gains would become a little harder and harder to come by. I'm sure there's still room left, but can you tell us about the runway left in terms of tender share gains?

Yes, a good question. It's interesting because, when you have a business like ours, which is different from what people would view as a traditional card business, all of our growth comes from signing new clients. And as a result, when you have 90%-plus of your new business coming from clients who have never had a program before, and you're spooling them up year after year after year, you'll always going to have this huge pool of new clients who are literally starting with zero tender share, and will have that pull ramp up to about 30% tender share after about three years. So you're always going to have that opportunity to grow tender share on the stuff that you've brought on in the prior three years.

In terms of the core, we're probably at -- what do you think, Charles, about one-third or so in terms of tender share? We have clients that are over 50%. So from that perspective, we think that with -- if you look at the retailer results over holiday and everything else, there's never been a more important time for them to put their shoulder into this data-driven targeted marketing space, which is what we do. So what we're seeing is a lot of interest and a lot of dollars coming into enhancing these programs, enhancing the platform, getting at the data, which to me would suggest that the tender share growth rate is going to continue for many years to come.

Great. Thank you very much.

Our next question comes from the line of Andrew Jeffrey of SunTrust.

Hi, good morning, guys. Thanks for taking the question. Ed, I wonder if I could just ask you a two-part question, expanding on sort of Ramsey's query on the strategic, but also more specifically with Epsilon. Is there any reason that Alliance Data shouldn't be more of a pure play? The card services business has consistently outperformed. You don't seem to be getting a lot of credit in the market for Epsilon or for AIR MILES, maybe BrandLoyalty notwithstanding. Maybe you can just talk about whether or not you've thought about card services as being more central business, and maybe the primary driver of your economic value? And as a corollary, to the extent you believe Epsilon is core, are you happy with low single-digit organic revenue growth? Given the value of that business, I would think it should be growing faster, if you're really -- it's really central to the long-term value proposition?

Yes, I mean, look, those are all reasonable questions. I think I can tell you from, if you go up to the Board level that we are, every year we go through the various discussions and the various iterations of what's next for Alliance, what should it look like, is it better the way it is today? Would it be better differently going forward? And that is an ongoing debate. I think that what I would like to see frankly is, we had all this noise, and a very tough challenge in Canada this past year.

You've been around the Company forever. You know Canada used to just constantly just rip it, year after year after year. I'd like to see how it does, once we retool the model, and get this thing back into growth mode. I think you're going to see some nice growth there, combined with BrandLoyalty.

Epsilon, same deal. I mean, Epsilon used to rip it pretty nicely for years, and the last couple years, frankly, it hasn't performed. And it seems it's a little bit of a whack-a-mole, of which business isn't performing. We think the last piece is this technology piece. If we can get that going at a very consistent level, then you're going to have almost half the Company, from a revenue perspective back consistently producing that 5%, 6%, 7% type growth that we've come to know and love.

What was nice about it is the fact that, when cards would go through some normalization, and the cycles would hit and you'd change the growth profile of the business, you would have the other half of the business that doesn't grow as fast through certain parts of the cycle, but is, usually grows very consistently. And you put the two together, and you have a very nice model with strong visibility through good times and bad. That's been the thesis all along, Andrew. Let us get Canada, and let us get Epsilon producing again. And then, we'll sit back and say okay, are we getting credit for it, or are we not getting credit for it? Obviously, if it's the latter, then that's a different discussion.

Okay. No, that's helpful. Thanks for making me feel old, Ed. (laughter) One quick follow-up. (multiple speakers) Yes, Charles, it looks like the yield has started to make a recovery. Is that something we should anticipate continuing in 2017?

I do. I think yields should be stable. Could be a little bit of positivity to it, based upon the trends we're seeing. But I basically think what's happened, Andrew, is some of the changes we made in mid 2015, what we consider card holder friendly have burned in, and now I would expect them to be stable, to maybe slightly increasing going forward.

Okay. Thank you very much, guys.

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

Thank you. Good morning. Cash flow utilization of, in 2017. You've already, now that you've bought 100% of BrandLoyalty, what do you -- first of all, what do you have in your guidance for buybacks, and what are your thoughts on utilization of the cash flow, what do you expect for cash flow in 2017, and how you plan to use it?

So M&A is something we always consider. I'll tell you right now, we don't really see anything that's pressing for us in 2017. What that means is we have a $500 million authorized share repurchase plan. That will be our top priority. As Ed talked about before, it could be best used early in 2017, which we think will happen. As the year progresses if we don't see any further M&A opportunity, we could increase the buyback, and be even more active. What I'd say, Bob, from a free cash flow standpoint, expect it to be up 7% to 10% in 2017 over 2016. Again, we'll have money we're keeping back at the banks. But with a little bit of slowing growth in AR, that will diminish a little bit. So what that means is free cash flow at the holdco will be up a little bit year-over-year, that enables us to be more active with the buyback.

Okay. How much of the buyback do you have built into the $18.50?

We considered the $500 million in terms of guidance.

Okay. And then, looking at your customer base in the card business, and you added -- I think last year was probably the most impressive, as far as quality of customers you've added in a single year. But looking at your base, there are a lot of retailers there, that are struggling. How do you -- I mean, are you concerned about the viability of more companies today than you have been in the past, and how is that built into your longer term thoughts on the card business?

Yes, another good question, I think. Look, it's no secret that a lot of the traditional bricks-and-mortar type clients had a tough go of it, over the past year, both as consumers shifted their spend outside of those verticals, as I said to sort of auto and home improvement, as well as more pure online players as well. So from that perspective, you will have a handful of clients that will not make it, and we've had usually a couple a year. The key thing here is, we don't lose money when -- if a client files, because our relationship is with the card holder, and we'll collect it out, but it does produce a grow-over issue in the subsequent years.

So that being said, what our game plan is, we will probably look to board a couple extra clients over and above what we normally do each year, in order to make up for what we think will be softer growth rates in the core. And that's sort of our hedge against that. Probably the biggest initiative that we are working on right now, which is going to be critical for the next several years is, we have always positioned our technology and our platforms to deal with hey, 90%-plus, 95% of the activity takes place in the stores.

And now what we're finding is it's rapidly shifting to online. And as a result, what we're doing is, we're stepping up our investments in that business to create what we believe will be the best suite of services for the digital channels for our retailers. So said differently, whether it's a retailer app where you would have the retailer, your favorite retailer on your phone, and that would be integrated very nicely with both payment and loyalty altogether, and have it as a seamless one-step transition and transaction, that's what we're after.

We also are after helping our retailers find new potential customers. So if you think about what Conversant can do with the SKU level information that we can get, they can actually go out to the digital channels, and find the most likely future customer for that retailer. So we want to move the model from really focused primarily on the bricks-and-mortar and more towards, we want to be the one-stop solution for all of the digital needs of the retailer, to not only drive online sales for existing customers, but then also find the best new potential customers for the retailer, and therefore help them grow. So that's going to be where all of the money is going.

Great. And just last question, on credit, your confidence in credit. What is driving the confidence in credit? What are you looking at, is it by program, what has caused the rise in credit loss? I think you said, you tightened credit somewhat back in 2015. Are you're looking at static [pull], or has the increase been driven by one or two retailers have driven an outsized portion of that credit, and now you'll see it stabilizing? What drives, what is driving the stabilization in the credit, and your confidence?

Yes, sure. I mean, I think having been in it for 25 years, we are pretty good at looking at vintages, and seeing what flows there. And so what we do every year is, or throughout the year is, we will go client by client, which is 155 different portfolios, and within those clients, vintage by vintage, and we will look at every single vintage, and see how they are performing. And again, what has happened is, you have a certain band of credit that you cater to. And coming out of the Great Recession, we had people who were signing up for the card who were at the very high end. As the recovery continued and people repaired their credit, we had more and more people come into the allowable band. And that's sort of what causes that normalization.

Again, as I tell a lot of folks, we are returning to a level that is comfortable for us. It's not like things are getting worse, that things are -- the consumer's fine. It's just we're normalizing out at a level that will help us optimize our earnings and tender share going forward. So it's a bottoms up, vintage by vintage, portfolio by portfolio, client by client look, And a year ago, we had a strong conviction that this would be sort of a 1.5 year, 2 year cycle to normalize. We're past the first 12 months, now we can see it, in terms of the delinquency flows. We can see it, this stuff normalizing, as we look out six months. So now the guesswork is kind of out of it, and now it's just a question of, let's get the next six months out of the way, and we're fully normalized.

Thank you very much. Appreciate it.

We'll take one more.

We have time for one more question. Our final question will come from the line of Sanjay Sakhrani of KBW.

Thanks. I guess, most of my questions have been answered, but just to follow-up on that last question that Bob asked. I guess, I hate the fast forward past 2017, given we have the whole year to go through. But it seems like when you look at that wedge chart, your delinquencies start to inflect, and kind of come down year-over-year. So when we look out to 2018, should we expect in the absence of any changes -- exogenous changes in the economy, that the charge-off rate has a downward bias, because of the lack of the seasoning from these vintage curves?

I think that's very possible, Sanjay.

Okay. (laughter) Okay. And then, when we think also about Epsilon, and this competitive tech offering that you guys are working on, I mean, could you just talk about the timing of when you expect that to gain momentum, and how we would see it manifest itself through the results over the course of the next year to year and a half?

Yes, I would expect the core of Epsilon, excluding technology will probably do mid single-digits all year. I would expect technology will still drift probably for the first four or five months as we begin to sign -- which is what we're doing now -- sign the new clients. But unlike last year, we're not going to have that big expense drag that we had to start out the year. So we're going to avoid that falling off a cliff at the beginning of the year that happened last year, because revenue was light, but expenses were heavy. So as I said, we're not going to have the same expense issue, and the question now is, just getting the momentum going back on the revenue side.

A final question on loyalty, and I know a lot of questions were asked about this before, but when I look at the revenue guidance on an adjusted basis, it's obviously flat. Is that just the continued residual impact of kind of higher redemptions seen in 2016, or is there something else kind of factored into that, such that it's a flat number?

It's two things, Sanjay. One is breakage is revenue, we took down breakage beyond 6%, so that hurts your revenue. The second is, you're basically normalizing the number of miles redeemed. You're going to see basically flattish miles redeemed compared to 2015, not 2016. So because you're going to have fewer miles redeemed, plus the fact you have less breakage revenue, that's what creates the flattish.

Okay, great. Thank you very much.

All right. I want to thank everyone for hanging in there. Again, apologies in terms of the noisiness of the quarter. But as you can tell, we know we've got a little more wood to chop in the first part of this year, but we are getting pretty jazzed up about the back half. And no question in terms of 2018, it's going to be a nice one. So that's what we're building for, and we'll talk to you next time. Thank you.

Thank you, ladies and gentlemen. This does conclude today's Alliance Data Systems full year 2016 earnings conference call. You may now disconnect.