Capital One Financial Corp (COF) 2017 Q2 法說會逐字稿

Welcome to the Capital One Q2 2017 Earnings Conference Call. (Operator Instructions) I would now like to turn the call over to Mr. Jeff Norris, Senior Vice President of Global Finance. Sir, you may begin.

Thank you very much, Liane. Welcome, everybody, to Capital One's Second Quarter 2017 Earnings Conference Call. As usual, we are webcasting live over the Internet. And if you want to access the call via the Internet, please log on to Capital One's website at capitalone.com and follow the links from there.

In addition to the press release and financials, we have included a presentation summarizing our second quarter 2017 results. With me today are Mr. Richard Fairbank, Capital One's Chairman and Chief Executive Officer; and Mr. Scott Blackley, Capital One's Chief Financial Officer. Rich and Scott will walk you through this presentation. To access a copy of the presentation and press release, please go to Capital One's website, click on Investors, and then click on Quarterly Earnings Release.

Please note that this presentation may contain forward-looking statements. Information regarding Capital One's financial performance and any forward-looking statements contained in today's discussion and the materials speak only as of the particular date or dates indicated in the materials. Capital One does not undertake any obligation to update or revise any of this information whether as a result of new information, future events or otherwise. Numerous factors could cause our actual results to differ materially from those described in forward-looking statements. For more information on these factors, please see the section titled Forward-looking Information in the earnings release presentation and the Risk Factors section in our annual and quarterly reports accessible at the Capital One website and filed with the SEC.

With that, I'll turn the call over to Mr. Fairbank. Rich?

Thanks, Jeff, and good evening, everyone. Capital One earned just over $1 billion or $1.94 per share in the second quarter, pre-provision earnings for the quarter were $3.3 billion and we continued to deliver year-over-year loan and revenue growth across our businesses. As we discussed last quarter, we continue to expect domestic charge-off rate for full year 2017 to be in the high 4s to around 5. We continue to expect full year efficiency ratio for the total company to be in the 51, net of adjusting items, and plus or minus a reasonable margin of volatility. And while the margin for error remains tight, based on what we see today, we remain well positioned to deliver 7% to 11% growth in EPS in 2017, net of adjusting items. All of our guidance excludes the potential impact of the Cabela's transaction.

Now I'll turn the call over to Scott to discuss second quarter results for the company. Scott?

Thanks, Rich. I'll begin tonight with Slide 3. Capital One earned $1 billion or $1.94 per share in the second quarter. In the quarter, we recognized $12 million of noninterest expenses related to our anticipated close in the Cabela's transaction. Excluding these costs, earnings per share was $1.96. We will continue to break out future deal and integration costs as well as the initial allowance build associated with the onboarding of the Cabela's asset. The closing of this transaction is still pending regulatory approval.

Pre-provision earnings increased 6% on a linked-quarter basis as we recognized higher revenues and lower noninterest expenses. Provision for credit losses decreased 10% on a linked-quarter basis as higher charge-offs, primarily in our commercial banking business, were more than offset by lower linked-quarter allowance builds. We have provided an allowance roll-forward by business segment, which can be found on Table 8 of our earnings supplement.

Let me take a moment to explain the movements in our allowance across our businesses. In our Domestic Card business, we built $155 million of allowance in the quarter. Two primary factors drove that build: balanced growth in the quarter and our expectation of rising charge-offs.

Allowance in our Consumer Banking segment increased by $36 million in the quarter. This increase was attributable to our Auto business where we build allowance for new originations. In the second quarter, we had about a 40 basis point impact to our auto charge-off rate from the initial effect of accounting changes in the timing of charge-offs, which I discussed over the last couple of quarters in our calls. Beginning in the third quarter and continuing throughout 2018, we expect these accounting changes to increase the annualized charge-offs rates by 15 to 20 basis points after which we affect -- we -- after which, the effect begins to reverse over time. Lastly, we had a $4 million release in reserves in our Commercial Banking segment as we released reserves associated with loans that moved to charge-offs, which was partially offset by reserve builds primarily focused in our Taxi Medallion business.

Turning to Slide 4. You can see that reported net interest margin was flat for the first quarter at 6.9%, and it was up 15 basis points on a year-over-year basis, fueled by strong growth in our Domestic Card business and higher rates.

Turning to Slide 5, I'll discuss capital. As previously announced, following the Federal Reserve's conditional non-objection to our 2017 CCAR plan, our board has authorized repurchase of up to 1.85 billion of common stock through the end of the second quarter of 2018. And we expect to maintain our quarterly dividend of $0.40 per share, which is subject to board approval. Our common equity Tier 1 capital ratio on a Basel III standardized basis was 10.7%, which reflects current phase-in. On a standardized fully phased-in basis, it was 10.6%.

And with that, I'll turn the call back over to Rich.

Thanks, Scott. I'll begin on Slide 9 with second quarter results for our Domestic Card business. Loan growth and purchase volume growth decelerated in the quarter but remained strong. Compared to the second quarter of last year, our ending loans grew $4.3 billion or about 5%. Average loans were up $5.8 billion or about 7%. Second quarter purchase volume increased about 7% from the prior year. Revenue for the quarter increased 7% from the prior year, in line with average loan growth. Revenue margin for the quarter was 16.6%. Noninterest expense increased about 3% compared to the prior year quarter.

Our Domestic Card business continues to gain scale and improved efficiency. The charge-off rate for the quarter was 5.11%, and the 30-plus delinquency rate at quarter-end was 3.63%. Both metrics were down a few basis points from the sequential quarter. Compared to the second quarter of last year, both metrics were higher as expected.

The impact of Growth Math remains largely in line with our expectations. As a reminder, Growth Math is defined as the upward pressure on delinquencies and charge-offs as new loan balances in our front book season and become a larger proportion of our overall portfolio relative to the older and highly seasoned back book. We continue to expect the impact of Growth Math will moderate in the second half of 2017 with a small tail beyond 2017. As Growth Math runs its course, we believe that our delinquency and charge-off rate trends will be driven more by broader industry factors.

Slide 10 summarizes second quarter results for our Consumer Banking business. Ending loans grew about 5% year-over-year. Growth in auto loans was partially offset by planned mortgage runoff. Ending deposits were up about 6% year-over-year, with a 4 basis point increase in deposit rate paid.

Compared to the second quarter of 2016, auto originations were up 14% to $7.5 billion, with strong growth in prime, near-prime and subprime. As we discussed last quarter, competitive intensity in the auto finance marketplace remains a bit muted, but continues to contribute to our growth. While we still see attractive opportunities for future growth, there are also reasons for caution in the auto industry, including expected declines in auction prices and an increasingly indebted consumer. Our underwriting assumes a decline in used car prices, and we've dialed back on some less resilient programs even as our overall originations have grown. As the cycle plays out, we continue to expect the charge-off rate will increase gradually and that growth will moderate.

Consumer Banking revenue for the quarter increased about 9% from the second quarter of last year, driven by growth in auto loans as well as deposit pricing and volumes. Noninterest expense for the quarter increased 5% compared to the prior year quarter, driven by growth in auto loans and an increase in marketing. Second quarter provision for credit losses was up from the prior year, primarily as a result of charge-offs and additions to the allowance for loan losses for the auto portfolio.

Moving to Slide 11. I'll discuss our Commercial Banking business. Second quarter ending loan balances increased 2% year-over-year, and average loans increased 4%. Higher average loans as well as higher noninterest income in our capital markets and agency businesses drove revenue growth of 9% compared to the second quarter of 2016.

Noninterest expense was up 11%, primarily as a result of growth, technology investments and other business initiatives. Provision for credit losses increased from the second quarter of last year, driven by charge-offs in the oilfield services and Taxi Medallion portfolios. The charge-off rate for the quarter was 80 basis points. Criticized and nonperforming loans rates were relatively stable in the quarter. The Commercial Bank criticized performing loan rate for the quarter was 3.9%, and the criticized nonperforming loan rate was 1.0%. Credit pressures continued to be focused in the oilfield services and Taxi Medallion portfolios. We've provided summaries of loans, exposures, reserves and other metrics for these portfolios on Slides 16 and 17.

I'll close tonight with a summary of key second quarter themes. Capital One continued to post solid year-over-year growth in loans, deposits, revenues and pre-provision earnings. We continued to tightly manage costs and improve efficiency even as we invest to grow and drive our digital transformation, and we continued to carefully manage risk across all our consumer and commercial banking businesses. Based on what we see today, we are affirming our guidance for full year 2017 Domestic Card charge-off rate, total company efficiency ratio and growth in EPS. All of our guidance is net of adjusting items and the potential impact of Cabela's.

As Scott mentioned, the Federal Reserve completed its CCAR process in the quarter. Our capital plan received a conditional non-objection, which requires us to resubmit a revised capital plan by December 28. We will resubmit our capital plan and are fully committed to addressing the Federal Reserve's concerns with our capital planning process in a timely manner.

Pulling up, we continue to be struck by the amount of change that's coming in the marketplace and the opportunity to capitalize on that. We have invested heavily in transforming our company and driving growth opportunities. We are well on our way to rebuilding our infrastructure with a modern technology architecture. And along the way, we are redesigning how we work. We are delivering resilient growth across our businesses, we are driving improving efficiency and we are building an enduring customer franchise. We continue to be in a strong position to deliver attractive growth and returns as well as significant capital distribution subject to regulatory approval.

And now Scott and I will be happy to answer your questions.

Thank you, Rich. We'll now start the Q&A session. (Operator Instructions) And if you have follow-up questions after the Q&A session, the Investor Relations team will be here available after the call. Liane, please start the Q&A.

(Operator Instructions) We'll take our first question from John Pancari with Evercore ISI.

Just want to ask about the vintages that you're seeing and the loss trends within the card vintages. I know you had indicated last quarter that your updated charge-off expectations incorporate your expectation for losses in the 2016 vintage to be on par with '15. Is that still the case? Or are you seeing any changes to that?

John, our expectations are the same, 2016 vintage on par with 2015 vintage.

Okay, all right. Yes, a follow-up to that was just on the Auto front. I know you had indicated that you do expect charge-offs in the Auto book to increase moderately over time, just consistent with the general pressure you're seeing in the business. Can you just help quantify the pace or help quantify that, what you mean me by moderately there? And then also, are you providing at a higher level at this point, given the worst things, severities, as used car values have pulled back?

John, why don't I start at the back end of that question, and turn it over to Rich. So in terms of our allowance process, we've been assuming kind of used car prices to fall and to continue to fall. So we have already been building that into our allowance, and I think that, that's something that we would expect to continue.

John, relative to the charge-off outlook, you could go back. And it's probably, for the last 3 years, I've been saying that we expect charge-offs to gradually increase in the auto business. And we're saying it again. Now -- and the reason for that is the same reason that's driven this conversation over these years, which is off of an exceptionally low base of charge-offs, the phenomenon of normalization we have fully expected. Now the interesting is Capital One's charge-offs that we had posted over time in general haven't gone up in that way. And there've been a couple of things going on there. The biggest thing has been sort of a gradual mix shift, up-market in that sense. The -- in particular, the higher growth that we've gotten at the higher end of the market has sort of offset that effect. But beneath it all, we look at things like the subprime business. There has been some gradual normalization going on. It's a natural phenomenon. Probably all in all, the amount of normalization has not been as high as we have ourselves internally forecasted because we've been pretty cautious about what we expect with respect to used car prices. And frankly, some of the competitive dynamics in the Auto business, I think, have kind of helped in -- competitive meaning the easing of some of the competition, particularly on the subprime side, I think has eased things. So if we look in hindsight at this 3- or 4-year period while we've been raising our estimates for losses, we think there's been a lot of value creation there, good opportunities for growth. The business is in a bit of an odd place with some easing competitive intensity but still the other dynamics that are associated with increasing potential for risk, things like used cards. But all in all, it's -- there's a good opportunity right now, but I think you should expect losses will gradually go up over time.

And we'll take our next question from Ryan Nash with Goldman Sachs.

We'll take our next question from Moshe Orenbuch with Crédit Suisse.

Rich, you talked a little about credit and the credit card business. Can you talk a little bit about the competitive environment from a growth perspective? I mean we've seen a number of the large banks are starting to actually show some growth acceleration. And so what are you kind of looking at that's different and kind of how do you see that playing out over the next year?

Yes, I mean -- Moshe, we -- first of all, we believe the opportunity to grow continues in the card business. What I would say is the exceptional opportunity has mostly run its course. And now there's a more, just -- there's just a good opportunity. Let's talk about the exceptional opportunity for a second. While a lot of people were not -- they weren't as heavily marketing and not as seeing the growth opportunity that we saw, we pretty much captured several years of outsized growth. And I think that has served us very well, even though we all know, of course, the upfront costs of that in terms of credit and allowance have been a little rough on the P&L. What I think has happened over time, how I -- the competition has definitely intensified, but it's not irrational. So let's talk about it in a couple of levels. First of all, at the -- where it's most -- competition is most obvious to everyone, who has like a television or anything else, is in the rewards space. And it is very clear that marketing levels are up, rewards levels on various cards are up. We've not only seen several players come forward with higher things like cash card rates. One of the striking things to me is what's happening on the co-brand side and the process of the renewal of co-brand deals. There has been a pretty significant cranking up of some of the rewards associated with those. The other striking thing happening, Moshe, on the competitive front at the top of the market is the early bonus, the giveaways associated with that. Now we -- that particular area, we worry a lot about. And we pretty much live inside the frontier of those early bonuses only because we have so much experience over the last 20 years on what happens when you get the hoppers, the teaser hoppers, the bonus hoppers, the various folks there. So I think we all have to keep an eye on what happens there. So it is clear that the stakes -- the table stakes are higher, the investment is higher and the competition is higher at the top of the market going after rewards. That said, we continue to be very pleased with our performance there, and we continue to -- winning in that space is about sustaining investment and about winning in product, marketing, service, digital experience, brand. There's a lot to that, but we very much note the competition. But we were all in, and we still like very much what's going on. The other thing I would comment, though, is about more the revolver side of the marketplace. I think people are -- you are seeing more growth there, but I'd put it in the rational category. We don't see really anything going on about pricing in that marketplace. We don't see underwriting standards being compromised. So I think it is at the more kind of normal phase in a market which is just people stepping up and investing more and trying to grow more. So in the context of all of that, I think it's still a reasonably healthy card marketplace. The growth opportunity we had before is not quite what it was, but we still like the opportunity.

And do you have a follow-up, Moshe?

Yes. If you just kind of extend some of that to Auto, you mentioned some easing in the subprime. Are you tilting more in that direction or more in the prime, in that business at this point?

Well, I think we're pursuing the same strategy that we -- that's probably a big word for this, the same approach and underwriting standards pretty much consistently across from the top of the market through near-prime and into the part of subprime that we play as well. The point I would make is that the competitive easing is more in the near-prime and prime part of the marketplace so that, therefore, the opportunity for growth is a little higher there. So it's, in a sense, same approach, same strategy, same underwriting, but a little bit different market dynamics depending on which part of the market we're talking about.

We'll take our next question from Ryan Nash with Goldman Sachs.

Can you hear me this time?

We got you, Ryan.

Okay, sorry about that. I was having headset problems. A question for Scott. Scott, you noted that the card reserve build was for both growth and the expectation for higher charge-offs, which I'm assuming you're trying to imply 2Q '18 charge-offs would be above 2Q '17. But just given what we -- everything you're seeing, competition, vintages, what you're seeing in terms of your delinquency improvement, I guess there was no change in the macro or the competitive environment. Any sense about how you would think about charge-offs for 2018 and/or when would you actually expect to see charge-offs begin to level off on a year-over-year basis?

Yes, Ryan, thanks for the question. Look, as we said last quarter, we thought that Q1 was the high point for reserve build in the Domestic Card portfolio. As Growth Math is starting to moderate in -- through the bank end of 2017, with a tail after 2017, the effect of Growth Math on the allowance is no longer going to be the biggest driving factor. And that's been -- over the last couple of years, that's been the really big driver. So what I would say with the allowance is, as we move through '17, increasingly, the allowance is not going to be really driven by kind of that upward pressure of Growth Math that we've been talking about, but it's really going to be driven by more just the overall industry factors. And where those go, I think, is something that we'll be talking about going forward, but I don't have a forecast for you for 2018.

Got it. And then maybe if I could just ask one quick follow-up. Rich, when you reiterated the EPS guidance, 7% to 11%, you noted again that the margin of error remains tight. Can you maybe just give us a sense of your confidence hitting the target? And what are the key swing factors at this point in time just in terms of being at above -- bottom of the range versus the top of the range?

So yes, thanks, Ryan. Look, I want to start by saying that I've stepped out of my character, and you might have wondered what happened to the Rich Fairbank that you knew because I haven't -- you'd have to go way back a lot of years ago to where we have given EPS guidance because, kind of to your point, I mean this is, in the end, the difference of large numbers, and it's subject to very big swings associated with allowance and other things that can happen. So we're -- we generally, and I want to make sure people know that, generally don't plan to be in the EPS guidance business. The reason that we kind of stepped out and did that last quarter was really more to show the market that while there's a lot of noise associated with some of these credit numbers, our story, our belief in Growth Math, the belief in what we could do on the efficiency ratio side, the momentum of the company and our ability to successfully get there, I think we felt pretty confident about that. But again, we cautioned, of course. And to your point, it's a pretty small range. So what are the biggest drivers of that? I mean, the biggest driver still always is credit. And in a world where everything's on balance sheet, and small changes in expectation lead to, you know this, into the whopping allowance effects, that is certainly something that is a -- probably at the top of the list for the biggest effect. On that one, we feel good about the trajectory of Growth Math, but we know that small changes can have effects. So we're certainly aware of that. I think that, that is the big one. I think beyond that, the -- I think the trajectory of our spending on things like marketing, the seizing of growth opportunities, the timing of progress on our continuing quest for efficiency will also be factors that play into this. And the other thing I would say, of course, that rates -- interest rates can -- they can always play a factor. But that's just a little window into our own dialogue with ourselves about this. But when we look at those things, while there's always risk associated with a guidance like this, we feel -- we like our chances, we like our momentum here, and we continue to maintain that guidance.

And we'll take our next question from David Ho with Deutsche Bank.

I just want to circle back on the underwriting standards that you may have tightened over the past year and, in large part, trying to get those vintages to perform in line with expectations. Have you seen or have you done an additional tightening for '17 vintages? Are you comfortable in your underwriting standards there? Obviously, it's hard to track FICO but, in terms of the mix, has that stabilized as well?

Yes, David. We -- one thing is, of course, this is hundreds of programs we're talking about and changes that we constantly do around the edges. But to generalize, it was around the beginning of 2016 that we started tightening around the edges, and we talked about that. And it is that tightening that had led us to be pretty confident that, in fact, 2016 vintage was going to come in better than 2015. You may recall we had talked about that and, certainly, had early indications consistent with that. Some of the slippage in the marketplace, and relative to our expectations, there was -- we're now more -- these are more on par with each other. Over the course of this journey, watching the marketplace, the competitive environment, the greater supply, and continuing to look at all of our programs and where credit was coming in, to your question, we have continued to tighten a bit around the edges. And important, but not only, but probably the majority contributor to our reduced growth is just our continued tightening. And particularly, the tightening that's going on is less around the originations side and more about the timing -- the size of lines and the timing of line increases under the philosophy of let's continue to really book the customers, take advantage of the window of opportunity. In our own time, we can choose when and how much to extend the credit lines. That's more of a kind of stored energy that we get there. So -- and by the way, in the marketplace, there's no doubt the marketplace become more competitive, so that's also contributed to our slowdown as well. But all of that said, I think that -- we have high expectations for 2017 coming in as a very strong vintage.

Right. And as a follow-up, what kind of card growth year-over-year are you comfortable with -- while still maintaining your '17 guide and obviously, your benign outlook for '18 in terms of growth not diminishing? Is it more in line with industry growth? How do you think about kind of the range there versus the kind of the industry average growth year-over-year?

Well, we're specifically not giving guidance year over -- for card growth. But I think you're asking what card growth do we need to kind of be able to maintain our EPS guidance. Was that -- were you linking those 2?

Yes.

Well, because the reason I say that is the amount of card growth is -- look, I think at this point, for the rest of the year, the amount of card growth is not that big a driver on the EPS. It'll have quite a bit of impact on next year's EPS and other things. But across the reasonable range of card growth that we would look at for the rest of the year, anything in that range probably wouldn't have that much impact on the EPS. Relative to next year's card growth, we're not going to give guidance on that, but I want to come back to my point that we're -- we continue to be all in, in terms of capitalizing on growth opportunities that are out there. They've gone from -- the growth opportunities have gone from exceptional to good.

And our next question comes from Sanjay Sakhrani with KBW.

This is Steven actually filling in for Sanjay. I just wanted to see if you can elaborate around the issues the Fed pointed out in the CCAR planning process. And how comfortable should we be that the problem will be remediated?

Yes. Thanks for asking the question. And as you might expect, we're not really in a position to talk about kind of the Fed process around CCAR. That's all guarded by confidential supervisory information that we're not able to disclose. So the things I can tell you, if you want to have -- the most that we can tell you is really to point you to the Fed CCAR report, and they're going to give you the best description of kind of the issues that they thought that there were. I will only just reiterate the point that Rich made, which is we are fully committed to meeting the requirements and resubmitting our capital plan within the time frames that the Fed has established for us.

Got it. And then the follow-up question is just around the subprime consumer behavior. Have you guys seen any changes around the behavior of them?

I would say, generally no. I -- we certainly note that the growth in revolving debt in subprime is higher than that in prime. By the way, the last few months, that's kind of settled down a little bit, but it's still higher than prime. So we certainly have our eye on that. Of course, that is also in the context of coming off of a base of a pretty big retraction -- retrenchment since The Great Recession. So we'll keep an eye on the data about the consumer -- I mean the economic -- economy metrics. But beyond that, we have not seen a lot of behavioral change. We have talked in prior conversations about the fact that there is some normalization generally going on. Second half of -- in the second quarter of 2016, we saw some, as an industry effect not just a Capital One effect, a little gapping out of some vintage performance that -- all would be consistent with the marketplace moving along and getting a little more competitive and consumers getting a little bit more indebted. But beyond that, we haven't seen anything unique at the subprime space, and that's why our strategy pretty much continues. We've just been a little cautious around the edges.

And we'll take our next question from Betsy Graseck with Morgan Stanley.

Two questions, one on how you're thinking about recoveries. We heard from some others recently that there is some pressure on pricing and recoveries. And I know you're not using third parties as much as some other folks are. But maybe can you give us a sense as to what you're seeing in the marketplace and how you anticipate that impacting you.

Betsy, yes, we don't sell as much of our charged-off debt as other people do. But we still do participate in that market, but just pulling up overall on recoveries. It's interesting, recoveries, we always tend to look at that independently of other metrics. But I've certainly seen over the years how recoveries move in tandem with a lot of other dynamics going on. So for example, in this journey that we've talked about of the industry's credit normalization, some of the things where even our own expectations about credit performance, we found things were -- came in a little higher than we had expected, recoveries was right in there as one of the little culprits in that. So we have seen generally some softening of recovery rates beginning in the third quarter of 2016, which continued in the early 2017. They were stable in the second quarter. But we'll have to keep an eye on recoveries. And in many ways, I think they will go in many ways as the overall credit marketplace goes.

And then the follow-up just on the Taxi exposure. I know you've provided us a spreadsheet showing us the exposure but wanted to understand how you're thinking about that, in particular, given that the criticized performing loan rate increased quite a bit in the quarter. Does that suggest that we're preparing to just build the reserve further as time goes on here?

Yes. Betsy, it's Scott, thanks for the question. So a couple of thoughts here. First of all, as you know, we've been disclosing the details about this Taxi portfolio for a few years because it's been an area that we wanted to make sure that our investors and others had a good insight into what our exposures were there. That's on Slide 17 in the materials. We've got about $500 million -- $580 million of taxi medallions on the balance sheet as of June 30, 17% allowance coverage ratios, so well provided for there. Of that, around $530 million is New York City taxi medallions. And I think that, at this point, it's well understood that with the arrival of Uber and other competition that there's a lot of pressure in the value of these medallions. And we've been seeing that pressure actually in revenues and in cash flows for several years. And as a consequence to that, we've actually been carrying and writing down our loans to the lower of fair value or carrying value for several years. So we've -- we're kind of well on our journey to making sure that we keep these loans marked at market prices. In Q2, we had a total provision of expense on Taxi of around $55 million, which was primarily charge-offs, but we also billed allowance, which was all based on kind of our view of the changes in the value of medallions that happened in the quarter. I will tell you that I've read the comments from other banks that have come out recently. I feel very comfortable with our marks and where we are on our portfolio. And I'd just say, kind of in terms of what we might see going forward, right now, we're working aggressively with our borrowers to restructure loans where possible. We're taking steps to make sure that we're protecting our collateral. The -- I think the allowance and our charge-off levels are indicative of kind of where prices are today. But there's certainly a risk that if this market doesn't stabilize that we could be subject to further write-downs if we see kind of fair values and prices drift further south.

And we'll take our next question from Bill Carcache with Nomura.

Can you remind us what percentage of the cards within your partnership business are store-only versus also general-purpose cards that have a Visa, MasterCard logo on them and can be used for out-of-store spend? And then are you seeing any kind of difference in customer payment behavior between the cards that are good at the partner store only versus the cards that can be used for general-purpose spending?

Bill, this is Scott. We don't actually break down in any detail kind of the partnership book from our broader card portfolio, so that's not an information that we're going to be able to provide you.

Okay. I guess maybe perhaps just if you could comment at a high level from your experience whether there's any customer kind of pecking order preference for, whether they're going to pay the private-label store card first versus the general-purpose card. If you can, any color on that at a high level would be helpful.

Bill, I -- we were not -- in The Great Recession, we were not big players in the partnership business. We did an acquisition since then but, of course, we couldn't resist but peek back and see what we could observe of in The Great Recession and under stress. I carry around an intuitive view, and I certainly have not seen anything too inconsistent with this that, in general, the -- these co-brand cards -- the defining thing that I would put on the co-brand cards that have a lot of out-of-store spending become like a primary spending vehicle. These things are top of the line, and I think that we would believe that they would lead the league tables in terms of resilience. We've all had a lot of conversation, both in our partnership business and, frankly, across our whole commercial business, about the woes of retail and what does that mean. And I think that if you look back to The Great Recession, I think that the private-label cards did seem to get hit harder. But one thing I would say in their defense is you move from a position of sharing economics to having the entire economics. And I still think most people are motivated by the desire to pay off and keep a good credit record. So look, I carry around the intuitive belief that our private label business probably runs a little bit more risk than the high-end co-brand programs, but that I haven't seen anything to cause us to say that I feel -- I think we still believe in these programs. You're going to get the vast majority of customers to repay. But it certainly is a risk.

And we'll take our next question from Chris Donat with Sandler O'Neill.

I wanted to shift a little bit to expenses. First with marketing, Rich, as you talked about going from an exceptionally good -- an exceptional opportunity to just a good opportunity, should we expect you to continue with sort of this level of marketing spend? Or does it get harder to attract the customers you want now that you're no longer in the exceptional opportunity for U.S. card?

Yes. Well, so a bunch of things to say to that great question. First of all, notice what we said that our dialing back in many ways is -- the dialing back of growth, the slowing of growth is, in many ways, driven by more conservative choices about the timing and magnitude of line increases, how much credit line is granted at the outset, those kinds of things, none of which really have anything to do with marketing. So we continue to be all in on the origination of accounts, and I don't see any change in that. And even, by the way, in dialing around the edges, dialing around the edges, those edges tend to be around approvals in the context of the same marketing. It's just when we get 100 customers in there, there might be a few more percent that might not make it through the filter kind of thing. So that is kind of point number one, the so what of which is our quest for a pretty sizable growth of accounts continues, and we feel pretty bullish about the prospects. The other thing is another thing that I've learned over the years of being in this business that sometimes investors feel like, "Okay. Well, there's going to be less growth, therefore, you don't have to spend money on marketing." Usually, it almost works sometimes the other way because other people step up. The table stakes go up. And unless we really say, "okay in a particular segment, we're just not going to compete," it -- they sort of have to pay more to get -- to stay in place kind of thing. Now there comes a point where things get so irrational, we have from time to time just pulled out of things. But my point about the card business, I still would call it more -- it's just something that's more competitive but generally rational, so we -- I think our investors should carry around the assumption that we're going to continue to be aggressive in our marketing, and I think it's -- while the marketing efficiency is -- compared with the last couple of years, the last 2 or 3 years that have been in the exceptional category, marketing efficiency is a little more in the "normal" category. But we're still all in on this marketing, and I think that will continue.

Okay. And then just a follow-up quickly with Scott on -- the salaries were down $88 million quarter-on-quarter. Was there anything elevated either in the first quarter or unusual in the second quarter?

Yes. In the first quarter, we had a lot of just kind of the normal taxes and such that happens in the first quarter where that starts to run out into the second quarter. So that just doesn't carry through, so nothing unusual there, and it's more or less consistent with kind of the seasonal trends.

And we'll take our next question from Rick Shane with JPMorgan.

Look, we've had an interesting discussion about the impact of ride share on Taxi Medallion prices. How do you manage risk in the Auto portfolio from ride share? You're probably, in the next year or 2, going to start to get back a lot of cars with a ton of miles on them. And I'm curious if that's something you're thinking about in terms of risks or how you sort of account for that in terms of underwriting.

Rick, look, I think you make a good point. I think the -- I think there are a number of factors that -- while I must admit I don't think we have put a real focus on, I don't remember being in a meeting where we're talking about your specific point. I think if you pull up and think about the Auto business and some of the longer-term issues, all the pressure from the tech companies of the -- on things like the ride-sharing you're talking about, even looking farther down the pike at what would be the impact of all the technology changes in automobiles and in the [limit], the driverless cars. But even on -- even before the day finally comes when we -- down the road, we're all doing our work in the backseat and there's no driver, long before that, the -- I think we all have to be very vigilant about what is the impact of technology change, so that there come to be -- I mean, generally, the quality of technology has sometimes allowed cars to last longer and has been a good guy in the Auto business. But the question is will there be a tipping point where the old cars just don't cut it and the new ones are so much better? So this, in fact, I just want to pause for a second and just kind of seize the moment, that one of the things that I don't -- I think banks don't do that well, I think Capital One did not do that well on things like the Uber story is pull way up across all of our lending businesses and ask what is the impact -- given that industry after industry is being revolutionized, what is the impact, especially, I think, in the commercial C&I business, of the revolution that's going on in our clients' businesses? And if we just go and make one loan at a time and do our nice underwriting standards, we could wake up and have a lot of rude surprises like we did in the taxi kind of business. So I don't have a great answer for your specific question, but I -- one of the things that we put a lot of energy into Capital One is pulling way up and instituting -- not only having the conversation, but instituting a kind of risk management process associated with the extraordinary revolution that's happening in our industries, in our clients' industries and in some of our core businesses. It's a great challenge.

Rick, one other thing I would just mention, we definitely, in part of our underwriting practice and assuming that auction prices continue to fall, is take into account the longer lives of cars. The -- our allowance is only 12 months, so it certainly isn't going to capture the shift in -- that you're talking about in terms of ride share. But part of the reason why we continue to really pay close attention to auction prices is there's a variety of risks that can impact that over time. And when we're originating these loans, we're thinking about kind of all those risks as we do that, which is a reason why we continue to plan for auction prices to move lower from here.

Got it. And again, my point is not are there going to be fewer cars on the road because we used them more densely or is there going to be a technology shift that basically makes old cars obsolete, I'm actually thinking about the more immediate, which is that I'm assuming you have a fair number of customers who are taking loans from you guys with the idea that they might be in the ride share business. And if the economics don't work out for them, those are cars that are very, very likely to come back and have very high mileage. And I'm curious if one of the things that we're seeing and, again, since you guys are so quantitative, if we're seeing higher mileage on the repos that are coming back in.

That, I -- off the top of my head, I can't -- I actually can't give you an answer to that one. I can do some follow-up, and we can circle back with you. But Rick, overall, we're taking into account all of those statistics in formulating kind of our expectations around auction prices and resale values. So I think that's kind of embedded in our assumptions.

And we'll take our next question from Brian Foran with Autonomous.

I've got one question on fees and then a follow-up on technology and digitalization. On fees, it's a pretty strong quarter. And in particular, service charges had been a drag for a long time and are now up year-over-year. And interchange revenue's up a little faster than spend volume. Can you maybe just describe is this any kind of inflection point? Or is it more a quarter-to-quarter noise? Just -- fees actually was a pleasant surprise this quarter, so just trying to figure out what drove that.

Yes. Thanks, Brian for the question. Noninterest income was up about 16% in the quarter, and there's a few things that are going on there. Noninterest income for sure is lumpier than net interest income. It certainly doesn't move quite as consistently with loan volumes. And so I'd just kind of go through and talk about a few of the drivers. So you're right that net interchange was a large driver. That contributed about $100 million in the quarter-over-quarter variance. And a few things there. We had higher seasonal purchase volume, which was favorable. And then, as usual, we have kind of the normally quarterly adjustments to rewards, liability expense, those kinds of things that run through there and can create some quarter-over-quarter volatility. And we saw that this quarter. We also, in the first quarter, had some U.K. PPI expense that ran through there. So that created just under $40 million of favorability when we look at the linked quarter. And then kind of -- I'd just kind of also point out that in our commercial business, that business had some lumpy fees that come in, and we did see around $25 million of higher quarter-over-quarter fees. Those are businesses that have a business model where they generate fees. And normally, we see them kind of -- a little bit, there's some puts and takes this quarter. They just happened to all kind of work in our favor, and so they all came in a positive way. But that's kind of the overall picture. I think they're all part of our business model. But I would say that we would expect that, that line item's going to continue to be a bit lumpy.

Look, Brian, I just want to add, you asked if there's an inflection point in the interchange thing. I think we -- that we don't think that. It bounces around a lot from quarter-to-quarter. This happened to be a quarter where there was strong interchange, but that's not necessarily a breaking point. I think that we continue to expect interchange growth and net interchange revenues to be below growth in purchase volumes.

And let us not forget, last quarter, I think the growth was negative. So this thing -- it's much better to just squint your eyes and look at like annual trends and sort of the relationship between purchase volume growth and interchange growth.

Maybe coming back to the closing remarks in the opening script you made, Rich, around digitalization. This is something, really for years now, you've been talking about pretty emphatically. And when you started, it was kind of -- I remember when you first said wrestle APIs, I had to google, and I was like, "Which call am I on?" But we've seen more and more banks come out and really start talking about what you were talking about 2 or 3 years later. So if I'm -- when I'm talking to investors, that helps make it like, "Okay, I see what he was talking about for so long is important." But I still find it's hard for people to kind of put some tangible parameters around it, how does digitalization and APIs and middleware actually translate to better revenue or better cost. So I'm sure there's some hesitance about telling your competitors everything you're doing. But I'm wonder if there's -- are there 1 or 2 examples you could give, backward looking, how this multi-year digital journey has started to produce benefits, either greater revenue or less costs or a combination of both?

Yes. Thank you, Brian. So first of all, we've been -- look, I mean, it was 20-some years ago, that same phrase I use now which build -- I used it way back then which is build an information-based technology company that happens to do banking, competing against banks that happen to use information and technology. And I'm very struck that I've got the same battle cry 20-some years into that. But the other striking thing to me, even though I think Capital One's pretty well positioned to do something like this, is just how staggering the amount of change that's going on and that will happen in the industry and the need to really transform ourselves across talent and infrastructure, the way we work, how we develop software where really there's nothing that's left unchanged. So where have we seen the -- where can you see the benefit? No one can disentangle what -- no one can prove what happens when your customers are a lot more delighted, how that affects things like growth and retention and things like that. But I just would say that -- and it's very easy, Brian, for you to -- I'm not going to rattle them all up here, there's a lot of various proof points associated with a dramatic increase in Capital One, the customer satisfaction, the -- go to the App Store and look at the ratings on the App Store, the -- between digital experience, the customer experience. Go to our own website and look at the ratings that and reviews that people put up on our own accounts there. For a variety of reasons, one of which is our digital transformation, we have had a tremendous momentum with customers that, that is pretty linked to the kind of exceptional growth that we have seen. Look, in the Auto business, I don't if you've seen some of the technology that we're employing in the Auto business associated with Auto Navigator. I would suggest take a look at that. That's pretty interesting. I can never prove to you what the contributions of delighted customers are on the growth front. Secondly, on the marketing front, there is an obvious and dramatic transformation going on in marketing and how it works and the growing role of digital marketing. Digital marketing itself has got the word digital in it. And it's about digital capabilities and, in fact, information-based strategies. That has been an important contributor to Capital One's success. The other thing that's going on is you've seen what's happening with our efficiency ratio, which got worse before it got better, but the important thing is that it's getting better by a lot. And when I started on this journey, I said that, "Look, I want to make one thing clear, so many banks are saying in their public statements, "We are going to self-fund our digital investment," I want -- and I've been saying to investors for years, and it's really 5 years actually this journey has been, not the 3 that you necessarily referred to, but over this journey, I said that I don't believe anyone can transform their company and be competitive in the future by self-funding this transformation. That is, you got to pay now and get benefits later. I also said that the biggest motivator of our digital investment is not a cost objective, it really is much better customer experience, a -- building a way more dynamic, well-managed, fast first to market, better controlled, all those other kind of benefits that come with this. All of that said, it's very clear that we are increasingly reaping important sizable benefits and efficiency as a result of years of digital investment. And the efficiency comes across distribution channels of retail, call centers, operations centers. It also -- the transformation in how we work has long-term benefits, many of which we've started to realize. So the interesting is the journey that was never motivated to be #1 for the sake of cost savings. Something where it costs a lot, the costs were a lot higher than the benefits at the outset, those relative meters are on their way to some significant change in position there. And the -- so it is the efficiency ratio would be the final place that I would point to.

And we'll take our next question from John Hecht with Jefferies.

And just one question. Going back a year, I think you were having better-than-industry card growth. I think it was somewhat balanced between new accounts and line extensions. You've now pulled back growth. And I'm wondering has the composition of growth changed? Is it coming more from one of those aspects or the other? And to the extent it's changing, what does that mean to the risk characteristics going forward?

Yes, John. So we had -- if you dial all the way back, I may get my years slightly wrong here, but I think around the '12, '13 time frame, we had what we call the brownout of line increases. And that was driven by a new regulatory requirement that required you to obtain income, a customer's income, before we could actually do a line extension. And implementing a whole way to obtain that data and so on led to a brownout in our line increases even as our originations were continuing at a reasonable level. Then what happened is the -- right around the time that we got the technology and the ability to address the regulatory requirement, that also coincided with sort of the big growth opportunity at Capital One. It contributed to, but also coincided with, so that we had a surge in line increases and then a surge in young customers who were also eligible for line increases. So I flagged over the especially '14, '15 time period that we were having a more than probably normal level of line increases in the overall mix of our business. Over the past -- I'm getting my time frames -- I'm only approximating here, but something -- I would argue around the beginning of '16, we got into pretty much an equilibrium between line increases and originations themselves. And we have generally been in that equilibrium ever since. To your point and to an earlier point that I said, if anything, our dial back's more on the line increase side than the originations side of late. So it is possible the mix between those 2 could move to the other side of equilibrium, but I don't think these changes are big enough to call that out. But I think the more important thing, which is what we called out in the '14, '15 time frame, that those were more than -- a little bit more than their share of line -- the normal share of line increases.

And our last question from tonight is from Ken Bruce with Bank of America.

The first question I have is -- just really relates to losses. And in the quarter, for the U.S. credit card business at 5.11. That was a pretty substantial improvement in the month of June just based on monthly reporting. Was there anything specific that impacted that June performance? And it had something on the order of like close to 50 basis point improvement month-over-month?

Ken, no, nothing that jumps out at us. One thing that -- well, certainly, we're delighted to see the 47 basis point sequential quarter improvement in charge-offs. But we always caution, don't get too carried away with any one month because, in fact, as a point in the opposite direction, and you obviously know this, the delinquencies, after several months of going down, actually went up a bit in the month. So they kind of went in the opposite direction. I think we looked at both of those. Look, we look very carefully at every month's worth of data. But I think that we should all just squint our eyes a bit and look at the bigger pattern and don't get too carried away with any one month of data. The big point is we still expect the Growth Math effects, measured in terms of year-over-year increase in delinquencies and loss rates, to moderate over the rest of 2017, to have a small tail in 2018. And in fact, we were just talking about this today that one of these quarters, we're actually going to say goodbye to the growth -- to Growth Math. In a sense, it's not going to really be newsworthy, and we're going to be back and rising -- our numbers rising and falling with the dynamics of the marketplace. But we are -- Growth Math quarters are numbered here before we retire it, which is a good thing. We look forward to that retirement party.

I'm sure. The -- there was a somewhat small shift, but one that if it continues could be -- have a substantial impact on that Growth Math. But there's a slight shift in the above 660 credit score versus the below 660 credit score. Is that something that is a specific strategy of Capital One to accelerate one versus the other? Or is that just kind of a by-product of what you discussed around trimming around the underwriting and so forth? And if I could just make that a 2-part question. You kind of answered Betsy's question around recoveries differently, but have you seen a change in the charged-off debt prices that are in the market?

On the charged-off debt prices, I do not have any recent -- I haven't been in a conversation associated with that. And I think that I would have heard about it if something was newsworthy there. But I -- don't rely on me for a latest read on that. The -- let me talk about the subprime mix. So our subprime mix has been -- first of all, our strategy as a company has stayed very steady over the years. And we've been going, as you know, right at the top of the market and all the way through to the higher end of the subprime marketplace. And over the last couple of years, basically during this growth surge, you have seen our subprime mix grow from -- I'm doing this from memory, but around 34% up to 37% last quarter, and it dropped down to 36%. First of all, I want to point out, in the world of all the precision we all live by, these are rounded to the nearest integer. And sometimes, the story is nothing more than a rounding story. Obviously, when you go up several hundred basis points that there's a trend there. And what we've said is the opportunity that we have seen in the subprime marketplace, the relative -- I'm now speaking in the years of 2014, '15 also, the lower -- a little bit lower competitive intensity there, all of that contributed to subprime's share growing a little bit. The strategy is really the same. The number has settled down a little bit. Cabela's is going to come along. And hopefully, everything works out perfectly there. And it comes in, and that will take a little bite out of the subprime mix as well. But the big story is, I think we see the opportunity to be across the board. You're probably right, a little bit more of our trimming around the edges has been in the higher risk areas or higher loss areas, so that just brings the subprime thing in just a little bit. But I -- my main point is the more things change, the more they really stay the same.

Thanks, Rich, and thanks, everybody, for joining us on the conference call today. Thank you for your continuing interest in Capital One. The Investor Relations team will be here this evening to answer any further questions you may have. Have a great night.

And that does conclude today's conference. Thank you for your participation. You may now disconnect.