發現金融 (DFS) 2021 Q2 法說會逐字稿

Good afternoon. My name is Stephanie, and I will be your conference operator today. At this time, I would like to welcome everyone to the Second Quarter 2021 Discover Financial Services Earnings Conference Call. (Operator Instructions) Thank you.

I will now turn the call over to Mr. Eric Wasserstrom, Head of Investor Relations. Please go ahead, sir.

Yes. Thank you, Stephanie, and good morning, everyone. Welcome to today's call. I'll begin on Slide 2 of our earnings presentation, which you can find in the financial section of our Investor Relations website, investorrelations.discover.com.

Our discussion today contain certain forward-looking statements that are subject to risks and uncertainties that may cause actual results to differ materially. Please refer to our notices regarding forward-looking statements that appear in our second quarter earnings press release and presentation.

Our call today will include remarks from our CEO, Roger Hochschild; and John Greene, our Chief Financial Officer. After we conclude our formal comments, there will be time for a question-and-answer session. (Operator Instructions)

Now it's my pleasure to turn the call over to Roger.

Thanks, Eric, and thanks to our listeners for joining today's call. We had a very good quarter that highlighted the strength of our digital business model and solid execution on our strategic priorities against the backdrop of continued economic improvement.

This quarter was characterized by 3 important events: First, our card receivables grew sequentially in May and June, causing our period-end receivables to be up quarter-over-quarter. This outcome, combined with the strength in consumer spending and our account acquisition, increased our confidence for moderate receivables growth this year and stronger growth in 2022.

Second, we benefited from a gain in our Payment Services segment. This gain is an outgrowth of a long commercial relationship and underscores our payments business' ability to forge innovative and lasting partnerships.

Lastly, we achieved a historic low in delinquencies, which resulted from consumers' strong liquidity position, our conservative stance on underwriting, and the proactive measures we took into the downturn to protect our credit quality. This outcome also supported our reserve release this quarter.

Turning to the quarter's results. We earned $1.7 billion after tax or $5.55 per share. These results include a $729 million onetime gain. But even excluding this gain, our results were very strong at $3.73 per share. The drivers of the quarter's strong results reflect the combination of our solid execution and supportive macro conditions.

Total sales were up 48% from a year ago, and 24% from the second quarter of 2019. Retail sales remained very strong, and there was significant improvement in T&E categories, which were hardest hit by the pandemic. Even travel returned to growth in June compared to (inaudible) levels. And sales volumes are accelerating. The 24% growth I cited is an increase from 15% in the first quarter relative to the same period in 2019.

We also see an attractive environment for account acquisition even in the face of heightened competition. We have removed nearly all of our pandemic credit tightening and have increased our marketing investment to align with these actions. These decisions supported new account growth of 26% over 2019 levels with strong growth among prime consumers as our differentiated brand and integrated networks support our strong value proposition, which centers on transparent and useful rewards, outstanding customer service and no annual fees.

While the current operating environment is broadly constructive, there are also some challenges. As we have highlighted before, the counterpoint of sustained strong credit performance is high payment rates, which in the second quarter were over 500 basis points above 2019 levels. We may be seeing evidence that payment rates are plateauing, and while we expect some moderation later this year, we believe payment rates will remain above historical levels for some time. Even so, we expect to strengthen our sales figures and the contribution from new accounts to drive loan growth through the back half of this year and accelerate in 2022.

As we have said in the past, we will invest when we see attractive opportunities, and the actions we took this quarter with increased marketing expenses and investments in technology and analytics were an example of that approach. These investments are consistent with our commitment to long-term, positive operating leverage and an improving efficiency ratio as they drive loan growth and enable a more efficient operating platform.

In our payments business, we benefited from a gain on our equity investment in Marqeta. This gain was the result of a relationship that began a decade ago, and we continue to see opportunities and innovative partnerships. I'm very excited about our investment in Sezzle that was announced last week, as we look to expand our partnership with the leading buy now, pay later provider. We also continued to grow our global acceptance presence and announced new partnerships in Bahrain and Portugal, adding to the 2 network alliances that we announced earlier this year.

Our debit business continued to build on its recent strength. PULSE volume increased 19% year-over-year and was up 33% from 2019 levels. In addition to the influence of economic recovery, this performance reflects the greater relevance of debit to many consumers through the pandemic period. Volume at Diners has also recovered to some extent and was up 41% from the prior year's lows. However, volume is still below prepandemic levels and may remain so for a period of time.

The strong fundamental performance of our digital banking model drove significant capital generation, which this quarter was also aided by our equity gain. We accelerated our share repurchases to $553 million of common stock, a level near the maximum permitted under the Federal Reserve's 4-quarter rolling net income test. We remain committed to returning capital to our shareholders. And going forward, our approach will be governed by the stress capital buffer framework.

On our call last quarter, we indicated that we hope to revisit our capital return for the second half of this year. And I'm very pleased that our Board of Directors authorized the new $2.4 billion share repurchase program that expires next March. We also increased our quarterly dividend from $0.44 to $0.50 per share.

With the current strength of the U.S. economy, I'm increasingly optimistic about our growth opportunities this year and beyond. Our value proposition continues to resonate with consumers. Our payment segment is expanding its partnerships and acceptance, and our capital generative model positions us for strong returns over the long term.

I'll now ask John to discuss key aspects of our financial results in more detail.

Thank you, Roger, and good morning, everyone. I'll begin with our summary financial results on Slide 4.

As Roger noted, our results this quarter highlighted the strength of our digital model, solid execution on our priorities and continued improvement in the macroeconomic environment. Revenue, net of interest expense, increased 34% from the prior year. Excluding onetime items, revenue was up 9%. Net interest income was up 5% as we continue to benefit from lower funding costs and reduced interest charge-offs, reflecting strong credit performance. This was partially offset by a 4% decline in average receivables from the prior year levels.

Excluding onetime items, noninterest income increased 29%, driven by the higher -- by higher net debt count and interchange revenue due to strong sales volume. The provision for credit losses decreased $2 billion from the prior year, mainly due to a $321 million reserve release in the current quarter compared to a $1.3 billion reserve build in the prior year, an improvement in the economic (inaudible) and ongoing credit strength were the primary drivers of the release. Net charge-offs decreased 41% or $311 million from the prior year.

Operating expenses were up 13%, primarily reflecting additional investments in marketing, which was up 36%, and high employee compensation, which was up 10%. A software write-off and a nonrecurring impairment at Diners Club also contributed to the increase. I'll provide more details on expense drivers and outlook later in the presentation.

Moving to loan growth on Slide 5. The ending loans increased 2% sequentially and were down just 1% from the prior year. This was driven by card loans, which increased 2% from the prior quarter and were down 2% year-over-year. Lower year-over-year card receivables reflect 2 primary factors: First, the payment rate remains high as households continue to have a strong cash flow position due to several rounds of government stimulus. Second, promotional balances were approximately 250 basis points lower than the prior year quarter.

While card receivables declined year-over-year, we considered the sequential increase to be an important data point reflecting continued momentum in account acquisition and very strong sales volume. The high payment rate remains a headwind to receivable growth, although we expect to see modest decreases in late 2021.

Looking at our other lending products. Organic student loans increased 4% from the prior year. We are well positioned as we enter the peak origination season. Personal loans were down 6%, driven by credit tightening last year and high payment rates. We are encouraged by continued strong credit performance in the portfolio and have expanded credit for new originations.

Moving to Slide 6. Net interest margin was 10.68%, up 87 basis points from the prior year and down 7 basis points sequentially. Compared to the prior quarter, the net interest margin decrease was mainly driven by a nearly 200 basis points reduction in the card revolve rate. Loan yields decreased 17 basis points from the prior quarter, mainly due to the lower revolve rate. This decline reflects the impact of increased payments as well as seasonal trends. Yield on personal loans declined 7 basis points sequentially due to lower pricing. Student loan yield was up 4 basis points.

Margin benefited from lower funding costs, primarily driven by maturities at higher rate CDs. We cut our online savings rate to 40 basis points in the first quarter and did not make any pricing adjustments during the second quarter. Average consumer deposits were up 6% year-over-year and declined 1% from the prior quarter. The entire sequential decline was from consumer CDs, which were down 9%, while savings and money market deposits increased 2% from the prior quarter. Consumer deposits are now 66% of total funding, up from 65% in the prior period.

Looking at Slide 7. Excluding the equity investment gains, total noninterest income was up $123 million or 29% year-over-year. Net discount and interchange revenue increased $102 million or 43%, as revenue from strong sales volume was partially offset by higher rewards costs. Loan fee income increased $20 million or 24%, mainly driven by higher cash advance fees with demand increasing as the economy reopens.

Looking at Slide 8. Total operating expenses were up $145 million or 13% from the prior year. Employee compensation increased $46 million, primarily due to a higher bonus accrual in the current (inaudible) versus 2020 when we reduce the accrual. Excluding this item, employee compensation was down from the prior year as we've managed headcount across the organization.

Marketing expense increased $46 million from the prior year as we accelerated our growth investments. We still see significant opportunities for growth, and we plan to accelerate our marketing spend through the year to drive account acquisition and brand awareness.

Information processing was up due to a $32 million software write-off. The increase in other expense reflects a $92 million charge on the remainder of the Diners intangible asset. Partially offsetting this was lower fraud expense, reflecting some of the benefits from our investments in data analytics.

Moving to Slide 9. We had another quarter of improved credit performance. Total net charge-offs were 2.1%, down 132 basis points year-over-year and 36 basis points sequentially. Net charge-off rate was 2.45%, 145 basis points lower than the prior year quarter and down 35 basis points sequentially. The net charge-off dollars were down $276 million versus last year's second quarter and $62 million sequentially. The card 30-plus (sic) [30-day] delinquency rate was 1.43%, down 74 basis points from the prior year and 42 basis points lower sequentially. Credit in our private student loans and personal loans also remained very strong through the quarter.

Moving to the allowance for credit losses on Slide 10. This quarter, we released $321 million from the reserves due to 3 key factors: continued improvement in the macroeconomic environment, sustained strong credit performance, with improving delinquency trends and lower losses. These were partially offset by a 2% sequential increase in loans.

Our current economic assumptions include an unemployment rate of approximately 5.5% by year-end and GDP growth of 7%. Embedded within these assumptions are the expanded child care tax credits and the benefit from the infrastructure fiscal package beginning in late 2021.

Looking at Slide 11. Our common equity Tier 1 ratio increased 80 basis points sequentially to 15.7%, a level well above our internal target of 10.5%. As Roger noted, we are committed to returning capital. The recent Board approval increasing our buyback and dividend payouts reflect that. On funding, we continue to make progress towards our goal of having deposits to be 70% to 80% of our mix.

Moving to Slide 12. Our perspective on 2021 continue to evolve as we see additional opportunities to drive profitable growth. We have increasing confidence in our outlook for modest loan growth in 2021 as strong sales and our new account growth should offset the higher payment rates. We expect NIM will remain in a relatively narrow range compared to the first quarter levels of 75%, with some quarterly variability similar to what we experienced this quarter. We anticipate a slight benefit from higher coupon deposit maturities and an optimized funding mix with yields affected by variability in the revolve rate.

Our commitment to disciplined expense management has not changed, and we remain focused on generating positive operating leverage and an improving efficiency ratio. For this year, we now expect nonmarketing expenses to be up slightly over the prior year, reflecting the higher compensation accruals and recovery fees. The increase in these expense categories is closely tied to the economic recovery. For example, the high level of consumer liquidity is supporting elevated recoveries. These recoveries have some costs associated with them, but are more than offset by lower credit losses.

Regarding marketing expenses, we expect this will step up more significantly in the second half of 2021 as we further deploy resources into account acquisition and brand marketing. With the continued improvement in credit performance, our current expectation is that credit losses will be down this year compared to 2020. Naturally, a material change in the economic environment could shift the timing and magnitude of losses. Lastly, as evidenced by our dividend increase and new share repurchase authorization, we remain committed to returning capital to shareholders.

In summary, we had another very strong quarter. We are well positioned for a positive top line trajectory given our sales trends and new account growth. Credit remains extraordinarily strong, and the economic outlook continues to improve. We maintained our discipline on operating expenses while investing (inaudible) returning organic growth opportunities. And finally, we continue to deliver high returns, allowing for enhanced buybacks and dividends.

With that, I'll turn the call back to our operator to open the line for Q&A.

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

Just wondered if you can give a little more color on the payment rate expectation. I guess just what do you see and what gives you confidence in the moderate decline expected for the back half of this year? And if you could talk about the -- how you weigh the risks that the payment rate may not moderate from here? And then secondly on that, are you seeing any differences in your FICO bands in terms of the payment rate behavior?

Yes, I'll take the call -- or the question, John. Thanks for that. So payment rates are, frankly, at a record high. So if you looked at our trust data, you can see that in June, they came in at about 29%, and that's -- frankly, that's an all-time high, at least as far as back as 2005. So what we're seeing when we look at the portfolio, now the portfolio, the trends are similar, but the overall payment rates a bit lower.

What we're seeing is a couple of factors. So one is a deceleration of the growth in payment rate. And second, we formed our expectation on the second half of the year based on all of the government programs that are out there. And most -- the most significant ones are expiring or have expired by the end of the third quarter.

So we expect that, coupled with the strong economic activity and lower savings rate that we're observing, to result in a moderate decrease in the payment rate certainly in the fourth quarter. In the third quarter, it could be flat to maybe even up a mild bit. But overall, our sense is that we've approached the peak of this, and it's going to begin to tail off.

And in terms of the mix by FICO band, it's pretty broad. Certainly, the higher FICOs skew transactor, so you'll have a higher payment rate. In some of the mid FICOs, you see revolvers turn transactor. But again, to John's point, we do expect it gradually to normalize.

Okay. Great. No, that's helpful. And then separately, I appreciate the commentary around the loan growth expectation, and you also expect some variability in your margin. So how -- what does that imply for your net interest income expectations? If you could elaborate a little bit on that in terms of the second half trajectory there and possibly going into 2022.

Yes. So I'll touch upon it. So the net interest income will follow loan growth. But as you know, in the fourth quarter, the build on loan balance is skewed towards the last 2 months of the quarter. So net interest margin is what I'll say the trajectory is changing in a positive direction, given the sequential loan growth we saw from first quarter to second quarter. But again, it's going to be in the single digits, lower single digits is the general expectation.

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

My first question is, I know everyone is sort of waiting for the consumer to relever and there's positive signs. But maybe, Roger, you could just speak to the competitive environment? I mean, obviously, everyone's chasing growth or wanting to grow. Or maybe you could just speak to how it's out there.

Yes. Thanks, Sanjay. As we've discussed in the past, the card business is always competitive. There was a nice little lull last year that we've been able to take advantage of and pick up even more market share. But I think this is returning to more normal, maybe a little heightened but more normal level of competition. You're seeing new products out there, increased marketing spend. But as I look at our value proposition, cost per accounts we're achieving and the returns, I feel very, very good about the marketing spend we're putting out there and what we're generating for it.

Okay. Great. My follow-up is on the Sezzle investment. Obviously, they're an up and coming buy now, pay later company. I'm just curious sort of if you could speak to that specific investment. And how you see that unfolding for yourselves, both in terms of the investment itself but also strategically inside of buy now, pay later?

Sure. So in terms of our buy now, pay later strategy, there are really 2 parts: The Sezzle investment is really being driven by the payment side. And that's similar to the investment we made in Marqeta a while back. Our set of network assets are very useful for many fintechs in terms of just an easier way to process payments and connect to merchants in a wide variety of forms. So I'd say that's the core of what we're doing with Sezzle.

We also believe that potentially, over time, there may be opportunities on the banking side. So in terms of our providing lending, not necessarily with Sezzle, but in the buy now, pay later space more broadly. And again, leveraging what we can do with unsecured lending and our direct merchant relationships. But I'd say, we haven't announced anything on the lending side of buy now, pay later at this time.

Your next question comes from Ryan Nash with Goldman Sachs.

John or Roger, can you maybe just talk about the expectations for monetizing the Marqeta gain? How much do you expect to reinvest versus used for repurchases? Is it already factored into the buyback? Second, can you maybe just remind us what the lockup is and your intention for the stake? And maybe just lastly, this increase in expenses that we have right now, marketing and nonmarketing investments, how much of these are being driven by the Marqeta gain? And should we expect these to be onetime in nature or should we expect them to stay in the run rate?

Okay. All right, Ryan. So the Marqeta gain was something that when we looked at the plan for the year, we didn't, frankly, envision that, that opportunity would turn out quite the way it did. So certainly a nice outcome. The spend that we're seeing isn't dictated based on that size of the Marqeta gain, but more about the broad economic opportunity that we're seeing right now to be able to drive positive growth. So frankly, we'd be spending at this level with or without the Marqeta gain.

The kind of the lock-up, those points are disclosed in the S-1, so you can reference that and get real specificity on it. And then finally, in terms of the expenses and are they onetime in nature. So within the presentation itself, we highlighted kind of 3 items that were significant. So one was the bonus accrual, unrelated to Marqeta again and more broadly a reflection of what's happening in the business and the outcomes we're seeing here in terms of generating positive returns.

The other item there was the Diners intangible, which we took the opportunity based on what we see as a change in the cash flows in Diners to fully impair that and took a $92 million write-off. And then the third item was a $30 million software write-off.

So all those are part of our normal, I'll say, hygiene factors. So the last 2 items were what I would consider relatively one-off, and I wouldn't include those in the kind of operating cash flows going forward. But as you're trying to get a sense for overall spend in 2021, we talked about accelerating marketing expense. So the way I would think about marketing specifically is that we anticipate it to return to 2019 levels, which would indicate a material acceleration in the second half of the year, which, again, is tied to the fact that we're seeing good origination opportunities to drive growth and long-term profitability.

Got it. And maybe as my follow-up, so you announced a new $2.4 billion buyback. But just given how high the levels of profitability are, you're still going to be well above the 10.5% level. So can you maybe just help us understand the strategy of getting the capital down over what time frame? Are you managing all in to CECL day 1? And then maybe, attached to that, Roger, you talked about potential lending opportunities in buy now, pay later. We've seen others take -- some go at it organically, some go via acquisition. What would you expect to be the strategy for Discover on a go-forward basis?

So I'll hit the capital point real quick. And then we'll flip it over to Roger to handle that second part of that question -- the second part of that second question, I should say. So the capital plan was based on a couple of different factors. So one, we want to be prudent in our distributions of capital and make sure that we -- organic -- that we allocate our capital to organic growth opportunities and then buybacks and dividends and then acquisitions as they appear. So that hasn't changed whatsoever.

The planning in terms of capital overall is that we are committed to that 10.5%. As we look at the impacts of the CECL capital relief transition, that's about 200 basis points. And then in the first quarter of next year towards the end of this year as we review the outlook for 2022 with our Board, we'll make recommendations with a couple of points in mind: the overall economic opportunity; the SCB and where we are with that, and there's plenty of room there; and then finally approaching the 10.5%.

Yes. And in terms of buy now, pay later, there are many sort of different things we could do. Clearly, going direct with the merchant relationships we have is one option we could work with partners. I think the market is still very immature, and that's why you're seeing many different combinations, and it will evolve over time. But we don't have anything specific to share at this time.

Your next question is from Don Fandetti with Wells Fargo.

Roger, kind of a bigger-picture question around technology spend. Just want to get your thoughts on where your head is right now, whether you're kind of making enough investments in AI and machine learning and staffing. Different financial institutions are at different points of the cycle.

Yes. It is a major source of investment for us. And I would say that the scarcest asset there is talent as opposed to technology dollars. And then part of it is just making sure you're investing to monetize the granularity you can pick up with enhanced data and analytics. So looking at the entire martech stack, your personalization abilities, what you can deliver through different channels. So it's a big focus.

I think we're probably the only major bank where the Head of Data and Analytics is a direct report to the CEO. And so I think that reflects the importance we see on that, and I feel very good about our level of investment.

Your next question is from Bob Napoli with William Blair.

Question, just to follow up, I guess on Roger. Your investments in like Marqeta and Sezzle and just the thought -- I mean, how are you working with Marqeta? How are these investments improving the product set, if you would, for Discover? And are there more of these types of investments? Or could they become acquisitions in the future?

Sure. So all of these investments start with a commercial relationship. We are not venture capitalists. We have no desire to get into that business. But where we have a commercial relationship with a company where we think there's a lot of potential and there is a way to invest and perhaps extract additional terms, maybe an exclusive arrangement or other commitments, those are the types of opportunities we pursue. And they're largely on the payment side. And we may even, given some of the opportunities we're seeing, accelerate that, although, again, it's going to be relatively modest.

In terms of acquisitions, given the valuations, I'm not sure I necessarily see doing acquisitions. And this approach around partnership and investment seems to work well. But we will look at it. John and our payments team have a very strong business development effort, and we'll consider acquisitions where it's a capability that will help us monetize our payments assets.

Great. Got a follow-up on spend. Your chart on Page 16, the acceleration through the quarter relative to 2019. I mean, are you surprised by the level of spend that we're seeing? And how do you think about that relative to the long-term trend? I guess if you take this quarter, you maybe had a 10% compound growth from 2019, but were you surprised by this and the acceleration through the quarter? And your thoughts on spend growth as you lap -- as you look at 2022 and onward.

Yes. I mean I think compound growth is a very challenging measure because it smooths an unbelievable cycle that I don't think any of us would have expected in 2019. So the year-over-year spend growth numbers are unsustainably high for the economy as a whole. The economy is growing above, I think, what economists would say its real level should be. And so you're seeing that pent-up demand from consumers. We expect it to normalize at a strong level, and we look to continue gaining market share given how we're positioned. But I don't think anyone is expecting retail sales to stay up 30% forever.

Your next question comes from Mark DeVries with Barclays.

I had a question about the reserve levels. Could you give us some sense of what you're kind of contemplating both from a delinquency and charge-off perspective in those reserves? I mean, I think trends have obviously been very benign and yet you're still almost 200 basis points above kind of your CECL day 1. What do you need to see to see those reserves come down more meaningfully?

Yes. Mark, thanks for the question. So when we approached this quarter, certainly, the economic data was improving. The portfolio is performing extremely well. And the charge-off trajectory, we revised our guidance and now believe it will be below the prior year. So all of those are very, very positive from a reserving standpoint.

We did put a bit of caution in the kind of reserve numbers as a result of the various government programs that are out there. So if you think about the eviction moratoriums, foreclosure moratoriums, various payment deferral programs and then the massive amount of government transfer payments that hit the economy, we felt like that those contributed to driving delinquencies and charge-offs below historical norms for our customer base.

So under CECL, we reserve for life of loan. And accordingly, we were waiting for those programs to run their course and the cash flows to kind of feed through the economy and impact our customers, and then be able to make, call it, a deeper change to the reserve levels. So as I said earlier, most of those programs ran their course in the third quarter. We obviously have the child care tax -- or the child tax credit. We don't expect that to be an impact.

So in the fourth quarter, we should begin to see more data that would allow us to take a different look at reserves and then into 2022, certainly. Our expectation is that the credit environment is very positive from a growth standpoint. And accordingly, we'll make reserves that kind of align with that.

Okay. Got it. And then just one more question. Do you have any other private investments that look like they could be in a large gain position just based on any kind of subsequent funding rounds and the valuations those are done at?

So we have some investments, yes. We try to get in early and build out the commercial relationship, as Roger alluded to. And we're hopeful that the combination of working with quality companies and developing deeper commercial relationships will help those folks be profitable and help us drive business in our payments area, while giving us an option value for that investment. So one would hope that there's more. But at this point, it would be way premature to speculate.

Your next question comes from Manu Srivareerat from UBS.

I guess you talked a lot about the benefits of having a differentiated product with flexible cash rewards. Now as things get back to normal and travel rewards become more relevant for consumers, are you seeing any impact to your utilization levels? Is there any indication that consumers might have a shifting preference for travel and my star rewards over cash rewards?

We've been successfully marketing cash awards against travel rewards and every other type of card out there for decades. So I wouldn't get too caught up in sort of the shift away from frequent flyer miles and in 2020. I think it points to sort of the stability and structure of our program that we didn't have to make a series of changes. We focus on continuously enhancing it.

A lot of people don't focus as much as we do on redemption, so the ability to redeem at point of sale at Amazon through our partner, PayPal. So we are confident that we can compete with the travel rewards programs. A lot of the miles cards really target the super prime transactors, which isn't -- ours is much more a lend-focused model. We've got a very good miles card ourselves that we market. But again, I feel really good that our rewards can compete against anyone out there.

Okay. And just as a follow-up, how should we think about the rewards rate as you think about growing loan balances, for example? Is there anything to that?

Yes. So I'll cover this one. So the rewards rate over time has been relatively stable. It's increased somewhere between 1 and 3 bps annually. We expect that the 5% programs and the rewards structure will continue that trend, while helping us to sustain a strong level of customer loyalty. So we look at that increase as a natural evolution of a good investment in our customer relationships.

Your next question comes from Rick Shane with JPMorgan.

Mark really covered what I wanted to talk about in terms of reserves, but I'd just like to follow up on that slightly. The observation was made that your reserve rate is about 2 points higher than it would have been on day 1. Day 1 was essentially an ideal economic environment, low unemployment, steady growth, et cetera. Should we look at day 1 as a destination for where the reserve rate will go, or more as a channel marker in terms of sort of an ideal operating environment?

Yes. Good question. So the short answer is that it depends on the economic outlook and the portfolio performance. I would say, to give a kind of view on the marker, when we look at the reserve rate back in, call it, January 1, we're at just over 6%. We're at just over 8% now. That difference is about $1.7 billion of reserves on the current loan book. So if we have a level of confidence in the economic environment, portfolio continues to be positive. And the work we've done on analytics drive, I'll say, positive credit performance.

There is an expectation in a benign environment that we should march back that way. No. I think it's premature to say we'll get there. There's a lot of factors that could change. But I look at that day 1 as a marker for a solid credit environment and a relatively strong performing credit book.

Great. Okay. That's very helpful.

Your next question comes from Bill Carcache with Wolfe Research.

Roger and John, can you guys help square the continued strength in customer acquisition that you expect on one hand, with payment rates remaining low for an extended period on the other? You mentioned, I think, John, the end of the pandemic-related federal forbearance program this summer as the catalyst for payment rate normalization. But why do you think payment rates will remain elevated after that support has ended?

Yes. I'll start it and then maybe Roger will add a couple of points on the back end. So certainly, if you look at, frankly, the projected GDP, that's above normal levels for a strong U.S. economy. And that's as a result of, frankly, all of the activity and the transfer payments I talked about earlier.

The -- I think there's a natural, I'll say, transition from where we are today to something that approaches historical norms. The savings rate is still high, but it's falling. And so that's positive. And in terms of overall sales activity versus payment rates, we do expect the sales levels at some point to moderate partly because of year-over-year comps and partly because of what's happening on economic activity.

But with that said, I would expect that payment rate, there's a bit of an inverse relationship. And what we're seeing from a growth standpoint is the sales activity, the new account acquisition, the payment rate as a bit of a headwind and then that revolve transactor rate aligning to payment rate. But despite all that, we've been able to grow from first quarter to second quarter. So as that payment rate moderates a bit, sales activity, I think, will be a lag to that. And we'll be well positioned to grow in the fourth quarter and beyond.

Okay. That's helpful. And if I may follow up on that. As we look out to next year, would you expect that process of -- that you've described in this -- sort of this payment rate and revolve rate normalization, should that provide an incremental tailwind to loan growth such that we could actually see your loan growth start to outpace your spending growth?

2022 is still about 6 months away, 5 to 6 months away. So we're working on it right now. I'll tell you this, I'm positive in terms of how the business is positioned and what we can expect from a top line standpoint as well as credit and expense. So those factors, I feel very good about, but to get into a level of specificity at this point is probably a bit early.

Your next question is from Meng Jiao with Deutsche Bank.

I wanted to (inaudible) student loans. As we head into the third quarter, it's probably the first time in over a year that probably many college students are on campus. I guess, are you doing anything differently in terms of how you're approaching the start of this particular school year? And sort of, is it fair to say that marketing to students is going to be a large piece of that accelerating marketing spend that you guys have commented on?

The third quarter is peak season for student loans, but all the marketing is done before they get to campus. Volume was suppressed last year because for many of our customers, we encourage them to take federal student loans and then use our private student loans to top off. And if their expenses were lower because they're paying tuition but not paying housing and food, et cetera, that had an impact on our volume.

On the other hand, we did benefit last year from a major competitor pulling out of the marketplace. So I would say we're optimistic with kids going back on campus, that this will be a strong year for originations and that we will continue to gain market share in the student loan product.

Got you. Great. And then secondly, I think you mentioned growing confidence in consumer. Quarter to date, have you seen, I guess, a continued acceleration in that spending? I know we're roughly 22 days into the third quarter. And then secondly, Roger, I think you (inaudible) spend trends are unsustainably high. I'm just curious if you had any further color as to when you expect that to sort of normalize.

I'll start with that one. I think it will take a while to normalize. There is a good amount of pent-up demand, but also pent-up liquidity in the form of the savings that have built up as well as just the open to buy that people have on their cards from sort of month after month of pretty high payment rate. So I think that will provide some support.

On the other hand, a huge amount of uncertainty in terms of what will happen with the Delta variant and other factors that can drive the economy. So we feel like we're well positioned, but unclear what the back half of the year will bring. In terms of what we're seeing so far this month, I would say travel continues to be constructive, but overall, not necessarily accelerating, but spend levels staying very strong.

Your next question is from Moshe Orenbuch with Crédit Suisse.

Great. I guess getting back to the BNPL, is your primary kind of revenue or growth kind of opportunity in the Sezzle on the lending side? Is that the way we should think about it?

No. You should think about it on the payments segment. So transaction processing revenue and sort of (inaudible) with payment processing and sort of connectivity to the merchant side. And so that's sort of -- our first entry into buy now, pay later has been more on the Payment Services segment as opposed to our direct banking side.

Got you. One of the things that you've talked about in the past about buy now, pay later is the question of whether it kind of encourages or discourages younger consumers from taking out a credit card. And I think you had said either on the last call or one of the presentations, that because of student lending brand of Discover that you haven't kind of seen that impact. Could you just talk a little bit about how you think about the value, I guess, of the student lending in terms of kind of -- and in terms of that marketing of the Discover brand?

Sure. We are very successful in the student card business. I think the student loans help, but also it has to do with the product, the value proposition and some very good marketing. And we see actually great credit performance from students, because contrary to popular belief, a lot of them are very responsible in how they handle their cards.

With student loans, I'd start by saying, we think it's a great business. It's profitable. I would say it is operationally complex, but it also gets our brand out there with parents as well as students when they're making a series of financial decisions. So there are not that many banks that are in it at scale, not that many, too, that have sort of the product set that we have to be able to leverage those relationships. So it's a key part of our banking strategy.

Got it. And just a quick follow-up on the reserving point. I guess I was sort of wondering if your reserves on day 1 included an expectation of a potential of a recession somewhere in the life of those receivables. Just as you think about it as we kind of enter 2022, would that be a lower probability just given the experience that we've had? Or is that (inaudible) think about it?

Yes. So when we did day 1, we looked at -- through the cycle of use. So what essentially that means is we didn't look at the trough of delinquencies and charge-offs, nor did we look at peak. So to extremely oversimplify it, a bit of averaging with the help of a bunch of data scientists. So that day 1 is a, call it, a normalized view of the credit environment.

Your next question comes from Mihir Bhatia with Bank of America.

I wanted to ask -- I wanted to go back to your comment on competitive intensity, and it certainly does feel like cash back cards has increased. I think you alluded to intensity being a little higher than maybe the average historically. So maybe just talk a little bit more about the dimensions of this competition. Is it spilling over beyond just higher direct rewards, like the 1.5% or 2% rewards.

Are you seeing any irrational signing bonuses, introductory offers. Spend $5,000 in 3 months, get $100, or anything irrational? Are you seeing anything on the balance transfer side? How is Discover responding to this increased competition? And is there -- is the increase in marketing spend in part because of this elevated competition?

Yes. So I'll start with the back half. The increase in marketing spend is driven really by the opportunity we see. So it is not by the competition. I do expect the cost per account to be modestly higher than, say, 2019 levels. 2020 was extraordinarily good. But that -- part of that has to do with just booking more new accounts, and the marginal ones tend to be more expensive.

I do think some of the rewards products out there will not be sustainable long term. You've seen 2% cash back offers come and go for a very long time, especially when we eventually get to a higher rate environment.

So far, balance transfer demand is still a bit suppressed. So we're not seeing the overly long 24-month 0% offers. You are starting to see some of those high, spend $4,000, get $400-type offers from some of the issuers that I think are most aggressive around growth. But again, nothing that I would say we haven't seen before. It tends to be different people at different times. But this is part of the card business and what we're used to competing against.

Great. And then just -- I just wanted to ask about your debit product. Any update on that? Has that -- where are we in terms of that being rolled out? Is there thoughts to expand the marketing or push behind it? Just any update on the debit product.

Yes. So given the excess deposits we had, we really pulled back on the vast majority of our deposit marketing, including that product. But we're very excited about the differentiation, having cash back. So we think we're well positioned against the challenger banks as well as traditional banks. And so we'll continue to ramp it up, probably more so towards the back half of the year and as we get into 2022.

Stephanie, I think we have time for one last question.

Your last question comes from Betsy Graseck with Morgan Stanley.

So I just had 2. One is on the marketing. I just wanted to make sure I understood what you meant by marketing expense anticipated return to 2019 levels. Are you saying that the back half of '21 will be at a run rate similar to full year 2019 or the back half of '21 is going to be similar to the quarterly run rate of 2019?

Yes. So think total year '21 to be similar to total year of '19. That would require substantial (inaudible) in the back half.

Got it. Okay. And then just separately on credit. I know a huge debate in the group here is around payment rates and how we're anticipating them to traject over the course of the next year or so. One other question specifically for you is that, if I recall correctly, you've got a homeowner SKU to your book, and home prices have clearly accelerated significantly, 15%, 20%, depending on location.

So I know you look at your customers routinely and check how they're doing financially. Are you seeing any signs of your customer base having refi-ed and materially dropped their monthly mortgage payment requirements, which would potentially keep your payment rates elevated for longer? Or how you think about this refi cycle and what it means for your customers and their financial health? And how that translates to payment rate for you?

Yes. We have -- and this cuts across past cycles. We haven't seen a link between sort of a refi boom and card payment rates. And so I don't have any expectation that would happen this time.

I would now like to turn it back over to Eric Wasserstrom for closing remarks.

Well, thank you very much for joining us. If you have any additional questions, the IR team is available all day and, of course, whenever. So thanks again, and have a great day.

Thank you. This concludes today's conference call. You may now disconnect.