Discover Financial Services (NYSE: DFS) Q4 2020 earnings call dated Jan. 21, 2021
Corporate Participants:
Eric Wasserstrom — Head of Investor Relations
Roger C. Hochschild — Director, Chief Executive Officer and President
John T. Greene — Executive Vice President, Chief Financial Officer
Analysts:
Moshe Orenbuch — Credit Suisse — Analyst
Sanjay Sakhrani — KBW — Analyst
Rick Shane — JPMorgan — Analyst
Mark DeVries — Barclays — Analyst
Betsy Graseck — Morgan Stanley — Analyst
Bill Carcache — Wolfe Research — Analyst
Don Fandetti — Wells Fargo — Analyst
Kevin Barker — Piper Sandler — Analyst
Bob Napoli — William Blair — Analyst
John Hecht — Jefferies — Analyst
Mihir Bhatia — Bank of America — Analyst
Meng Jiao — Deutsche Bank — Analyst
Dominick Gabriele — Oppenheimer — Analyst
Presentation:
Operator
Good morning. My name is Maria, and I will be your conference operator today. At this time, I would like to welcome everyone to the Fourth Quarter and Fiscal Year 2020 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.
Eric Wasserstrom — Head of Investor Relations
Thank you, Maria, and good morning, everyone. Welcome to this morning’s call. I’ll begin on Slide 2 of our earnings presentation, which you can find in the Financials section of our Investor Relations website, investorrelations.discover.com.
Our discussion today contains certain forward-looking statements that are subject to risks and uncertainties that may cause the actual results to differ materially. Please refer to our notices regarding forward-looking statements that appear in today’s press release and the presentation.
Our call today will include remarks from our CEO, Roger Hochschild; and John Greene, our Chief Financial Officer. As we conclude our formal comments, there will be time for a question-and-answer session. During the Q&A session, please limit yourself to one question; and if you have a follow-up question, please get back into queue, so we can accommodate as many participants as possible.
Finally, I’d like to extend a tremendous thank you to Craig Streem for all of his help and support over the past few weeks and for his friendship and guidance over the past many years. So, thank you, Craig.
And with that, I’d like to turn the call over to Roger.
Roger C. Hochschild — Director, Chief Executive Officer and President
Thanks, Eric, and thanks to our listeners for joining today’s call. I too want to add a farewell to Craig, and thank him for many years of service to Discover, dating all the way back to great support in our original spin-off from Morgan Stanley.
Our strong fourth quarter results were the capstone to good performance in a very challenging year, proving the value and resilience of our digital banking business model. While the economic impacts of the coronavirus pandemic were expansive, our business stood up to these challenges and we are in $799 million after-tax for the fourth quarter and over $1.1 billion for the full year.
Our fourth quarter results underscore the capital generation of our model. While our revenues were down 4% year-over-year, our outstanding credit performance combined with the actions we took to reduce our funding costs and control our expenses, enabled us to exit the year with a 30% ROE in the fourth quarter.
Looking back on the full year, our operating results highlight the strength of our business and the execution of our team. We proactively adapted to the many ways in which the pandemic has altered our operating environment, including changes in consumer spending patterns, repayment trends and borrowing habits.
Discover has always provided best-in-class customer service and this did not change with the pandemic. While other issuers faced significant challenges with long hold times, there was no disruption to our outstanding service as we leveraged our digital capabilities and our 100% U.S.-based customer service. We kept average hold times under five minutes through the trough of the downturn and they quickly returned to normal levels of under one minute.
Additionally, our operational flexibility allowed us to keep our employees safe with nearly all employees continuing to work-from-home since mid-March.
Similarly, our history of conservative credit management and the resilience of our prime customer base positioned us well as we entered this period of economic stress. We also took quick action at the onset of the pandemic to mitigate credit risk by tightening new account underwriting criteria, reducing promotional rate offers and pulling back on credit line increases, and we added $2.4 billion to reserves as the macroeconomic outlook deteriorated.
As expected, the tightening of our underwriting criteria combined with elevated payment rates impacted our loan growth and our total loans decreased 6% year-over-year. However, we took substantial market share in private student lending and gain share in card lending.
Our digital bank operating model combined with disciplined expense management has historically generated industry-leading efficiency, but in 2020, we took additional action to control costs and delivered on our commitment to cut $400 million from planned expenses despite absorbing several one-time items that have occurred during the year. Excluding the one-time items, expenses were down nearly 2% year-over-year.
We achieved this while still investing in analytics, automation and core technology capabilities to support long-term growth and efficiency improvement. For example, we’ve invested in new analytics to optimize acquisition marketing to our core customers and we enhanced customer engagement through more targeted offers and interactions.
Our payments business was also well positioned heading into last year’s sudden economic decline. Our PULSE debit business had a strong year with volume up 10% from 2019, reflecting the impact of stimulus funds and higher transaction sizes as debit became a critical way of procuring goods during the pandemic.
Our Discover proprietary network was down 2% for the year and exited the fourth quarter at plus 5%, in line with card sales volume. While Diners Club volume declined, reflecting the impact of the pandemic on global T&E spending, we signed five new international network partners as we continue to expand global acceptance.
As we look into 2021, we believe there is a significant amount of uncertainty around the timing and shape of the economic recovery and some of the impacts from last year’s downturn have yet to be fully realized. For this reason, we anticipate deterioration in our credit performance in the back half of this year.
Nonetheless, we see reasons for optimism. After bottoming in April, our card sales steadily rebounded throughout the year and returned to growth in the fourth quarter. We saw improvement across all categories, especially in grocery and retail, which now represent more than half of our sales mix. This favorable sales trend has also continued into the first half of January.
In this environment, we will remain disciplined on expense management but also committed to making investments for growth and efficiency and anticipate higher marketing expenses relative to 2020 levels.
We also anticipate resuming a more normal pattern of capital return. Earlier this week, our Board of Directors approved a new $1.1 billion share repurchase plan and we may begin buybacks as soon as the first quarter, subject to the Federal Reserve’s limitations.
In conclusion, 2020 presented us with unparalleled challenges but our business model has well positioned, and I couldn’t be more pleased with how the team here at Discover responded. I’m confident that the actions we took strengthened our business and will drive long-term value for our shareholders and customers.
While there is still uncertainty around the pace of the recovery, I’m optimistic that 2021 will be a better year.
I’ll now ask John to discuss key aspects of our financial results in more detail.
John T. Greene — Executive Vice President, Chief Financial Officer
Thank you, Roger, and good morning, everyone. I’ll walk through our fourth quarter results, starting on Slide 4. We earned $799 million in net income or $2.59 per share. These results included several one-time expenses, totaling $137 million. Excluding these, EPS would have been $2.94.
There were a number of factors, both positive and negative, that influenced our performance during the year. Importantly, our results for 2020 reflect proactive management of our funding and operating cost and our conservative approach to credit management. These factors helped offset the revenue impacts of elevated payment trends and lower sales volumes.
However, we’re seeing some positive signs with a return to sales growth in the quarter and continued expansion of our net interest margin. In the fourth quarter, net interest income was down 2%, reflecting a 5% decline in average receivables and lower loan yield. This was mostly offset by decreased funding costs driven by lower market rates and management of our deposit costs.
Non-interest income was 14% lower, driven by higher rewards costs from strong engagement in the 5% category this quarter, a one-time write-off of certain real estate facilities and lower loan fee income also contributed to the year-over-year decrease.
The provision for credit losses was $305 million lower than the prior year due to meaningful improvement in credit performance. There was no change to reserve levels in the current quarter, whereas the prior year included an $85 million reserve build.
Operating expenses increased 8% year-over-year driven by one-time items related to software write-offs, charges for penalties and restitution, and costs of a voluntary early retirement program. Compensation expense also contributed to the increase. These were partially offset by lower marketing costs and decreased professional fees. Excluding one-time items, expenses were down 4% from the prior year.
Turning to loan growth on Slide 5. Total loans were down 6% from the prior year driven by a 7% decrease in card receivables. Lower card receivables were the result of three factors. First, payment rates continue to be elevated. While this has reduced loan balances, it has had a favorable impact to credit performance. Second, promotional balances have continued to decline due to actions we took early in pandemic to tighten credit. And third, an increase in transactor activity.
Looking at our other lending products. In student loans, we had a strong peak origination season, increasing market share in leading to 7% loan growth. Personal loans were down 7% as a result of actions we took early in the pandemic to minimize potential credit losses.
Moving to Slide 6. Net interest margin continued to improve, up 34 basis points from the prior year and 44 basis points from the prior quarter to 10.63%. Compared to the third quarter, lower deposit pricing was the primary driver of the improvement in net interest margin. We cut our online savings rate another 10 basis points from the third quarter as we continue to actively manage funding costs lower. Online savings rates, 12 month and 24-month CDs are all down to 50 basis points.
Lower interest charge-offs and a favorable mix of promotional rate balances also contributed to the margin expansion from the prior quarter. These factors offset the impact of the high level of balance sheet liquidity that we are currently carrying.
While this is depressing, our net interest margin today, it should benefit margin as we deploy this liquidity into future loan growth.
Average consumer deposits increased 18% from the prior year. We continue to see steady demand for our consumer deposit products, which were up $1 billion from the prior quarter.
All of the growth from the third quarter was from indeterminate maturity accounts, which allows us to immediately capture the benefit of deposit rate decreases.
Turning to Slide 7. We continue to optimize our mix — our funding mix. And our goal remains to have 70% to 80% of our funding from deposits. We also have an opportunity to benefit from higher rate funding maturities over the next couple of years. Both of these items are expected to benefit net interest margin in future quarters.
Looking at Slide 8. As Roger noted, we delivered on our commitment to reduce planned expenses by at least $400 million during the year, and I’m pleased to say that we accomplished this goal despite several one-time expense items that occurred in 2020.
Total operating expenses in the fourth quarter increased $94 million or 8% from the prior year, driven completely by $137 million in one-time expense items. Excluding these, operating expenses would have been down $43 million or 4% year-over-year. Yet, even as we remain disciplined on costs, we continue to invest in analytical capabilities that we expect will drive future growth and efficiency improvements.
Looking at some of the individual line items. Employee compensation was up $57 million or 13%, driven by a one-time item related to a voluntary early retirement program as well as staffing increases in technology and higher average salaries and benefits.
Marketing and business development expense was down $75 million or 32%. Most of the reduction was in brand marketing and card acquisition as we aligned marketing spend and tightened credit criteria in response to this changing economic environment.
Professional fees decreased $22 million or 10%, mainly driven by lower third-party recovery fees related to courts operating at limited capacity.
On Slide 9, you can see our credit metrics for the quarter. Once again, credit performance was very strong and total charge-offs below 2.4% and credit card net charge-offs nearly 2.6%. The card net charge-off rate was 78 basis points lower than the prior year, while the net charge-off dollars were down $180 million or 28%.
Compared to the third quarter, net card charge-off rate declined 82 basis points. The 30-plus delinquency rate was 55 basis points lower than last year and increased 16 basis points from the prior quarter.
Credit also remained strong in our private student loan portfolio with net charge-offs down 31 basis points year-over-year. Excluding purchased loans, the 30-plus delinquency rate improved 40 basis points from the prior year and 9 basis points compared to the prior quarter.
In our personal loan portfolio, net charge-offs were down 147 basis points and the 30-plus delinquency rate was down 29 basis points year-over-year.
To wrap up credit, this was another quarter of strong performance across all of our lending products driven by the actions we’ve taken in underwriting, line management and collections, the resiliency of our prime customer base and the impact of government stimulus.
We believe that losses will increase in the second half of 2021 and into 2022. However, the timing and magnitude of losses could be impacted by any additional government assistance or material shifts in the economic environment.
Moving to the allowance for credit losses on Slide 10. We held allowance flat to the prior quarter. While the macro environment has improved, the outlook remains uncertain.
Similar to our past approach, we modeled several different economic scenarios. The primary assumptions in our economic model were a year-end 2021 unemployment rate of about 8% and GDP growth of about 2.7%. We did not include any additional stimulus beyond what has already been provided to consumers.
Our economic scenarios also consider the increasing number of COVID cases, the timing of the vaccine rollout and the recent increase and unemployment claims.
Turning to Slide 11. Our Common Equity Tier-1 ratio increased 90 basis points sequentially to 13.1%, remaining above our 10.5% internal target and well above regulatory minimums. We have continued to fund our quarterly dividend at $0.44 per share.
Regarding share repurchases, as Roger indicated, earlier this week, we received authorization from our Board of Directors to repurchase up to $1.1 billion of common stock. And we intend to begin share buybacks in the first quarter, subject to the Federal Reserve regulatory limitations.
Moving to Slide 12 and some perspectives on how to think about 2021. We anticipate modest positive loan growth for the year. We see opportunities for customer acquisition, but the level of growth will be dependent on payment rate trend and the timing and pace of the broad economic recovery.
With respect to our net interest margin, as I previously mentioned, we expect we will continue to see benefits to margin from lower deposit rates and maturity of higher rate funding. We also remain committed to disciplined expense management and we’ll continue to invest to drive profitable long-term growth and efficiency improvement. This includes increased marketing investments for new customer acquisition as well as investments in advanced analytics and technology capability.
In terms of credit performance, as I’ve already noted, we expect losses to increase in the second half of 2021 and remain elevated into 2022.
Finally, our capital allocation strategy has not changed, and we remain committed to returning capital to shareholders through dividends and buybacks.
In summary, we had a solid fourth quarter with strong credit performance across the portfolio, net interest margin expansion driven by lower funding costs, flat reserves; and excluding one-time items, operating expenses were down year-over-year.
I’m very pleased with our execution throughout 2020 and Discover’s efforts to protect employees and provide uninterrupted best-in-class service to our customers.
The quick actions we took at the onset of the pandemic and our investment in core capabilities protected and strengthened the Discover franchise in 2020.
The challenges we faced, demonstrate the resiliency of Discover’s digital banking business model; and while uncertainty remains, we are well positioned for growth as our economy continues to recover.
With that, I’ll turn the call back to our operator, Maria, to open the line for Q&A.
Questions and Answers:
Operator
Thank you. [Operator Instructions] We’ll take our first question from Moshe Orenbuch from Credit Suisse.
Moshe Orenbuch — Credit Suisse — Analyst
Great. Thanks, and welcome, Eric. And if Craig is listening, it’s been a pleasure working with you too. I guess both, Roger and John, you talked about the idea of increasing marketing expenses into 2021. Could you just kind of flesh that out a little bit, talk some about how much you’d like to grow accounts? What you’re looking for? What signals would drive you to step on the gas a little harder or pull your foot off? Because you’ve seen some of your competitors already start to spend the marketing. We don’t know yet what that’s generated. But maybe just talk about that a little bit.
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes. Thanks, Moshe. So, you saw how we’re thinking about the expense base for 2021. And then, I’ll talk more specifically around marketing and new customer acquisition and growth.
So, in terms of expenses, the business is committed to driving positive operating leverage over the mid-term. Now, opportunities in 2021 will dictate how much marketing dollars we ultimately end up spending for new customer acquisition. But in terms of the overall expense base, there will be incremental dollars for marketing, new customer acquisition and third-party recovery fees. As the courts reopen, we expect that those fees will increase consistent with the level of recoveries that we hope to achieve on bankrupt accounts.
So, outside of those areas, we’re targeting to keep expenses flat across the business. Now, certain accounts may go higher, certain may go lower, but the way you can think about it is growth initiatives, incremental spend, the balance of the income statement in terms of expenses will be flat to down.
In terms of new account growth, what we’re targeting is mid-single digits. Maybe if we’re fortunate, we see some opportunities, upper single-digit account growth. And we hope that will translate into increased loan balances. We didn’t give any specific guidance on loan balances because of the broad economy uncertainty there; and frankly, repayment trends have been pretty remarkable; and with it, potential new round of stimulus, that could further increase the prepayment trends.
So, that’s how we’re thinking about growth and expenses. So, if there’s a follow-up, we could take a quick follow-up. If not, we’ll head on to the next question.
Moshe Orenbuch — Credit Suisse — Analyst
Great. Thanks very much.
Operator
Our next question comes from the line of Sanjay Sakhrani of KBW.
Sanjay Sakhrani — KBW — Analyst
Thanks. Good morning. And my congratulations to Eric and Craig as well. Wanted to drill down on the credit quality and reserve assumptions. I understand sort of the reversion to the mean assumption given the underlying unemployment rate. However, I think Moody’s has been improving their laws, their unemployment assumptions and that rate is declining sort of in the second half of this year and into 2022.
I’m just curious sort of how to put your assumption that the loss rates will get to the reserve levels — those embedded in the reserve assumptions? And at what point do you reassess that? Is there something sitting inside your portfolio that’s leading you to be more conservative? Thanks.
Roger C. Hochschild — Director, Chief Executive Officer and President
Okay. Thanks, Sanjay. So, I’ll take this one as well. So, first, let me start off by saying we’re very pleased with the portfolio’s performance. So, the resiliency of our customer base has been remarkably strong. So, in terms of the modeled assumptions — so in my script, I talked about 8% unemployment at the end of 2021. I realize that’s higher than where we are today from reported number. And we also assumed GDP growth of 2.7%. Now, there are some folks forecasting an increase in GDP as well.
So, as we thought about the reserves and the positioning of the balance sheet, there’s a couple of things that we took into account. So first is the overall unemployment numbers. There’s about 10.7 million people out of work. There’s another 7.3 million people that aren’t included in the unemployment number due to the fact that they haven’t actively worked — looked for work in the past four weeks.
So, to me, where we are from a life of loan loss reserving standpoint made perfect sense is conservative. We didn’t feel like we had enough data points at this point, given the level of uncertainty, to make a material change to the absolute level of reserves in the fourth quarter, given a life of loan reserving assumption. But I can say this, we’re going to continue to look at our portfolio and the macro environment.
We’re going to look specifically at the trajectory of unemployment and the type of unemployment. So, we’re seeing unemployment levels transition from service workers to white collar workers, who would more likely be representative of our customer base and the impact of stimulus.
So today, we’re well positioned, and we’re going to continue to reassess it, first quarter, second quarter and into the second half of next year and make appropriate adjustments.
Sanjay Sakhrani — KBW — Analyst
Thank you.
Operator
Our next question comes from the line of Rick Shane of JPMorgan.
Rick Shane — JPMorgan — Analyst
Good morning, and congratulations, Eric. And Craig, when I saw the — I’m hoping you’re listening. When I saw the voluntary early retirement, I hope you’re a reasonable chunk of that number.
Sorry. When we think about what you’ve said in terms of marketing and adding new accounts, that makes sense. There will presumably be a lag in loan growth as you add accounts. Historically, when you started to grow the portfolio, again, it’s been led by wallet share gains and line limit increases.
I’m curious when you think you might take that sort of brown out alignment in increases off? And how we should think also about rewards rate as you think about wallet share going forward?
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes. So, I’ll start it on the rewards rate. We tend to keep our rewards program very stable. It provides a lot of value. And so, while you’ve seen competitors make dramatic changes, we feel like our leadership position in cash rewards serves as well.
So, we have in the past talked about a low single-digit increase in rewards rate due to structural changes. That will likely continue. But beyond that, we feel very good about where we’re positioned and the competitiveness of our program.
In terms of growth, it’s always been a mix of new accounts and stimulating the existing portfolio. And so, it will remain that going forward. In terms of what would get us to loosen credit, I think many of the indicators that John talked about, right? So, getting better line of sight in the direction of white collar employment is probably the most critical one.
Rick Shane — JPMorgan — Analyst
Got it. Okay. Thank you, guys.
Operator
Our next question comes from the line of Mark DeVries of Barclays.
Mark DeVries — Barclays — Analyst
Yes. Thank you. I’ve got a follow-up to Sanjay’s question. Clearly, the credit performance you’ve experienced so far is much better than you would have expected, just given all of the different macro assumptions. And so, my question is, what do you think you need to see in the data before you feel comfortable releasing reserves?
John T. Greene — Executive Vice President, Chief Financial Officer
Okay. Thanks Mark. So first will be the performance of the portfolio. So, what’s happening on specific roll rates, are they holding or are they deteriorating, consistent with what our modeled expectations are? Broad macro is going to be important. We also do know that if there’s another round of stimulus, that will — which isn’t baked into our reserve assumptions– we do know that that will have an impact on a couple of factors — certainly, delinquency, repayment rate and ultimately charge-offs and required provisions. So, we’ll keep an eye on that.
Roger and I both talked about white collar employment levels and [Speech Overlap] will also be new jobless claims. So, putting those factors together with a strong overview on how the portfolio itself is performing, will be the key factors in determining what we do with reserves in 2021.
Mark DeVries — Barclays — Analyst
Got it. Thank you.
Operator
Our next question comes from the line of Betsy Graseck of Morgan Stanley.
Betsy Graseck — Morgan Stanley — Analyst
Hi, good morning.
Roger C. Hochschild — Director, Chief Executive Officer and President
Good morning.
Betsy Graseck — Morgan Stanley — Analyst
I had a couple of questions on the buybacks that you announced. I just wanted to understand if the buyback announced — like, what kind of CET1 you’re thinking about when you are putting forth that buyback estimate?
And then the second question that’s related to that has to do with whether or not you’ve embedded reserve releases in your estimates for that? I mean, because of the four-quarter trailing, federal right now is — kind of circular reference there. So, wanted to understand how you’re thinking about reserve releases and what your target CET1 is?
John T. Greene — Executive Vice President, Chief Financial Officer
Okay. Thanks. So, we continue to target 10.5%. In terms of how we thought about buybacks this year, the first piece I would say is, we wanted to ensure we’re prudent with our capital, given the level of uncertainty. We do have the Fed constraints in terms of the four-quarter average of net income. And so, the first quarter and — well, the second quarter and first quarter of 2020 are impacting that calculation for the first quarter of ’21. And then, we’re also thinking about the CECL transition impact, which will be somewhere between 200 and 250 basis points as we think about CET1.
So ultimately, we didn’t want to be out at the far edge of the buyback envelope. We felt that $1.1 billion was an appropriate return of capital given a level of uncertainty. And that will take a dent out of, what I’ll say is, the excess capital that we have right now. But our earnings power will be really important and that will give us an opportunity to reassess that in 2022 as well.
Betsy Graseck — Morgan Stanley — Analyst
Okay. Thanks. And then just separately, a follow-up question here on how you’re thinking about the interchange rate? I mean, there’s been a couple of — there’s been some pressure on it recently after years of improving. I just wanted to understand, is the recent behavior more of a short-term phenomenon, less T&E? Or is there something else going on that we should be thinking about remodeling out that one?
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. Yes. So, the fourth quarter did come down a little bit. And we can largely point to mix as there was a strong pull away from the traditional online — or excuse me, the brick-and-mortar retailers to the online retailers, which certainly impacted it. But from our perspective, very well aligned with our 5% categories in the fourth quarter, which also drove incremental sales through our card. And ultimately, we think it will translate into other forms of revenue, specifically interest income as balances revolve.
Betsy Graseck — Morgan Stanley — Analyst
Okay. So, this quarter was primarily the 5% cash back on the dot-coms that got obviously utilized very, very fully?
John T. Greene — Executive Vice President, Chief Financial Officer
Yes.
Betsy Graseck — Morgan Stanley — Analyst
Okay. Thanks.
Operator
Our next question comes from line of Bill Carcache of Wolfe Research.
Bill Carcache — Wolfe Research — Analyst
Thanks. Good morning. I’d like to echo my congratulations to Eric and Craig. I wanted to ask about the net interest margin. And I believe that 10.63% is the highest we’ve ever seen. Can you talk about whether the trough is behind us? How sustainable this level is and the extent to which you see room — margins to actually expand from here, given the room for deposit re-pricing and remixing towards lower cost deposits that you see from here?
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. Sure. Sure, Bill. So, from the trough, our second half NIM improved by nearly 80 basis points, incredible. So, obviously, we can’t run that trajectory in perpetuity. So, a way to think about that is we ended the fourth quarter at the 10.63% that you just mentioned.
My view is that we still have some opportunity on deposit pricing, especially, if there’s another round of stimulus because that will put a lot of liquidity in the market and the competition for deposits will further abate.
Now, how much, ultimately, room we have there, uncertain. Is it 10, 20 or more bps, to be determined. But my view, at least 10 bps, very conservative view.
The maturity profile, we included that information in the deck to allow folks to be able to model through some improvements that we’ll see there. And then — so frankly, some things that are pushing against net interest margin, especially in the second half of the year. If the credit losses do accelerate as we’ve indicated, that will put some dampening pressure on net interest margin.
So, broadly speaking, we do see further room for expansion there. But I would jump from the fourth quarter number and do those steps I just mentioned to get to a reasonable way of thinking about the balance of the year.
Bill Carcache — Wolfe Research — Analyst
Got it. Thanks, John. Very helpful.
Operator
Our next question comes from the line of Don Fandetti of Wells Fargo.
Don Fandetti — Wells Fargo — Analyst
Hi, good morning. Rogers, if you could comment on what you think about the sort of long-term growth rate of the card business? If it’s been impacted by — you have a lot of new areas like personal loans, you have buy-now pay-later. Is this just on the edges or do you think that this has some impact on the growth of the card industry? And could you be wrong on that, I guess? And could some of these new initiatives and FinTechs be more impactful?
Roger C. Hochschild — Director, Chief Executive Officer and President
I certainly could be wrong. So, I’ll say that upfront. But we have not seen either of those have an impact on card loan growth. And you’ve seen that in the past, whether it’s been a home equity loan boom, where a lot of people are doing cash out refis and using that to pay down debt.
A lot of consumers seem to carry a level of credit card debt that they are comfortable with, certainly our base, and they tend to revert back to that amount. And so buy-now pay-later is the most recent trend that’s out there. We’re looking at that very carefully and have yet to be able to see an impact on our revolving loan balances.
So, in my decades in this business, there’s always something that’s going to kill off credit cards. But so far, the growth trajectory of the industry remains solid.
Now. it is a mature business. But we also — as we think about our growth, it’s a combination of where the industry goes, but also our ability to take share from our competitors and capture a disproportionate of student, young adults who are coming into the industry. And so, we feel good about that.
Don Fandetti — Wells Fargo — Analyst
Okay. Thank you.
Operator
Our next question comes from the line of Kevin Barker of Piper Sandler.
Kevin Barker — Piper Sandler — Analyst
Good morning. Just wanted to follow-up on your — your guidance for losses to increase in the second half of 2021 and then likely remain elevated into 2022? I mean, are you envisioning, just given the current economic environment, that these are likely that we’re going to have a plateau in 2022? Or is it going to be like a slow decline after we peak in the second half of 2021, just given what your view is on the economy and how things are playing out so far?
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. Okay. So, at this point, there’s a bit of both art and science in terms of modeling kind of peak charge-offs. And what we’ve seen as we’ve gone through this pandemic is the peak has continued to push into future periods. And essentially, that’s what we’re seeing today.
It’s hard to believe, given the level of absolute unemployment and those folks that are outside of the employment ranks that aren’t in the unemployment number, that there isn’t going to be some material impact to credit and charge-offs at some point.
The roll rates we’re seeing right now in terms of — from aging buckets, one to the next, are incredibly strong, which is positive. So that means, by itself, there can’t be an acceleration of charge-offs in — at least in the first four months of the year. Beyond that, you would expect, given the unemployment numbers that the roll rates will deteriorate, charge-offs will increase and continue to increase until there’s absolute stability.
So, we’re seeing a peak in late 2021 and that could carry through into 2022 and then moderate as the towns go back and the economic — the broad macros improve.
So, that’s how we’re thinking about it. I’m not sure if we’ve got it 100% right. If we don’t, we’re going to adjust accordingly.
Kevin Barker — Piper Sandler — Analyst
Okay. And then the follow-up on your comments about targeting to keep expenses flat across the business, is that relative to account growth? Or is that saying, we’re — year-over-year absolute expenses will remain flat in 2021?
John T. Greene — Executive Vice President, Chief Financial Officer
Yeah. So, I want to be careful here. It wasn’t absolute expenses. So, what I tried to do is distinguish between those expenses that will help us accelerate growth. And we expect those to increase. Those that aren’t targeted to accelerate growth, we expect will remain flat to down. We expect — if you go through kind of line items of the expense base, salaries and wages, we’ve done some things this year to level that off, including the voluntary early retirement program.
We have activated our procurement organization around third-party spend, and we’ve driven a lot of productivity through that. We have looked at every single line item on the expense base and we’re making determinations on whether or not those expenses will help us drive long-term growth. If the answer is yes, maybe will increase. If the answer is no, they’re going to be flat to down.
Kevin Barker — Piper Sandler — Analyst
So just to be clear, those expenses that are driving growth, should they be in line with account growth or be above or below, just dependent upon what you’re seeing underlying the business?
John T. Greene — Executive Vice President, Chief Financial Officer
So maybe one way of looking at that, for card, new accounts as an example, we expect our cost per account to be below what we saw in 2019. And that’s with a tighter credit box and reflects the benefits we’re seeing from some of our investments in advanced analytics as well as just the differentiation and appeal of our product.
Kevin Barker — Piper Sandler — Analyst
Thank you very much.
Operator
Our next question comes from the line of Bob Napoli of William Blair.
Bob Napoli — William Blair — Analyst
Thank you. Good morning. I think I’ve said goodbye to Craig, like at least eight times over the last few decades. I don’t think it’s going to be the last one for some reason, but good luck.
Roger, just — and so the world has come Discover’s direction, if you would, I think the digital banking, branches banking structure that you have and then you have the unique asset of the network, obviously. But what are you working on? There’s a lot of changes. While the markets come your direction, there’s a lot of new businesses, direct banks, digital banks and development of companies like Venmo or private companies, Chime. Are there things that you’re doing as you look at this to be on offense to expand the ecosystem of your products and services to try to get direct deposits to get more of, let’s say, transaction banking accounts as well? What are you doing on the banking side? With all of the innovation in the market, where is Discover investing?
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes. So, great question. If the world is coming your way, you’ve got to keep moving to stay ahead. And so, that is our focus. We’ve always stood for innovation back to our founding and inventing credit card rewards, but more recently with everything from the FICO score on statements, ability to freeze your account. And so, you can rest assure that that focus is still there on a pipeline of customer-driven innovation across all of our products.
Specifically, in the deposit side, we think there’s a lot of opportunity to get into transaction accounts. It will be a while before they become a material part of our funding base. But with our low-cost direct-to-consumer digital model as well as the advantage we have from being exempt from the Durbin interchange caps because we own our proprietary network, we’re uniquely positioned for a bank over $10 billion.
And so, it isn’t maybe the primary focus right now, just given the excess level of fundings, but it is a critical initiative. And we feel good about our ability to compete both against traditional branch-based banks, but also against any of the new FinTech players.
Bob Napoli — William Blair — Analyst
Okay. Thank you. I’d love some color on what else you look — you’re thinking about there. But as you look at the — your customers — the customers that you’re adding, is there any change in the demographic mix of the new customers you’re adding? There’s a lot of times we are getting commentary or questions around, well, the millennials are not going to borrow on their credit cards the way others did. And there are other new forms of credit. Is Discover getting the same share of those younger customers? And are you keeping them? Do you feel there’s anything to the thought that the millennials will be less likely to use credit cards? And, I mean, if so, are you looking at other products like buy-now pay-later?
Roger C. Hochschild — Director, Chief Executive Officer and President
So, in terms of millennials, based on the data we see, we’re either the leading or one of the leading underwriters for college students. And the brand is incredibly strong. We have college students and young adults that appreciate sort of the leading digital functionality, as well as some of the innovations I talked about. And we’re seeing very similar usage patterns as we saw in prior generations of customers.
So, we’re very excited about the growth there. And I think being in the student loan business and the second largest originator, helped to get our brand out there in front of the next-generation of consumers.
Bob Napoli — William Blair — Analyst
Okay. Thank you. Appreciate it.
Operator
Our next question comes from the line of John Hecht of Jefferies.
John Hecht — Jefferies — Analyst
Morning. Congratulations to Eric and Craig, as everybody else has said. And thank you guys for taking the questions.
Maybe follow-on to Bob’s question, but in a different way. I mean, you guys have tightened — over the past several quarters, you’ve had substantial net paydowns. I’m wondering, has your kind of back book composition changed in a good or bad way, or a positive or, I guess, negative way based on those patterns?
Roger C. Hochschild — Director, Chief Executive Officer and President
I would say nothing dramatic. And part of the advantage we had, we have been tightening for a number of years coming into this. We felt like we were late cycle and talked about that with you guys on the call. Clearly, we didn’t expect it to end the way it did in early 2020. But that helped us from having to take some of the magnitude of changes, but I think some of our competitors did.
So, we try and be consistent in how we run the business. And so, we’ve targeted the same prime consumer. And I think the drove [phonetic] have not seen any dramatic shifts in terms of our composition, either with the new accounts we’re booking or our existing portfolio.
John Hecht — Jefferies — Analyst
Okay. Great. Thanks. And the second question is private student lending. I think you guys referred to some market share gains in the recent periods. I know there’s been some shifts in terms of other big banks that are exiting that segment. And then there’s a new administration and then maybe some changing policy or some thoughts about potential changing policy. Maybe just some commentary given your momentum there and your outlook there given those factors?
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes. So, I would say there’s always a lot of discussion about what might happen in Washington about student loans. I would say, keep in mind, that over 90% of student loans are the federal student loan program. And so, that’s where a lot of the attention is focused, very different animal in terms of the, quite frankly, lack of underwriting of that product and the losses they experienced compared to how we go to market.
So, we feel really good about the business, clearly benefited from one of the larger players stepping back. But we believe we would have gained share even if they hadn’t. And so, it reflects the fact that the brand is well-positioned. It resonates with consumers and we take the same approach in terms of customer experience and differentiation with the student loan product as we do on the card side.
John Hecht — Jefferies — Analyst
Great. Thank you guys very much.
Operator
Our next question comes from the line of Mihir Bhatia of Bank of America.
Mihir Bhatia — Bank of America — Analyst
Hi. Maybe just staying with some of your non-card products. I was wondering if you could talk a little bit about just the outlook in competitive intensity you are seeing for some of the — whether it’s student loans, personal loans, even just on the network side of your business? I know there has been a focus to grow some of that too. So maybe just talk a little bit about what you’re expecting from those businesses as we head into 2021? Thank you.
Roger C. Hochschild — Director, Chief Executive Officer and President
Sure. So, I’ll start on the payment side. Always, very intense competition. In the payment side, we compete largely against two very large players. So, especially in debit, it’s really head-to-head competition for merchant routing day in, day out. We don’t expect that to change. But I feel good about the products we have, and we have a great team on it.
In terms of other products, we talked about student loans. For personal loans, we have modestly widened credit on that. That was the product we tightened the most, just given the volatile in the downturn. We’ve loosened up, I would say, marginally, and feel good about what we’re originating, positioned a little differently than most. We’ve always had a relatively narrow credit box for that, and those loans are, sort of, bigger ticket debt consolidation primarily. But I would say across all of our products, given the returns we get, these are all highly competitive, very challenging markets, and that sort of occurs day in, day out.
Mihir Bhatia — Bank of America — Analyst
Thank you. And then just if I could quickly follow-up on some of your NIM comments. I know you mentioned the funding side of the balance sheet, optimizing that. Is there also an opportunity a little bit to optimize on the asset side of the balance sheet? Maybe you were running with a little bit of excess cash in 2020 given the downturn, or is that fairly well-optimized already? Thank you.
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes, thanks. I’ll jump in on that one. Yes, we do have some excess liquidity right now, and there is an opportunity to continue to move that forward. Now we’re going to, I’ll say, gauge that based on the level of asset growth because asset growth will consume that liquidity. And we’ve built a plan that assumes a level of growth. So. that’s one point.
The other piece is around deposit pricing and how we price the deposits coming in, turning into cash. And then in terms of balance sheet positioning, we are mildly asset-sensitive right now. So, in a rising rate environment, that will also be beneficial to net interest margin.
So quite honestly, the liquidity, I think, will take care of itself over time. And the positioning of the balance sheet in terms of asset sensitivity, very, very positive to be accretive to net interest margin in a rising rate environment.
Mihir Bhatia — Bank of America — Analyst
Understood. Thank you.
Operator
Our next question comes from the line of Meng Jiao of Deutsche Bank.
Meng Jiao — Deutsche Bank — Analyst
Good morning. Thanks for taking my question. I just wanted to get a sense on how you guys are thinking about deposit growth specifically, both direct-to-consumer and the broker deposits? I guess for DTC, it’s now, I guess, 62% of total funding. I believe, previously you mentioned a longer-term target of 70% of the funding stack. Is that hard to still hold? Or do you expect DTC deposits to be even higher as a percentage of the funding stack going forward?
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. So, we’re targeting 70%, 80%. So, 62%, you’re correct on the number. So direct-to-consumer, our proposition has been very, very positive. We don’t compete on the basis of price, which has been a good thing in terms of helping us to modulate some of the liquidity that we have. But also, the fact that deposits continue to grow shows that there is a level of loyalty and trust with the Discover brand.
In terms of the broker CDs, we actually use that almost as a valve of sorts, right? So as our funding needs increase, we’ll go more heavily into brokered CDs. As they decrease, we shrink it. So that’s the way we’ve managed it traditionally. It’s going to continue to be a liquidity channel for us, but a, what I’d say, less important channel over time in terms of total quantum of deposits.
Meng Jiao — Deutsche Bank — Analyst
Great. And then a second question. Just broad based, is there, I guess, anything structurally different in regards to releasing reserves under CECL than the prior method of looking at allowance reserves?
John T. Greene — Executive Vice President, Chief Financial Officer
Not so much structurally. I mean, we’ve got a thorough process that considers all the elements of CECL under GAAP. The one thing I would say is the life of loan reserving does require more modeling and, frankly, a greater level of judgment given how far out into the horizon you’re projecting losses.
So, there’s certainly a very strong governance element. There’s a science to it, and then there is a level of professional judgment or art to it as well. So — yes, same could be said for incurred, but the horizon is much more difficult given timing of what we’re trying to project.
Meng Jiao — Deutsche Bank — Analyst
Great. Thanks for taking my questions.
Operator
And ladies and gentlemen, we have time for one more question. Our final question will come from the line of Dominick Gabriele of Oppenheimer.
Dominick Gabriele — Oppenheimer — Analyst
Thanks so much for taking my questions. Can we just think about a potential windfall of excess capital from reserve releases? Excuse me. If that was to happen and the economic situation persisted the way it is today and you felt comfortable releasing the reserves, can you talk about the breakdown of how you would use that excess capital? I mean, when tax reform came, then there was a flood of capital. Companies started talking about we’ll do half, about a third for growth, a third for capital return, a third for investment in technology, something like that. Can you talk about how you think about those pieces, should a flood of capital come your way? Thanks.
Roger C. Hochschild — Director, Chief Executive Officer and President
Yes. I’ll let John talk about capital return, but I would say, our business does not let itself to rapid deployment of capital, right? We market on a consistent basis, sort of flooding the market in a given quarter based on the amount of capital we have, I don’t think makes sense from a long-term standpoint. The same holds true for technology, right? A lot of it is about spending smart, not just putting huge amounts of money.
So, I have a real hard — if we do have a quarter with a big reserve release, there may be some things. If the margin in terms of investments in the business, that by and large it will fall to the bottom line. And I’ll let John pick it up there.
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. And so, Dominick, I do appreciate your optimism regarding a flood of capital as a result of reserve releases and a powerful economy. Our priorities actually remain the same, so in terms of how we think about capital and allocating the dollars first to growth then to dividend and share repurchases, and then the last priority would be small M&A, I think, bolt-on capabilities or certain niche products that we think will drive long-term shareholder value. So, no change there.
We go through an annual capital planning process here as most financial services institutions do. And we share the outlook with our Board and our priorities. And obviously, there’s some regulatory constraints that we manage too as well. And then we’ll make good long-term decisions to generate profitable growth and shareholder return.
Dominick Gabriele — Oppenheimer — Analyst
Great. Thank you. That makes a lot of sense. And then if we just think about the — if you look at the NIM in particular and the — this quarter, and the interest charge-off reduction in the quarter, that had a big impact on the yield. And so, could you just talk about the balance between, let’s say, that interest and fee charge-off even just reverting to normalized levels, not including the spike of losses, let’s say? But has that sort of just normalized over 2021 versus some of the benefits you have on the interest expense savings that you’re doing, given all the — what you’re doing there? Maybe if you balance those two against each other, could you still see kind of NIM expansion or levels in 2021 versus stable to improving versus 2020? Thanks.
John T. Greene — Executive Vice President, Chief Financial Officer
Yes. We — Dominick, we do see opportunities for NIM expansion even in the face of increased interest charge-offs as the portfolio matures and contends with some of the economic stress. But the numbers in terms of quantum, I’m probably not going to get into that level of detail on the call here. But I will go back to what I said earlier in one of my questions in terms of how to think about it. So, we do see an opportunity for NIM expansion. And that — some of that will be tapered by credit and the impact of delinquencies. But even contemplating that, there will be a level of expansion.
Dominick Gabriele — Oppenheimer — Analyst
Excellent. Thanks so much.
John T. Greene — Executive Vice President, Chief Financial Officer
Sure. Very good.
Eric Wasserstrom — Head of Investor Relations
All right. Well, thank you all very much for joining us. Anyone who has additional questions, please give us a ring. Emily and I will be here to answer questions. And have a great day.
Roger C. Hochschild — Director, Chief Executive Officer and President
Thanks, everybody.
Operator
[Operator Closing Remarks]