John Greene: Yes. Yes. So I’m going to run through the primary drivers. So first would be the Fed rate changes in the second half of ’22 as well as what we’ve anticipated either two or three increases in 2023. Second impact is the yield on our investments, which is improving with the increase in the rate environment. And then, the third piece has been some pricing actions we took in the consumer banking products. So, think about the non-card products. So offsetting that would be kind of the cost of funding. So DTC and external funding costs have increased. And then we’re also anticipating an impact from credit, all of which the net of those gives us a high level of confidence that certainly, we’re going to see peak NIM in the first quarter and then stepping down from there through 2023.
Operator: Thank you. Our next question comes from Mark DeVries with Barclays.
Mark DeVries: I have one more credit question for you. I know you don’t generally give out guidance more than one year out, but I think some of the commentary around the charge-off guidance has some implications for 2024. I just wanted to try and clarify I mean if you look at the guidance range, it seems to imply you kind of exit 2023 at a charge-off rate at 4% to 4.5%. I think, John, you indicated 2024 is kind of a peak year? Should we expect — is it reasonable to assume that that’s implying kind of a charge-off rate north of 4% for 2024?
John Greene: Yes. So you were right on your call here. We gave a range for 2023 of 3.5% to 3.9%. I talked about the curve and what we think will happen to the curve and the slope of that. So, Mark, as a matter of prudence, I think that’s probably as far as I’m going to go here.
Mark DeVries: Okay. Fair enough. Thanks for that color.
Operator: Thank you. Our next question comes from Betsy Graseck with Morgan Stanley.
Betsy Graseck: Maybe a slightly different way to address this question is you perhaps could give us some color on where you have seen your fully seasoned vintages peak in terms of net charge-off rates and around what kind of month within the seasoning path that happens in a range of months that would be helpful to understand.
John Greene: Yes. So typically, we’ll see it around 18 months. And it varies based on credit quality. So, the highest credit quality. So, I think FICO would typically peak a little later. And then the weaker, I’ll say, the weaker credits typically peak a little bit earlier. But on average, I think, about 18 months or so.
Betsy Graseck: And the level that you’ve been seeing, it would be helpful to understand how historically, the vintages that you want to write to are trajecting in terms of peak. And maybe if you could comp 2022 vintage in 2021, what you’re seeing there would be helpful?
John Greene: Yes. So — the first part of that answer would be it would depend on the vintage. So if you go back to our 2020 vintage and remember, there’s COVID, right, we were locked down. We ceased underwriting kind of the near prime and lower prime and concentrated on upper prime that vintage will season at a peak loss level below what Discover historically has done. If you look at ’21, ’22, we were essentially back to an underwriting standard consistent with history. And you can use that information to get some level of comfort around what can be expected in ’24 on this vintage.
Betsy Graseck: Okay. Because you’re basically saying ’22 is a normal — is exhibiting behavior that is more normal pre-COVID type of vintages?
John Greene: It is. Yes. Yes. The one difference that I think is important for folks to codify in their minds is that we’re coming off an abnormally low base, right? So the entire portfolio is normalizing. We’ve talked about that consistently actually, since the beginning of last year that we thought the portfolio was normalizing. And what you’re seeing here in the 2023 guidance is essentially the portfolio normalizing.
Betsy Graseck: The current reserve ratio level is consistent with this normalization whereby peak losses hit in ’24?
John Greene: Yes. Otherwise, my controller would have taken an exception to our reserve process.
Betsy Graseck: Right. And this is — I know we’re talking about card, but is this the same kind of expectation across the other asset classes as well, student and personnel?
John Greene: Yes. Although what we’re seeing in personal loans is, again, loss rates below historical norms. Payment rate is beginning to normalize. And we had talked about the fact that we perhaps overcorrected on that product in terms of underwriting, in terms of the pullback. We pulled back significantly. So I expect some seasoning and normalization there. But again, we’re very, very confident about the loss performance of that product. We understand where it is on the payment priorities for folks. So we’re going to be mindful of the economy on that. And student loans, yes, that’s normalizing. We did have and likely we’ll have a little bit of impact when we see the full impact of the student debtors on the government programs having to pay back loans, but it’s underwritten to a high standard. 80-plus percent have cosigners. So, we feel very comfortable about that product as well.
Operator: Thank you. Our next question comes from Rick Shane of JPMorgan.
Rick Shane: Look, this is an interesting milestone where the reserve rate is 658 basis points. It basically is apples-to-apples seasonality versus CECL day one and up 50 basis points. I’m curious when we think about your economic outlook and how you build a CECL reserve where you compare to CECL day one on a like-for-like basis, would you build the same allowance? Or have you made adjustments and then compare your economic outlooks in each of those points in time, please?
John Greene: Yes. So good question, Rick. So we’ve referenced CECL day one in the past, but I’d like to remind folks, day one was first time we rolled this new standard out. We were using new models. They’ve been tested extensively. And the macros were late cycle with higher unemployment levels. So as we look at kind of where we are today or as of the fourth quarter, right, 6.58 in terms of total loss reserve rate. That seems appropriate based on kind of what we’re seeing in the macros and how the portfolio is performing. So, do we specifically reference day one only from the standpoint of where it was back on January 1, 2020, to where it is today, but we don’t use that as a decision point whatsoever.
Rick Shane: John understood. I’m more curious that if you like were you’ve described that the models have evolved. And I think that, that’s fair, and I think everybody appreciates that. What I’m asking is, on a like-for-like basis, would — do you think that reserve rates are lower today using the same assumptions as you refine them versus CECL day one?
John Greene: Yes. No, no, they’re not lower. We’re looking at the portfolio performance. It’s performed extraordinarily well. We’re seeing a bit of seasoning now, as you would expect in this type of product. And the macros are contemplating a minimum level of increase in unemployment of at least 0.5% and more likely 1%. So what you’re seeing here is a CECL reserve for the quarter that reflects those macros.
Operator: Thank you. Our next question comes from Bob Napoli with William Blair.
