Citigroup Inc. (C) Earnings Call Transcript & Summary

Hello, and welcome to Citi's Fixed Income Investor Review with Chief Financial Officer, Mark Mason; and Treasurer, Mike Verdeschi. Today's call will be hosted by Tom Rogers, Head of Fixed Income Investor Relations. [Operator Instructions] Also as a reminder, this conference is being recorded today. If you have any objections, please disconnect at this time. Mr. Rogers, you may begin.

Thank you, Natalia. Good morning, and thank you all for joining us. As Natalia mentioned, I'm joined this morning by our Chief Financial Officer, Mark Mason; and our Treasurer, Mike Verdeschi. In a moment, Mike will take you through the fixed income investor presentation, which is available for download on our website, citigroup.com. Afterwards, Mark and Mike will be happy to take your questions. Before we get started, I'd like to remind you that today's presentation may contain forward-looking statements, which are based on management's current expectations and are subject to uncertainty and changes in circumstances. Actual results and capital and other financial condition may differ materially from these statements due to a variety of factors, including the cautionary statements referenced in our discussion today and those included in our SEC filings, including, without limitation, the Risk Factors section of our 2019 Form 10-K. With that said, let me turn it over to Mike.

Thank you, Tom, and good morning, everyone. On today's call, I will cover a number of topics. First, I'll briefly discuss our year-to-date 2020 operating results. Second, I will cover recent balance sheet trends. Third, I'll review our issuance program. And finally, I'll discuss our continued strong liquidity and capital position. Slide 3 summarizes our results for the third quarter and year-to-date 2020. Year-to-date, we reported net income of just over $7 billion, which included a $10.8 billion increase in credit reserves under the CECL framework. On Slide 4, we show average balance sheet trends over the past 5 quarters in constant dollars. On this basis, we have grown our balance sheet by approximately 13% over the last year as we continue to support our consumer and institutional clients as they manage through the COVID-19 health crisis while also maintaining a strong balance sheet. We continue to see strong growth in deposits across both GCB and ICG. Loans declined primarily driven by a modest decline in GCB as a result of lower consumer spending. Trading-related assets and liabilities grew, reflecting a continuation of heightened client activity relative to the prior year. And long-term debt increased as we built liquidity to support our clients and maintain the firm's robust liquidity profile. Slide 5 presents trends in our loan portfolio on an average basis in constant dollars. Total Citigroup loans decreased 1% year-over-year. In our consumer franchise, average loans declined 4% year-over-year, reflecting the impact of lower consumer spending in our cards business, higher payments by customers given fiscal stimulus and a continuation of the pause in proactive marketing. In our institutional franchise, average loans grew 2% year-over-year. Average loans in Corporate Lending grew 4% year-over-year but were down 14% sequentially, reflecting significant repayments as we assisted our clients in accessing the capital markets. Private Bank loans increased 6% largely driven by residential real estate lending to our high net worth clients. TTS loans decreased 7%, reflecting softness in underlying trade flows and the continued low level of spend in commercial cards. And loans included in Corp/Other continued to decline driven by the wind-down of legacy assets. On Slide 6, we show credit quality trends in our GCB and ICG loan portfolios. In ICG, credit quality remained broadly stable. In the third quarter, nonaccrual loans declined by roughly $400 million sequentially to $3.6 billion or 93 basis points of total loans, reflecting write-offs and repayments across the portfolio. In GCB, credit trends remain broadly stable to improving this quarter given high levels of liquidity in the U.S., lower spending and the benefits of relief programs. However, we do expect losses to begin to rise next year as government stimulus and other programs roll off and unemployment remains elevated. At an aggregate level, we are likely to see losses peak towards the end of 2021, with losses in Asia and Mexico peaking a little earlier than the U.S. That said, we believe we are well prepared for expected credit losses across both our consumer and institutional portfolios. As of the third quarter, we held credit reserves of $28.9 billion. Of this amount, $3.1 billion represents our management adjustment for economic uncertainty, which factors in a downside scenario that is more adverse than our base case. And as we noted on our earnings call earlier this month, given the lifetime nature of the CECL methodology and the conservative nature of our management adjustment, it is now more likely than not that we'll see reserve releases and our ACL come down in 2021 to offset future losses and as we continue to progress through the crisis, assuming our base case holds. Turning to Slide 7. We show trends in average deposits over the past 5 quarters in constant dollars. Total average deposits increased 19% from the prior year period, reflecting continued client engagement as well as the elevated level of liquidity in the system. Now let me cover our parent benchmark debt issuance program on Slide 8. So far this year, we have issued approximately $20 billion of parent-level benchmark debt across a variety of tenors, including the inaugural $2.5 billion affordable housing bond we priced last week. Going forward, we'll continue to maintain the flexibility to issue a mix of tenors, currencies and structures as we prudently manage the liquidity profile of the firm and support our clients. On Slide 9, let me cover our issuance, maturity and redemption expectations. As I just mentioned, we have issued approximately $20 billion year-to-date, and we expect our total issuance to be around this level. However, we will remain opportunistic given the uncertain environment. And we will continue to be flexible around optimizing our funding through opportunistic redemptions. On Slide 10, we show the composition of our long-term debt outstanding. During the third quarter, our total long-term debt declined by approximately $7 billion to $273 billion as we allowed our bank debt to mature given strong deposit growth. On Slide 11, we provide an update of our LCR metrics and drivers. Our average LCR increased modestly to 118% this quarter. Turning to Slide 12. Let me summarize our key regulatory capital metrics. Our CET1 capital ratio increased to 11.8% primarily driven by net income. And our SLRs were 6.8% and 6.6% for Citigroup and Citibank, respectively. Moving to our last slide. Let me summarize several key points. Year-to-date, we have performed well and demonstrated the significant earnings power of the franchise. We earned roughly $7 billion of net income year-to-date despite a $10.8 billion increase in credit reserves. We maintain a strong capital position with a CET1 capital ratio of 11.8%, close to 200 basis points above our regulatory minimum requirement, an SLR of 6.8% and a surplus above our TLAC requirement. And we also maintain a strong liquidity position with an average LCR of 118%, over $950 billion of available liquidity resources and we expect to be in compliance with the NSFR when the rule is effective. Our capital and liquidity profile remains strong and well positioned to withstand potential pressure if economic conditions deteriorate. And we continue to have the flexibility to deploy these resources in ways that deepen and expand our client relationships and support the broader recovery. Before we move on to Q&A, let me touch briefly on LIBOR. We are continuing to prepare for the transition away from LIBOR by the end of 2021 by working with our regulators and various industry working groups as well as through the work streams we've established within Citi. Since the second quarter, we have transitioned the discounting of centrally cleared euro and U.S. dollar interest rate derivatives to the euro short-term rate and the secured overnight financing rate, respectively; announced the adoption of the newly published ISDA protocol for existing IBOR derivatives transactions; and increased our virtual client communication efforts, including outreach regarding these new industry-led protocols and solutions. With regard to the subset of our preferreds which we have mentioned on previous calls, while we are continuing to evaluate alternatives, we don't have any further updates at this time. And with that, Mark and I will be happy to answer your questions.

[Operator Instructions] Your first question is from the line of Scott Cavanagh with APG.

I just wanted to applaud your efforts on the affordable housing bond. We view that as a great step in the right direction there.

Thank you.

So moving away from the notable consent order issue, just -- sorry, I don't know what...

Getting a lot of feedback here. There we go.

Yes. I can hear.

Okay. That's better.

So when we talk about the last topic you brought up, the LIBOR transition, could you talk about your efforts to transition both on the asset and liability perspective? And then when you think about the outreach to your holders, whether it's the preferreds or unsecured, how are you thinking about that, addressing potential language, whether it's consent, fees, exchanges? How are you thinking about addressing that?

So Scott, it's Mike. A couple of thoughts on the preferreds. So as we've talked about in the past, there's a number of things that we're evaluating. We talked about a couple of options, including an amendment or even an exchange. But at the same time, we're working with the industry on broad-based solutions. So we think it's important to have that evaluation being done in parallel. So for the benefit of everyone, not jumping ahead with some transaction or some amendment makes sense to us given the broader industry efforts. We're well aware of the transition time line. And our view on these securities are unchanged, that we have a desire to address them. More broadly, in the transition, when we look at the assets and liabilities, obviously, there's a number of work streams and efforts underway. That includes looking at derivatives. We've talked about the ISDA protocols. So that is a significant effort and something that is an important step forward in dealing with the transition in the derivatives book. And then, of course, when we look at the broader portfolios, we're thinking about that transition away from LIBOR. We've talked a lot about the replacement rate of SOFR and how that operates as more of a risk-free rate. And so we're having to be thoughtful about the pricing of those loans, both internally and how we think about that for our clients. So a lot of the work is around, again, that client outreach, things around the operational capabilities, things around the pricing of that activity. So a lot of that work continues around those efforts, broad-based transition planning. But again, there will be some important steps along the way that we're taking with those industry-wide efforts.

Okay. And then a follow-up question, talking about regulation. So we had the final rule for the TLAC holdings. How should we be thinking about this regarding how meaningful it is for, not only a Tier 2 deduction, but rather from when you think about your trading desk and the risk allocation and the disincentive to hold GSIB bonds and the potential for more volatility of these bonds because of this rule, could you kind of give us a scope of what is historical levels in regards to this new minimums? And is this something that will be meaningful?

Sure. So generally speaking, as we look at this final rule, it's substantially consistent with the proposal. And we believe the rule does allow for sufficient flexibility in market-making activities. And that's both in BAU and even during heightened stress periods where you may be carrying a higher level of inventory. So if you look at the past year, that exception for market-making activities was still sufficient. So when we look at our Tier 2, obviously, we're carrying substantially above where our target is. We've talked about a target of roughly 200 basis points. And at this point, our Tier 2 is closer to 240 basis points. In terms of the monitoring, of course, we're always going to be closely evaluating the balance sheet, the use of resources of the balance sheet into different businesses and in different activities. And really, this would be no different. Of course, we're going to have a lens on this activity and understanding the use of resources. So our read is that we think that exception for market-making is actually quite sufficient.

[Operator Instructions] Your next question is from the line of Brian Monteleone with Barclays.

Regarding the consent order, can you talk a little more granularly about the gap analysis, the kind of scope of activities that are being analyzed? What areas do you think are most pressing? And kind of can you give us some color on, I guess, how you think this is going to make Citi a safer bank?

Sure. So I mentioned a couple of things. So one, we obviously got 3 consent orders. So one was the civil money penalty, the $400 million. And the other 2 were largely aligned and focused on our risk infrastructure and controls. And there were a number of areas that were highlighted in both. One was enterprise risk, data, controls, compliance, et cetera. And as you pointed out, the consent order referenced the need to develop and state plans, identify where we are today and close the gap, identify the gap and close the gap. We've been clear on a couple of things. One, that we've got a sense of urgency to get after this. Two, I think I've been clear that I consider this an investment, an investment in risk infrastructure and controls. And three, as an investment, I would expect improvement in our processes, enhanced technology and digital capabilities, automated controls, straight-through processing, et cetera, and as a result, greater efficiency in the way we do business and in our ability to leverage our data and deliver products and services that are more innovative and in a faster way to our clients. And so I highlight that to say, this is an investment that we are making in our platforms and in our operations. To be a little bit more specific, there are operations that we still have today that have manual processes and steps that are part of them, whether it be loan origination and processing, where we have people who are responsible for checks and balances and controls throughout that process as we're trying to move towards more updated and more automated platforms. As we make these investments in things like those loan operations, we will be able to reduce and eliminate manual steps and replace that with an improved technology and process and/or improved data coming into our systems so that you need fewer people to actually reconcile and provide the checks and balances along the way. That's going to make the process more efficient. That's going to ensure that we have more accurate information without having those manual steps to get there. And that's going to allow for us to get a better product faster on the other side of this. And so investments required, obviously. We have a sense of urgency to get after those investments, not only because of the consent order but because of what they mean in our ability to do business and serve our clients and win.

And then maybe a follow-up question for Mike. On Slide 21, you have the surplus TLAC and long-term debt, which I believe is including the temporary benefit around the leverage exposure. Do you have what that surplus would be without that temporary benefit? And as you think about kind of managing your outstanding TLAC over the next 6 months, are you managing this buffer as reported and assuming that the temporary relief gets extended or are you managing it more towards an ex relief basis?

Sure. Certainly, managing the funding profile and the TLAC assuming that, that exception goes away at some point. And it is set to expire at the end of the first quarter. So yes, as you look at our TLAC and the long-term debt as a percentage of leverage, that gives us a $34 billion TLAC cushion today. If you take out or as you would let that exception expire, it basically takes away roughly $18 billion. So it brings down your TLAC to, call it, $16 billion, all else equal at that time. In the past, we've talked about a targeted range of $7 billion to $8 billion. So at $16 billion, it's still double the targeted range that we're comfortable running at. And that actually doesn't include the issuance that we had just done last week so. But we are definitely managing the profile with a long-term view as if that expires.

Your next question is from the line of Robert Smalley with UBS.

A couple of questions, first one on commercial real estate. Could you talk about your exposures there as we see concerns about the impact of the pandemic on the sector with respect to construction, mortgages and also any other kind of secondary exposure you might have to businesses or families who derive most of their income from real estate? That's my first one. Second one, just a data point, could you give us your VAR number for the quarter? It's usually a Q number, but I haven't seen it reported yet. And then third on preferreds, I think you left a preferred outstanding to float, the 6 1/8%. It's got LIBOR plus 447 reset. Could you talk about why you elected to do that? We're seeing some other of your peers issue fixed for life securities in the $25 market at 4 3/8%. Is 400 something that you're okay with paying? What's the kind of Mendoza Line there on the back end?

Great. Well, there's a lot there. Why don't I take the first 2. And then, Mike, I'll let you chime in on the third one. So first, look, let me start by just kind of reiterating something that I've stated before, which is, remember, our strategy is to focus on large multinational clients. They tend to be high investment grade in terms of that focus. We've been very disciplined about that. We've got a disciplined risk framework, and we manage very tightly to our risk appetite. With that said, we've got a long history in many of these sectors. Many of the sectors that we talk about as high risk industries today, we have a long history in them. We have a long history with the clients that we serve in those specific areas. Now with commercial real estate specifically, as of the end of the quarter, our total credit exposure -- corporate credit exposure to CRE was about $58 billion. About $42 billion of that was funded, and that would represent about 6% of our outstanding loans. Additionally, we had approximately 80% of that exposure as of the quarter end was to borrowers in the U.S., and 73% of that was investment grade. When we look deeper into the CRE portfolio, about 57% of the exposure relates to exposures that are secured by mortgages or underlying properties or in well-rated securitization exposures. And a little bit more than half of that is in Community Reinvestment Act-related lending. About 25% relates to unsecured loans to large REITs and about 80% of the exposure there is rated investment grade. And a little less than 20% relates to our CRE exposure in the Private Bank, getting to a part of your question, of which 100% of that is secured by mortgages or underlying properties. And about 78% of that was rated investment grade again at the end of this quarter. Final point, I guess, I'd make is that, as you know and as we've mentioned before, we feel as though we are well reserved as a firm. And as I look at the allowance for credit losses for the CRE portfolio, they're about 2% of funded loans and about 4% on the funded non-investment-grade loans. So again, if you think about our total ICG business, it has a funded-to-loan ratio of 1.82%, you see that in our presentation, a higher percentage tied to this CRE portfolio. On your question related to VAR, our average VAR decreased in the quarter by about $48 million and is down to about $100 million, $104 million or so at the end of the quarter. And that decrease is largely due to the market volatility coming down and as well as the normalization of credit spreads. Mike, you want to take the last question there?

Sure. So in terms of the preferreds, as we've talked about before, as these securities become callable, we'll evaluate, one, the need for that capital to remain outstanding; and then the second, how do we think about the economics if we do need that capital, of calling that structure and reissuing versus leaving it outstanding. And keep in mind, these structures, this one in particular is callable quarterly. So we will continue to evaluate the structure and all the structures that are callable around whether it makes sense to leave outstanding or call and reissue. And I think that series that you picked out in particular is pretty close for consideration. But we'll continue to evaluate all the structures as we normally do.

Are there any considerations that from the outside we may not be seeing that go into your calculation as to calling or not calling or anything that you're weighting more that we may not be considering?

No. I don't think so. I mean, of course, when we do these, we'll look at the price. We'll include fees in our thinking and other factors. We've talked about the transition away from LIBOR in the language. So that's another factor we think about. But I don't think you're all missing anything.

Your next question is from the line of Arnold Kakuda with Bloomberg Intelligence.

You and your peers largely kept the allowance for loan losses flat in 3Q after a big 1/2 increase. And I hear you on the, if the base case holds, the 2021 reserves might go down. But over the past few weeks, we've seen U.S. COVID cases rise. And so my question is, it might be too early to talk about a change in your potential economic assumptions versus September 30, but are the future reserve builds or releases, are they more dependent on the outlook for the longer-term outlook or maybe even 2021, '22 than kind of the near term?

Yes. Look, again, we've got about $28.9 billion in reserves. And if you look through what we did in the third quarter, we had a slight build, but the significant portion of that was tied to our downside scenario. And so when we looked at kind of the base scenario given the volumes we saw in our consumer business, we actually saw some improvement in the reserve levels. And it was, in fact, the assumptions that we made around our downside scenario that drove an increase, the $800 million increase in our management adjustment, which went up to about $3.1 billion. I highlight that to say, as we look at the variables that go into our models, it was, in fact, our estimates for a downside, worse unemployment, slower recovery in the way of GDP, a situation where there was no vaccine in 2021, the developed market equities decline meaningfully, so a pretty significant downside scenario. We probability weight that, and that was the more meaningful driver in the slight reserve build and the increase in the management adjustment. I'd say to the other part of your question, there's still a fair amount of uncertainty that remains just kind of given the rise in cases that we're seeing, given the election that's out there, given Brexit, given the lack of clarity around stimulus. But that said, we feel good about the reserve levels we have. We feel good about the downside scenario that's been factored in and the probability associated with that. And we'll just have to see how things kind of play out. It's too soon to suggest otherwise.

Okay. Got it. And then switching gears. Citigroup, your June stress test results were better than peers. Your CET1 declined less than peers, right? And I think that's a great achievement. It's really a testament to all the hard work that you've put in to the transformation of Citigroup since the financial crisis. However, since then, you've announced a CEO change and then followed by the consent orders. So how should we think about the upcoming stress test results kind of in light of some of the recent changes that we've seen at Citigroup?

Well, yes, I think I would separate the 2. And so you're right. We have had, I think, very strong performance coming straight through 2019. And we've continued to perform well through this crisis, an unprecedented crisis, and continue to maintain capital strength and liquidity and serve our clients. We are having a transition of our CEO. After a long tenure here, Mike has announced his retirement, and Jane will take over in February. And I think that will create an opportunity for us to continue, I think, a path of continued strong performance under Jane's leadership. And we're very much looking forward to that with Mike having brought us to a very, very good point, I would say, in the way of our strategy and our financial performance. We're in a crisis and so the regulators have decided to request a resubmission from the industry. And we, like others, have spent the past almost 45 days working on that resubmission, crafting our own scenario of what this stress could look like and comparing that to the 3 scenarios provided by the Fed. And we'll make that submission. And I think in December, they've talked about perhaps sharing the output from that, but they haven't spoken to how they would intend to use that. And so we don't have any additional clarity as to how the results of that would be used. The final point I'd make is that, we've built a resubmission that reflects the environment that we're in and our best sense for what we would need to continue to invest in our franchise to ensure strong performance going forward. And we go into this next submission with an SCB from the last one that was 2.5%, and that was floored. So our underlying SCB was something closer to 1.9%, including the dividend. So I think we go into it from a position of relative strength given the environment.

Your next question is from the line of Hima Inguva with Bank of America.

Scott asked my question on TLAC holding, so that's great. I'm going to switch up a bit. And if you don't mind, I'm going to ask you about the election that we have coming up. And market seems to be thinking a Biden presidency with a Democratic Senate. I'm not asking you to opine on the outcome of the election. I'm more interested in knowing your thoughts around the implications of one way or the other. If you think we could have a more stricter regulatory regime going forward if it were to be a Biden presidency or if it is status quo, do you expect anything, I guess, to move forward in the Senate? And then to that, if you see a split Congress? Any other color that you could provide in terms of the detail of how the Senate may shape up, again, not the outcome. How regulation may shape up depending on how the turnout may be, really appreciate it.

Yes. Hima, as you can imagine, I mean, I'm not going to comment kind of on the politics and the potential outcomes of an administration change. I think that would be difficult to do at this stage, and there's so many factors involved in that. I think there are a couple of things that remain true almost regardless of how it plays out. And one is, it's an obvious one, I think, but we need a resolution to this health crisis that we're managing through. And we need a vaccine that's put in place, one that works and one that can be broadly distributed. We need a stimulus program to help bridge consumers through this and in order to kind of starve off some of the unemployment, increased unemployment that we're seeing and slower growth of GDP. Certainly, we potentially could see tax implications from it, but that is an uncertainty. And if that plays out with a Biden administration, I talked on the earnings call as to what the implications of that would be. I think we, like many other corporations, will have to kind of take a look at what this crisis means for our going forward, right? So we're working with clients on a regular basis as they think about their supply chains and things of that sort, as they think about their real estate footprint and things of that sort. And we, too, will think about things like our real estate footprint or the future of work, for example. And so there's a fair amount of uncertainty that's out there. I can't speak specifically to the implications of the political outcome. What I can say is that our focus on many of the things you've heard us talk to before, whether it's investing in this franchise or execution against our strategy or improving profits or improving our return on and return of capital will still be there. And we'll still have a sense of urgency for continuing down that path.

Okay. Great. And if I could, if I may ask a follow-up on that, on the real estate footprint or maybe more importantly about future of workplace. Do you envision a scenario where you see your employees primarily working remotely or maybe doing a hybrid work setup? Can you share any thinking around that?

Yes. Look, I'm of the belief that there's always going to be -- it's important to have some population kind of in the office if for no other reason, training. And that's how a lot of knowledge gets transferred through and to our employees. And so I don't, I certainly don't envision a world where everyone is working remote at our firm. I do think that we're still learning from this remote working pilot, if you will. It's a pretty extensive pilot. But we're still learning from that. And I do think it will inform different parts of the organization and different parts of the operations as to what will work in the office and what can work remotely. But it's still soon to tell. At this point, we're still at a little bit north of 60% of our population globally working remotely. And so it's still -- and by the way, it's working, right? We're still doing business. We still had, I think, a very good quarter in the third quarter despite a significant part of our population having to work remotely. And we're still covering our clients effectively. And so we're still learning there, Hima. I don't have an end strategy to point to yet, but it is something that we continue to analyze and assess.

Your next question is from the line of Nikhil Khosla with JPMorgan.

I'll take another stab at the LIBOR transition and preferreds, and I appreciate your prior comments on the subject. Just thinking about the transition away and preferreds with LIBOR back ends, when you say the consent to exchange or whatever shape or form that move away from LIBOR takes for the preferred, are you contemplating a possible move to a CMT plus usual back end in that consent or exchange or is it only going to be into a SOFR plus credit spread plus the back end kind of a structure?

Sure. Thanks for the question. And as we look at these structures, we're, of course, going to take the feedback of our investors and where there's demand. And yes, we've seen back-end SOFR. We've used that. Part of that is certainly supporting the transition away from LIBOR and into SOFR. That being said, we know there's been structures out there using CMT, and that demand has been good for CMT. And so as we look at needs in the future, that will certainly be a consideration.

And for the outstanding ones, which were issued back in the day which have LIBOR back ends, I'm just curious about, as you think about this consent to exchange you plan to do to correct the LIBOR language, can you think about moving those bonds to -- exchanging them to a CMT structure versus SOFR as well?

Yes. So again, anything that we would come to market with, we would consider what are our options for doing a new transaction. I think with the exchange, we would have to -- if that was a scenario that we wanted to go ahead with, we would have to think about what qualifies as an exchange of a like security. But I think as we do new issues or call bonds and come to the market, CMT is something we would certainly consider given the interest in that structure. And I think in exchange, we would just have to evaluate whether that would be an appropriately like exchange.

Understood. Appreciate that. And a second one on the business side. On the card side, we've seen a couple of pure-play card names. One came out and guided to much higher number of delinquencies from people coming off forbearance. And we had another name which kind of did not see that. So just curious to see what you're seeing in the 2 parts of your card businesses. As people come off forbearance, how many are kind of rolling into that delinquent bucket and between delinquent buckets? And has anything really changed in the last month or so since the stimulus seems to be on the back burner here?

Yes. Sure. Look, it's interesting. We have continued to see good performance from our consumer card customers. And we've seen that both in payment rates. We've started to see purchase sales improvement just in terms of -- they're still down year-over-year for certain, but not down as much as they had been in the prior 2 quarters. And as you've heard us say before, a lot of this is a byproduct of the stimulus but also the discipline of the consumer in deciding to channel that liquidity towards continuing to pay their cards or card balances. In the U.S., to answer your question more specifically, we've seen about 80% of those that were on a forbearance program roll off program. And of those, about 80% continue to make payments in our U.S. card business. In our Mexico business, when we look at customers there, we see probably 91% have rolled off by the end of the third quarter. And about 84% of the cards customers have remained current after rolling off, and about 95% of the PL or personal loan customers and payroll customers have remained current after they've rolled off. And so we've seen continued resiliency. You look at the delinquency rates that we show, they're relatively stable. And as we mentioned, we're now kind of expecting losses to peak really closer towards the end of 2021 given the way the delinquency buckets have been trending.

Your next question is from the line of Gary Kessler with Goldman Sachs.

One more on LIBOR, particularly some -- we're starting to see some issuers address, whether it's in Europe or Federal Farm Credit Bank here with some exchanges. And when you think about the regulatory piece or the legislative piece, and there's been items written on things being held up in Albany, and you've got Delaware involved here with preferred. I mean what is sort of like the optimal scenario? You guys have been more upfront than most on at least willingness to discuss it. So how should we think about what the optimal scenario is for pursuit of updated fallback language. And we're drawing towards the end of the year. We're going to be in crunch time as we cross into January. So I think we're all looking for clarification because this is obviously a big issue. And as a money center bank, member of ARRC committee, we'd love to get more clarity from all of you. But you guys have more of a willingness to speak. So we're sort of looking for you guys to be the market leader here. So I mean what can you tell us maybe a little bit more extensively on that front?

As I've been talking about for the securities that we've been talking about addressing, we've been looking at a parallel process, meaning that working with the industry on broad-based solutions but also looking at what we could potentially do on our own. I do think the more consistent that the industry tackles some of these items and think about prefs broadly, I think the better off we are because you create that consistency. And I think as you've seen with ARCC making suggestions, I think the reason why they've been effective is because you are bringing together the broad industry. You're coming up with proposed solutions, and you're rolling that out in a largely consistent fashion. I do think the industry is best served with that consistent approach. And I think, again, it's a complex transition away from LIBOR given the scope of it, given the amount of products it impacts. Obviously, when you think about derivatives, there's a protocol focus. Therefore, that is meant to create some consistency. Of course, on the loan side, there's repapering efforts and so where you can get consistency there with fallback language. So my view is that creating consistency, bringing the industry together to help try to recommend and solve some of these topics is how the industry is best served.

Just from a timing perspective, I mean, it's obviously very complicated. So when is it too late? I mean when as investors do we need to start worrying that we're not getting there fast enough? And what can we do to be partners in this?

Yes. It's a good question. Of course, the focus that we've had, that the industry has had is preparing to transition away from LIBOR by the end of 2021. So I think that's a good way to think about it. So I would say that's where our focus has been. More broadly, as you think about this transition away, there's market liquidity that has to develop. And this, again, is away from prefs. But market liquidity developing with the recent move away from LIBOR to SOFR discounting, as an example, hopefully brings more liquidity into the system, more participation in those rates. And of course, the transition is a function of the bank's capabilities but also the investors and the clients we serve. So it's very broad-based. It really needs that broad sponsorship to transition away successfully as an industry. So there's a lot here. And I think just everybody doing their part and working towards that transition will serve everyone best.

Your final question is from the line of Jesse Rosenthal with CreditSights.

Quick follow-up question on the topic du jour, LIBOR and the preferreds, kind of getting at the question a different way. As mentioned that you do have this proposed legislation at the New York State level. I think it's also circulating at the federal level. So just wanted to kind of get your view on the possibility or the feasibility even of a legislative fix here and then if there's anything that we should be monitoring on that legislative outcome.

Look, I realize there's a lot of appropriate questions around this, and it's hard to put a probability on whether that happens at New York or some other level. I am encouraged by the dialogue occurring and looking at that broad-based solution. I do think it's sensible to look at this holistically. But it's very hard to put a probability on how that gets solved and whether the legislative solution will go through and at what level that gets done. So sorry, I wish there was more that we can say with a degree of certainty. But again, as I said before, broad-based solution is the best, and it's good to see this dialogue occurring.

Fair enough. And then just a quick follow-up on just if you were to pursue an exchange or just, in general, if any issuer pursues an exchange. Are there any potential regulatory issues or concerns with losing capital treatment there if you're specifically exchanging a preferred that has less than 5 years to the first call?

Look, I think there's a variety of topics that we've had to cover, tax, regulatory. We don't believe that would present a problem for us. So we've done a lot of work in this space. There's more work to be done. But given the nature of why we're doing such work around this in terms of a consideration, we don't believe that would be problematic.

That concludes the question-and-answer session. Mr. Rogers, do you have any closing remarks?

I just like to thank everyone for attending the call. And of course, if you have any follow-up questions, feel free to reach out to me in Investor Relations. Thanks, everybody.

Thank you.

Thank you, ladies and gentlemen, for your participation. You may now disconnect.