KeyCorp ($KEY)

Okay. Up next, we have KeyCorp, and we're delighted to have with us today, Clark Khayat as CFO, and now Head of Tech and Operations of KeyCorp. Clark, thanks so much for joining us.

Thank you for having me.

So Clark, you've been now been CFO for a little over 3 years. Over that time, the stock has around doubled. You've increased your NII guide for about 2 years in a row and capital has clearly become a strength. Now you're taking on an expanded role at Key, your leading tech and operations in addition to finance. So I want to start there, and then we'll move into the rest of the business. But in your new role, what are your top priorities as you take on this new responsibility and how closely integrated our finance tech and operations today?

Yes. So first, I'd say what a time to take over technology. So the last 3 or 4 months have been pretty wild and dynamic. And I'd say, look, two, maybe two major things which are more I think, strategic than tactical at this point. The first would be just learning as much as I can, not just about what's going on in a technology world, but how we do it at Key, how we think about it. I do think historically, the CFO role has been pretty tightly aligned on technology, obviously, things like funding and sorts and project approvals. So I always knew what's going on, but getting into the details is obviously more important. And then the second really is just around ensuring we have the right alignment across the organization so that the work we're doing in technology and operations is supportive of what the business is trying to accomplish, and I think that's going to be more and more important over time, given how quickly things are changing and the types of things banks like Key want to do with technology going forward. So lots of learning. It's been an energizing experience to sort of roll up your sleeves and get into something else. And hopefully, we can deliver the next 3 years like the last 3 years.

Perfect. So well, you mentioned it's an exciting time to be doing this. And obviously, the reason for that is AI and all the changes going on there. So can you talk about how technology is shaping the -- how AI is shaping the technology strategy to Key?

Yes. So I think like when we talk about AI, it's important. I think the question you get a lot, which is sort of this is what's your AI strategy? And I think it really has to be how is your business strategy enabled or advanced with AI. I think that's really the better way to think about it. And I would say I'd break it into two components. One, probably feels like every company in the world, which is like lots of exploration, we'll say 60-ish proofs of concept out there today, all of them we should probably do, they're all pretty incremental in marginal. So think about the thing that makes, Manan, work 4% more efficient or get you to do three things more productively, like on the margin, that's all really valuable stuff, it improves employee productivity, improves employee experience, client experience, things like that, but they're not like I'm not going to come to the next conference and say, look at how much value we generated from those activities because they're very sort of individual a little bit here and a little bit there, absolutely valuable. They probably don't cost a lot. I'm not sure they drive massive change.

And difficult to measure as well.

Yes and difficult to measure, right? And then you go to the next piece, which is how do you actually get scaled value out of these tools? And that's where I think we, in some ways, looked into two components that fit really well together. Over the last 4 or 5 years, we've had a very intense focus on end-to-end in a handful of areas. So really understanding everything that starts with the client that goes literally all the way back through the organization regardless of who's part of the organization it is, which sometimes can be challenging in larger companies. The second is having almost all of our data and applications in the cloud, which just allows you to spin up things very quickly to develop at pace and as importantly, to tie -- to develop tie a data model to that end-to-end process, which are really the core requirements of applying AI effectively. So we've got a couple of areas. You'll hear us talk about them probably more as we go through the year and stand these up where we think we can get end-to-end scaled value out of AI. And some of it will be customer experience and pull-through in productivity, and some of it will be more automated, less manual processes. It's all the stuff we talk about. We're just trying to do it in three to four places at size. The one that sort of really sits in my new world is going to be software engineering, and I think about that in three pretty basic ways. One is have the ability to develop appropriately on behalf of business strategies. The second is to have the governance and tooling and platforms to allow the businesses to use these things, right? You have people in the business who want to build also. And you don't have to be a coder to build anymore, which is cool and terrifying at the same time. And then the third is applying those practices to the engineering team and getting us much better at running the shop more effectively, more productively over time. And despite my CFO hat, I don't view this as an expense thing. I view it as very comfortable spending the dollars we're spending, we just want to make them more productive.

It's all about productivity there.

Yes.

And then I guess on the data side, you have all your data in the cloud already, and then it's about making sure you have the right data quality. Is that a process? It's all about...

And we would call these operational data zones as sort of our phrase and it's get the operation really well-defined end-to-end and match the data that helps you run and manage that process. So we try to pull data, and we haven't done it everywhere, right? But in the places where we're most advanced and I think most AI-ready, we have these processes well defined. They have the appropriate operational and risk metrics and then they have the data that supports all that capability. And that's where you can just apply AI with some pace. And it can be things like just, again, creating automation, it can be creating or replacing software. So there's a bunch of different ways to apply it. And the hardest thing to date has been like, well, which platform do you pick and which model and how do you get access to the appropriate model at the appropriate time. And it's all moving -- there's a big announcement this morning, right? Literally, you wake up and something's changed from yesterday. So it's getting enough sort of capability and traction quickly to make progress while retaining enough flexibility that you can change when the world is changing.

And so is it for now a productivity opportunity and an expense opportunity? How are you thinking about it from a revenue perspective?

When I think productivity, a lot of times, that's what it is. It's, for example, how do you make the client onboarding experience smoother and easier and the outcome of that is we get to revenue sooner. So I used to run our payments business and the comment I used to make is, hey, if you make a loan, you give the clients the money on day 1, they have the money. You sign a payments contract, like the work starts the day you sign the contract. You're not seeing flow on that day because usually, you have to -- there's an implementation, there's a rewiring and then there's an optimization of usage, right? If you can pull that forward, make that easier, make the APIs and the interconnectivity of that simpler and faster and instead of 6 months or 4 months until the pipes are running, they're running in a month. And I can optimize them faster and get more flow there, you will see revenues. So that's probably the easiest way to think about it, but there's a handful of ways where we're trying to make our client and employee interaction much more dynamic and much faster in a way that allows us to just onboard sooner.

So it looks like there's a lot of opportunity there. One of the questions that comes up in client meetings is just the importance of scale with AI, right? Because it feels like the benefits of going to the largest across different industries. But as you think about super regional banks like Key relative to maybe some of the money center banks, I guess, how do you think about the puts and takes about AI there?

Yes. So I mean, I think if your view is I have to have the best toy at all times, that is really expensive. And I think the question we have to ask ourselves in a lot of cases is what is good enough to accomplish what we're trying to do. And in many cases, the tool that might not be the most current is still pretty well fit for purpose to do what we needed to do. And that's sort of my point of if you just wait to get fluent on the most current thing, you're out of date next week and then you have to start over versus saying I have a platform that's working for me. Those tools are good enough for what I need to do right now. And then it's our job to make sure the chassis is interchangeable with the new tool, but the core of like, how am I getting this productivity, that toolkit has to be like established at some point. So I do think there's benefits to just getting in the game, even if it's not necessarily at the most extreme version, and if you think about the biggest thing in the last 3 months has been how you handled Mythos and the cyber risk and all those. One, I think the industry, and I give people a lot of credit here, does a very good job when there's an industry-wide threat of sort of pulling together and sharing best practices. The second thing is you have massive vendors who are sending you the vulnerability patches out and saying like put this in the system, right? So you're getting a lot of outside-in help on how to accomplish this. And then even if you're not using the most current model, the lab model is pretty good at doing this stuff. And when you put the three together, you can manage it. So again, we're really sensitive to not showing up on a call and saying, our expenses just got blown out because we didn't really -- we weren't tracking our tokens, right? And to date, we really haven't had that issue. And I think going to the cloud when we did, which is 4 or 5 years ago, we've got some muscle built around managing consumption. So we've got a FinApps team that tracks how much cloud consumption we're using and who's using it and what are they using it for? That's the same level of rigor you need to be watching tokens and just make sure that you've got a process to deploy them to people who know how to use them and then that you're moderating that usage to make sure not that you're not spending money that you're getting value for the money you're spending.

And then you brought up cyber, I guess, how much time are you devoting to that now relative to what you were doing before? And how much more worried are you about it given Mythos and all the headlines there?

Yes. Well, before I was like talking about it, now I'm living it a little bit more. So lots more time, again, I think a lot of what happened in the last 3 or 4 months has driven that just there's a little bit of fear of the unknown. I think now that we've seen the tools, and we've seen some of the output, I think people are getting more comfortable that it's manageable. And again, I think it's been a fairly well-orchestrated kind of unified response, which I think is important to the stability of the system. So I'm thankful that people have done that. I'd say with AI every day, I am -- there's something that gets me really excited about it and something that terrifies me about it, and that's probably going to be the way it is. So the more you know, the more you know. And again, it's like, wow, that's incredible. And then you're like, well, that's incredible. That's something to watch for. So it's sort of high alert for both reasons.

Got it. Okay. And maybe staying at the intersection of tech and operations, how are you thinking about the implications of stable coin into organization for the banking industry? And then can you talk about how you see that impacting, whether it's deposits, payments or earnings durability over time?

So this one is interesting because when you think about blockchain and stablecoin, they've been around a long time. And then you think about like how fast AI is progressing, and they're sort of on different paths. I think we've been a believer at Key since 2015, maybe that cross-border could benefit from blockchain. So -- and I think the advent of stablecoin versus not backed crypto is a big part of that, right? So I think it does create some safety and security to that process that probably wasn't resident a decade ago. But that use case, I think, is just immensely logical. I think, of trust disbursements, I think of escrow management, right. There's a bunch of these things that with programmable money in smart contracts absolutely should be out there. And so -- and I think they will get traction. And when you think about Key, we do spend a lot of time thinking about trying to be as close to the edge as we can on payments capabilities. And I think those to me are very resident with payments or our fee-based businesses. So I would imagine, as those really emerge you will see us playing in that pool. We're in every consortium discussion out there around tokenized deposits and there are a lot of them. I think it's all learning at this point. And I think at some point, you have to decide that there are going to be three or four of these or is there going to be one, and I'm not sure we know the answer to that. So we're just trying to make sure we're exposed and engaged in all of it. And I'm sure whatever the final outcome is, we'll play appropriately. In terms of durability of earnings over time, I think it all comes down to how extreme adoption of stablecoin is and what it means for deposits, and you can take it to logical extremes, which I don't really see happening in the near term. I think it will be more likely just part of the ecosystem. And the -- a lot of the use cases, frankly, for day-to-day interactions are less compelling in the U.S. as they are in other countries. So I don't necessarily see that maybe getting as much traction as soon as it could or has in other places. But I do think it's just going to be part of the ecosystem going forward. And I think we're most likely to see it valuable in some of these kind of fee-based client platform opportunities.

And what do you think about -- I guess, once you get tokenized money market funds and you have agentic AI and you have the ability to move your money around a lot faster, what do you think happens with deposit costs over time?

Yes. I mean I think a lot of those abilities are there today. Like if you're a rate optimizer in deposits, you have all the tools you need to rate optimize. The question is, do you want to spend the time doing it. And I think there's like some percentage of the population that does that, right? And my equation on -- because if you go back 3 years, the question we used to get all the time is how come all deposits don't have a 100% beta. And my response would always be like, well, are your deposits all priced at the absolute maximum at all times? And the answer, of course, is not. So I think there's, again, some -- whether it's AI or stablecoin because, frankly, to me, those have the same potential impact to the broad system, which is if there are fewer deposits available and if they are more expensive, I think that impacts credit access over time, right? And if you don't have enough credit, I think that is really the way the U.S. economy works. So I think we all probably have to think about that more broadly than the individual pieces. But if I think about our hybrid commercial accounts today, we're doing a version of rate optimization for clients. So we offer that. They use it because it's convenient to them. We actually don't have to pay them 100% of the index, right? Because it's a convenience we provide.

There's a cost to that....

So again, I think there's elements of it that happen today. I don't see it as a near-term issue. But again, the more capability that exists and the less movement friction that exists over time, the more people will begin to use this. And I think we'll all just have to watch and see what that means. But I see it also as this is an industry thing, it's not a key bank.

Of course. Yes. And you bring up an interesting point that the cost of credit might go up as well, and that has ramifications for the rest of the economy. It has ramifications for borrowing for interest rates, everything else. So there's a lot that would have to move around that if it does happen.

Correct.

Okay. Perfect. Okay. So let's get to your CFO hat. Well, let's maybe talk about Monitor in the second quarter. Anything you'd like to call out for 2Q so far?

Yes. So look, I think another sort of solid quarter as we kind of march back to the targets we've shared. I think we're seeing, again, good loan growth of about $1.5 billion quarter-to-date. So continues to be solid, and that's a mix of new client activity and some additional uptick in utilization from the first quarter. That, I think, will drive NII on the order of maybe 3% quarter-to-quarter. So again, we're just -- we're continuing to see that build as we expected. Fees, I think we spoke on the first quarter call of investment banking debt placement sort of $175 million to $180 million for the quarter. That's kind of where we see it coming in. That gets us up 6% to 7% in the first half from last year. And I get very comfortably on pace for our mid-single digits growth year-over-year. Again, I think a big question there is, do you see the middle market M&A component pull through? We haven't seen as much of that yet. And so that's still something we're watching, but the pipelines continue to be very strong and conversation is very, very consistent. Expenses, we're going to see a little bit of a pickup in the second quarter, call it, 3.5% to 4%. That's not the trajectory for the year. So there's a pickup really off first quarter from some of the banker hiring we've done from general merit pools from some benefits cost and that some of those are market-based. So as the market has been strong, we adjust those up. I think we'll see those plateau in the second half. So they'll be up but not to the same degree, and that gets us very comfortable with the kind of 3% to 4% expense guide for the year. And then credit, still pretty benign. I think charge-offs, we feel good about that kind of 40-ish basis point number. After 4, 5 quarters consistently of every credit metric improving, you might see a little bit of plateauing or maybe even an uptick in one or two. I think that's really around some of the rate movement and some of the softness pockets that are really driven by oil, whether that's ag or consumer products companies or transportation right, that are just feeling a little bit of that pitch. I don't think that drives through to loss in the near term, but we have some fairly conservative triggers on credit metrics, and we'll follow those. So if I stand back and look at that, I'd say the year, we just -- we feel really good about the guidance for the year, and we'll continue to march forward.

And just to be clear on the quarter guide, the NII numbers that you gave and the expense numbers you gave, those are Q-on-Q?

Correct. Yes, if I did say that, that was the intent. So thank you for clarity.

That's great. Okay. Very clear. So let's get into some of the other drivers. So as we think about NII, loan growth has been picking up nicely. You just mentioned there is more utilization that's happening. I think you also spoke about embedding an appropriate level of conservatism in the guide? And how have these loan trends progressed, whether it's in -- on the C&I side or whether it's on the consumer side, I know you're running off parts of the portfolio. So can you paint the picture for us on the loan side?

Yes. So -- our commercial continues to be very strong. So as I mentioned up. So to be up roughly $4 billion from year-end if you put the first and second quarter to date together, that has been I'd say the most positive is if you look at our middle market business, it's broad-based, it's market, it's industry. There's nothing. It's just that the teams have been productive broadly. And that's across all markets, newer teams, older just that business is really operating pretty consistently. And then really in the larger areas, utilities and power, where we have some strength that has been -- continues to be a source of strength. And it's just -- I don't know where that stops given the demand for power in general. So I think we've been well positioned for that just given how focused we've been in that industry for a while, and we're continuing to see that come through. I think some of the other balance sheet dynamics with rates where they are, we're seeing not quite as much resi pay down as we would have expected at the beginning of the year, which makes sense that mortgage replacement rates are higher. Similarly, on the commercial real estate, less paydown activity and less refi. And a lot of times, that refi goes off the balance sheet to some permanent solution that we'll place. So we would see a little bit of loan balances sitting there on top of what we're originating. At this point, utilization, which was up well in the first quarter, up again strong in the second quarter has actually come down a little bit from its peak earlier in the second quarter. So that's always going to be something we're watching. And I think at this point, we feel comfortable with our guide for the year. But if it continues at this pace, we'll revisit it as we get probably early into the third quarter.

So I guess with more growth on the C&I side, slower paydowns, I guess the other side of the balance sheet is how you fund that growth. We were just talking about industry deposit competition offline. Can you talk a little bit more about what you're seeing quarter-to-date there? Are there any changes in the deposit competition side?

Yes. So we're -- I think we're a little bit different in that. We have pretty seasonal flows in the first quarter and first half of the year. We tend to trough in May. We did that again and then we start to build up from there. We would -- and we're seeing that. So we'd expect second quarter averages to be pretty close to first quarter averages on deposits. We'd expect end of period to be at or above where the first quarter was. So we'll see that dip. And then we expect it to build throughout the back half of the year. So if that plays out the way we see it happening, we feel very good given that growth, the loan-to-deposit starting point, which tends to be lower than others. And the remixing, even though it's lower, we're still remixing out of residential mortgage, that's allowing us to recycle some funding. If loan growth got stronger or those deposits didn't come on at quite the same rate, then we'll look back at what the funding profile looks like. I feel like we have lots of ample funding, whether it's client deposits or wholesale funding and the question it just comes down to cost and efficiency. So we'll look at that when we need to. We haven't felt that yet. And just to put maybe a finer point on it, we've been at kind of 350 on our front book in consumer since March of last year, we haven't moved that up. We have others who are market-by-market doing different things. And we haven't felt the need yet to do it if some of those balance sheet dynamics change, then we'll obviously look at that and make an adjustment. And I think if there's a hike at some point, then you probably do need to make a change. But we're watching that literally daily, and we'll start to make some adjustments if we need to.

Yes. I was just going to ask you about that. I guess, if there is -- even if there's a possibility of a hike later in the year, are you seeing deposit competition ramp up for the industry? I get that you might not have to act because you have a lot of flexibility on your balance sheet. But are you seeing more competition in the industry? We've heard a few banks talk about more competition in the Midwest. Are you seeing any of that out there in the market?

I mean, look, we are seeing higher rates in certain markets. The Midwest would be one of them. I think there are probably some exogenous or unique reasons for that happening that aren't necessarily broad. But look, if rates are going to stay higher or they're going to get higher, historically, that tends to drive fewer deposit dollars and those dollars are more expensive. So we all have to be prepared for that in advance. And I would say the rigor that we put in, particularly on deposit management since 2023, we've kept going even though things have felt better because it's just -- we think it's the right way to run the bank. So we're watching this, and I think you made the point earlier, like you can't wait until things change to change them. You got to really do some of this in advance. And we'll put all those factors together, and we'll do what we need to do to make sure we're in a good spot.

So maybe to round out that conversation as we think about NII and NIM, you just reiterated the NII guide for the year. You also have guidance out there for NIM to hit about 3.25% plus by the end of 4Q '27. Is that still kind of how you're thinking about it? I know the Street is actually not quite there, but is that still how you're thinking about it?

There wouldn't be anything I'm seeing at this point that would cause me to change that. If I think about our interest rate position, we've been neutral now for a while, which I think is right, given some of the uncertainty. And we've just been talking about it 3 or 4 months ago, there were how many cuts were coming, and now it's more likely than not a hike. So I think that neutrality has served us pretty well. And when I really break that apart, we're slightly liability sensitive on the front end. We're asset sensitive in the belly of the curve and 40 basis points of 3- to 5-year rate increase gives us a little bit of reinvestment rate that I think more than offsets or offsets pretty comfortably whatever the front end would cause us to feel in year. So that's, again, why we feel pretty good about the guide in a variety of situations because we've got a little bit of offsetting plays on the curve.

Got it. Okay. Let's talk about these diverse set of fee businesses, I think right after earnings, you announced you'll be able to acquire Clearwater, which is a European investment bank. Can you talk about the rationale for that deal?

Sure. So this one, I think, hopefully, is pretty straightforward. One, the value of our business is ensuring that our sell-side clients get the opportunity to get in front of the broadest group of potential buyers. And we've seen now for years, Europe being a place that has a lot of appetite. And vice versa, when we're doing buy-side opportunities, we want the ability to showcase opportunities that are not just domestic. We've had a referral relationship with Clearwater now for 5 or 6 years. It's been quite productive. So in terms of acquisitions when you think about it, this is like people we've been working with pretty closely for a while. We know them. We've got good working relationships. We've seen the value of how this can work back and forth. And so it was really just a formalization of something we've been doing. And to me, it just -- it was kind of a natural thing that we've sort of talked with them about over time and just this year was really the right time to take a further step at that. So to us, it's -- we've talked about these boutique deals as sort of pseudo organic extensions of how we run the business. This one is literally can you formalize this thing we've been doing with them now for some time. So I feel like it's -- this one is about as natural as probably an acquisition could be.

Got it. All right. So -- and then you also spoke about on the M&A side, middle market M&A not quite picking up just yet. What do you think you need to see for that to pick up? Because you said pipelines are pretty good, right? So...

Yes. I mean they were record high earlier in the year. They've stayed -- I don't know if they stayed exactly at that level, but they're pretty close. I do think that is more about rates than maybe the large ticket M&A because those deals are going to be more stock for stock or those are just companies that are going to transact when they've decided they're going to transact. And the shape of the curve or the absolute rates maybe don't matter as much. I think we're also seeing -- we see a lot of sponsor activity and more and more, I don't know what the percentage is, but the concept of moving a -- an asset portfolio company to a continuation fund versus selling it if it's not the right market is becoming more prevalent. So I think we're trying to figure out exactly like what is the thing that pushes it. And I think it tends to be -- it will be certainty of some kind, right? And we've -- I think we've continued to be a version of uncertainty now for whatever it is 6, 7, 8 quarters. So if rates are going to be higher, people will figure it out. As long as they know that if rates are going to be lower, people will transact. If we don't know, then people tend to want to wait and see.

Is there a window within which people are trying to transact or not quite. It's just more about certainty?

I would say it's probably more about certainty unless it's some lifetime -- we got to get out of this fund at some point. But again, that's partially why these continuation vehicles have become popular because they can bridge that without exiting completely.

Got it. All right. Let's talk about capital. So the capital story continues to be a good one. You announced a $3 billion buyback authorization intra-quarter and you still have higher capital levels than peers. So what -- I guess, the question is before we get into maybe buybacks and capital deployment, have you seen tangible benefits from holding that higher level of capital, whether it's in client conversations or in any other way?

I mean, for sure, I was thinking about -- like I'm not sure these are tangible. Maybe they are. I have to think about that for a minute. But I think our ability to attract quality people to the platform and our ability to get our people out and prospecting for new clients is completely different post August of '24 than it was before. We have been coming off this diet. We were telling people -- I mean, Chris was out telling everybody like we're going on offense. But if you're coming out of that, you're like, well, maybe but I want to make sure that whatever balance sheet I have, I have for my best clients, so I don't want to use it somewhere else. The minute we got capital and we started sort of fixed our liquidity issues, fixed our capital issues and get our earnings back in place. I think people just have a level of confidence that we can go out and serve clients the way we want to and the way we need to. And I think you've seen it in client growth and loan growth over that time. So -- and we've done a really -- I mean, we hired -- we said 9-plus percent new bankers last year to the platform. I think that would have been hard to do if our capital was in a worse position.

Yes, I'd call that tangible. Fair enough. And then as you're thinking about capital deployment from your when you think about the level of capital, right, like Moody's recently placed a company in review for a potential upgrade. And then you also spoke about 100 basis points benefit of CET1 from Basel Endgame. So how does that impact how you're thinking about either buybacks or any incremental capital deployment from here?

Yes. So one, I think we're always going to do whatever we can with good clients. So we want to make sure that we're sticking to our relationship strategy, but we want the balance sheet to be open to good clients and new clients. I don't suspect at this point, we'll push on the dividend very hard. I think that yield is sort of right in line with people. It feels right. And then for us, I think the buyback component is something that is more consistent and sort of methodical than anything. We've been asked like, do you just use that 100 basis points on day 1. And I just -- that, to me, feels particularly in times of uncertainty is like not the most judicious safest decision versus telling people like you can expect us to just continue to do this over time until we get to that range in a way that we're very comfortable. So I think you'll see more of the same, and we're on track to do the $1.3 billion of buybacks that we talked about this year, and we'll just continue to sort of work through that authorization over time.

Got it. And it's a relative game as well, right? So if everyone is freeing up capital at the same time, you don't want to do everything on day 1.

Yes, correct. And if things go sideways, you'll be asking me why I did all those things on day 1.

I hear you.

I don't want to answer that question.

I hear you. All right. Great. So let's put it all together. You're well on your way to achieving the 15% plus ROTCE target by 4Q '27 and then you have 16% to 19% in the longer term. What are the biggest drivers that get you from the medium term, 15% plus to that 16% to 19%?

I think, one, it's getting -- I mean, this is just the math, right? It's getting the balance sheet efficiency to the right level, which we continue to have fixed asset repricing. We'll have that for some time. But if you get that to the right place and you're always going to have a little bit of it, I think you see NIM in a much more comfortable place. It's good quality relationship loan growth because that often is the driver to our fee businesses, right? The more new clients we have and the more credit we have out there, we tend to do better on payments and payments, commercial deposits and capital markets. And I think that combination is the right return profile and then continue to invest and grow our wealth business. And behind that, just manage our expenses appropriately, which I think we've done a good job. So again, just to reiterate, when I think about what we can do with things like AI, it's not spending less. It's getting more out of the dollars we're spending. And I feel like we're -- we've got some ways to do that, that will support that kind of return.

Got it. All right. With that, we're out of time. Clark, thanks so much for joining us.

Good to see you.

Thank you.