KeyCorp (KEY) Q3 2026 Earnings Call: Record Revenue, Higher NIM and Improved Credit

Key Takeaways

• KeyCorp reported Q3 2026 earnings of $0.51 per share, up 24% year over year, while record quarterly revenue exceeded $2 billion, an increase of 8%.
• Net interest margin expanded 6 basis points sequentially to 2.95%, supported by commercial loan growth, fixed-asset repricing, deposit growth and lower reliance on wholesale funding.
• Average deposits increased $3.6 billion, or 2.4%, from the prior quarter. Client deposits rose $2.8 billion, while total deposit costs increased only 3 basis points to 1.66%.
• Period-end C&I loans grew $3.3 billion, or 5% sequentially. Line utilization increased 200 basis points to 33%, its highest level in more than three years.
• Credit quality improved. Net charge-offs were $89 million, or 32 basis points, and the nonperforming asset ratio fell 19 basis points to 55 basis points.
• Management tightened full-year 2026 net interest income growth guidance to approximately 10%, raised average loan growth guidance to approximately 5%, and lowered expected net charge-offs to 37–39 basis points.

Core Financial Results

Business and Operating Performance

Average loans increased $1.5 billion sequentially. Average C&I loans rose $2.3 billion, or 4%, partially offset by continued runoff in lower-yielding consumer loans. Of the $3.3 billion increase in period-end C&I balances, management said approximately $2 billion came from higher utilization.

Commercial loan growth was broad-based across industries and geographies. Management cited new client acquisition, capital expenditure, inventory investment, smaller acquisitions, and demand from utilities, power and renewable-energy clients. Commercial clients increased 3% year over year, while relationship households grew 2%.

Average deposits rose $3.6 billion sequentially, exceeding loan growth and allowing KeyCorp to reduce wholesale funding. The company attributed the increase to commercial operating balances, deeper relationships with existing clients, and deposits from newly acquired banking clients.

Investment banking and debt placement fees reached $214 million, up 27% sequentially, driven by M&A and syndication activity. Investment banking pipelines increased 5% sequentially and 20% year over year, despite strong transaction completion during the quarter.

Trust and investment services income rose 10% year over year, supported by higher market values and positive net flows. Cards and payments revenue increased 8%, while deposit service charges grew 5%. Assets under management reached another record.

KeyCorp spent $1 billion on technology in 2026. Its AI initiatives are focused on commercial onboarding and servicing, commercial originations, consumer servicing and software engineering. Management said these programs remain at an early stage and are intended to improve throughput, cycle times, controls and employee productivity.

The company repurchased $320 million of common stock during Q3. Year-to-date repurchases exceeded $1 billion, leaving approximately $250 million to complete the targeted $1.3 billion program for 2026.

Management Guidance

Management expects remaining 2026 loan growth to be fully funded by client deposits. It also expects incremental client deposit growth beyond its previous second-half target of $3 billion.

KeyCorp maintained its target for a net interest margin of at least 3.25% in Q4 2027. Management said approximately 70% of the required expansion could come from the repricing of about $24 billion of securities, swaps and residential mortgages, with an estimated weighted-average yield pickup of approximately 165 basis points.

The company also remains focused on achieving and sustaining ROTCE above 15% by year-end 2027, on the way to its longer-term target range of 16%–19%. Management expects 2026 to produce record revenue and pre-provision net revenue and said KeyCorp is positioned for another year of outsized revenue and earnings growth in 2027, while noting that detailed 2027 guidance will be provided after year-end.

Risks and Areas to Watch

Deposit competition remains intense, and management expects deposit pricing pressure to increase following the September rate increase. KeyCorp expects its cumulative deposit beta to move toward the mid-50% range over time, with a lag during the first several quarters.

Management identified transactional commercial real estate as a weak area because higher capital costs are causing deals to be repriced or delayed. Middle-market M&A activity also has not normalized, with rate volatility and uncertainty over future rates affecting sponsor activity.

KeyCorp said geopolitical and macroeconomic uncertainty remains elevated. The bank is managing its balance sheet to remain broadly neutral to interest-rate movements over a 12-month horizon.

Direct data-center exposure is less than $1 billion and is primarily to hyperscalers. Management said it remains cautious about direct lending into data-center construction because of the sector’s rapid growth.

Analyst Q&A Highlights

• Deposit growth: Management expects deposits to grow at least as fast as loans through 2027. Additional deposit growth could support further reductions in wholesale funding.
• Commercial lending: Higher utilization reflected client spending on capital projects, inventory and acquisitions. Management expects Q4 earning-asset growth to moderate from Q3 but said it has not seen broad erosion in loan demand outside transactional real estate.
• Margin expansion: Fixed-asset repricing is the largest component of the path to a 3.25%-plus NIM. Deposit growth, deposit costs and the ability to repay wholesale funding are the main remaining variables.
• Capital allocation: Management maintained its 2026 repurchase target but said future buybacks will remain the final priority after supporting clients, investing in people and technology, and maintaining the dividend.
• Credit quality: Management attributed resilience to disciplined underwriting, an originate-to-distribute model, limited leveraged-loan growth, a super-prime consumer portfolio and industry-specific commercial expertise.
• AI deployment: AI is intended to support relationship banking rather than replace it, including by reducing paperwork and streamlining commercial credit, collateral and onboarding processes.

Full Earnings Call Transcript

Complete Earnings Call Transcript

Management Remarks

Operator

Good morning, and welcome to KeyCorp's Third Quarter 2026 Earnings Conference Call. [Operator Instructions] As a reminder, this conference is being recorded.

I would now like to turn the conference over to Troy Gates, KeyCorp's Director of Investor Relations. Please go ahead.

Troy Gates

Thank you, operator, and good morning, everyone. I'd like to thank you for joining KeyCorp's Third Quarter 2026 Earnings Conference Call. I'm here with Chris Gorman, our Chairman and Chief Executive Officer; Clark Khayat, our Chief Financial Officer and Head of Technology; and Mo Ramani, our Chief Risk Officer.

As usual, we will reference our earnings presentation slides, which can be found in the Investor Relations section of the key.com website. In the back of the presentation, you will find our statement on forward-looking disclosures and certain financial measures, including non-GAAP measures. This covers our earnings materials as well as remarks made on this morning's call. Actual results may differ materially from forward-looking statements and those statements speak only as of today, October 20, 2026 and will not be updated.

With that, I will turn it over to Chris.

Christopher Gorman

Thank you, Troy, and good morning, everyone. We reported third quarter earnings of $0.51 per share, up 24% year-over-year, reflecting the strength of Key's relationship-driven franchise the disciplined execution of our strategy and the momentum we continue to build across the business. We reported record quarterly revenue of over $2 billion, which represents 8% year-over-year growth. PPNR growth of $60 million on a linked-quarter basis and importantly, return on tangible common equity that exceeded 15% as we continue to make significant progress with respect to achieving and sustaining a 15%-plus return on tangible common equity by year-end 2027 on our path to achieving our current long-term target between 16% and 19%.

We continue to deliver on the commitments we have made. Our relationship-focused business model continues to differentiate key and was evident again this quarter. Our deposit franchise continues to perform well in a competitive environment. Average total deposits grew $3.6 billion, while interest-bearing deposit costs increased only 3 basis points during the quarter. Overall funding costs rose just 1 basis point. Net interest margin expanded 6 basis points sequentially to 2.95%, reflecting the continued fixed asset repricing and funding optimization. Our capital markets team delivered strong results with investment banking and debt placement fees increasing 27% sequentially with growth in M&A and syndications.

Asset quality remains strong. Our net charge-off ratio improved to 32 basis points, and our nonperforming asset ratio improved 19 basis points to 55 basis points. During the quarter, we repurchased $320 million of common stock, keeping us on pace to achieve our full year share repurchase target of $1.3 billion. We remain comfortable with our current capital levels with strong organic capital generation, we will continue to take a measured approach to capital management.

As a reminder, capital priorities remain unchanged. First, we will continue to support our clients and prospects. Second, we will continue to invest in people and technology. Third, we will maintain our strong dividend. And finally, we will return capital to shareholders through share repurchases. While we continue to successfully complete tuck-in acquisitions like Clearwater, traditional old bank M&A is not a priority. The economic environment remains constructive, and our momentum continues to be strong. We are seeing healthy client engagement and activity across our businesses. We continue to organically grow clients.

In the third quarter, commercial clients increased 3% and relationship households increased 2% from the prior year. Commercial loan growth remained strong and broad-based by both industry and geography. Period-end C&I loans increased 5% sequentially, reflecting both new client acquisition and a 200 basis point increase in line utilization to 33%, the highest level in over 3 years. The line utilization increase is reflective of clients continuing to invest in their businesses. Our consumer trends remain strong and resilient as our clients continue to benefit from a healthy labor market and the wealth effect. We continue to see strong performance and are gaining share across our priority fee-based businesses, investment banking, commercial payments and wealth management.

In the third quarter, these businesses collectively grew 12% compared to the prior year. Investment Banking pipelines are up 5% sequentially despite significant pull-through in the third quarter and remained at historically elevated levels, supported by record M&A pipelines. While middle market M&A activity has yet to normalize, we continue to see significant client engagement and remain confident in our expectation for mid-single-digit investment banking fee growth this year. In commercial payments, Total gross payment fees increased 8% compared to the prior year. We continue to scale embedded banking and are encouraged by our trajectory.

In wealth, assets under management reached another record this quarter. Wealth remains a significant opportunity for us as we are less than 10% penetrated with respect to our current base of mass affluent households who bank with Key. We continue to invest in people and technology. We are attracting top talent and continue to hire frontline bankers. Our $1 billion technology budget this year has supported new product and service capabilities while allowing us to position the company for the future.

We continue to embed AI in enterprise-wide end-to-end processes, currently targeting 4 domains: commercial onboarding and servicing, commercial originations, consumer servicing and software engineering. Earlier this year, Clark assumed the leadership of our technology organization, creating even tighter alignment among our business strategy, operating model and technology investments. Under his leadership, we are accelerating the deployment of AI across these domains to simplify processes, increase productivity and help our teammates spend more time delivering advice and solutions to clients.

While we remain in the early stages, we are encouraged by the opportunity in front of us. As we look ahead, the geopolitical and macro outlook remain uncertain. At Key, we remain focused on managing what we can control and remain well positioned to perform through a range of potential economic scenarios. While our results are strong, we acknowledge there is uncertainty in the broader market. we are actively managing the risks and maintaining discipline throughout the bank. We are well positioned for a range of potential rate scenarios and continue to intentionally manage our balance sheet to be fairly neutral from an interest rate perspective.

In summary, we are proud of our third quarter performance and the sustained momentum across our business. These results would not be possible without the contributions from each of our teammates. Thank you for your dedication that will result in what we expect to be both a record revenue and record pre-provision net revenue year in 2026. We are currently in the midst of our planning process, and we'll have more to say on 2027 at year-end. But with our strong trajectory and healthy pipelines, I believe we are well positioned to drive another year of outsized revenue and earnings growth in 2027.

With that, I'll turn it over to Clark.

Clark Khayat

Thanks, Chris. As Chris highlighted, continued execution of our relationship-based strategy drove strong performance in the quarter. For the third quarter, revenue exceeded $2 billion. PPNR reached a record level and we improved returns, inclusive of improving credit quality.

Starting on Slide 4. We reported third quarter earnings per share of $0.51. Revenue was up 8% year-over-year, while noninterest expenses increased by 5% when excluding the FDIC special assessment in the year ago quarter. Tax equivalent net interest income increased 10% year-over-year and 4% sequentially. Noninterest income increased 6% year-over-year. Provision for credit losses of $61 million included $89 million or 32 basis points of net charge-offs and a reserve release of $28 million. The net release reflected improved portfolio performance scenario impacts and continued remix to higher credit quality relationships, partially offset with some qualitative build to reflect ongoing economic and geopolitical uncertainty. We remain well reserved for a range of economic outcomes. Return on average tangible common equity increased 260 basis points year-over-year to 15.1%.

Moving to the balance sheet on Slide 5. Average loans were up $1.5 billion sequentially. Average C&I loans increased 4% or $2.3 billion, partly offset by the runoff of low-yielding consumer loans. On a period-end basis, C&I loans increased by $3.3 billion or 5%. As Chris mentioned, growth came from a combination of broad-based new client acquisition and an increase in utilization.

Turning to Slide 6. Average deposit balances increased $3.6 billion or 2.4% sequentially, exceeding loan growth and allowing us to reduce wholesale funding. Client deposits, which exclude brokered CDs, grew $2.8 billion or 1.9% driven by growth in commercial operating balances. Last quarter, we communicated expectations for $3 billion or more of client deposit growth in the second half of 2026. And based on the growth in the third quarter, we have nearly achieved that. While we continue to monitor deposit dynamics in a higher rate environment, we now expect incremental client deposit growth beyond the original $3 billion, and we expect our loan growth for the remainder of 2026 to be fully client deposit funded.

Total deposit costs increased 3 basis points sequentially to 1.66% and total funding costs rose just 1 basis point. If market expectations materialize as currently estimated, we expect beta in the rate hiking cycle to be roughly symmetrical to the rate easing cycle reaching a mid-50s cumulative beta over time with the beta lag in the first few quarters.

Turning to the drivers of NII and ME on Slide 7. Taxable equivalent NII was up 4% and net interest margin of 2.95% increased 6 basis points from the prior quarter. The increase was driven by commercial loan growth, funding optimization via strong client deposit growth swaps and securities repricing at higher rates as well as an additional day in the quarter. We continue to manage our balance sheet to a fairly neutral interest rate position over a 12-month horizon, which we view as prudent from a risk management perspective given continued rate uncertainty and our focus on driving performance through the cycle.

On Slide 8, noninterest income increased 6% year-over-year and included the negative impact of a $6 million Visa related charge in the current quarter. Investment banking and debt placement fees were $214 million, a 27% increase sequentially driven by M&A and syndication activity. Trust and investment services income grew 10% year-over-year, reflecting higher market values and positive net flows. Cards and payments and service charges on deposit accounts increased by 8% and 5% year-over-year, respectively, both reflecting strong commercial payment activity. Corporate services fees were stable compared to the year ago quarter as higher loan commitment and client FX fees were offset by lower hedging activity.

Commercial mortgage servicing fees were $53 million. At quarter end, we renamed primary for special service are approximately $735 billion of commercial real estate loans of which about $270 billion is in special servicing. Active special servicing third-party assets were flat sequentially at $10 billion, about half of which is office. We continue to expect $50 million to $60 million of commercial mortgage servicing fees for the fourth quarter.

On Slide 9, third quarter noninterest expenses were $1.2 billion, up 2% sequentially. The increase was driven by investments in people and technology, higher incentive compensation related to strong fee growth and an additional day in the quarter. We continue to expect full year expense growth of 4%. A portion of our technology investment is focused on leveraging AI to create a more efficient and scalable operating model. As Chris mentioned, in addition to broad-based adoption of AI tools, we selected 4 end-to-end processes against which to apply these capabilities at scale. The confluence of well-understood processes with supporting data and an opportunity to reimagine client and employee experiences has proven to be a valuable recipe for AI tools. We are still early in the process, but are optimistic that we can generate growth and efficiency while improving risk and controls.

In commercial onboarding and servicing, we're focused on delivering a faster client onboarding experience which feeds our time to revenue generation while simplifying client support. In consumer servicing, we're deploying AI to automate high-volume tasks, reduce fraud and enable faster issue resolution. In commercial originations, we're streamlining the credit and collateral processes to improve client and banker experiences and accelerate cycle times. And in software engineering, we're transforming how we build, deliver and support technology through AI to achieve greater speed, quality, resiliency and efficiency. In summary, I'm energized by the opportunities ahead and what AI can unlock across Key.

Turning to Slide 10. Credit quality is broadly improving is our focus on disciplined client selection, prudent underwriting standards and deep domain expertise in helping clients navigate the uncertain environment was evident this quarter. Net charge-offs were $89 million or an annualized 32 basis points of average loans. As expected, nonperforming assets decreased $191 million or 19 basis points sequentially and at 55 basis points is the lowest since late 2023. The decrease was driven by broad-based portfolio resolutions. The criticized loan ratio was stable at 4.9%.

Moving to Slide 11. Our CET1 ratio was 11.1%, and our March CET1 ratio was 9% at current. In the start of the year, we've repurchased over $1 billion of shares and approximately $250 million remaining to complete our targeted $1.3 billion of repurchases in 2026.

Moving to Slide 12. We are updating our 2026 guidance. We are tightening our full year net interest income growth outlook to approximately 10% compared to the prior guide of 9% to 11%. We now expect average loan growth to increase approximately 5% compared to the prior guide of 4% to 5%, and average commercial loan growth of approximately 10% compared to the prior guide of 8% to 10%. We see full year net charge-offs coming in the range of 37 to 39 basis points compared to the previous guidance of 40 to 45 basis points.

The outlook reflects the improving credit trends we've seen in our portfolio to date. We now expect to generate a net interest margin of approximately 3% in the fourth quarter of 2026, reflecting average earning asset growth of approximately $1.5 billion to $2 billion from the third quarter. The updated guide is primarily a function of continued commercial loan growth and preferred capital optimization actions that are accretive to EPS and returns but are slightly diluted to them.

Compared to our guidance in the second quarter, we now expect fourth quarter earning assets, excluding AOCI, to be approximately $2.5 billion higher. This increases NII and returns, but reduces NIM approximately 2 basis points. We will continue to execute our relationship-based strategy to drive attractive returns from new and existing clients. With respect to preferred capital, we redeemed $525 million of Series D preferred stock in September and we're planning to opportunistically issue new preferreds in the second half of the year. Based on prevailing market conditions, however, we've not yet made the decision to issue new preferreds. We'll continue to watch the market and act accordingly to optimize our capital structure.

As a result, we'll see lower preferred dividend payments in 4Q, which from an earnings standpoint will be partially offset by slightly lower NII. The net result is tremendous on EPS, but will impact NIM by about 1 basis point. All other guidance remains unchanged. In summary, we continue to see strong momentum in our business through the fourth quarter and expect to deliver on our commitments while navigating a dynamic and evolving environment.

With that, I'd like to now turn the call back to the operator to provide instructions for the Q&A session. Operator?

Operator

[Operator Instructions] Our first question today is from Manan Gosalia from Morgan Stanley.

Question-and-Answer Session

Manan Gosalia

Clark, you saw a nice acceleration in average deposit growth this quarter. Can you unpack what drove that deposit growth? And what goes into your expectations for that mid-50s deposit beta that you outlined? And actually, maybe if you can layer on to that, given that the deposit growth was higher than loan growth this quarter, how much opportunity is there to reduce the reliance on wholesale borrowings over the next few quarters?

Clark Khayat

Yes. Thanks, Manan. So as you noted, we grew deposits by about $3.6 billion on average in the quarter with $2.8 billion of that being client deposits. And as you noted, deposit costs were only up about 3 basis points. So on the second quarter call, we noted we had some visibility into seasonal deposit flows, primarily commercial operating deposits and that view supported our guide at the time of $3 billion of client deposit growth in the second half.

We obviously saw that in more in the third quarter alone, driven not just by the seasonal growth, but by deepening deposit relationships with the existing clients and adding deposits from new to bank customers, which would have come on with loans in the first half of the year. So overall, very pleased with the trajectory on the deposit side. I'd say the trends we're seeing now continue to be encouraging, which gives us confidence that we can fund our 4Q loan growth with client deposits. Given the rate hike in September, we don't expect the same modest price increase of 3 basis points, but we do feel like we can effectively manage the cost along with the balanced growth. Competition at this point is certainly intense, but it has been competitive all year, and we haven't really seen that intensity increase yet.

Increased industry loan demand could ratchet that up, although we'd expect some of the loan growth for us to moderate actually in 4Q. To be clear, I'd say we aren't taking a victory lap here, and we're definitely not getting complacent. The deposit game requires consistent efforts so we're not going to let off the gas at all. I would say our focus on primacy, including deposits and commercial payments over the last decade plus is serving us well. Our visibility here is based on name-by-name bottoms-up analysis and that's really what's driving our confidence. So to your point about kind of deposit growth outpacing loan growth in the quarter. We mentioned we expect to fund that loan growth with deposits in the fourth quarter. And I think going forward through '27 would expect to grow deposits at or better than loan growth.

Manan Gosalia

Got it. And then maybe on the flip side on the loan growth side. The C&I growth was really strong this quarter at up 5% Q-on-Q. I think you touched on this in your remarks a little, but can you expand on what's driving that strength, especially given that the utilization rate is up 200 basis points Q-on-Q. How sustainable is that? And how should we think about that into 2027?

Christopher Gorman

Yes. Manan, it's Chris. So it's broad-based. As I mentioned, it's broad-based from a geographic perspective. It's broad-based from an industry perspective. We're growing new clients, which obviously is helpful. The other thing that obviously is driving it is people are investing in their business. And we've been looking forward to that happening for some time, and you're seeing people spending money on CapEx, people buying small tuck-in acquisitions and we're expanding the customer base. So it's pretty broad-based.

The one area that I would point out where we've had really significant growth, and it's an area that we're really strong in is our utilities, power and renewables. So you can imagine with all the discussion about the shortage of electrons in this country and the shortage of distribution are good clients that are investment-grade utilities are also making some pretty significant investments.

Operator

Our next question is from Scott Siefers from TD Securities.

Robert Siefers

Clark, I was hoping you could please help to put sort of a finer point on the margin progression. It sounds like there will be some sort of just some noise in the fourth quarter and then you'll also be impacted by the strong loan growth you're seeing, but you're sticking with the 325 plus margin by 4Q '27, which is great. Maybe once we get through that sort of transitory noise in the fourth quarter, maybe if you can help sort of bridge the gap between now and end of '27, sort of what keeps you confident that you can still get there over the next several quarters?

Clark Khayat

Sure. Thanks, Scott, and welcome back to the call. So first, maybe let me just zoom out a little bit and start with the broader objective of ROTCE. So I'd just remind you the most important target we have out there for the fourth quarter '27 is delivering 15-plus percent ROTCE. And based on what we're seeing today and the performance we're delivering, we remain highly confident we're going to achieve that. We are also maintaining, as we said, and you noted the $3.25-plus NIM for 4Q '27, so that remains intact as well. And I'd just note that when we first offered that target, the environment was quite different. And we did note at the time that we thought we had multiple paths to achieve that, and we still believe that to be true.

So as we build the branch I'd say the most valuable opportunity for sure is the fixed asset repricing opportunity between now and the end of '27. We have about $24 billion of security swaps and resi mortgages were only off and given the current view of rates and our forward starting swap positions, we see a weighted average yield pickup of about 165 basis points there. So if you consider that say that alone drives roughly 70% of the NIM expansion required to hit that target. So more mechanical than it was a quarter ago, and that's just a function of the rate environment.

The second driver would be client deposit growth, which we just covered a little bit. But as we've discussed, we've seen great client deposit growth at attractive rates, and we feel we can fund our loan growth going forward with deposits. Third, and I'd say probably more of an expansion of a second. We think we can actually grow deposits. Client deposits faster than loans here, which allows us to pay down wholesale funding and again, would be accretive to that NIM expansion there. And those deposit benefits are going to be partially offset by rate hikes given our modest liability sensitivity on the short end of the curve. But net-net, we think those components get us there.

I think it's important to remember, we're managing the balance sheet to be fairly neutral to rate moves here and the pressure on deposits from higher rates often carries with it better reinvestment rates on the fixed asset repricing. So there's a little bit of a trade in both directions there. And as I noted, the move up in rates, particularly over the last month or so has made achieving the target more mechanical than it was just a quarter ago. So hopefully, that provides you a little bit of detail.

Robert Siefers

Yes. That's perfect. And then maybe, Chris, one for you. So investment banking, great to see the better pull-through. I guess the change since last quarter is maybe there's been a little bit I guess, just this little air pocket in overall activity industry-wide. So maybe just sort of your updated thoughts. It was good to hear you reiterate the full year expectations for investment banking, but maybe sort of where are we? Where do you feel more and less confident than you may have felt, say, 90 days or so ago?

Christopher Gorman

Sure. So in this business, I never really, Scott, as you know, focus a whole lot on one quarter or the other. The business just broadly has a lot of trajectory. We've -- as you know, it's a business that's a mature business that we've had for some time. We have a public goal of getting to $1 billion a year by 2028. We're very comfortable that we can hit our mid-single-digit number. Pipelines continue to be strong. The pipelines in spite of the pull-through are up 5% on a linked-quarter basis and 20% year-over-year.

Now the one piece of that, and I'm still comfortable with our guidance, by the way, the one piece of that, that we have to watch is the real estate piece because transactional real estate, the cost of capital is such a significant part of those deals. Every piece of transactional real estate right now is getting retraded. I personally think that we will get through this because I think once people come to grips with the fact that we're in a higher for longer environment, I think people will transact. But that's the one wildcard. But I feel -- as you know, I feel good about the business. We've been building it for a long time, and we've got a bunch of great clients.

Operator

Our next question is from Ebrahim Poonawala from Bank of America.

Ebrahim Poonawala

Maybe just Chris, following up on, I think you mentioned power utilities driving loan growth contributing meaningfully. It sounds like that is sort of a second derivative of when we think about the whole AI CapEx and complete build-out. Maybe talk to us around when you like your visibility into the amount of growth that's been driven AI-related, second, third derivative versus non. And what I'm trying to get to is, and I'd love your perspective if we see a stalling in AI financing in the markets and things slow down, does it meaningfully then slow down when you think about just customer activity across the board for Key and your peers?

Christopher Gorman

That's a great question. There's a lot in that. Let me kind of -- let me unpack that a little bit. The first part of the question is secondary and tertiary derivatives of the AI build-out. I would contend that the investment cycle for these publicly owned utilities is one that's literally decades long, and we've been efficient in the amount of power that we've had. So I don't think that stops. Is part of it driven -- on the margin driven by AI? Yes. I mean, it's obviously, those are big consumers of power.

With respect to our book of business, as it relates to data centers specifically, we have less than $1 billion of exposure. And that exposure is directly to the hyperscalers as opposed to directly to people just building out the data centers. I do think there are long tentacles to the build-out. And I think if you think about, for example, trying to buy a diesel generator. Right now, there's a waiting -- you'll have to wait 24 or 36 months in order to get one. So I do think the tentacles of the build-out are fairly long it's hard for me to calculate for you exactly what the primary, secondary and tertiary factors are there. But I do think -- I think the tentacles are long.

Clark Khayat

First, I think that's well said. And maybe a couple of other points from a risk management perspective. One, we do have internal sector limits. So utilities would, of course, fall into that. These relationships typically span over 3 decades. So again, we've been very careful with clients that are tried and true. We're not really focused on merchant power generation. So again, we're looking for contracted power. And again, to Chris' point, really, there's a need for electrons regardless of the AI build-out. Things like our largest sponsors are less than of our portfolio, again, take-or-pay contracts and the like. So again, we feel good about the exposure. And again, just a reminder, there are internal limits.

On the AI front, too, we did do a bit of an exercise relative to what sectors could be, what I call adjacent. So we did kind of carve out what could be higher impact, what could be kind of neutral. And then what are some areas to watch. And again, in terms of high impact on what I call AI adjacency outside of the data center point that Chris made, that number is very small, less than $1 billion. So potential disruption there in business services. So we are looking at it from multiple angles, and we do feel comfortable with the exposure.

Ebrahim Poonawala

That's helpful. And maybe just a separate. I know you talked about capital markets looking into over the fourth quarter. But just in terms of momentum in capital markets, if there are investments you've made that should help as we look into next year? And maybe, Chris, give us a mark-to-market on sponsor activity. I think you mentioned about 40% of the business is exposed to sponsors are higher rates, more sort of geopolitical headlines is all that again serving as a wet blanket on activity there?

Christopher Gorman

Yes, I do think that mid-market, which as you correctly point out, sponsors are a big part of is yet to normalize. I'll give you a statistic deals under $1 billion this year are down 11%, whereas deals over $1 billion are up 12%. However, as you well know, there's an inverse relationship between your cash-on-cash return and the holding period. I think I'm optimistic that those deals will start to come out. And we're obviously -- when we talk about our pipeline, these are where we have a signed engagement letter and people are spending a lot of time on it.

So I do think -- and I think once -- obviously, when you have dislocations in the market like we had in the debt market over the last month, that obviously puts things on hold. But once people come to grips that we're in a hire for longer, pick your tenure, 5%, 6%. I think people will transact there. So yes, it's still depressed. Yes, it's in the backlog. And yes, I think it will start to free up.

Clark Khayat

And maybe just to respond to your question about investments Ebrahim, just a reminder that we've added a fair amount of bankers over the last 2 years, most to hit the ground running pretty well. And then obviously, we closed Clearwater here in the quarter. So we're continuing to build capability by market and industry.

Operator

Our next question is from Mike Mayo from Wells Fargo.

Michael Mayo

Just more on the commercial loan growth, I'm trying to separate what the industry tailwind from what's the drawdown, which a lot of banks are not seeing those loan utilization improve and kind of new business as a result of your hiring. So I see like kind of 3 different parts of that. I know you hired a lot of commercial bankers, and remind us how much you hired, what you're still hiring and how much they contributed to the loan and deposit growth.

Christopher Gorman

Well, there's a lot in that, Mike. Let me start with the final part of the question. So last year, we increased our number of bankers on the ground by around 10%. And we monitor very closely their ramp up. We've always said that it takes 12 to 18 months for bankers to be successful. We were very pleased in that the cohort of middle market bankers that we hired last year, 90% of them brought in a piece of business that doesn't mean they were profitable as a higher, but probably in a piece of business in the first quarter. So we're pleased with that trajectory. And we -- as you know, we've hired groups of people very successfully and integrated them.

With respect to kind of bifurcating what is utilization and what is other loan growth, if you look at the end of period, $2 billion of our $3.3 billion in loan growth was utilization. And I think that utilization plays out in a few areas. One is, as I've said for some time, I expected people to start to go a little longer on inventory in the environment that we're in. I think we're starting to see that. I think some of the utilization was actually people actually buying smaller businesses, and you can see strategic moves there. And also, we've talked about ever since the big beautiful bill in July of 2025 that we expected people to be investing more in CapEx. So those are some of the points.

Now as far as Key being an outlier in terms of utilization, a portion of that might rate is probably below significantly below probably many of our peers. And the reason we're able to do that is because we can lend to investment-grade companies because we can do so much, so many other things for them, payments, hedging, advice, et cetera. So that's just a bit of the texture of kind of the puts and takes around utilization.

Michael Mayo

Do you -- is that an area that you're focused on, that you're worried about? That's a growth area? Maybe is it just the early days of that whole process? Or how do you kind of back that in with your experience in wholesale banking for several decades?

Christopher Gorman

Sure. So it's not an area that we're focused on, Mike. As I mentioned, we have about $1 billion of data center exposure, but that data center exposure is really to the hyperscalers. It's not necessarily to the people building the data centers per se. My sense of it is, having done this for a while, anytime you see growth, it's such a pace I think it's time to be a little cautious. And so we're cautious in terms of lending directly into the data centers at this point.

Operator

Our next question is from Chris McGratty from KBW.

Christopher McGratty

Clark, on the capital ratios. The -- obviously, the move in rates adjusted ratios a bit this quarter. Could you just speak about appetite for buyback? I know I hear you in the fourth quarter with the $1.3 billion. But as we contemplate Basel III with great uncertainty, and improving growth. How do we think about the toggle between the 3 over the medium term?

Clark Khayat

Yes. Thanks, Chris. So first of all, I'd say we feel very good about our current capital position and our ability to manage that. I think with [ Mark Capital ], which we've been talking about now for several years, it's really half the equation and we haven't been including the benefit of the Basel III changes, which I think we all feel are highly probable to arrive early in '27 with implementation of '28. So as a reminder, when those come in on the risk-weighted asset reduction, we get roughly 100 basis points of additional benefit to the capital ratio in that point puts us squarely back into our range.

So we'd expect not be overly concerned about the current position. We've talked about being willing to be below that 9.5% to 10% range on a limited basis. And I wouldn't describe that as kind of spending in advance of that Basel III, but I think we'd be shortsighted not to include it given the timing and probability of that materializing. So I think that's sort of the broad base we did reiterate, as you noted, the fourth quarter buybacks to get to the [ 13% ] for the year. I think we'll have more conversation in January about how to think about the go forward, but I would really consider buybacks as Chris went through the capital priority stack as managing sort of at the end of the discussion around capital ratios and capital return.

So I don't know that we'll give the same kind of clarity on expectations for buybacks going forward post '26. We are doing a few things, I think, in the AFS to HTM realm that will insulate us a bit from further moves up and happy to impact that as well. But I think you can expect us we've been a low user of HTM historically for a variety of reasons. I think some of those are changing. So you can see us -- you will see us move up the HTM percentage of our portfolio. But again, I think the capital priorities are pretty clear, and the buybacks to us are sort of the last decision we're making in that process.

Christopher McGratty

Okay. Great. And then on the operating leverage front, I hear you on the investments in AI and what you're doing there. Any notable expense projects over the medium term that would interrupt the momentum you have on the operating leverage, Chris, I heard you talk about next year being another good year for PPNR and revenue growth. But anything on the expense horizon that we should be mindful?

Christopher Gorman

No. We have -- we're going to continue to invest in people and in technology. Our technology budget has gone from $800 million to $900 million to $1 billion. But obviously, we're generating a lot of capital, and there's nothing -- there's no huge expense out there, Chris, on the horizon. We're going to keep focusing on organic growth and keep -- continue to build the business brick by brick.

Clark Khayat

I do think the one reminder is just the full year impact of Clearwater, which we talked a little bit about recently, that will hit, obviously, for year-over-year or so it will look a little bit higher than otherwise would be planned, but we don't see any reason why kind of post '27, we couldn't be managing to sort of 3% annual expense growth.

Operator

Our next question is from Erika Najarian from UBS.

L. Erika Penala

Chris, this is struck -- I was just strucked by something you said in prepared remarks. You mentioned another outsized revenue growth year in '27 and fully appreciate you're in the middle budget season. But if I just unpack the components, you're looking for 18 basis points of NIM expansion 4Q '25 to 4Q '26. Your guidance implies 25 basis points of margin expansion from 4Q '26 to 4Q '27. And so if I think about it, and you're looking at 10% NII growth for '26, is that the right ballpark to think about '27? Or is there sort of an asset growth slowdown that we should consider?

Christopher Gorman

A couple of things. One, you're kind of ask the question, but we'll obviously provide guidance for 2027 at the end of the year. However, I think Clark mentioned something that I think is really important, and that is of the increase in margin, 70% of it is now mechanical based on the $24 billion repricing in the last quarter of this year and next year. And obviously, we've got a lot of momentum in our business, and we're going to continue to grow the business. And so we'll have more to say on specific targets after the fourth quarter.

Clark Khayat

But I think, Erika, you're picking up on something we've been talking about and the industry has been talking about, which is the trade-off between good quality loan growth and NIM. And obviously, we've made, as others have some trade-offs to go into what we believe will be very accretive long-term client relationships at the expense of NIM, but not at the expense of returns, EPS or NII. I think your comment about asset growth slow down is probably the right place to be focused and that trade-off is something that likely continues to present itself. But with the visibility we have now, we think we can manage that combination.

L. Erika Penala

So just to follow up on that. Obviously, deposit growth has been impressive and there's a seasonal aspect. As we think about slowing earning asset growth, how should we think about year-over-year deposit growth progression? In other words, you hired a bunch of bankers. You clearly made the investments. And so how should we think about the progress year-over-year on deposit growth sort of eliminating for seasonality?

Clark Khayat

Yes. I think the way to consider that is when you make a loan to a client. We have a theory here that we've espoused for the last decade or so, which we referred to as kind of payments attachment, which means when you make a loan the expectation to the bankers, you're going to get the deposit, you're going to get the payments with it. I think we do a very good job with that, but it doesn't necessarily happen on day 1. It takes a little bit of time to move accounts to get things changed. So some of the third quarter growth you would have seen would have been off new clients in the first half. And it's hard to predict exactly how long that time frame takes.

But even if we had a little bit of new loan slowdown in '27. I think you'd still see some pull-through from the '26 production in terms of commercial deposits.

Operator

Our next question is from John Pancari from Evercore ISI.

John Pancari

Just back to the loan demand side, I mean you saw the pretty good step up in utilization in the quarter, but you also cited the expected moderation in loan growth in the fourth quarter. Can you just elaborate a little bit more on what areas that you expect that moderation? And then also, are you beginning to see any erosion around loan demand at all amid the rise in rates? And is that what's influencing your expectation on that front in any way?

Christopher Gorman

I would say we're not seeing an erosion in loan demand with the notable exception John, that I made earlier, and that's on transactional real estate. I think some of those deals are clearly slowing down. No, we're seeing -- look, there's significant loan demand out there, and we're out talking to our clients. Loan demand is -- it's never -- it doesn't grow evenly quarter-to-quarter. But we feel good that there's a lot of demand out there. And keep in mind, we're doing this at a time when we're distributing about 70% or 82% of the capital that we're raising. So there's a fair amount of activity out there, and we're being selective, but we've got a lot of opportunity.

Clark Khayat

And John, we guided, just to remind you of earning asset growth in the fourth quarter of about $1.5 billion to $2 billion just relative to where we were in the third quarter, that's a little bit of a step-down. And some of that, I think, to Chris' point, will get placed in the market.

John Pancari

Okay. No, that helps. And then on the deposit side, I know growth came in better than your expectation and you're confident in the ability to fund upcoming loan growth with deposits. How should we think about the mix and noninterest-bearing component of that? And on the beta front, I know you're looking for the mid-50s over time, how does that play out in the near term in terms of deposit cost progression as we look at the next few quarters from here, I know you're up about 3 bps in the third quarter.

Clark Khayat

Sure. I think, look, deposit competition or composition, sorry, on the commercial side, I think it's going to look more interest-bearing to noninterest-bearing from a reported standpoint, just given our use of hybrids, and we continue to use that as a vehicle to sort of capture a full client relationship. But I think we'll see primary growth in operating accounts, which is allowing us to manage the overall expense. So I think that commercial loan growth will come on kind of consistent with the portfolio rate we see today, obviously subject to indexed changes from rate hikes. So we feel like it will be a representative sample there broadly.

On the consumer side, we have started to pick some places to raise some rates and do some targeted deposit raising, which we've had some early success on and we're doing that to raise some balances but not more broadly raised the overall consumer deposit rate. So we feel pretty good about our early returns there. And based on how that materializes, I think we'll have the opportunity to do a little more of that. But we also still really aren't at the full offer to some of the competition, but we have picked up our MMDA promo and CD rates in certain markets in targeted ways in the last quarter. And again, we've seen some positive response to that.

So I think it's not going to be all commercial, but we'll start to see a little bit of consumer growth come on that we think will be valuable. And that commercial growth, we think, will be kind of representative of what it looks like today on a portfolio basis.

Operator

Our next question is from David Chiaverini from Jefferies.

David Chiaverini

You mentioned about being neutral to interest rates and the forward curve. Does the path to $325 million -- you mentioned about 70% to 75% of it is mechanical. Does that mean with being rate kind of neutral, does that mean loan growth is the main swing factor? Can you unpack that a bit?

Clark Khayat

Sure. I think it's really loan growth and deposit growth and the cost of the deposit growth. So I fail to fully answer the last question. I do think we'll see kind of low 40s beta here in the fourth quarter, and that over the next -- probably first or second quarter of '27 would start to move up to mid-50s obviously depends on if there are not additional rate hikes in the year. But we'd expect over sort of 2 to 3 quarters to get to that 50 to mid-50s number. If that were to be more loan growth, or less deposit growth or more expensive deposit growth, I think that is the combination of things that we're watching really closely.

But the visibility we have now makes us feel like we can fund that loan growth with deposits, we can potentially pay down some wholesale funding with deposits and we can effectively manage the cost of those deposits within the scope of where rates are. And there's that natural trade-off, as I mentioned, between higher rates on the front end, delivering better reinvestment rates in the counter if rates go down. So we think there's a little bit of a natural hedge in that neutrality.

David Chiaverini

Very helpful. And my follow-up, in your prepared remarks, you mentioned about with the AI use cases, you mentioned 4 pillars. I was interested in the comment about commercial originations as being part of that AI use case. Normally, when I think of commercial banks, it's relationship banking and somewhat insulated from AI. Can you walk through how you're using AI for commercial originations?

Clark Khayat

Yes. If you think about you think about particularly complex commercial relationships, they tend to be -- there's a lot of decision-making interaction and work that goes into that process. And we just think there's opportunities to streamline that to make them more informed to improve the borrower experience and the banker experience in something that historically been has been maybe a little bit more analog in nature because they tend to be, to your point, person-to-person relationships.

We just think we can build a chassis underneath that, that is AI-enabled and streamlined. And then that is not at the expense of the relationship. It is to augment the relationship. So I think it's to reduce paper and a lot of back and forth and make that process more streamlined.

Christopher Gorman

So David, for example, you asked about commercial loans, but if you think about commercial onboarding and servicing, we completely reengineered that process over the last 3 years which now opens it up to use AI. And our goal there is to have twice throughput and half the time, and frankly, have it be a better client experience and a better experience for our employees. That's kind of how we think about it. And these 4 domains that we're talking about are enterprise-wide. So it involves reengineering the whole process and then using AI as an enabler.

Operator

Our next question is from Gerard Cassidy from RBC.

Gerard Cassidy

Coming back to the investment banking comments about sponsor activity, and we all recognize that the uncertainty in the economy and the geopolitical risk, the political risk in this country. Those are constants. We hear about them for years. So putting them aside for a moment, what do you -- what has to happen to get that activity in the middle market M&A to really pick up? Or is it the companies themselves that are in these portfolios are just not worth what the sponsors think they're worth and they may never come back?

Christopher Gorman

Well, I think part of it is, I'm sure there's some element of sort of bid ask, what is the value, what's the exit strategy, what's the timing? I also think I personally think one of the big constraints has been not only the volatility in interest rates, but the volatility in the forwards of interest rates. If, for example, you thought interest rates if you thought there were going to be 5 cuts. And in fact, there were no cuts, and there were -- there was an increase. I think that kind of puts people on the sideline.

I think once people have a clear view of where rates are or where they think they're going to be over a period of time, Gerard. I think that frees it up. Another thing that has changed a bit, though, are these continuation funds. And so some of the algorithms as to years holding period. I think some of those have changed with some of the continuation funds. Having said that, there is a very significant backlog out there that at some point is going to have to clear.

Gerard Cassidy

Very good. And I might regret asking this question. I asked that on one of your competitors' calls. But can you guys give us some insights on credit quality? Has there been a structural change in the way the industry and KeyCorp in particular is underwriting credit today because kind of it's so resilient, and you guys have obviously some meaningful improvement this quarter. But across the board, the banking initiatives credit was very strong now.

We're certainly not naive enough to suggest that in a down economic cycle, credit is not going to get worse. We know that. But it just seems that things are really quite good. And has the industry been more conservative or the financial prices driving the regulators to derisk the industry? What are your guys thoughts on that?

Christopher Gorman

So I'll give you some perspectives on what I think drives it industry-wide and then some things that I think are idiosyncratic to us. I think in the global financial crisis, there were a lot of lessons learned. And so I think if you think about here at Key, where we built the underwrite to distribute model, including in real estate. That was a big fundamental change for our risk profile. So that's the lessons learned, the low financial crisis. I think the advent of the private credit market were in the private credit market really dominated the small leverage deals for some time, that's what a $3 trillion market.

I think that would have been in the banking system, and it's not in the banking system. So I think those are 2 kind of fundamental things, Gerard. The other thing I would say that's idiosyncratic, the key is on our consumer side, as you know, we don't have a big card portfolio, and we have a super pont portfolio at funding. Our average FICO score is 766. And and currently, it's 787. So on the consumer side, we have a super prime book. And then lastly, which I think is a huge advantage that is idiosyncratic to key. Is this notion of going to market by industry groups I think is really, really helpful because I think it enables you to decipher and discern among the people in the industry. So I think you put all that together, I think those are the things that have changed.

Clark Khayat

I might add one just data point, Gerard. If you go back to when I joined Key in 2012, I think our leverage book was about $2 billion I think today, it's $2 billion. We're obviously a much larger bank. So that's not where I think the banks broadly have seen the growth. That growth is happening. It's just not happening, I think, in some of the traditional banking spaces.

Unknown Executive

Gerard, great question. I think the only [indiscernible] other thing I would add is I think, again, from a supervisory perspective, the financial risks continue to be a major focus for them. So again, tying our stress testing to our underwriting to our portfolio limits and having that all be one ecosystem. We continue to add to that sophistication as well as, again, connecting adapts with data across the enterprise as well. So we just continue to be expected to trend the bar, and we're seeing -- we're also just continue to uplift these programs.

Operator

Our next question is from Matthew O'Connor from Deutsche Bank.

Matthew O'Connor

You guys are obviously focused on growing the commercial side of the loan book and the results there have been really good. Just remind us on the consumer side, you've been running down there? And where are those balances kind of level out? And are there any areas that you're kind of leaving into or would like to be better at over time in the consumer lending side?

Christopher Gorman

We would have said 2 quarters ago that the quarterly runoff would have been about $600 million given where rates have gone, the current runoff is about $400 million. And what we're doing is we're basically redeploying out of mortgages that are very, very safe mortgages because they're made to their jumbo mortgages to first mortgages to doctors and dentists, but we're redeploying that in our C&I portfolio, where we have a really good track record of having a complete relationship and driving a bunch of revenue. We're going to continue to do that.

The one area that we've tasked our consumer folks with doing is given some of the metrics I just gave about our consumer customer. Many of our customers have significant equity in their homes. And what we're doing is we're building a digital interface that's as good as for home mortgage -- I'm sorry, for home equity, that's as good as our home equity -- our home mortgage app. And so that will be an area that you'll see grow in 2027.

Matthew O'Connor

Okay. That's helpful. And then just one clear question on the fees. The corporate services income was down Q2 and flat year-over-year. Just remind us kind of what -- well, one, what drove that? And then the interplay of that and say loan growth like as loans grow, is there like a contra there? Or what that relationship is, if there is one?

Clark Khayat

Yes. Thanks, Matt. Generally, you would see corporate services increase with loan growth. So what's in there would be foreign exchange fees, derivative fees and some loan fee categories. In the quarter, we saw derivatives or client hedging fees come down, and I think that's really just reflective of the rate volatility. So we might see that come back. I think fourth quarter, we'd expect it to be pretty flat to fourth quarter of 25%, but we did see a dip there in direct client activity of lending and again, just driven by sort of an unwillingness of clients to lock in rates with the current volatility. And I would say that's sort of flip side of our spot and forward starting swaps are coming on at pretty attractive rates and clients aren't wanting to take the other side of that in some cases.

Christopher Gorman

Yes. And the other piece of it, of course, we do a lot of hedging around oil and gas. And as you can imagine, there's not a huge -- a lot of activity in terms of people hedging oil and gas right now.

Operator

Our next question is from Ken Usdin from Autonomous Research.

Kenneth Usdin

Clark, just wanted to ask you, you talked about being slightly liability sensitive to the front end of the curve. And I wonder if you can kind of talk us through that and incremental swap adds you've made, does that get you to the point where you want to be in terms of balance? And how does that affect the future rate sensitivity when you consider all of that you added this quarter and the forwards that are still coming on?

Clark Khayat

Yes. So the short-end rate sensitivity or liability sensitivity, if you -- really, when we talk about that, is if you think we're getting to the kind of mid-50s terminal betas that we talked about, the breakeven is kind of low to mid-40s. So we're going to be a little bit liability sensitive on the front end of that deposit data. But again, as we noted, assuming no material change in the shape of the curve, we would assume we get some of that back on the reinvestment rate. I think we feel, again, quite good about the neutral position we're in.

I think the steeper the curve is, the better that is for us, the flatter it is, obviously, is the inverse. And we've effectively been replacing swaps at this point as they roll off and as we noted, replacing them at much higher levels right now. And we've done a little bit of maintenance sort of out into '28 and '29, where we saw some portfolio gaps on the swap side. But generally speaking, I think we're pretty much in maintenance mode and feel good about maintaining that neutrality.

Operator

That concludes our Q&A session. I would now like to pass the conference call over to CEO, Chris Gorman, for any closing remarks.

Christopher Gorman

Thank you. We appreciate your interest in Key. If you have any further questions, feel free to reach out to Troy Gates and the rest of our IR team. Thank you.

Operator

Thank you. This concludes today's call. Thank you for joining. You may now disconnect your lines.