← back to rankingKEY · KeyCorp
Banks - Regional · mkt cap $23.0B · calls: Q1 FY2026 vs Q4 FY2025
38.0 conviction · conf-adj 35
conf 3/10 partial
enthusiasm:18.0 · trend:8 · quantifies:5 · impact:0 · under_radar:0 · credibility:0 · business_impact:4 · disruption:0 · commitment:0 · confirmation:3
Enthusiasm latest 6 / prev 5 (rising)
Key’s AI story shifted from generic efficiency and call-center wins (Q4) to named internal use cases plus a wealth/mass-affluent personalization pitch and explicit GenAI-linked utilities lending (Q1), with Clark now overseeing tech/ops. Management does not quantify AI-specific revenue, savings, or ROTCE—only broader tech spend—and Clark concedes ROI is hard to measure, with benefits framed as capacity and avoided future investment. Credibility is moderate on operations/risk themes, aspirational on wealth AI, and conservative on AI infrastructure credit risk.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $11.2B · net income $1.8B · net margin 16.3% · diluted EPS 1.52
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: 0.0% · next-FY EPS uplift: -6.48% · vs analysts: unclear · priced in: high (model's call-read: medium; verdict above is the hard-data one used for ranking) · confidence: 3/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
Total tech investment ~$1B (FY2026)
cost | ~$1.0B | Investment OUTFLOW, not a quantified return. $1,000M / $11,188M = 8.9% of revenue spent on tech (spend, not uplift). Incremental vs prior-year tech/ops midpoint $850M (mid of disclosed $800–900M) = $150M. No revenue gain or cost saving attached. If fully expensed near-term: after-tax drag = $150M × (1−0.21) = $118.5M; eps drag = 100 × (−$118.5M) / $1,829M = −6.48%. Sits inside guided opex; overlaps claim 2 — counted once in aggregate. | 0 | -6.48 |
Tech & ops spend $800–900M last yr -> $1B this yr
cost | $800–900M -> $1B | Incremental SPEND ~$100–200M (mid $150M), not a saving. After-tax @21%: ~$150M × 0.79 = ~$118.5M = −6.48% of $1,829M NI as a DRAG (range −4.32% at $100M to −8.64% at $200M). No offsetting AI saving quantified, so no positive EPS effect bookable. Same $150M as claim 1 — not double-counted in aggregate. | 0 | -6.48 |
Additional $100M customer-facing tech (FY2025)
cost | $100M | $100M / $11,188M = 0.89% of revenue. Spend aimed at engagement/topline ('easier to bank at Key'), but NO revenue figure, take-rate, or conversion disclosed → cannot size topline. Already embedded in FY2025 actual base (rev $11,188M, NI $1,829M), so FY2026 forward increment from this line = $0 → rev 0, eps 0. | 0 | 0 |
Mass affluent: 1.15M households, <10% penetrated
engagement · soft | 1.15M customers; <10% penetrated | >90% runway, but no revenue-per-household, fee level, or take-rate disclosed and AI's role is qualitative ('more to say'). 1.15M × (unknown $/hh) = unsizable. Real opportunity, zero dollar anchor. | | |
AI/data-center & AI-adjacent credit exposure de minimis
other · soft | de minimis; no $ disclosed | Credit-RISK context (loan book), not a revenue/earnings driver. Explicitly no dollar figure. Reassurance, not uplift. The utilities/power-buildout lending remark is supplier-adjacent but is a forward statement with no number. | | |
Assumptions: FY base: revenue $11,188M, net income $1,829M, diluted EPS $1.52, net margin 16.35%. Incremental tech spend uses disclosed $800–900M → $1B trajectory; midpoint prior $850M → $150M incremental FY2026 (single-counted in aggregate). Tax 21% on cost items. Investment treated as near-term fully expensed opex (no capitalization disclosed). No incremental AI revenue or quantified saving from management → incremental margin not applied; would default to 16.3% if any were disclosed. $100M FY2025 customer-facing spend embedded in actual base → FY2026 increment $0. EPS sized vs adjusted/consensus basis (~$1.48 consensus, $1.52 reported — aligned); denominator positive, not thin. Forward qualitative statements (underwriting savings, margin expansion, utilities lending) excluded — no $ anchor.
Top line: No quantifiable AI topline uplift for FY2026: no $ revenue, bookings, or NII target tied to AI. Spend intensity only — FY2026 $1B tech ≈ 8.9% of $11,188M revenue. The customer-facing $100M and 1.15M mass-affluent (<10% penetrated) opportunity are genuine revenue intents but carry no disclosed revenue figure, take-rate, or revenue-per-household, so they cannot be sized without inventing numbers. Management explicitly deferred quantification.
Bottom line: Every hard figure ($1.0B tech spend, $800–900M→$1B trajectory, $100M incremental) is INVESTMENT, not a saving or return. The only anchored bottom-line effect is a near-term cost drag: ~$150M incremental tech spend → ~$118.5M after-tax → −6.5% of $1,829M NI (−6.48% EPS) if unoffset and sitting inside guided opex. The $100M FY2025 line is already in the base (0% forward). Offsetting productivity/cost-avoidance benefits (underwriting, automation, margin expansion) are stated as future and unsized — sizing an uplift off them would mean fabricating a return.
The ~$1B tech spend is part of guided opex consensus is built on, so it is priced. There is no disclosed AI BENEFIT number sitting above the consensus EPS trajectory (0.77→1.02→1.48) to be either priced-in or missed — the upside is real-but-unquantified, not above-consensus-by-the-math, and no numeric gap can be demonstrated either way (hence medium, not high or low). The disclosed ~$150M incremental spend (−6.5% EPS if fully unoffset) is modest vs $1,829M NI and likely embedded in guidance; reported FY2025 EPS $1.52 vs consensus $1.48 (+2.8%). Hard claims are spend/runway, not upside.
MODEL CONSENSUS (impact)
partial
Agree: no quantified AI revenue/savings; only spend. Net is a ~-6.5% EPS drag, not uplift. Reconciled X's anchored arithmetic with Y's conservative priced-in/confidence.
Conflicts reconciled
- est_eps_uplift_pct: X=-6.48 vs Y=null -> used -6.48 because both derived the same ~6.5% after-tax spend drag; recorded the anchored figure
- soft (spend claims): X=false vs Y=true -> used false because the $ figures are disclosed/anchored (soft means unanchored, not 'no benefit base')
- est_rev_uplift_pct: X=0 vs Y=null -> used 0 because no AI revenue claim was quantified = zero hard uplift
- priced_in: X=high vs Y=medium -> used medium because no AI benefit number sits above guided-opex consensus to be priced or missed
- confidence: X=5 vs Y=3 -> used 3, lowered on conflict
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | – | – |
| EPS uplift % | – | – |
| Priced in | medium | – |
| vs analysts | unclear | – |
| Confidence | 3 | – |
| Top line | No quantifiable AI topline uplift for FY2026. The customer-facing $100M and the 1.15M mass-affluent (<10% penetrated) opportunity are genuine revenue intents, but neither carries a disclosed revenue figure, take-rate, or revenue-per-household — so they cannot be converted to a % of the $11.188B base without inventing numbers. Management explicitly deferred quantification. | – |
| Bottom line | No quantifiable AI EPS uplift. Every hard figure ($1.0B tech spend, $800-900M->$1B trajectory, $100M incremental) is INVESTMENT, not a saving or return. Modeled as spend they are EPS headwinds (e.g. the $100M = ~$79M after-tax = ~4.3% of $1.829B NI), and the offsetting productivity/'money-saving' benefits (loan underwriting, process automation) are stated as future, unsized. Sizing an uplift off these would mean fabricating a return. | – |
| Reasoning | Consensus EPS already steps 0.77 (FY23) -> 1.02 (FY24) -> 1.48 (FY25), and the ~$1B tech spend is part of guided opex consensus is built on, so the spend is priced. There is no disclosed AI benefit number sitting above that trajectory to be either priced-in or missed — the upside is real-but-unquantified, not above-consensus-by-the-math. Hence medium, not low: no numeric gap can be demonstrated either way. | – |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
Total technology investment (FY2026): approximately $1 billion (2026, bottomline)
“Lastly, we are investing approximately $1 billion in technology this year that will give us new product and service capabilities and deliver better outcomes and experiences for those we serve.”
Technology and operations spend trajectory: $800 million to $900 million last year to $1 billion this year (2025–2026, bottomline)
“Our investments in tech and ops have gone from $800 million to $900 million last year to $1 billion this year.”
Incremental technology investment (customer-facing): $100 million (2025 full year, bottomline)
“We invested an additional $100 million in technology, focused on customer-facing capabilities that make it easier for our clients to bank at Key.”
Mass affluent household opportunity (AI penetration context): 1.15 million customers; less than 10% penetrated (as of March 31, 2026, topline)
“With a mass affluent household opportunity of 1.15 million customers, we remain less than 10% penetrated, implying a significant runway going forward.”
AI/data-center credit exposure: fairly de minimis (data center); no dollar AI-adjacent figure disclosed (Q1 2026, both)
“our data center exposure is fairly de minimis. We have also looked at what we have called AI-adjacent type exposure and worked with our board on that as well.”
PAST (realized)
- Q4 FY2025 — Chris Gorman: We've done it in certain areas, like our call centers, in certain areas like internal things.
- Q4 FY2025 — Chris Gorman: I think we're doing well with respect to implementing AI, but there's a lot more that we can do.
- Q1 FY2026 — Clark Khayat: I would say where we—and I think others—are seeing benefits is in efficiency and capacity, but it is really showing up more in avoidance of future investments.
- Q1 FY2026 — Chris Gorman: We are already investing heavily in our wealth platforms.
CURRENT (now)
- Q1 FY2026 — Chris Gorman: As it pertains to AI, we are focused on a few thematic use cases that will enhance client experiences, accelerate credit decisioning, increase technology productivity, and strengthen risk and security monitoring.
- Q1 FY2026 — Chris Gorman: Clark has assumed an expanded role to lead our technology and operations organization… as we leverage AI to grow our business and better serve our clients.
- Q1 FY2026 — Clark Khayat: our data center exposure is fairly de minimis… AI-adjacent type exposure… very, very well controlled and monitored.
- Q4 FY2025 — Chris Gorman: We're investing heavily in AI and technology… tech and ops… $1 billion this year.
FORWARD (guidance)
- Q1 FY2026 — Chris Gorman: We will have more to say on that in the future (AI for mass affluent).
- Q1 FY2026 — Clark Khayat: If we can do that, then you start to see the scale of the platform and the benefit of that cost avoidance in a real way. Then we can come back and say, we spent these dollars, we created these improved processes, and they drove this level of margin expansion.
- Q4 FY2025 — Chris Gorman: opportunities to really rethink our entire business. For example, loan underwriting and processing… apply a lot of technology. It'll be money-saving.
- Q1 FY2026 — Chris Gorman: Utilities and power… massive build-out that is required for GenAI (loan-growth opportunity).
TRACK RECORD — PROMISE vs DELIVERY
—/100 (no quantified promises) no-quantified-promises 6 calls reviewed
Across six KeyCorp earnings calls (Q4 FY2024–Q1 FY2026), management discussed technology spend, cloud migration, digital/analytics capabilities, and—in Q1 FY2026—thematic AI use cases (credit decisioning, productivity, risk monitoring), but never issued a quantified AI/automation promise with both a numeric target and a timeframe or milestone. Credibility on AI delivery cannot be scored from this transcript set.
PRICED-IN (REFINED)
HIGH (already in)Est. revisions rising · Fwd P/E 27.5 · EV/Sales 3.5x
AI claim maps to Cards And Payments, Trust And Investment Services, Service Charges On Deposit Accounts
Price targets step up clearly (lastMonthAvg 27 > lastQuarterAvg 25 > lastYearAvg 22.95) and forward EPS jumps from ~$1.02 (FY2024) to ~$1.48 (FY2025), so consensus is already baking in strong recovery/growth even as buy/strongBuy counts have edged down recently. At 27.5x next-FY EPS and ~3.5x EV/Sales, KEY trades rich versus typical regional banks, implying much of the efficiency/earnings-upside narrative is in the multiple. AI-driven gains would most plausibly flow through fee businesses like Cards And Payments and Trust And Investment Services, where the market is already paying up for higher future earnings.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
3Q4 FY20242Q1 FY20253Q2 FY20253Q3 FY20253Q4 FY20255Q1 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
Five quarters of cloud and tech spend only; Q1 FY2026 first named AI themes without metrics or products.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
5/10 qualitative impact moderate medium-term · mixed evidence
Where AI matters: ops/credit automation and wealth cross-sell
KEY is a real adopter—call-center automation, themed credit/risk/productivity use cases, and ~$1B tech spend—but management admits AI ROI is unmeasurable and offers no AI-linked revenue or savings; mass-affluent AI is aspirational with 1.15M households and <10% penetration but no dollar take-rate.
Caveats: ~$150M incremental tech spend is a near-term EPS drag with no offsetting AI savings disclosed; Utilities/data-center lending is AI-buildout exposure, not own-business AI upside; Wealth/mass-affluent AI remains narrative until monetization and ROI are quantified
AI DISRUPTION / CANNIBALIZATION RISK tailwind · 3/10
GenAI pressures commoditized fees and digitized wealth advice but does not automate away relationship commercial lending, spread NII, or trust/AUM economics; KEY’s push is mainly defensive cost/credit capacity, not billable-hours deflation.
OPTIONS / MARKET STRUCTURE
option liquidity: good
proxy inputs — dollar-ADV $286M · beta 1.055 · px $21.18
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.
CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions adding, management language 4/10 measured.
INSIDERS selling 1 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) adding as of 2026-03-31: 117 new / 101 closed positions; 463 increased / 269 reduced; institutional ownership -2.46pp; +14 net 13F holders
MGMT LANGUAGE 4/10 measured AI barely discussed; brief thematic focus and future tense, no AI metrics, timelines, or delivered results.
commit “As it pertains to AI, we are focused on a few thematic use cases that will enhance client experiences, accelerate credit decisioning”
commit “as we leverage AI to grow our business and better serve our clients”
hedge “a few thematic use cases that will enhance client experiences, accelerate credit decisioning, increase technology productivity”
VERBATIM AI QUOTES
“As it pertains to AI, we are focused on a few thematic use cases that will enhance client experiences, accelerate credit decisioning, increase technology productivity, and strengthen risk and security monitoring.”
— Chris Gorman, Q1 FY2026
“Before turning it over to Clark, I am pleased to announce that Clark has assumed an expanded role to lead our technology and operations organization, in addition to his role as CFO. We look forward to the contributions he will bring to our technology and operations teams at a pivotal and exciting time as we leverage AI to grow our business and better serve our clients.”
— Chris Gorman, Q1 FY2026
“Lastly, we are investing approximately $1 billion in technology this year that will give us new product and service capabilities and deliver better outcomes and experiences for those we serve.”
— Chris Gorman, Q1 FY2026
“Utilities and power continue to be an area of huge opportunity when you think about both renewables and the massive build-out that is required for GenAI.”
— Chris Gorman, Q1 FY2026
“So it is a great question. The answer is the funding for these data center buildouts are in the capital markets and also at some of the banks. And, you know, we have been in the power business for a long time. And as a consequence, I think we do it pretty well. There are all kinds of nuances in these deals. Who pays for the cost overruns, for example, etcetera, etcetera. Who has the right to do what under a bunch of circumstances. We feel very good about the loans that we have. But these loans are both in the capital markets and in the banking system.”
— Chris Gorman, Q1 FY2026
“Yeah, and, Chris, I just might add that our data center exposure is fairly de minimis. We have also looked at what we have called AI-adjacent type exposure and worked with our board on that as well. And, again, very, very well controlled and monitored. We are really not chasing a lot—again, these larger projects or hyperscalers. And, again, it is very well managed.”
— Clark Khayat, Q1 FY2026
“That is a great question, and it is very timely because I spoke to our big producers in this business as recently as Tuesday morning at their sales conference. I think there is a huge opportunity to use AI. We are already investing heavily in our wealth platforms. And I think as you think about serving that many customers, there is huge opportunity for AI. We will have more to say on that in the future, but that is a perfect application. Many of these customers are rather homogeneous in their needs. And I think we are armed with perfect information because it is all running through the bank. And I think harvesting more detailed information so we can do a better job of serving these customers that already know and trust KeyCorp and have their money on some other platform where, as you can imagine, they are not getting incredible service—just because it used to be that if you had $5 million, you got incredible service everywhere. Now, as you know, the number is a lot higher. And so this is a huge opportunity for us.”
— Chris Gorman, Q1 FY2026
“Well, we would have to get to that first, and then share it because, as you know, Gerard, these are pretty hard to measure. I would say where we—and I think others—are seeing benefits is in efficiency and capacity, but it is really showing up more in avoidance of future investments. And so that is hard for me to come and say, "Hey, Gerard. I did not spend these dollars I may have otherwise spent." The way I think we really need to demonstrate that is to scale some of these platforms, which has been a theme of Chris's now for—I do not know—as long as I have known him. If we can do that, then you start to see the scale of the platform and the benefit of that cost avoidance in a real way. Then we can come back and say, we spent these dollars, we created these improved processes, and they drove this level of margin expansion.”
— Clark Khayat, Q1 FY2026
“We're investing heavily in AI and technology. Our investments in tech and ops have gone from $800 million to $900 million last year to $1 billion this year. I think we're doing well with respect to implementing AI, but there's a lot more that we can do. We've done it in certain areas, like our call centers, in certain areas like internal things. But there's opportunities to really rethink our entire business. For example, loan underwriting and processing. We can look at those whole horizontal areas and apply a lot of technology. It'll be money-saving. And, by the way, it'll give you a better experience for our clients.”
— Chris Gorman, Q4 FY2025
“We invested an additional $100 million in technology, focused on customer-facing capabilities that make it easier for our clients to bank at Key.”
— Chris Gorman, Q4 FY2025
ANALYST QUESTIONS ON AI
Q (Q1 FY2026, Gerard Cassidy): How can you guys embrace AI to penetrate that client base and make it even more profitable because you are using AI?
A: Chris Gorman: huge opportunity to use AI in mass affluent/wealth; already investing heavily in wealth platforms; homogeneous customer needs and bank-held data are a perfect application; will have more to say in the future. Clark Khayat (follow-up on measuring AI ROI): benefits are hard to measure; showing up in efficiency, capacity, and avoidance of future investments; proof will come from scaling platforms and eventual margin expansion, not yet disclosed.
Q (Q1 FY2026, Gerard Cassidy): Do you think we will ever get to the point where outsiders like folks on this call could actually measure, for your dollar of spending in AI, it actually incrementally led to a 50 basis point of ROTCE improvement?
A: Clark Khayat: would have to achieve and then share; hard to measure; benefits today are efficiency/capacity and cost-avoidance of future spend; need platform scale before tying spend to margin expansion.
Q (Q1 FY2026, Ebrahim Poonawala): When we think about the AI data center loans that are being made right now, is most of that distributed in the capital markets or syndicated to banks? Is there risk tied to this data center spending on bank balance sheets at KeyCorp and broadly?
A: Chris Gorman: funding is in capital markets and at some banks; Key has long been in power and feels good about its loans, with deal-structure nuances. Clark Khayat: data center exposure is fairly de minimis; reviewed AI-adjacent exposure with the board; very well controlled; not chasing larger hyperscaler projects.