BNY Mellon
The financial sector's clearest case of operating leverage alongside large-scale AI deployment, though management does not attribute the results to AI alone. Q2 2026, reported July 15, delivered record revenue of $5.70B (+13% YoY), diluted EPS of $2.45 (+27%), 606bps of positive operating leverage, a 39.8% pre-tax margin and 31.3% ROTCE, with AUC/A of $62.6T (+12%). Management raised 2026 revenue-growth guidance to about 10-11%, and headcount fell about 7% YoY while noninterest expense rose 7%. The CEO credits the new commercial and platform operating models; the CFO described AI spend as modest relative to a roughly $4B technology budget and the headcount decline as an output of everything rather than an AI target. Earlier disclosures put 218 AI solutions in production (Q1 2026), over 40% of new code AI-authored and the Eliza platform in daily use by about half of staff. Net interest income (+20%) and higher market values also drove the quarter.
Scenarios
Custody and asset servicing are process-heavy and repetitive, so agentic AI trained on proprietary servicing data compounds. Q2 showed revenue +13% with headcount down ~7% — if that ratio persists through a full rate cycle, the margin gain is structural rather than cyclical.
Part of Q2 came from NII (+20%) and market values, both cyclical, and management does not attribute the guidance raise to AI, so the AI share of operating leverage is asserted rather than isolated. Custody fee compression is structural, and if peers reach similar AI efficiency the gains flow to clients through pricing rather than to margins.
Key Factors to Watch
- ●Q2 2026 (Jul 15) — record revenue $5.70B (+13%), EPS $2.45 (+27%), 606bps operating leverage, 39.8% pre-tax margin, ROTCE 31.3%
- ●FY2026 revenue-growth guidance raised to ~10-11%; headcount down ~7% YoY (CFO, Jul 15)
- ●218 AI solutions in production (Q1 2026) and >40% of new code AI-authored — the most specific AI production metrics among large US banks
- ●NII +20% and higher market values drove part of Q2 — AI's contribution to operating leverage is not isolated in disclosures
Score History
| Date | Score | Direction | Note |
|---|---|---|---|
| 2026-10-10 | 7.2 | Positive | W41 rotation refresh — score holds 7.2, no dim change. Rebased on Q2 2026 release (Jul 15, not Jul 22 as previously logged) — revenue $5.70B +13%, EPS $2.45, 606bps operating leverage, AUC/A $62.6T, FY revenue guide ~10-11%, headcount -7%. Corrected framing — the guidance raise is not explicitly attributed to AI; CFO calls AI spend modest vs a ~$4B tech budget. No AI event after Aug 22. |
| 2026-08-22 | 7.2 | Positive | Score 7.0->7.2 (moat_durability 7->8). Q2'26 call (Jul 22) raised FY26 revenue-growth guidance from ~6% to 10-11% with a 400bps positive operating-leverage target, explicitly attributed to AI integration; ~40% of software now AI-written and headcount down 7% YoY alongside record revenue - AI operating leverage is now a demonstrated structural margin advantage, not pilot talk. |
| 2026-04-25 | 7.0 | Positive | Score 6.9→7.0 (aam 8→9). Q1 2026 disclosed 218 production AI solutions (4x YoY), >40% of new code AI-authored, 50% of annual account plans AI-drafted, 70% of restricted-party payment screening reviewed by AI. EPS +42% with 290bps positive operating leverage; NVIDIA SuperPOD on-prem added. |
| 2026-04-10 | 6.9 | Positive | Score 6.8→6.9 (rounding drift correction — dimensions unchanged, formula recomputed to match stored value) |
| 2026-03-08 | 6.8 | Positive | Score 7.0→6.8 (formula reweight: sb 0.25→0.15, are 0.20→0.25, md 0.20→0.25, dr 0.20→0.25, aam 0.15→0.10) |
| 2026-03-08 | 7.0 | Positive | Initial assessment from batch 9 blind spot review |
Finance Peers
Last researched: 2026-10-10
This is research and analysis, not financial advice. Scores reflect AI impact potential, not investment recommendations.