GS, JPM, MS: Wall Street AI Won't Replace Investment Bankers

Goldman Sachs CEO David Solomon said publicly this week that the Wall Street he runs is "not going to have an AI jobs apocalypse." The Bloomberg Odd Lots interview is the latest in a string of public-banking-CEO commentary calibrating expectations about how artificial intelligence reshapes investment banking, trading, and broader financial services. The market-cap-leader investment banks — Goldman Sachs (GS), JPMorgan Chase (JPM), Morgan Stanley (MS) — are pricing this calibration differently. Their multiples and forward growth assumptions tell you which CEO commentary the market actually believes, and which path AI will follow at Wall Street firms.

The actual CEO calibration happening

In the past three months, five investment-bank CEOs (Solomon, Dimon at JPM, Gorman/Pick at MS, Moynihan at BAC, Fraser at C) have publicly addressed AI's impact on banking employment, costs, and competitive dynamics. The unified message: AI is augmentation, not displacement, in the foreseeable horizon.

Two specific dynamics drive this calibration:

1. The "junior banker tasks" most candidates for AI automation are themselves the source of mid-career talent development. Wall Street's career ladder relies on junior bankers learning the business through repetitive analytical tasks. Removing those tasks creates a pipeline problem — there's no clear replacement model for developing senior bankers in 2030.
2. Client relationships, not analytical work, remain the source of investment-banking value. AI can produce a deal model faster and more accurately than a first-year analyst. AI cannot replace the senior banker's relationship with a corporate CEO that determines whether a deal happens. As long as M&A and capital markets revenue depends on relationship-mediated decisions, the headcount needed to maintain those relationships is roughly stable.

The implication: AI improves productivity per banker but does not fundamentally reduce banker count. For investors, this matters because the productivity gain shows up as margin expansion, not headcount-driven cost compression.

Data points

drillr terminal snapshot (June 3, 2026):

The forward revenue growth divergence (all four names show large negative consensus expectations) reflects normalization from elevated capital markets activity in 2024-2025 toward longer-term run-rates. This is not AI-induced — it's mean reversion. The market is pricing relatively similar negative consensus across the cohort despite divergent 1-year returns.

GS specifically has outperformed materially over twelve months (+73.7%) — primarily from sustained M&A advisory and equity capital markets pipeline strength. JPM and BAC are roughly flat (-6.6% and -4.7% YTD) — reflecting interest rate compression and consumer/middle market exposure. MS at +18.4% YTD has captured both M&A franchise strength and asset management growth.

The CEO calibration commentary specifically addresses the long-run pipeline: if AI doesn't compress headcount, then capital markets revenue normalizing to lower run-rates means margin compression rather than per-employee productivity gain. The differential margin profile (JPM's 28.6% EBITDA vs MS/GS at 22.5%) suggests JPM has structural margin protection from diversification while GS/MS rely more on cyclical capital markets revenue.

{
  "hint": "A clean infographic showing the investment banking value chain with three stages labeled left-to-right: 'Relationship building', 'Deal sourcing & execution', 'Client servicing'. Above each stage, small icons represent: human banker icons for relationship and client servicing (in dark blue), and AI icons (gold) for deal sourcing/execution. A horizontal bar at the bottom labeled 'AI productivity gain margin expansion vs headcount neutral'. Plain white background, business publication editorial aesthetic, no decoration.",
  "aspect": "16:9",
  "style": "minimalist editorial value-chain infographic",
  "alt": "Investment banking AI value chain showing relationship building stays human while deal sourcing execution sees AI productivity gain leading to margin expansion not headcount reduction",
  "caption": "Wall Street AI value chain — productivity gain in execution, relationship work unchanged"
}

Analysis: positioning across three scenarios

Three scenarios for the major investment banks over 12-24 months.

Scenario A — Margin expansion materializes as CEO commentary implies. AI productivity gains drive margin expansion of 100-200 bps across the cohort. Forward revenue growth consensus stabilizes at current negative levels but margins expand. GS, JPM, MS all re-rate modestly upward. Implied 12-month total returns: GS +10-15%, JPM +5-10%, MS +12-18%, BAC +8-12% (plus dividends).

Scenario B — AI displacement happens faster than CEO commentary suggests. Productivity gains drive substantial headcount reduction over 5-7 years; near-term margin expansion is sharper than expected. Cohort re-rates higher faster but with regulatory/political backlash risk. Implied 12-month: GS +20-30%, JPM +15-20%.

Scenario C — Capital markets revenue normalization materializes faster than expected. M&A advisory revenue + ECM/DCM revenue both decline materially. AI productivity gains insufficient to offset top-line pressure. Cohort multiples compress. Implied 12-month: GS -10-15%, JPM -5 to +5%, MS -10-15%, BAC -5%.

The asymmetric profile favors Scenario A with B as upside optionality. For positioning, GS and MS provide the more concentrated capital markets exposure; JPM and BAC offer more diversified revenue sources. The CEO calibration commentary effectively reduces the Scenario C downside (no AI-driven sudden margin collapse expectation) while maintaining upside (margin expansion possible).

A barbell strategy: 40% GS + 25% JPM + 20% MS + 15% BAC captures the cohort upside while managing concentration risk. The adjacent fintech disruption context provides a different threat vector — and the parallel AI capex absorption thesis clarifies that AI investment dollars are flowing primarily into infrastructure, not banking productivity tools — CME, ICE, NDAQ face structural exchange-business risk while traditional investment banks face only productivity-driven margin questions. The investment bank cohort sits in a relatively defended position.

What to watch

• Q2 2026 earnings (mid-July): Watch operating margin sequencing across the cohort. Material margin expansion above prior quarters confirms CEO commentary; flat margins suggest productivity gains are slower.
• Investment banking headcount disclosures: Quarterly headcount commentary in management discussion. Sustained flat or modest growth confirms the "augmentation not displacement" narrative.
• AI-specific commentary in earnings calls: CEO and CFO discussion of AI investment and productivity gains. Numbers above expectations could trigger re-rating; below would signal slower path.
• Junior banker hiring announcements: Wall Street first-year recruitment classes for 2027 start. Hiring above 2025 levels confirms talent-pipeline calibration; hiring cuts signal accelerated displacement.
• Investment banking fee pool data (Refinitiv, Dealogic): M&A advisory and ECM fee pool growth trajectory. Sustained levels support the cohort thesis; meaningful decline activates Scenario C.

---