Scott Bok Explains What Investment Bankers Actually Do All Day

Former Greenhill CEO Scott Bok examines whether AI will displace investment bankers, emphasizing that disruption depends on the specific tasks bankers perform (origination, execution, strategic advice, and client relationships). He argues AI can automate analytical and routine execution work but is less likely to replace high-value client-facing, negotiation and strategic functions. For portfolios, expect gradual, role-specific efficiency gains and potential margin pressure in commoditized execution businesses, with outperformance possible for banks that successfully augment senior bankers with AI.

Analysis

AI will automate a large share of routine investment-banking production — financial modeling, comparables screens, first-pass diligence and pitch creation — on a 12–36 month cadence, but it does not obviate credibility, deal orchestration or regulatory navigation. Expect 40–60% of analyst-level hours to be replaceable by tooling in the near term, producing immediate cost/timing pressure on banks that compete on junior-hour intensity.

Second-order effects are the key alpha opportunity: fewer analyst hires will shrink the training pipeline, creating a talent cliff 3–7 years out where senior bankers become scarcer and command higher spreads on advisory fees; law firms, accounting boutiques and compliance teams will see workflow compression but higher demand for judgment-intensive services. Banks that own proprietary transaction and client data will compound advantage — their models improve faster and create sticky platform economics that benefit cloud/AI vendors and consulting integrators.

Tail risks that would slow or reverse disruption include regulatory/audit crackdowns on model explainability and liability from hallucinated diligence (weeks–years), and visible deal mishandlings that re-price the value of human oversight overnight. A second reversal catalyst is coordinated adoption by large bulge players who can amortize AI investments; if they move first, boutiques lose the window to capture commodity-driven fee pools.

Strategically, position for owners of client relationships and AI infrastructure rather than labor arbitrage. Over the next 12–36 months, favor vendors and consultancies that embed into bank workflows and boutiques or platforms that monetize senior coverage — avoid balance-sheet-heavy models that rely on large junior teams and high fixed-cost staffing models vulnerable to immediate efficiency substitution.

Key Decisions for Investors

• Long Microsoft (MSFT) 6–18 months: buy shares to capture multi-year cloud and Copilot-like contracts with large banks. Risk: slower enterprise procurement or regulatory blocking; Reward: 20–40% upside if banks accelerate cloud migration and pay-for-service models.
• Pair trade — Long Evercore (EVR) / Short Goldman Sachs (GS) 12–36 months: buy boutiques with lean senior-led models that capture fee uplifts from commoditization and short bulge-bracket banks reliant on scale junior execution. Risk: bulge-bracket reinvestment and scale advantage; Reward: target 1.5–2x spread if boutiques expand market share.
• Long Accenture (ACN) or buy ACN 9–18 month calls: exposure to consulting and systems integration revenue as banks implement LLM pipelines and compliance tooling. Risk: project delays and margin pressure; Reward: 25%+ upside from multi-quarter deal ramp.
• Short recruiting/staffing exposure (e.g., temporary staffing proxies) 6–24 months: tactical short on firms whose revenues are concentrated in entry-level placement for financial services. Risk: underestimates stickiness of some roles; Reward: capture downside as banks cut junior intake and move to tooling for throughput.