What It Takes to Build One of the World's Biggest Banks

Summary of What It Takes to Build One of the World's Biggest Banks

by Bloomberg

1h 2m•January 26, 2026
Listen to Episode

Overview of What It Takes to Build One of the World's Biggest Banks

This Odd Lots episode (Bloomberg) features Bill Demchak, CEO of PNC Financial, on how scale, strategy, regulation, technology and risk management shape modern banking. Demchak explains PNC’s ambitions to be one of the handful of nationally relevant U.S. banks, how the firm grows organically and by acquisition, how it modernized its tech stack to enable AI and automation, and what key structural issues (discount window mechanics, stablecoins, Basel rules, credit-card caps) mean for banks, customers and regulators.

Guest

  • Bill Demchak — CEO, PNC Financial Services Group
  • PNC: a national bank (sixth-largest U.S. bank), coast-to-coast retail presence, roughly one-sixth the size of JP Morgan. Core focus: retail deposits, middle-market / smaller large corporates, traditional commercial and capital markets products.

Key takeaways

  • Scale remains a dominant competitive advantage in banking. Large banks win by ubiquitous presence, product breadth, and balance-sheet capacity.
  • PNC aims to be “one of the five or six” banks that control U.S. retail — ambition is relevance, not vanity size.
  • Physical branches still matter regionally: PNC uses a branch-density threshold (~7–8% share of physical presence) as a tipping point for market dominance; they are expanding into high-growth metros (e.g., Houston, Dallas, Miami).
  • M&A is strategic, not at any cost. Market perceptions of PNC as an acquirer have depressed its valuation, but Demchak stresses disciplined economics and timing (sellers don’t sell in strong markets).
  • Successful integrations = mechanical (data migration, account conversion) + cultural (training, branch buddies, preserving frontline staff). Data migration opens accounts on PNC products over conversion weekends.
  • Tech investment was decisive: post-2008 replatforming into cloud-native, microservices enabled modern automation and AI use-cases.
  • Practical AI adoption: requires clean, single-source data; PNC identified 171 AI use cases (1.4bn addressable spend in care/ops), prioritized 5 live pilots; expects significant productivity gains (another ~30%).
  • Discount window and liquidity plumbing are opaque and cumbersome. Banks pre-position collateral (securities repo vs. pledging loans) to ensure access; reforming mechanics and stigma would improve system resilience.
  • Policy risks: a 10% cap on credit-card APRs would effectively collapse the current credit-card economics — hurting rewards, fees, lines and probably credit availability.
  • Stablecoins paying yield should be regulated like money-market vehicles; mixing payment and investment functions raises systemic questions.
  • Basel “Endgame” proposals need better math; regulators should fix model transparency and calibration to avoid perverse risk incentives.
  • Private credit both competes with and partners with banks; PNC invests in private credit funds to keep client relationships while sharing credit risk.

Topics discussed

Scale & retail strategy

  • Branch-density metric (~7–8% share) predicts market outperformance.
  • PNC building ~300 branches in new markets (with plans to continue adding).
  • Retail deposits are sticky, low-cost funding that underpins corporate lending.

Mergers & acquisitions / integration

  • Market penalizes likely acquirers; PNC prefers disciplined, economically sensible deals.
  • Integration blueprint: import client data, open accounts on PNC systems, deploy physical pods (equipment/signage), heavy training and branch-buddy support; keep client-facing employees when possible.

Regulation & policy

  • GSIB-focused regulation favored large banks in some crisis scenarios; Demchak argues competition matters and rules shouldn't unduly wall-off activities.
  • Discount window: mechanical frictions, slow processes, and stigma reduce its usefulness; PNC prepositions collateral (including physically archived loan docs) to ensure access.
  • Basel Endgame: original draft used flawed math; expect recalibration and more transparent modeling.

Technology & AI

  • PNC replatformed after the 2008 crisis: consolidated data centers, moved to cloud-native microservices.
  • AI use-cases prioritized around automation (document reading, care-center workflows, trust document processing), not just generative chat toys.
  • Legal constraints complicate AI underwriting (adverse action disclosure rules require specific reasons for denials).

Credit & markets

  • Credit cycle: currently healthy in Demchak’s view, though many firms are over-levered rather than fundamentally broken; fraud and poor underwriting remain constant risks.
  • Private credit: banks co-invest with private funds to retain client wallets and fee streams while distributing credit risk.

Silicon Valley Bank (2023) lessons

  • SVB failed from interest-rate/asset-liability mismanagement (long-duration securities + deposit negative convexity), and lack of collateral pre-positioning; PNC gained deposits and opened many new accounts during the run.

Notable quotes / pithy lines

  • “The ambition isn’t about size. The ambition is about being relevant.”
  • “Once you get over 7% branch density in a market you tend to outperform.”
  • “We’re not going to do deals at any cost — that’s a terrible assumption.”
  • On AI: “In order to utilize AI, you need clean data, defined data with a single source of truth.”
  • On proposed 10% credit-card cap: “You’ll just simply shut down consumer credit in the U.S.”

Practical implications / recommendations

For banks

  • Invest in consolidated, cloud-native architecture and clean single-source data before scaling AI.
  • Prioritize disciplined M&A economics; integrate acquisitions with strong people-first programs.
  • Pre-position collateral to manage liquidity risk and test discount-window access procedures.

For regulators & policymakers

  • Reassess discount-window mechanics and stigma to restore efficient intraday liquidity markets.
  • Treat yield-bearing stablecoins as money-market-like instruments (appropriate liquidity and disclosure rules).
  • Ensure Basel proposals use transparent, accurate math to avoid unintended risk migration.

For consumers & investors

  • Be skeptical of headline policy proposals (e.g., 10% APR cap) — they can have major secondary effects on product features and credit availability.
  • Understand bank valuations may reflect perceived M&A risk; listen to strategic discipline rather than headline growth ambitions.

Data & small facts called out

  • PNC is roughly 1/6 the size of JP Morgan.
  • Branch density tipping point cited: ~7–8% physical presence share.
  • PNC announced plans to build ~300 branches in new markets and expects to add hundreds more over time.
  • PNC identified ~171 AI use cases across $1.4bn addressable care/ops spend; ~40% of that is addressable, with five priority projects live.

Bottom line

The episode explains why scale matters in banking but reframes scale as “relevance” achieved through a mix of strategic branch expansion, client-focused commercial coverage, disciplined M&A, deep tech investment and prudent liquidity management. PNC’s modernization (cloud + AI pilots) and operational playbook (data-driven integrations, branch training, prepositioned collateral) are presented as concrete answers to the competitive and regulatory challenges facing banks in 2026.