GCC Intelligence Series AgentShift: Capital Markets Reading Guide Capital Markets · v2 Feasibility · Wholesale Bank Archetype
GCC Intelligence Series · Capital Markets · v2 Feasibility

AgentShift — Capital Markets

A Strategic Decision Enablement Framework for Agentic AI Transformation spanning both parent-bank and GCC universes for Capital Markets — mapping AI agent impact, jurisdictional anchoring, and a five-way decision (Keep onshore / Agent-Direct / Wait-then-Agent / Hybrid / GCC-Lift) for every role. Global Wholesale Bank — Capital Markets archetype: Global wholesale arm; G-SIB regulator stack (prudential + conduct + market); CCP clearing.

The Core Framework

Every role is scored on three dimensions: Transaction Intensity (T) — repetition, rule-density, volume; Decision Complexity (D) — judgment and ambiguity; and Regulatory Blast Radius (RBR) — fiduciary and regulator exposure. Higher transaction + lower decision = greater agent impact; RBR dampens the score on regulator-facing work (IPV, Model Validation, Surveillance).

Each role falls into one of three restructuring zones:

Agent-Core— agent IS the worker, 60–80% FTE reduction
Agent-Augmented— humans lead, agents on subtasks, 30–50% gain
Human-Core— agents are peripheral tools, 10–20% uplift

Feasibility Classifier — the strategic question: for every role across the whole bank, where should the work actually live?

Keep Onshore— jurisdiction / client / regulator anchor
Agent-Direct (skip GCC)— high impact + production-ready, automate at parent
Wait-then-Agent— hold onshore briefly, agentise in 12–24 mo; offshore detour does not pay back
Hybrid (GCC + Agent)— GCC as transformation lab, agents augment from day 1
GCC-Lift (classic)— labour arbitrage still wins, AI peripheral

What’s Inside

8 Diagnostic Layers: Agent Impact, Transaction, Decision, RBR, Readiness, Offshoring, Feasibility, Mode, plus HC Detail — each recolors and re-aggregates the treemap.

Two Role Universes: 76 GCC roles (captive ops, 1000-FTE baseline) + 30 Parent-bank archetypes (Global HQ/NYC front office, 5000-FTE baseline). Toggle the Location filter to rewire the bridges.

3 Transformation Layers: Savings Map, Capacity Released, Upskill Potential — each triggers a People + Economics dual bridge.

Feasibility Layer: 5-way classifier driven by geo-lock, client proximity, RBR, readiness, and impact. Renders a Parent → GCC bridge (Parent FTEs → Keep → Agent-Direct → Wait → Movable to GCC) and three-stage economics (Parent cost → after-GCC cost → after-agent cost, in USD $M/yr).

60 upskill pathways mapped to CISI, ICMA, CFA, FRM, PRMIA, ISDA, ISLA, CSC, ICA — CM-specific.

Adjustable sizing: GCC headcount, Parent FTEs, and base CTC (USD) drive all bridge math.

Start Exploring ↓
106 roles: 76 GCC + 30 Parent
Archetype: Global Wholesale Bank — Capital Markets
5-way Feasibility classifier
3-stage economics: Parent → GCC → Agent
Layer
Transform
Low
High
Location Value Chain Domain
▶

AgentShift Capital Markets — Methodology & Layer Definitions

v2 extension — Strategic Decision Enablement Framework: v1 answered "of the work already in the GCC, how much will AI eat?" v2 answers the bigger question — "for every Capital Markets role in the bank, given AI's trajectory, where should the work actually live and when does it move?" Two universes loaded (76 GCC + 30 parent-bank archetypes spanning HQ city and NYC). A five-way feasibility classifier resolves into Keep Onshore · Agent-Direct · Wait-then-Agent · Hybrid · GCC-Lift. The Feasibility layer renders a Parent → GCC bridge and a three-stage economics bridge (Parent cost → GCC cost → Agent cost in USD $M/yr).

Agent Impact

Composite of T and inverse D, dampened by RBR. Green (low) to red (high).

Transaction (T)

Repetition, rule-density, volume.

Decision (D)

Judgment, ambiguity. Color inverted.

RBR

Fiduciary and regulator exposure. Multiplicative dampener.

Readiness

Tool maturity for the workflow.

Offshoring

Regulatory locality + client proximity + time zone + talent depth.

Feasibility

The strategic decision per role.

Keep: anchor
Agent-Direct: skip GCC
Wait-then-Agent: hold & agentise
Hybrid: GCC + agent day 1
GCC-Lift: arbitrage

Restructuring Mode

Three zones from Agent Impact.

Savings Map

Net savings = capacity × cost band − agent cost.

Capacity Released

FTEs freed by automation.

Upskill Potential

Redeployability composite with CM-specific upskill targets.

Cost model: Base CTC (USD) configurable. GCC cost bands 1x–2.6x; parent FO 11x–75x. Agent cost ratios 0.18x–0.40x. GCC after-move cost approximated at 0.20x of parent run-rate for movable work. Hybrid post-agent factor 0.60x of GCC; GCC-Lift post-agent factor 0.85x of GCC. Archetype: Global wholesale bank; G-SIB regulator stack. All figures illustrative.