Why the next edge in investment banking is a system that never stops watching the market — and what the consulting consensus still misses.
The major consulting firms have reached a rare consensus: artificial intelligence is remaking the deal business. Yet most banks are bolting that intelligence onto a workflow that remains fundamentally episodic. The larger prize is a system that is always on — one that maps an industry in granular detail, tracks it from both the demand and supply sides, links every company to its strategic priorities, and predicts where the next transaction will form.
Where the consultants agree
The direction of travel is settled. McKinsey estimates that generative AI could create $200 billion to $340 billion in annual value for banking, and now frames AI-enabled origination as a strategic prerequisite rather than a productivity play — surfacing risk and capital insight to the front line in near-real time. Bain describes AI as the lifeblood of M&A, reporting that adoption among dealmakers more than doubled in 2025, with roughly 45 percent of more than 300 executives surveyed now using it, most heavily in sourcing, screening, and diligence.
BCG’s 2025 M&A report calls embedded AI and advanced analytics a critical edge, beginning with the identification of high-potential targets. Deloitte finds that about 86 percent of deal organizations have integrated generative AI, two-thirds of them within the past year.
The infrastructure is following the thesis. Bain acquired — and later divested — Sutton Place Strategies specifically for its granular industry taxonomy and its ability to map the relevant buyer universe and unify an otherwise fragmented sourcing environment.
The gap the consensus hides
Here is the uncomfortable part. For all that adoption, AI in most banks is still a point solution — a faster way to read a data room, draft a memo, or screen a list once a process is already live. The workflow beneath it has not changed. It remains episodic: the bank mobilizes around a deal, runs its analysis, and stands down until the next mandate appears.
The cost of that posture shows up in coverage. Sutton Place Strategies’ origination benchmark found that private equity firms saw, on average, fewer than one in five of the relevant transactions in their own target markets. Even disciplined originators miss the large majority of what is happening in their sectors. No one wins a mandate on a deal they never saw forming.
The next edge is not a better tool inside the deal. It is a different unit of analysis: the market, watched continuously.
From watching deals to watching markets
An always-on system rests on four capabilities. The first is a detailed, living industry taxonomy. Generic sector codes flatten the very distinctions that drive deals; a real map resolves an industry into its segments, sub-segments, and adjacencies, and stays current as the sector moves.
The second is coverage of both sides at once. Demand and supply are tracked together — who is acquiring and why, who is consolidating, where capital is concentrating, and which assets are quietly coming into play. A view of one side is half a thesis.
The third is linkage: every company tied to the taxonomy and to its own strategic priorities, not just the names in a current process. When each operator, sponsor, and strategic is positioned against the map and its stated intentions, the question of who would logically buy or sell an asset becomes a query rather than a week of analyst work.
The fourth is prediction. The system scores which companies are most likely to transact — as sellers and as buyers — before a banker is in the room, turning origination from reactive to anticipatory and handing execution a buyer universe that already carries its strategic rationale.
Why it matters now
The consultants have described the destination; few have operationalized the road to it. Bain’s taxonomy assets, Deloitte’s adoption curve, and McKinsey’s framing of origination as a strategic prerequisite all point the same way — toward continuous, predictive, full-market intelligence. What the market still lacks is a workflow rebuilt around that idea rather than decorated with it.
That is the principle behind Espalier. We treat origination and execution as one continuous system rather than two separate events. The same always-on intelligence that anticipates where a transaction will form also equips the banker who runs it: the full universe of logical counterparties, ranked by strategic fit, and a narrative grounded in how the sector truly creates value.
The banks that win the next cycle will not be those with the fastest tools inside a live deal. They will be the ones that never stopped watching the market.
