When I stepped into a business unit CFO role overseeing more than $50 billion in combined exposure, the first thing I heard from corporate was blunt: “It’s the Wild West.” The line had a different take. They felt corporate didn’t understand the business. Both sides were partially right, which was the whole problem. Nobody was looking at the same numbers at the same time, and without a shared view, every conversation turned into a negotiation about whose version of reality was more credible.
I came to believe banks don’t fragment because people refuse to work together. They fragment because the economics governing decisions are incomplete at the moment those decisions get made.
Parallel math
Every major function inside a bank runs its own legitimate financial discipline. The Lending Department protects spread and structure while the Treasury Department is focused on liquidity and funding cost. The Risk Department owns capital while the Finance Department holds the line on return stability and forecast credibility. All of them are rational, necessary, and internally consistent.
The trouble starts when those logics never get reconciled into one system before commitments are made. Instead they’re sequenced: Pricing clears a hurdle, credit confirms risk tolerance in a separate room, funding gets reviewed in a different forum, and capital intensity surfaces in portfolio analytics weeks later. No policy is violated, but the enterprise absorbs trade-offs nobody explicitly debated, and those trade-offs accumulate in the balance sheet long before anyone names them.
One reason the pattern persists is because isolated accuracy creates a false sense of control. Pricing looks disciplined relative to market, risk grading is conservative, and capital models are technically sound. Every slice withstands scrutiny on its own terms.
I remember sitting in a review where a deal cleared every hurdle we tracked at origination. Clean. Six months later, when capital intensity and deposit behavior were layered in, the enterprise return told a completely different story. The system had never evaluated the full equation at once, not because anyone failed, but because nobody owned the composite view. Those disconnects accumulate as modest forecast adjustments and pockets of unexpected capital consumption. Over a few quarters, variance becomes a recurring feature rather than something you explain away.
Making the relationship the unit of value
The shift with the most impact was reframing the relationship, not the deal, as the unit of economic value. Once we did this, loans, deposits, fees, capital consumption, and risk stopped living as parallel metrics. A lower-spread loan could be rational when anchored to durable operating balances, and a pricing concession might be justified by long-term relationship expansion. But the reverse kept showing up, too: Relationships that appeared profitable at origination sometimes revealed structural inefficiency once you layered in funding behavior and capital usage over time.
Linking all of this to the financial plan was uncomfortable. Some relationships that looked strong transactionally didn’t hold up in aggregate, and people who’d championed those relationships weren’t thrilled to see the numbers reframed. The tension was productive, though. It gave us something concrete to discuss instead of trading anecdotes about client potential.
What changed in the room
The first time we brought relationship-level economics into the Asset Liability Committee (ALCO) in a fully integrated way, the questions got sharper. The nature of debate shifted. We stopped arguing about whose numbers were correct and started examining the same numbers from different angles. Over time, this changed how people prepared. Leaders showed up expecting prior assumptions to be surfaced, and relationship managers anticipated questions about duration and capital intensity before walking in the door.
Early in the repair effort, we’d spoken openly about collaboration and shared accountability. The words were right, but the measurement systems weren’t aligned yet, and behavior followed the math. Culture shifted later—without a town hall, without a reorg—once relationship economics began governing decisions consistently. People responded to what was measured and revisited, and the walls came down because the economics finally gave them a reason to.
Adapted from “Beyond Pricing: Disciplined Performance. Real Impact."