When your core banking system is so deeply integrated that exit would halt operations, you are not negotiating with a vendor. You are paying rent to a landlord.
I watch banks sign contracts that run hundreds of pages, with terms that favor the provider in every dimension that matters. Exit clauses are punitive to the point of impossibility. And they sign anyway, because the alternative is organizational paralysis.
This is presented as normal. Procurement teams call it "enterprise software." Executives call it "strategic partnership." But the structure is simple. The bank depends on the core for operation. The core provider depends on the bank for revenue. These are not symmetric dependencies.
When one party cannot credibly threaten to leave, negotiation becomes performance art. Pricing increases arrive on schedule. Feature requests enter multi-year roadmaps. Support degrades to minimum viable. And objections are met with the same implicit response: you cannot afford to exit.
Why This Persists
The structure did not emerge from predation. It emerged from integration.
Core systems became infrastructure because they needed to become infrastructure. Manual operations could not scale. Compliance demanded centralized records. The core solved real problems, and solving them required depth.
Integration created dependency. Dependency created switching costs. Switching costs created concentration. Concentration created pricing power. And pricing power, once established, reinforces itself.
Vendors did not need to collude. They simply needed to survive long enough for integration depth to eliminate competition. Banks that chose poorly decades ago are still running those cores, not because the systems are good, but because replacement is prohibitive.
Why It Matters Now
The cost was tolerable when change was slow. A ten-year contract was reasonable when technology shifted every decade. Integration complexity was manageable when compliance requirements were stable.
Those conditions no longer hold.
Technology shifts are measured in quarters. Compliance demands multiply faster than systems adapt. Customer expectations are set by companies that deploy weekly. Competitive pressure comes from entities not carrying the structural weight of a locked core.
Banks with inflexible cores cannot move at the required speed. They cannot deploy products quickly. They cannot integrate with emerging platforms efficiently. They cannot experiment without multi-quarter cycles.
This is not a technology problem. It is a leverage problem. The bank does not control the foundation on which it competes. Strategic decisions are constrained by a vendor with no incentive to prioritize the bank's needs over its own revenue optimization.
Opportunity cost accumulates. Competitors move faster. Customers migrate. And leadership watches the gap widen, knowing that closing it requires escaping a dependency they cannot afford to exit.
The Misunderstanding
The common narrative blames bank leadership. Better vendor selection. Harder negotiation. Earlier modernization.
This misidentifies the problem.
Infrastructure dependency creates conditions where vendor identity becomes irrelevant. Once a core is deeply integrated, switching to another locked vendor does not solve the structural problem. It resets the timer on the same trap.
Negotiation requires leverage. Leverage requires credible alternatives. When exit is prohibitive, alternatives are not credible. The vendor knows this. The bank knows the vendor knows this.
Banks believe they are customers. They are tenants. And tenants do not set terms.
What This Means
The structure is stable but not permanent. Systems optimized for extraction eventually face pressure from entities not subject to the same constraints.
Neobanks launch without legacy cores. Embedded finance bypasses traditional infrastructure. Regional banks explore cooperative ownership models. Technology companies with platform experience begin asking whether banking infrastructure could be rebuilt on different terms.
None of these have displaced incumbents. But their existence signals that extraction equilibrium is not inevitable. It is contingent on the absence of credible exit.
When exit becomes possible, leverage rebalances. When leverage rebalances, contracts change. Not from altruism. From competition.
The question is not whether this happens. The question is whether incumbents drive it or get displaced by it.
Banks that recognize the landlord dynamic have options. Cooperative models. Modular architectures. Regulatory frameworks mandating portability.
These are not easy. They require capital, coordination, and time. But they are structurally different from negotiating with a landlord.
The banks that do not act will continue paying rent. The amount will increase. The service will not improve.
And the contracts will keep getting longer.
