The Toll Collector Problem

5 min read

The Intermediary That Refuses to Die

Most "enterprise blockchain" is a rebranding exercise. Take an intermediary that charges rent for a function the protocol performs natively, wrap it in distributed ledger language, and sell it back to the same customers at the same margin. Permissioned chains where one company controls the validators. "Blockchain" that is really just a slower database with identical trust assumptions. The toll collector does not disappear. It gets a new logo.

A general-purpose execution network (XRPL, Ethereum) processes any transaction without gatekeeping. A purpose-built enterprise ledger (Corda, Hyperledger) restricts who can participate and who validates. The distinction is the difference between a public road and a private toll bridge.
Concept

The Toll Collector Pattern

Banks charge $25 for a wire transfer. The actual cost of moving bits across a network is a fraction of a cent. The $24.99 spread is the toll. Clearinghouses charge basis points to stand between buyer and seller. The actual settlement is a database update. Big 4 auditors bill $800/hour to verify transactions that exist on a public ledger. The verification is reading data that anyone can read. Market makers charge the bid-ask spread to provide liquidity. Automated market makers do the same thing at the protocol level for near-zero cost. Every one of these intermediaries performs a function that blockchain infrastructure handles natively: value transfer, settlement, verification, liquidity provision. The intermediary's business model depends on that function remaining expensive and opaque.

  • $0.0002 per XRPL transaction vs. $25 per wire transfer
  • Trust lines replace custodial accounts. No bank needed to hold your balance.
  • AMMs replace market-making desks. Liquidity is a protocol feature, not a service.
  • On-chain governance replaces quarterly auditor visits at $800/hour.
  • The toll exists because the function was historically difficult. The function is no longer difficult.
Example

Enterprise Blockchain's Graveyard

IBM Food Trust launched in 2018 to track food supply chains on a permissioned Hyperledger network. IBM controlled the validators, charged subscription fees, and required participants to use IBM infrastructure. It replaced a paper intermediary with an IBM intermediary. Walmart was the anchor client. IBM shut it down in 2022, along with TradeLens, its shipping logistics chain built with Maersk. Same pattern: centralized operator, subscription fees, proprietary infrastructure. JPMorgan Onyx processes billions in repo transactions daily on a permissioned network JPMorgan controls. It is JPMorgan's database with "blockchain" in the marketing copy. R3 Corda, backed by a consortium of banks, built a messaging system that banks control, validate, and permission. Each of these projects preserved the toll collector. They moved the tollbooth from a physical office to a server rack, kept the same pricing, and called it innovation.

Deloitte published that "76% of enterprises are exploring blockchain." Translation: 76% hired Deloitte to run a pilot program that preserves Deloitte's consulting fees. The pilot produces a report recommending further consulting.
Concept

The Real Test

Evaluating any blockchain project requires one question: does it remove the intermediary, or does it rebrand the intermediary? A general-purpose execution network (XRPL, Ethereum) runs permissionless. Anyone can transact. Validators are distributed. No single entity controls the ledger. A purpose-built enterprise ledger runs permissioned. The operator decides who participates, who validates, and what transactions are allowed. The operator IS the intermediary. The cost comparison tells the story. A cross-border wire transfer costs $25-50 and settles in 1-3 business days. An XRPL payment costs $0.0002 and settles in 3-5 seconds. The wire transfer price includes the toll. The XRPL price reflects the actual cost of the operation.

  • Is the network permissionless or does one entity control access?
  • Are validators distributed or does the operator run them?
  • Does the project reduce costs to protocol-level rates or maintain legacy pricing?
  • Can participants transact without the operator's approval?
  • If the operator disappears tomorrow, does the network continue functioning?
Scenario

Applying the Test

A company pitches you a "blockchain-based real estate platform." They tokenize properties on a private chain they operate. They charge 1.5% origination and 0.5% annual management. They control which properties list, which investors access the platform, and how secondary trading works. Run the test. Who controls the validators? They do. Can you transact without their approval? No. If they shut down, do your tokens work? No. Does the cost reflect protocol rates? No, it reflects legacy asset management pricing. This is a REIT with a blockchain landing page. Compare: properties tokenized on XRPL with trust lines, governed by on-chain rules, tradeable on any DEX, custodied in your own wallet. The platform is a front-end. The protocol is the infrastructure. If the front-end disappears, your assets still exist on-chain. That is the difference between a toll collector and an execution network.

Summary

The toll collector pattern charges for functions that blockchain performs at the protocol level. Enterprise blockchain projects that preserve centralized control, proprietary infrastructure, and legacy pricing are not blockchain innovation. They are intermediary preservation. The test is simple: does the project remove the tollbooth or relocate it? General-purpose execution networks eliminate the toll. Purpose-built enterprise ledgers rebrand it.

Key takeaway

Most enterprise blockchain replaces one intermediary with another. The real test: does the project remove the toll collector or just give it a new name? If one entity controls the validators, sets the pricing, and gates access, it is a database with blockchain branding.

Take the quiz