Ask most engineers why a financial institution wants blockchain and you'll get an answer about decentralization. Ask the actual auditors at Horizon Capital, and you get a much simpler answer: they wanted to stop taking the numbers on faith.
The problem was never the ledger itself
Horizon Capital's risk data was accurate. The five departments generating it were competent. The problem was that reconciling those five sources each quarter was a manual process that took eighteen days and produced a report auditors had no way to independently verify — they were trusting the reconciliation process, not checking the underlying data.
The reframe that mattered
This wasn't a data quality problem. It was a verification problem. The data was fine; proving it was fine to a skeptical third party was the actual bottleneck.
What a permissioned ledger actually changes
Every department writes to the same Hyperledger Fabric ledger instead of maintaining a private system that gets reconciled later. Each entry is cryptographically signed and timestamped, so an auditor doesn't have to trust DreamImp, Horizon Capital, or any single department — they can verify the chain of entries directly.
The result wasn't really about blockchain
Reconciliation time dropped from eighteen days to two. But the number that actually changed the relationship with regulators was audit findings: eleven per year, down to zero, because for the first time auditors could check the underlying data instead of just the summary report.
The lesson that generalizes
If you're evaluating blockchain for an enterprise use case, the right question isn't "do we need decentralization." It's "does a third party need to independently verify something they currently have to trust us about."