The BaaS Reset: What Synapse's Collapse Means for Fintech Treasury
When $85M in customer funds can't be reconciled, it exposes a treasury infrastructure problem hiding inside a banking problem.
In June 2024, Synapse Financial Technologies filed for Chapter 11 bankruptcy. By the time the dust settled, approximately $85 million in customer funds were stuck in a reconciliation nightmare between Synapse’s ledger and the ledgers of its banking partners, primarily Evolve Bank & Trust.
Customers of fintech apps built on Synapse’s middleware couldn’t access their money. Not because the money was gone (though some of it might be). Because nobody could agree on where the money was.
Let me say that again: a financial technology company whose entire purpose was to track the movement of money between parties could not produce an accurate accounting of where the money sat.
Most coverage of the Synapse collapse treats it as a Banking-as-a-Service story. A cautionary tale about middleware risk, regulatory gaps, and the fragility of the BaaS model. That framing is correct but incomplete.
Synapse’s collapse is also a treasury story. Perhaps the most important one in fintech in the past five years. Because the core failure - the inability to reconcile multi-party fund flows in real time - is not unique to Synapse. It’s endemic to how fintech infrastructure has been built.
What Actually Happened
The mechanics are worth understanding because they reveal a structural problem, not just a company-specific one.
Synapse operated as middleware between fintech apps (Yotta, Juno, Copper, and others) and partner banks (primarily Evolve). When a consumer deposited money into a fintech app, the money flowed through Synapse’s infrastructure to a bank account at Evolve. Synapse maintained a ledger tracking which dollars belonged to which end consumer. Evolve maintained its own ledger.
In theory, these ledgers should match. Synapse’s record of “Customer A has $500” should correspond to Evolve’s record of $500 allocated to Customer A’s benefit in the pooled account.
In practice, the ledgers diverged. Over time, small discrepancies accumulated. Transaction timing differences, failed transactions, edge cases in the reconciliation logic, and possibly errors in how funds were allocated between sub-accounts all contributed to a growing gap between what Synapse’s ledger said and what Evolve’s ledger said.
When Synapse went bankrupt, there was no mechanism to resolve the discrepancy. A court-appointed trustee, Jelena McWilliams (former FDIC chair), spent months trying to reconcile the two ledgers. As of early 2025, tens of millions remained unresolved. TabaPay, a payments company, ultimately acquired Synapse’s technology assets, but the customer fund reconciliation remained a separate, ongoing legal and operational challenge.
The fundamental problem: two parties maintained independent ledgers of the same money, with no real-time reconciliation between them, no independent source of truth, and no regulatory framework requiring one.Why This Is a Treasury Problem
The standard narrative frames Synapse as a banking supervision failure. The FDIC didn’t regulate Synapse directly because Synapse wasn’t a bank. Evolve was regulated, but the OCC and FDIC oversight didn’t extend to the accuracy of Synapse’s middleware ledger. There was a supervision gap.
That framing is accurate. But it misses the deeper issue.
The Synapse failure is a treasury infrastructure failure. Specifically, it’s a failure of three capabilities that should be foundational to any system that holds or moves money.
Failure 1: No Real-Time Reconciliation
In a properly architected treasury system, reconciliation happens continuously. Every transaction is matched against corresponding records in real time. Discrepancies are flagged immediately, not discovered months later during a bankruptcy proceeding.
Synapse’s architecture relied on batch reconciliation. Ledgers were compared periodically, not continuously. Small errors that would have been trivially easy to resolve in real time compounded over months into an $85M mess that required a former FDIC chair and a bankruptcy court to untangle.
This is not a novel technology challenge. Real-time reconciliation across multiple ledgers is a solved problem in treasury infrastructure. Banks do it internally. Payment networks do it. The technology exists. It just wasn’t implemented in the BaaS middleware layer because nobody required it.
Failure 2: No Independent Source of Truth
Synapse had a ledger. Evolve had a ledger. When they disagreed, there was no independent third-party record to arbitrate.
In traditional treasury, this is what custodians provide. When a company holds assets at a custodian bank, the custodian’s records serve as an independent source of truth. If the company’s internal records disagree with the custodian’s records, the custodian’s records are authoritative.
The BaaS model introduced a new entity (the middleware provider) between the customer and the bank without establishing who maintains the authoritative record. Synapse believed its ledger was correct. Evolve believed its ledger was correct. Both couldn’t be right.
Independent treasury visibility, the ability to verify your cash position without relying on a third party’s ledger, would have prevented or at least dramatically limited this failure.
Failure 3: No Cash Position Visibility Across the Stack
The fintech apps built on Synapse had no independent view of where their customers’ money actually sat. They trusted Synapse’s API to tell them. Synapse trusted its own ledger. Nobody had end-to-end visibility.
This is the equivalent of a CFO who knows their cash position only because their bank told them, with no independent verification, no multi-source aggregation, and no reconciliation against internal records. No serious treasury operation runs this way. But that’s exactly how the BaaS stack was architected.The Regulatory Response
Regulators moved quickly after Synapse. The response is reshaping the BaaS landscape in ways that every fintech platform should understand.
Bank-level scrutiny of partnerships. The OCC and FDIC have significantly increased scrutiny of how banks manage their fintech partnerships. Evolve received a consent order. Other BaaS-active banks (Cross River, Blue Ridge, Coastal Community) have faced heightened examination. Banks are now required to demonstrate that they can independently verify the accuracy of middleware ledgers.
Direct accountability for fund reconciliation. Regulators are pushing toward a model where the bank, not the middleware provider, is ultimately accountable for knowing where customer funds are. This shifts the architecture. Banks need independent reconciliation capabilities, not blind trust in a middleware partner’s API.
BaaS model reevaluation. Some banks have paused or exited BaaS partnerships entirely. Others are demanding significantly more robust technology and operational controls from their fintech partners. The era of “move fast and partner loose” in BaaS is over.
The net effect: BaaS is not dead, but it’s being rebuilt with treasury-grade infrastructure requirements. Real-time reconciliation, independent verification, and end-to-end cash visibility are moving from “nice to have” to regulatory expectations.
What This Means for Fintech Platforms
If you’re a fintech platform that holds customer funds, processes payments through banking partners, or relies on BaaS infrastructure, the Synapse aftermath has three direct implications.
Implication 1: Banks Are Prioritizing Risk Over Growth
The BaaS banks that remain active in the market are now running significantly more rigorous due diligence on fintech partners. Partnership timelines are longer. Compliance requirements are more demanding. Banks want to see that their fintech partners have robust reconciliation capabilities, independent cash tracking, and the ability to produce accurate, auditable records at any point in time.
If your treasury infrastructure is a spreadsheet, this is a problem. Not because spreadsheets can’t reconcile (they can, slowly). But because you can’t demonstrate to a banking partner that your reconciliation is continuous, automated, and auditable. The spreadsheet is a liability in banking partnership conversations.
Implication 2: Independent Treasury Visibility Is Non-Negotiable
The lesson from Synapse is stark: if your view of your cash position depends entirely on a third party’s ledger, you are one ledger error away from an existential problem.
Fintech platforms need independent treasury visibility. This means direct bank connectivity (not just middleware API access), real-time balance verification across all accounts and partners, automated reconciliation that flags discrepancies immediately, and an auditable record of every transaction from source to destination.
This isn’t paranoia. It’s basic treasury hygiene. The only reason it wasn’t standard practice in fintech is that the BaaS model made it seem unnecessary. Synapse proved otherwise.
Implication 3: Real-Time Reconciliation Becomes Table Stakes
Batch reconciliation, matching ledgers daily or weekly, was acceptable in the pre-Synapse world. It’s not acceptable anymore.
When money moves in real time (and in the era of faster payments, instant settlement, and 24/7 rails, it increasingly does), reconciliation needs to happen in real time too. A discrepancy that’s caught in minutes is a rounding error. A discrepancy that compounds for months is an $85M catastrophe.
Real-time reconciliation requires treasury infrastructure. Not spreadsheets. Not periodic bank statement imports. Automated, continuous, multi-source reconciliation with immediate exception handling.The Opportunity: Treasury Infrastructure as the Trust Layer
Here’s the constructive read on the BaaS reset: it creates an opportunity for treasury infrastructure to serve as the trust layer in fintech.
Pre-Synapse, the trust model in BaaS was implicit. Fintechs trusted their middleware providers. Middleware providers trusted their banking partners. Banking partners trusted their internal systems. Nobody independently verified the full chain.
Post-Synapse, the trust model needs to be explicit and verifiable. Someone needs to provide independent, real-time, auditable verification that the money is where everyone says it is.
That’s treasury infrastructure. Not as a back-office function. As the core trust layer that makes multi-party financial systems work.
The platforms that build (or buy) this capability won’t just be compliant. They’ll have a competitive advantage. In a market where banking partners are scrutinizing fintech partnerships more carefully than ever, the platform that can demonstrate treasury-grade reconciliation, visibility, and controls will win partnerships that competitors can’t.
The Macro Context: Why This Matters More in 2026
The Synapse collapse didn’t happen in a vacuum. The macro environment makes its lessons more urgent, not less.
Tariff uncertainty is increasing cash flow volatility. The universal 10% tariff increase is disrupting cross-border trade flows. The IMF projects a 1% US GDP hit. For fintech platforms with international operations, this means more unpredictable cash flows, more FX volatility, and more pressure on treasury visibility. In volatile environments, the margin for reconciliation error shrinks.
Ray Dalio’s “capital wars” are accelerating. The dollar’s share of global reserves has declined from 71% to 58%. Multi-currency operations are becoming the norm, not the exception. Each additional currency adds reconciliation complexity. Each additional banking partner adds a ledger to verify. Treasury infrastructure that was adequate for a single-currency, single-bank operation is insufficient for the multi-currency, multi-partner reality.
Cross-border volumes are growing. $290 trillion in cross-border payments by 2030. More money moving across more borders through more intermediaries. Each intermediary is a potential reconciliation failure point. The Synapse problem, scaled globally, is the financial infrastructure challenge of the next decade.
The $300B SaaS wipeout shows that software moats are fragile. AI is eating through enterprise software categories. The vendors that survive are the ones providing genuine infrastructure value, not just interface convenience. In fintech, genuine infrastructure value means treasury-grade reliability: reconciliation, visibility, controls, and auditability.
What Smart Platforms Are Doing
The fintech platforms I talk to that are ahead of the curve are doing three things in response to the BaaS reset.
Building independent cash visibility. They’re not relying solely on their banking partner’s API for balance information. They’re establishing direct bank feeds, independent reconciliation engines, and multi-source verification. If their banking partner’s ledger and their own ledger disagree, they know about it in minutes, not months.
Upgrading reconciliation to real-time. Batch reconciliation is being replaced by continuous, event-driven reconciliation. Every transaction triggers a matching process. Discrepancies are flagged and escalated immediately. The goal is zero unresolved exceptions at any point in time.
Treating treasury as a strategic function. The era of treasury as an afterthought is ending. Post-Synapse, treasury infrastructure is a banking partnership requirement, a regulatory expectation, and a competitive differentiator. The platforms that invest in it now will be the ones that banks want to partner with, that regulators trust, and that customers feel safe using.
The Bottom Line
Synapse’s collapse exposed a truth that fintech had been ignoring: you can’t build reliable payment infrastructure without reliable treasury infrastructure. The two are inseparable.
The BaaS model assumed that middleware could abstract away the complexity of multi-party fund flows. It couldn’t. Not without real-time reconciliation, independent verification, and end-to-end visibility. Those are treasury capabilities, and their absence is what turned a company failure into an $85M customer fund crisis.
The BaaS reset is painful. It’s also productive. It’s forcing fintech to build the treasury infrastructure it should have built from the start. The platforms that embrace this moment and invest in treasury-grade infrastructure will emerge stronger. The ones that treat it as a compliance checkbox will be the next cautionary tale.
In a world of rising tariffs, multi-currency complexity, and increasingly sophisticated financial regulation, treasury visibility isn’t optional. It’s the foundation everything else sits on.
Jeff Forkan is the founder of TreasuryPath, an AI-native treasury platform. He previously built payments infrastructure at Gusto and cross-border systems at Currencycloud and Visa. He writes Headless Banking, a newsletter about fintech infrastructure, treasury, and the future of money movement.