When suppliers become competitors: the BaaS power struggle

# When suppliers become competitors: the BaaS power struggle

The hook: a Saturday night with no bank

On a Friday in May 2024, roughly 100,000 fintech customers across apps like Yotta and Juno went to bed with a checking balance. By the weekend, some couldn't access it at all. The company sitting between them and their money, a Banking-as-a-Service (BaaS) middleware provider called Synapse, had collapsed into bankruptcy, and nobody, not the fintechs, not the sponsor banks, not the bankruptcy trustee, could fully reconcile who owed what to whom. Estimates of the shortfall in customer funds ranged as high as $85 million (as reported by bankruptcy court filings, 2024).

Synapse wasn't a bank. It was infrastructure: the plumbing that let fintech apps plug into real, FDIC-insured banks without becoming banks themselves. That plumbing was supposed to be neutral. It wasn't. And its failure exposed something MBA students need to understand structurally, not just as a scandal: in BaaS, the "supplier" sits close enough to the customer relationship that it can, and increasingly does, become a competitor.

Mapping the chain: who does what in BaaS

Banking-as-a-Service lets a non-bank company (a fintech, a retailer, a payroll app) offer bank-like products (accounts, cards, payments) without a banking license. Four layers typically exist:

• Sponsor banks: FDIC-insured (Federal Deposit Insurance Corporation) chartered banks that legally hold the money and carry regulatory responsibility. Examples: Evolve Bank & Trust, Cross River Bank, Column Bank, The Bancorp Bank.
• BaaS middleware providers: technology layers that connect fintechs to sponsor banks via APIs, handling ledgering, KYC (Know Your Customer, the identity-verification process required under anti-money-laundering law) and transaction routing. Examples: Synapse (defunct), Unit, Treasury Prime, Bond (acquired by FIS).
• Fintech distributors: the consumer-facing brand. Chime, Current, Yotta. They own the customer relationship, the app, the marketing.
• Regulators: the OCC (Office of the Comptroller of the Currency), the Federal Reserve, and the FDIC, which oversee sponsor banks, plus the CFPB (Consumer Financial Protection Bureau), which oversees consumer protection.

The textbook assumption is that middleware is a "dumb pipepipeAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.Voir la définition complète →": interchangeable, low-margin, non-strategic. Synapse proved that assumption wrong, and expensively so.

Why the "neutral supplier" model breaks down

Middleware providers control three things that make neutrality hard to sustain:

1. The ledger. Synapse, not the sponsor bank, kept the master record of which end customer owned which dollar in the shared bank account. When Synapse's systems and Evolve's systems disagreed, nobody could prove whose numbers were right. This is a classic supplier power problem: whoever holds the system of record holds leverage.

2. Multi-homing across fintechs. A single middleware provider serves many fintech clients simultaneously. That gives it aggregate visibility, and aggregate risk, that no single fintech client has. It also means the middleware provider's failure is correlated across the whole client base, unlike a single fintech's failure.

3. Proximity to the regulatory relationship. Sponsor banks answer to regulators for BaaS activity across their whole fintech portfolio. After Synapse, and after enforcement actions against sponsor banks like Cross River and Evolve for weak compliance oversight (Federal Reserve and FDIC consent orders, 2024), banks pulled compliance and ledgering functions in-house rather than trust a middleware layer. That is a supplier being disintermediated by its own upstream partner.

The result: sponsor banks started building their own BaaS platforms (competing with the middleware layer), and some fintechs started applying for bank charters or partnering directly with banks, bypassing middleware entirely. The "neutral" layer got squeezed from both sides.

Power dynamics: who actually has leverage now

Think of this as a value chain where power has historically flowed toward whoever bears regulatory risk and owns the customer, not whoever adds technical convenience.

| Player | Source of power | Vulnerability |

|---|---|---|

| Sponsor banks | Regulatory license, capital, legal liability | Reputational and enforcement risk when partners fail; balance-sheet exposure |

| Middleware (BaaS) | Technical integration, multi-client scale | Non-bank, no charter, no deposit insurance in its own name; replaceable |

| Fintech distributors | Brand, customer acquisition, data | Depend entirely on someone else's charter to operate legally |

| Regulators | Enforcement power, can shut down programs | Reactive; often intervene after failure, not before |

Post-Synapse, sponsor banks reasserted power by insisting on direct ledgers and tighter oversight of middleware partners. Some, like Cross River, began offering more integrated, bank-controlled BaaS stacks, effectively competing with the very middleware companies they used to rely on. That is the "supplier becomes competitor" dynamic in reverse: the upstream player (the bank) moved downstream into the middleware business it once outsourced.

Meanwhile, large fintechs with scale (Chime, SoFi) reduced dependency by acquiring bank charters (SoFi bought a bank charter via Golden Pacific Bancorp in 2022) or diversifying across multiple sponsor banks, reducing single-point-of-failure risk.

Where the margin actually sits

BaaS economics are thin and contested. Roughly (industry estimates, 2023 to 2025, figures vary by contract and are not standardized publicly):

• Sponsor banks typically earn fee income plus interest on pooled deposits held on their balance sheet (this is the biggest single revenue lever for banks in these arrangements).
• Middleware providers earn per-account or per-transaction fees, often in the range of low single-digit dollars per account per month, thin enough that scale is the only path to profitability.
• Fintech distributors capture the customer lifetime valuecustomer lifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.Voir la définition complète →: subscription fees, interchange revenue (the fee merchants pay on card transactions, typically shared back to the card issuer and program), and cross-sell into lending or investing products.

Simple worked illustration (illustrative, not sourced from real contracts): a fintech with 500,000 accounts paying a middleware fee of $0.70/account/month generates about $4.2 million a year in middleware revenue from that single client. If the middleware provider has thin margins per client, it needs many such clients simultaneously to be viable, which is exactly the concentration risk that made Synapse's failure so systemic to its client base.

This thin, transaction-based margin is precisely why middleware providers are tempted to move up the chain (toward owning ledgers, data, or even direct bank relationships) rather than stay a passive pipepipeAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.Voir la définition complète →. Passive pipes don't generate enough margin to survive.

The regulatory response: closing the neutrality gap

Regulators have started treating BaaS arrangements as a systemic oversight gap. Key developments:

• The FDIC proposed rules (2024) requiring banks to maintain their own records of who owns end-customer funds in custodial accounts, rather than relying solely on fintech or middleware records (see the FDIC's proposed rule on custodial deposit accounts).
• The OCC and Federal Reserve issued joint guidance emphasizing that sponsor banks remain fully responsible for third-party risk management, even when a middleware provider is technically at fault.
• The CFPB has signaled interest in whether BaaS fintech customers receive adequate disclosure that their "bank account" actually depends on a chain of three or four companies, not one.

The direction of travel: regulators are pushing accountability back onto the chartered bank, the entity they can actually supervise and penalize, rather than accepting "the middleware did it" as a defense. This strengthens sponsor banks' leverage over both middleware and fintech distributors, since banks now have compliance incentives to control the stack more tightly.