The Nested Sender Nobody Registered

Roles: the 400 merchants are the Originators - they authorize the entries and own the underlying obligations. Cascade is a Third-Party Sender because it acts as an intermediary between those Originators and the ODFI and transmits entries on their behalf. Harbor is a nested Third-Party Sender: it sits between the merchants and Cascade and never touches Northline directly. Northline is the ODFI and remains liable to the network for every entry in the flow.

Gaps:

• No Originator-level due diligence or credit review. A $5M limit for "Cascade" is not an exposure limit for each Originator, and a limit that cannot be applied to the party creating the risk is not a control.
• No identification of Cascade's TPS and nested TPS status in Northline's records or in its reporting to NACHA - an undocumented nested sender is the definition of an unmanaged relationship.
• No flow-down. Northline's agreement binds Cascade only; nothing requires Cascade to impose the same obligations on Harbor and Harbor's merchants.
• No audit rights and no file-level visibility into originator identity, which makes both credit monitoring and return-rate monitoring impossible.

All three of Cascade, Harbor and Northline have work to do, and liability does not travel down the chain with the files.

• Cascade must underwrite and monitor its Originators, including Harbor-originated ones, set limits for each, and terminate Originators producing unacceptable returns. "Small" is not an exemption; small merchants with high unauthorized rates are precisely the population worth watching, because they are cheap to set up and easy to use as a front.
• Harbor must do the same for its merchants and give Cascade what Cascade needs to meet its own obligations.
• Northline cannot rely on those promises. It needs return rates broken out by originator, limits enforced at that level, and audit rights that reach Cascade and, by flow-down, Harbor.

Unauthorized returns are the highest-signal metric in the set: an R05/R10 cluster is evidence that the underlying authorizations are defective or that the channel is being used to test stolen account data. That justifies immediate suspension of the Harbor channel pending evidence of valid authorizations, not a line item in a quarterly review.

Northline funds first. As ODFI it settles with the ACH Operator, so returned debits create an immediate debit position against Northline whether or not the merchant, Harbor or Cascade can pay. Originator credit risk lands on the ODFI's balance sheet, not on the TPS's.

Contractual recourse then runs Northline to Cascade, Cascade to Harbor, Harbor to the merchant, in whatever order the agreements provide. Because the merchant is insolvent and the money is gone, recovery is likely a fraction of the loss.

The design lesson is that an exposure limit is only real when it is backed by something Northline can reach: a reserve or holdback sized to the returns a merchant can generate, a guarantee from Cascade, collateral on demand if returns spike, and real-time visibility into merchant-level activity. A single limit for the TPS meant Northline was effectively lending unsecured to 400 businesses it had never evaluated - and it had no contractual right to look at any of them.