News analysis ·

ACH Credit-Without-Debit Gaps: Who Absorbs the Loss

UMass researchers found ways to spend ACH credit before the matching debit fails. We trace how the money moves and which bank is left holding the loss.

Green duotone close-up of a 1934 handwritten deposit slip with a Paid stamp and a printed note that cheques are not available until collected.
Photo: Unknown author / museumsvictoria, Public domain (cropped, duotone)

Researchers at the University of Massachusetts argue attackers can spend money that arrives over the ACH network before the matching debit has actually succeeded, Payments Dive reported. The study describes three ways to exploit that timing gap. This article looks at how an ACH pull moves between two banks, and which party is left with the loss when the debit bounces.

What the UMass study found

The paper was written by UMass Amherst professor Taqi Raza and Ali Malik, who is working on his doctorate in the same university's computer science and electrical engineering programme. It looks at transfers that run under the rules of Nacha, the body that operates the US automated clearing house, or ACH, and sets the standards banks follow when they move money between each other.

The authors call the problem credit without debit. Money shows up as usable at one bank while the debit that was supposed to fund it fails at another. They sorted the attacks into three types, which they named creditbleed, creditmint and creditreset.

"Our goal is to strengthen financial security," Raza said, according to Payments Dive. The university did not name the banks involved, saying the work is about weaknesses in interbank transfers rather than particular institutions. Raza said the findings reached Nacha through the banks, and that the issues seem to be addressed in the updated Nacha specification.

How the money moves in an ACH pull

An ACH debit is a pull. A customer at Bank A asks Bank A to collect money from an account they say they own at Bank B. Bank A, the originating bank, sends the request into the ACH network. Bank B, the receiving bank, then checks the account and either pays or sends the debit back as a return.

The weak point is timing. Nacha rules allow the originating bank to make the incoming money usable as provisional credit before Bank B has decided whether to pay. Provisional means the credit can still be taken back. If Bank B later refuses, Bank A has to reverse an entry that its customer may already have spent.

Each of the three attacks uses that same gap in a different way:

  • Creditbleed: the attacker pulls money from a restricted account at another bank. The first bank shows spendable funds until the restricted bank finally reverses the transfer.
  • Creditmint: when linking an outside account, banks often send small test deposits to confirm ownership. Attackers steer those deposits into restricted accounts that cannot send money back, and repeat the trick to collect cash.
  • Creditreset: the attacker pays a credit card bill from a debit-restricted account. The card limit goes up, the issuer's collection then fails, and the attacker spends the extra room.

In every case the money that leaves the system is real, while the money that was meant to cover it never arrives.

Who carries the risk when the debit fails

This is the part that matters most for banks. When a debit comes back, the receiving bank has done its job: it refused to pay from an account that could not pay. Under Nacha's stated position below, the decision to let credit be spent early sits with the originating bank or its originating customer. If the attacker has already withdrawn the money, whoever made that call is exposed to the loss.

Nacha's position, as a spokesperson told Payments Dive, is that a returned ACH debit can come back within approximately two banking days, for reasons such as insufficient funds or a frozen account. Nacha said whether to let funds be withdrawn before that return window closes is a decision for the originating bank or its originating customer, weighed against how much risk each is willing to take.

Put simply, the network sets the clock, and each bank decides how much it trusts the money before the clock runs out. A bank that releases funds at once is choosing to carry the return risk. A bank that holds funds until the window passes is safer, but slower for honest users. For card issuers, creditreset shows the same choice in a different place: raising a limit on a payment that has not cleared is lending against money that may never come.

Why the stakes are rising

ACH is a large and growing rail. Citing an October Nacha release, Payments Dive said third-quarter ACH Network payments reached 9.3 billion, up 5.9% from a year earlier, with value up 10.2% to $25.6 trillion. More volume means more pulls running through the same provisional-credit window.

Large payment firms are also spending on fraud defence. Visa said in a May review that over the final six months of last year it spotted roughly $1 billion worth of scam payments, and later bought fraud-prevention firm BioCatch for $2.4 billion.

What to watch

The first thing to watch is how banks apply the updated Nacha specification that Raza says appears to address the reported problems. The second is funds-availability policy: whether originating banks shorten how fast they release ACH credit, or add checks on restricted accounts before treating a debit as good. The third is account linking, where test deposits are a cheap and common ownership check that the creditmint method turns against the bank.

Sources

Written by AI from the sources above and published automatically: before release, a script confirms every figure against the sources and checks for copied text. No editor reviews it. Treat it as background reading, not financial advice.

← All posts