As of June 2026
For a century, settling your own payments at the Fed meant one thing: a full master account, or a bank in the middle to do it for you. The Fed's payment-account proposal creates a third path, but the design is defined by what it removes.
The account grants access to the rail. It does not grant a bank’s balance sheet. That difference turns liquidity management from a back-office function into the operating model.
Proposal date
May 20
the Federal Reserve asked for comment on a special-purpose payment account in 2026.
Access
Fedwire / FedNow / NSS
eligible institutions could clear and settle through Federal Reserve payment services.
Cushion
None
no discount window, no intraday credit, no interest on balances, and controls to prevent overdrafts.
What the proposal changes
The Federal Reserve’s May 20, 2026 proposal would establish a payment account for legally eligible institutions to use for clearing and settling payments. The proposal does not widen who qualifies for Federal Reserve accounts or services. It changes what a qualifying institution can get: access to payment services without requiring the same full account profile that carries bank-like support.
What the proposal withholds
The limits are the architecture. Payment-account holders would not access the discount window, would not receive intraday credit, would not earn interest on balances at a Reserve Bank, and would use automated controls that prevent overdrafts. Closing balances are capped by expected payment activity, with the board memo describing a maximum closing balance framework up to one billion dollars.
The sponsor bank was doing more than routing
A sponsor bank did not just pass the payment through. In many operating models it absorbed timing gaps, covered settlement windows, and created a liquidity cushion between outbound and inbound money. Remove that layer and the payment account holder inherits the rail’s operating burden without the cushion that used to sit underneath it.
FedNow makes that burden sharper because it never closes. If a payment can leave at 2 AM on a holiday weekend, the prefunding model has to be alive at 2 AM on a holiday weekend. Misjudge the position and the payment fails. No sponsor bank sits underneath it. The Federal Reserve proposal deliberately withholds that cushion.
Direct Fed access reads like the prize. It is the exam. The firms that pass will be the ones that can hold an intraday cash position on a real-time rail with no bank underneath them.
