FASB Proposed Three Tests That Turn USDC Into Cash on the Balance Sheet. Agent Payments Just Got Its Accounting Bridge.
On Tuesday, August 18, the Financial Accounting Standards Board proposed guidance that lets certain stablecoins sit under the "cash and cash equivalents" line on a corporate balance sheet under US GAAP. Comments are open until November 19. That is a smaller sentence than the story. Corporate treasurers already touch stablecoins. The question the proposal answers is whether those holdings show up next to your money market funds or under "digital assets, mark to market" three schedules to the right. Move that line one column to the left and the entire conversation about agent-payment settlement changes shape.
I have been writing about x402 and AFTA for the better part of a year on the working assumption that the last mile was going to be a legal and accounting fight, not a technical one. Yesterday one side of that fight moved.
The Three Tests
The proposal is narrower than the coverage suggests. A stablecoin does not automatically qualify. The issuer, and the holder, have to be able to answer three questions in the affirmative.
| Test | Requirement | What It Actually Requires |
|---|---|---|
| Redemption | At par, on demand | Direct issuer redemption right into US dollars within one business day. Secondary market liquidity does not count. |
| Reserves | 1:1, segregated, liquid | Full backing in short-term, highly liquid US-dollar assets (Treasury bills, insured bank deposits). Crypto and gold reserves are disqualifying. |
| Attestation | Independent, annual | Annual independent verification of the reserves. This is a real audit standard, not a signed letter from the issuer. |
Read that as three doors an issuer walks through. Any one of them closed and the stablecoin stays classified as an intangible digital asset, which is where FASB put it in ASU 2023-08 back in December 2023.
Who Passes, Who Does Not
I want to be careful here because a final rule is months away and issuers will change reserve composition to fit. On the proposal as written on Tuesday:
| Stablecoin | Verdict | Where It Sits |
|---|---|---|
| USDC (Circle) | Passes cleanly | Cash and short-duration Treasuries at BNY Mellon, monthly attestations by Deloitte, direct issuer redemption right on Circle Mint. |
| RLUSD (Ripple) | Likely passes | Same shape as USDC on reserves and redemption. Attestation program still maturing. |
| PYUSD (Paxos) | Likely passes | Paxos trust structure and monthly attestation cover the mechanism. |
| USDT (Tether) | Fails as written | Reserves include commercial paper, secured loans, and Bitcoin. Attestation is quarterly, not an audit. Direct redemption gated to verified corporate accounts above a size threshold. |
| DAI, FRAX, USDe (algorithmic or crypto-collateralized) | Excluded | Reserve composition disqualifies them under the "liquid US dollar asset" language. This is not a gray area. |
The line runs cleanly through Circle. That is not an accident. Circle spent the last four years building a compliance-first product and paying the higher operating cost that comes with holding reserves in T-bills at a custodian instead of chasing yield in commercial paper. Yesterday the standards-setting body wrote the rulebook around that operational choice.
What Actually Changes on the Balance Sheet
Today a treasurer holding, say, $50 million of USDC has to record it as an intangible digital asset under ASU 2023-08. Fair-value accounting each period, gains and losses hitting the income statement, footnoted separately from the cash line. Any auditor reviewing the 10-K sees a small crypto position, even if the CFO manages the balance as operational float that clears in seconds.
Under the proposed rule, that same $50 million sits inside cash and cash equivalents alongside your money market funds and 30-day Treasury bills. No fair-value adjustments. No separate footnote category. The cash conversion cycle now includes a settlement layer that clears on Base in two seconds at a cost measured in cents, and the balance sheet reflects that reality instead of penalizing it.
This is the difference between an experiment and a treasury policy.
Why This Is the Layer Agent Payments Was Missing
Coinbase reports 169 million x402 payments across 590,000 buyers and 100,000 sellers in the protocol's first year. Cloudflare and AWS both wired x402 into their edge networks in July. The AFTA network we run with TerminalFeed and other federated publishers settles receipts in USDC on Base. The technical layer is done.
The bottleneck was never technical. It was the CFO conversation. Try this pitch to a Fortune 500 controller: we want to fund an autonomous agent's operating float with $200,000 of USDC on Base, and let the agent settle micropayments to sixteen upstream data providers without our treasury team touching each transaction. The technical answer is easy. The accounting answer, until yesterday, was that the $200,000 sits under intangibles, gets marked every quarter, and shows up in the audit as a digital-assets line item that has to be explained.
The FASB proposal moves that number into the cash line, where it lives operationally. That does not make autonomous agents popular overnight. It removes the single loudest reason a controller had to say no. We wrote about why we chose USDC on Base the first time from the technical side. The accounting side, which is the harder pitch to enterprise buyers, just got materially easier.
The Structural Split
Two categorizations coming out of this proposal, and both matter more than the accounting detail suggests.
The first is the split between cash-equivalent stablecoins (USDC, likely RLUSD, likely PYUSD) and everything else. A US corporate treasurer choosing which stablecoin to hold operationally now has an accounting reason to prefer the compliance-first set. That is a demand floor for Circle. It is also a demand floor for the payment rails that route USDC natively, which is why the Stripe and OpenRouter deal from Sunday reads slightly differently in this light. A billing rail with USDC settlement on Base and Solana and Tempo, wrapped inside an accounting envelope the CFO can approve, is a different product than an inference gateway with an experimental payment layer bolted on.
The second split is between issuers that can meet the tests and issuers that have to change their reserve model to meet them. Tether has an obvious economic incentive to reallocate reserves into short-term Treasury bills and commission a real audit before the comment period closes. The economics of that shift are not small. Commercial paper returns more than T-bills, and Tether's float earns real money. The choice Tether makes over the next 90 days tells you whether the offshore stablecoin market wants to rejoin the onshore accounting world or is content to be a different asset class.
The Fine Print
Two caveats worth naming. First, this is a proposal, not a final rule. FASB is taking comments until November 19, and the final Accounting Standards Update likely lands in Q1 2027. Circle and Tether and every trade group with a position will file letters. The rulebook you plan against today may not be the rulebook that ships.
Second, the classification does not automatically confer regulatory approval. A stablecoin being a cash equivalent on a 10-K does not answer whether the OCC, the SEC, or the CFTC wants a payment stablecoin regulated as a bank deposit, a security, or a commodity. The Clarity Act is still stalled in the Senate. The accounting layer landing first is interesting on its own terms because it changes the operational calculus without waiting for the legislative fight to resolve.
Three Signposts
What I am watching between now and the November 19 comment deadline:
One, whether Tether announces a reserves reallocation into short-term Treasury bills and a full audit before the deadline, or waits to see if the rule gets watered down. A reallocation is the Tether-passes-the-test path. Silence is the two-tier stablecoin market path.
Two, whether any of the Big Four accounting firms publish interpretive guidance ahead of the final rule. The Big Four's early read sets the safe answer every listed company's Q1 2027 close will be graded against. Deloitte, given its role as Circle's attestation partner, is the one to watch.
Three, whether Coinbase or Stripe prices an enterprise-facing USDC treasury product that packages cash-equivalent classification as the pitch. Both have the distribution to move fast. Circle has the incentive to help them. If a productized treasury float on USDC shows up inside 60 days, this proposal accelerates from an accounting change into a go-to-market moment.
Our Take
Regulation by accounting standard is underrated as a lever for market structure. A ruling that looks narrow (which stablecoins qualify as a cash-equivalent line item on a US 10-K) reshapes the operational default for every corporate treasurer weighing whether the technology is ready for the balance sheet. The classification argument has always been the harder one to win. Yesterday it got a lot easier for USDC, and by extension for the settlement layer every agent-payments builder has been assembling on top of it.
The wire (x402) is done. The chain (Base, Solana, Ethereum) is done. What was missing was a legal and accounting conversation that ended with the auditor saying "put it on the cash line." That conversation just got materially easier for exactly one stablecoin family, and Circle spent four years earning the position. The compliance-first product got the compliance-first ruling. The rest of the industry now has to decide whether to change to match it, or accept that it lives one column to the right.
Practical read for anyone building agent payments today: if your settlement layer is not already USDC-native, this is the quarter to make that choice. The 60-day window between now and the comment deadline is the one where you can still ship without an accounting conversation being the first blocker your enterprise buyer raises. We stay on the story through the November 19 deadline and the Q1 2027 rule finalization on the originals index.