Tether Holds $114.96B in Treasuries. Washington Is Building Its Replacement.

Tether's most recent reserve report — self-published, unaudited — lists $187.75 billion in total reserves against roughly $183.64 billion in liabilities. The difference, the entire excess buffer, is $411 million. That is 2.24% of liabilities. For an instrument that clears a meaningful share of global crypto liquidity and functions as de facto dollar infrastructure across several emerging markets, the shock absorber is thin. Two weeks later, a second data point surfaced with considerably less evidence attached. According to unnamed sources, the Trump administration is weighing an offshore stablecoin program designed to expand dollar usage abroad and support demand for US Treasuries. Three agencies are named: Treasury, State, and the DFC. There is no charter, no named partner, and no code. Check the logs, not the tweets. So let's open them. Stablecoins get filed under "infrastructure," which is a category error. There is no consensus mechanism to audit, no validator set to stress, no throughput bottleneck worth arguing about. Mint, redeem, transfer — that layer has been commoditized for years. When a stablecoin fails, it does not fail from a gas inefficiency. It fails from a duration mismatch, a custody gap, or a disclosure that only ever existed as a PDF. Here is the distinction most readers skip. Tether publishes an attestation, not an audit. An attestation offers limited assurance on a single point-in-time snapshot. A full audit tests internal controls and liability completeness. The gap between those two instruments is where every historical Tether controversy has lived. If the reporting says "reserve report," assume the weaker instrument. The policy side is equally soft. "Considering" is not "announcing." A program with three named agencies, zero named companies, and one anonymous source is the classic shape of a trial balloon — a signal released to measure reaction before any commitment exists. In my own monitoring framework, single-source policy reporting gets a confidence floor, not a confidence band, because the same channel that leaks a trial balloon also leaks the version designed to be denied. The choice of DFC matters more than the others. The DFC is a development finance institution. That is not monetary policy; that is geo-economics. It points the program toward markets where dollar demand is high and bank coverage is thin. One structural detail matters for everything that follows. Tether is not a US entity. Its issuance sits offshore, outside the perimeter that US regulators can inspect directly. That single fact explains why a state program would be framed as "offshore" rather than as domestic licensing: it is an attempt to build the same capability inside a jurisdiction that can read the books. Start with the balance sheet. $114.96 billion sits in direct US Treasury holdings — 61.23% of reserves. The remaining 38.77%, roughly $72.79 billion, is disclosed as a single total with no breakdown. Historically that bucket has contained commercial paper (since retired), secured loans, precious metals, and bitcoin. In my own audit work on reserve-backed instruments, I have never seen a stress test survive when more than a third of the collateral is a footnote. Note the denominator problem. Tether's reported liabilities are group-level and are not identical to USDT circulating supply. Any ratio computed against them carries that imprecision. The 2.24% excess figure is directional, not exact — but it is far too small to be treated as a shock absorber of last resort. Compare the buffer to the regime that governs money market funds, which are required to hold far thicker liquidity cushions precisely because they redeem at par on demand. Tether operates a structurally similar promise — par redemption — under a materially thinner reserve requirement and no equivalent supervisory floor. The promise is the same. The safety rail is not. The revenue model explains the incentives. Tether earns interest on reserves and pays holders nothing. Roughly all of its income is interest income — a carry trade. This is not a Ponzi; it is asset-backed, and the assets are real. But it is a floating-rate business. With 61.23% of reserves in short-dated Treasuries, profit is levered directly to the front end of the curve. Illustratively, a 100 basis point decline applied to the Treasury portion compresses gross interest by roughly $1.15 billion annualized. A cutting cycle shrinks the margin without shrinking the operational surface. Which brings us to the function Washington apparently wants to formalize. When a private offshore entity holds $114.96 billion of sovereign paper, it is performing quasi-sovereign financing. Tether is a marginal buyer of bills, an offshore-dollar node in the Eurodollar chain, sitting outside the Fed's direct perimeter. That is precisely why a state would want a version it controls. Ecosystem position matters here. Tether's moat is not cryptographic. It is network effect: USDT is the base pair on most centralized venues, collateral across DeFi, and the working dollar in economies with broken banking. Migration cost is high and switching is slow. The proposed program, by contrast, has an empty niche — no liquidity, no integrations, no users. It would start at zero on every axis except one: state backing. The headline is the most dangerous artifact in this story. Two independent facts — a program under consideration, and a reserve disclosure — were placed side by side. Readers reflexively fused them into "the government is partnering with Tether." The reporting says the opposite: no partnership exists. Correlation is not causation, and adjacency in a headline is not a contract. The likelier reading is exclusion, not endorsement: Washington routing around the largest offshore issuer rather than adopting it. If the program ships, the beneficiaries are more plausibly US-domiciled, audit-ready issuers than the incumbent. The second blind spot is governance. The industry chants that code is law, but Tether's issuance contracts carry freeze and blacklist authority, and the upgrade path runs through a small set of admin keys held at iFinex. That is the exact structure DAOs claim not to have — upgrade rights concentrated in a multisig. Code is law; hype is just noise. And here, the admin keys are the law that actually binds. The signal is not in the headline; it is in the follow-through. Watch four things: whether "considering" becomes a legislative draft, who appears on the partner list, whether the next reserve report breaks out the $72.79 billion, and whether USDT trades at a persistent discount in OTC markets. The program, if it arrives, will not replace Tether. It will decide which issuer the next decade of dollar expansion runs through.