Tether: Treasury Infrastructure at Sovereign Scale
Tether holds $122B in direct US Treasury positions and $141B in total Treasury exposure as of December 31, 2025 (BDO Italia year-end attestation). That direct holding alone exceeds the sovereign Treasury position of Germany, the UAE, Spain, or Australia. The company is roughly 300 employees, no external investors, and reported approximately $10B in profit for 2025, down from $13B in 2024. None of that is the company's primary product. The product is the dollar-pegged token used by hundreds of millions of people in countries where local financial rails operate at a fraction of the throughput the US system takes for granted. This report argues the right frame for Tether in 2026 is macro-infrastructure: a privately-held money-market-fund-shaped entity that has become a meaningful marginal buyer of short-duration US sovereign debt, with the systemic risks that scale produces.
The Scale
The reserve composition tells the structural story. As of the year-end 2025 BDO attestation, approximately 82% Treasuries, 10% money-market funds, 5% repo agreements, with the remainder in gold (127.5 metric tons), Bitcoin (96,184 BTC), corporate bonds, and secured loans. Cash and bank deposits sit at roughly 0.04% of total assets ($64M against $192.8B). Short-duration sovereign paper backs the product. Bank deposits play almost no role.
The composition shift is real. In 2021, 49% of reserves were commercial paper and only ~3% was actual cash. Post-2022 regulatory pressure forced a wholesale rewrite of what backs the token. The current attestation supports a different risk profile from the one most coverage assumes.
Per BDO Italia year-end 2025 reasonable-assurance attestation. Excess reserves of ~$6.3B sit on top of $186.5B in token-backed liabilities.
The economics resemble a money-market fund more than a payments company. Tether takes in dollars, invests them in short-duration T-bills, and keeps the yield. The difference: a money-market fund passes most of that yield back to its investors. Tether keeps all of it. Tether does not monetize ordinary USDT transfers. Issuer fees apply to direct minting and redemption (0.1% in some cases, with minimums), but secondary-market peer-to-peer and exchange transfers generate zero revenue. The choice was deliberate: in 2014, the founders debated whether to take 1 to 10 basis points per transaction the way Visa and Mastercard do, and chose zero in favor of adoption.
The Macro Read
Brookings flagged stablecoin issuers as a meaningful marginal buyer of US Treasuries in late 2025, ranking the category behind only a handful of foreign sovereigns during one recent measurement period. Tether alone holds more US Treasuries than several G20 nations.
The read for TokenIntel users: stablecoin issuer demand is now part of the plumbing of short-duration US sovereign debt markets. A material disruption to USDT (regulatory shock, run scenario, redemption cascade) would force the unwind of $122B in T-bills into the secondary market. That is not a crypto risk in the traditional sense. It is a macro-liquidity risk that touches the same instruments money-market funds, banks, and central counterparties use as collateral.
BIS researchers have separately found that stablecoin use correlates more strongly with remittance costs and transactional needs in emerging and developing economies than Bitcoin or Ether use does. McKinsey's 2025 estimate puts identifiable real payment activity in stablecoins (B2B settlement, remittances, card-linked spending) at roughly $390B annualized, against total stablecoin onchain flows of approximately $33T. The gap is enormous and worth keeping in mind when reading "trillions in stablecoin volume" headlines as evidence of adoption. Most of that flow is exchange and arbitrage activity. Real-economy commerce accounts for a small fraction.
The Trust Gap
Tether publishes quarterly attestations prepared by BDO Italia. The quarterly reports provide limited assurance under ISAE 3000. The year-end reports (Q4 2024 and Q4 2025) are stronger reasonable-assurance engagements with more rigorous testing. Neither is the same as a full set of audited financial statements in the public-company sense. BDO reviews Tether's assertions about its reserves and reports whether those assertions are materially misstated. It does not produce the kind of full financial audit institutional allocators typically require.
The trust deficit is real and earned. In 2019, the New York Attorney General found that Bitfinex (Tether's sister exchange) had taken $850M from Tether's reserves to cover losses from a payment processor whose funds were seized by authorities. Tether settled for $18.5M. The CFTC separately fined Tether $41M for misrepresenting that USDT was "fully backed" during a period when it was not. At one point, Tether's website quietly revised its language from "100% backed by USD" to "100% backed by our reserves, which may include affiliated entities."
That history is what people mean when they call Tether opaque. The 2025 reserve position is dramatically different from the 2021 one, but the audit gap persists. Tether hired a new CFO in early 2025 from LetterOne, reportedly specializing in contentious audits, which signals staffing for an eventual Big Four engagement. Until that engagement produces a published full audit, professional allocators will continue to discount everything else by some non-trivial margin.
The S&P Global stability score downgrade to 5 (the lowest reading) in late 2025 is the cleanest external read on this. S&P specifically cited rising exposure to higher-risk assets (Bitcoin, gold, corporate bonds, secured loans) at 24% of reserves, up from 17% a year earlier. The asset-mix drift is visible in the BDO attestation. Whether you read that as prudent diversification or as creep up the risk curve depends on your priors.
Rate Sensitivity
The 2024-to-2025 profit decline (roughly $13B to $10B, a 23% drop) is the cleanest evidence of the rate-cycle dependency in Tether's business model. The math:
| Scenario | Effective yield | Direct T-bill income | vs 2024 baseline |
|---|---|---|---|
| Fed rate ~5.25% (2024 baseline) | ~4.8% | ~$5.9B | baseline |
| Fed rate ~4.25% (current 2025) | ~4.0% | ~$4.9B | −$1.0B |
| Fed rate ~3.25% (200bp cut scenario) | ~3.0% | ~$3.7B | −$2.2B |
| Fed rate ~2.25% (400bp cut scenario) | ~2.0% | ~$2.4B | −$3.5B |
Indicative only. Computed against the year-end 2025 direct Treasury position of $122B; effective yield assumes a roughly 3-month duration ladder. Real-world numbers vary with Treasury reinvestment cadence, MMF and repo income, gold and BTC mark-to-market, and any incremental venture-portfolio gains or losses.
A 200bp cumulative rate cut against the year-end 2025 position would reduce direct T-bill income by approximately $2.2B annually. Material but not existential. The hard-asset hedge (gold and Bitcoin holdings, which tend to appreciate in rate-cutting environments) provides a partial offset. The venture portfolio is the more interesting hedge: roughly $20B in proprietary investments (segregated from USDT reserves, funded from retained profits) covering AI, energy, telecoms, agriculture, media, and physical financial infrastructure. The portfolio is too diffuse to evaluate as a single thesis, but the strategic logic is clear: an asset mix that does not depend entirely on the Fed funds rate.
Concentration Risks
Chain concentration on TRON. Approximately 44% of USDT supply (~$82B) sits on TRON. TRON also handles roughly 65% of all sub-$1,000 USDT transactions. The chain is dominant in retail-scale dollar movement. Any sustained operational, governance, or regulatory issue at TRON propagates directly into Tether's largest distribution rail. Tether's wallet-development-kit cross-chain routing is designed to neutralize this over time, but until adoption is broad the concentration is real.
Issuer concentration in stablecoin distribution. Per Visa onchain analytics, USDC has overtaken USDT in adjusted transaction volume for the first time since 2019. Volume share over the period: USDT 87% (2019) to roughly 36% (2026); USDC 13% to roughly 64%. This figure is well-cited but worth verifying against Visa's primary source before quoting in subscriber-facing content. The volume flip does not translate cleanly into a profit threat: Circle surrenders an estimated 60% of revenue to distribution partners (Coinbase alone received over $900M from Circle in 2024). Tether owns its distribution channels organically and is now buying physical on-ramps directly (chains of bodegas, kiosks, prepaid-credit shops in emerging markets). The two issuers are playing different games, and Tether's game is more profitable by an order of magnitude.
Liquidity stress in a redemption scenario. Tether's cash position is near zero (0.04% of assets); the model relies on T-bill secondary-market liquidity and Cantor Fitzgerald's ability to liquidate same-day. In 2022, coordinated short-selling triggered $7B in USDT redemptions within 48 hours and $25B within 20 days. Tether honored every redemption. That stress test was at $80B in circulation. At $186B today, the stress test has not been re-run.
Regulatory bifurcation. The launch of USA₮ on January 27, 2026 (federally regulated under the GENIUS Act framework, issued by Anchorage Digital Bank, custodied by Cantor Fitzgerald) creates two structurally separate Tether products. USDT remains the offshore product. USA₮ is the onshore institutional product. Bo Hines, former Executive Director of the Presidential Council of Advisors for Digital Assets, runs Tether USA₮. The strategic logic is sensible. The risk is contamination: by launching a highly regulated onshore product, Tether implicitly acknowledges USDT is not built to the same standard. If institutional markets start treating the two products as proxies for each other's reputation, opacity around USDT can pressure USA₮ trust by association.
What to Watch
Five concrete watch items for the next 6 to 12 months:
- Big Four audit announcement. The 2025 CFO hire signals staffing for an eventual full-financial-statement audit. The first published audit from a Big Four firm would re-rate Tether's institutional acceptance materially. Until then, the trust gap persists.
- Reserve composition drift toward 24%+ in higher-risk assets. S&P's downgrade rationale tied directly to this trend. If the share of non-cash-equivalent reserves continues climbing toward 30%+, the rating cuts deeper, and the macro-infrastructure framing weakens.
- USA₮ supply growth and onshore institutional uptake. Adoption among US-licensed payment providers, custodians, and trading venues. Either it scales as the regulated alternative to USDC for Tether-distribution-network users, or the duplication of effort suggests USDT can not be defended onshore.
- TRON chain-share trend. If WDK cross-chain routing actually pulls volume off TRON, Tether's chain-concentration risk decreases. If TRON's share holds at ~44% or grows, the concentration risk compounds.
- Sovereign-Treasury comparison drift. Brookings flagged the issuer category as marginal Treasury buyer. If Tether's direct T-bill position grows past the $150B level, it crosses into the top 15 of all reported holders globally and the macro-infrastructure framing becomes harder to dismiss.
Closing Thoughts
The dominant frame for Tether in crypto coverage is still circa-2019: a stablecoin issuer with reserve transparency questions and a regulatory cloud. That frame produces incomplete analysis of the 2026 entity. The 2026 entity is a money-market-fund-shaped business holding $122B in direct US Treasuries, generating roughly $10B in annual profit at a 23% rate-cycle drag, distributing through hundreds of millions of users in jurisdictions where the dollar is the savings product, and quietly buying physical on-ramp infrastructure across Latin America, Africa, and Asia.
The reserve question is real. The audit gap matters. The S&P downgrade is a defensible signal. None of those concerns negate the macro-infrastructure read. They are the macro-infrastructure read. A privately-held entity with sovereign-scale Treasury exposure, no full audit, and a redemption mechanism that depends on T-bill market liquidity is the new shape of stablecoin systemic risk. The right policy response, the right regulatory framework, and the right institutional positioning are still being worked out across multiple jurisdictions.
For TokenIntel users: USDT is not a TI signal asset. There is no buy-hold-sell read. The relevant questions are upstream: how does Tether's reserve composition drift affect the Stablecoin Liquidity Score? How does TRON chain concentration affect counterparty risk for assets that route through USDT? How does the Treasury-holdings position affect the macro-liquidity read TI uses to set its regime classifier? Those are the inputs that matter for TI's framework, and this report is calibration material for them.