Over the past 90 days, Tether added more than 30 million new users. USDT supply, meanwhile, grew by only $446 million. Let that tension sit for a second. If 30 million people were actually adopting a currency, you would expect the float to balloon. It didn't. And yet Tether's stablecoin market share climbed above 60%. Those two facts shouldn't live in the same headline.
The traditional read is that Tether is simply too big to keep growing at the same rate. Fair enough. But I've spent the last decade watching protocol balance sheets, and the more interesting explanation is that adoption is happening at the edges, not in the treasury. New users aren't buying USDT to park wealth; they're using it to move money and leave. That is utility, not loyalty. It's also the exact situation where accounting opacity matters least—until it matters most.
On July 31, Tether released its Q2 2026 report. The headline numbers: USDT issuance around $184.6 billion, total assets $187.751 billion, total liabilities $183.642 billion, and token-related liabilities at $183.622 billion. Assets exceed liabilities by about $4.11 billion. Net operating profit landed near $1.5 billion, driven mostly by U.S. Treasury and repurchase positions. Tether also said it reduced secured loan exposure by $2.38 billion—a 15% cut—while adding 14 tonnes of gold to bring its vault over 146 tonnes. BDO prepared the report. CEO Paolo Ardoino said USDT remains fully backed despite volatility in gold and Bitcoin.
Let me translate what this report really is. It is a balance-sheet snapshot taken at one moment in time, prepared by an outside accountant under agreed-upon procedures. That is not an audit. An attestation can confirm the pieces are there. It doesn't test what happens when the pieces lose value. It doesn't stress-test a run. And it doesn't answer the question that matters for global money: what do users actually own? Tether says “fully backed.” What it doesn't say is “fully governed.”
It's worth sitting with the word “full” for a moment. Paolo Ardoino is right that assets currently exceed liabilities. But “fully backed” is a point-in-time phrase. It doesn't include a stress-testing scenario. It doesn't include a liquidity ranking. It doesn't tell you how long it would take to convert 146 tonnes of gold into usable dollars during a global bank holiday. It doesn't tell you what happens to the secured-loan book when the collateral managers lose their nerve. You can be fully backed and still be temporarily illiquid; those two states are not identical. For a global money system, illiquidity is indistinguishable from insolvency.
This distinction was theoretical in 2020. It is no longer theoretical. In a huge part of the world, USDT is the banking account people never had. It's how they escape inflation, cross-border payment fees, and arbitrary national controls. Thirty million new users isn't a vanity metric; it is a signal that stablecoins have become civic infrastructure. But civic infrastructure should not be a private monopoly's revenue stream. That's the tension no attestation resolves.
One insight hides in plain sight: the profit engine. Tether earned $1.5 billion in a single quarter, mainly from U.S. Treasuries and repos. Let me spell out the mechanism: USDT holders deposit real dollars with Tether. Tether gives them a token with no interest. Tether then buys government bonds and earns yield. The yield goes to Tether, not to the people who provided the money. This is not a bug; it is the business model. It is exactly how banks treat depositors, except that stablecoin users have no deposit insurance and no legal claim to the yield. Every underbanked person using USDT is quietly subsidizing a profit machine.
And here is where my audit background makes me pause. In 2017, I reviewed over 40 early Ethereum projects for governance weaknesses. The ones that looked best on paper were often the most dangerous because the backers were skilled at narrative, not engineering. When I read that Tether cut secured loans by 15%, my first reaction isn't relief; it's arithmetic. A 15% reduction implies roughly $13.5 billion in secured loans still remain on the books. “Secured” is a comforting word. It doesn't tell you how the collateral is priced, who controls it, or whether it can be sold quickly in a global dash for exits. Most importantly, it doesn't tell you what happens to those loans if the collateral's value is exactly what's crashing.
Then there's the gold. Adding 14 tonnes to reach more than 146 tonnes sounds like responsible reserve management. Gold is one of the oldest stores of value in human history. But gold is also a cumbersome asset for a stablecoin. It doesn't earn yield, it isn't instantly transferable in a digital liquidation, and its value can swing violently in the same week as crypto. Why buy gold if your dollar token is backed by Treasuries? Because gold is what institutions buy when they expect volatility, not just in markets but in the political and regulatory system around those markets. This looks less like a reserve strategy and more like an escape plan.
There is one more detail that only an industry veteran would notice. Tether says it is continuing to advance the audit process with a Big Four accounting firm. Notice the verb: “advancing.” Not “completed.” If a Big Four audit was imminent, the words would be different. Every quarter, this promise gets mentioned. Every quarter, it stays a promise. Maybe that's because a full audit would require revealing too much about counterparties, the loan book, and operating agreements. Or maybe the big firms have decided the reputational risk still outweighs the fee income. Either way, the market has learned to treat a future audit as a credit card for the present.
The other strange number is the user growth versus token supply. 30 million new users with only $446 million of extra issuance suggests a tiny average holding size. A lot of people are moving $20 or $50 at a time. In a sense, this is wonderful: stablecoins are finally everyday money. In another sense, it reveals fragility. The users who trust USDT most are the ones least able to absorb a failure. The users who understand the trust structure are often the ones holding Tether short. That asymmetry should worry anyone who believes in open finance.
Now, the contrarian part of me wants to push back on the cynicism. The traditional financial system has no real-time public disclosure. Banks routinely lend out 90% of depositor money and call it liquidity. The Federal Reserve balance sheet is a full-employment program for quantitative analysts, yet no one demands the Fed publish a real-time breakdown of every counterparty. Tether, by contrast, has produced an independent attestation, a $4.11 billion cushion, and a pile of Treasuries that would make most central banks blush. By any reasonable standard, Tether is doing more transparency than the system it's replacing. If we demand impossible perfection from Tether, we might push users toward unregistered schemes with no reserves at all.
But this pragmatic argument has a hard limit. When a stablecoin holds more than 60% market share, it stops being a product and becomes a systemic node. A single failure would take down every bridge, every exchange, every lending market built on USDT. Decentralized finance was supposed to distribute risk across a network. Instead, we've centralized trust into a single balance sheet and called it permissionless.
I keep repeating one sentence to myself in these moments: democracy isn't a transaction where every voice holds weight. It's a structure, not a receipt. Tether's report is a receipt. It tells you the assets are there. It doesn't tell you who gets to vote when the protocol changes. It doesn't tell you whether the people using USDT as a lifeline will ever get a seat in the room where governance decisions are made. “Fully backed” is doing a lot of work in this report. “Fully accountable” would be doing much more.
Here is the question I want you to carry into the rest of this cycle: Can a private dollar genuinely be a public good? Tether has solved the engineering problem—it can issue a stable token at scale. It has solved the balance-sheet problem—it holds more assets than liabilities. What it hasn't solved, and may never solve, is the moral problem. If reserves are real but representation is absent, we have simply recreated a bank and given it a token ticker.
I'm not asking Tether to become a charity. I'm asking all of us to stop equating a clean balance sheet with an open system. The path forward isn't just more audits; it's more participation. It's stablecoin governance that actually responds to the people who use it. It's reserve policies that are encoded, not merely reported. It's an industry that realizes democracy isn't a transaction where every voice holds weight—it's an ongoing argument, and the argument should include the unbanked, not just the auditors.
By 2027, we will know how far the Big Four audit promises will actually go. We will know whether stablecoin monopolies can coexist with the decentralization ethos that gave them life. The next bear market won't be caused by a chart crossing a moving average. It will be caused by the moment a trustworthy-looking balance sheet meets a real-world rush for the exit. Until then, maybe the smartest thing we can do is stop asking “Is USDT safe?” and start asking a harder question: “Who gets to decide what safe means?”
I don't want to pretend Tether is uniquely evil. It is not. It has done what the market demanded: build a dollar bridge for places that need one. But democracy isn't a transaction where every voice holds weight—it is a practice, not a utility. Bridges become toll roads. Toll roads are not emancipation. The moment we confuse convenience with liberation, we have lost the plot.
I don't have a clean answer. But I know this much: stability isn't a static number; it's a living relationship between promise and redemption. Tether gave us real numbers. Now I'm still waiting for a mechanism.
