Tether Reserve Buffer Explained: What the $7.95B Surplus and Q3 2026 Report Mean for USDT Holders

 

The excess reserve buffer is reserve assets minus liabilities, measuring how much issuer equity sits ahead of token holders' claims. It is neither profit nor the reserve total. In Tether's third-quarter figures published on October 30, those four numbers were reserve assets of $191.451 billion, liabilities of $183.496 billion including $183.470 billion of issued tokens, excess reserves of $7.955 billion, and net operating profit of $1.5 billion for the quarter. The buffer equals 4.33% of liabilities, and assets cover liabilities at 104.33%.

 

What matters most this quarter is not that the buffer nearly doubled from $4.11 billion to $7.955 billion but where the increase came from. Of the $3.84 billion rise, roughly $700 million came from gold appreciation and roughly $2.47 billion from bitcoin's rally, together $3.17 billion or 82.5% of the increase. A buffer whose purpose includes absorbing declines in gold and bitcoin therefore expanded mainly because gold and bitcoin rose, and if those prices reverse the buffer contracts on both sides of the equation at once. All figures are as of September 30, 2026, attested by BDO.

 

 

 

Cash-like reserves, meaning Treasury bills, repo and cash, totalled $139.58 billion or 72.9% of assets, down from 74.9% at the end of June, and down $1.06 billion in absolute terms because overnight reverse repo fell $9.11 billion while bills rose $7.88 billion. Cash-like assets cover 76.1% of liabilities, leaving a gap of roughly $43.92 billion filled by $19.08 billion of gold, $8.37 billion of bitcoin, $14.18 billion of secured loans, $6.00 billion of other investments and $4.18 billion of public equities. Bitcoin alone equals 105.2% of the buffer, gold 239.9%, and the two together 345.1%, meaning a combined 29.0% decline in both would erase the buffer entirely. On unit economics, Tether's quarterly net operating profit per dollar of token outstanding was 0.8174%, or 3.27% annualised, 13.0 times Circle's net income on the same basis.

 

 

These four get used interchangeably while answering completely different questions, and separating them is a precondition for judging what the buffer actually means.

 

Reserve assets of $191.451 billion answer how much the issuer holds. Liabilities of $183.496 billion answer how much the issuer owes token holders, of which $183.470 billion is issued tokens, or 99.99%, with the small remainder other liabilities. Excess reserves of $7.955 billion are the difference, answering what remains after holders' claims are satisfied in full, and because they rank behind holders they constitute a first-loss layer. Net operating profit of $1.5 billion answers how much was earned in the quarter, making it a flow where the other three are stocks.

 

Comparing a flow against a stock requires labelling the basis. The $1.5 billion quarterly profit equals 18.9% of the $7.955 billion buffer, a flow-to-stock ratio meaning that rebuilding the entire buffer at the quarter's earning rate would take slightly more than five quarters. It is not a yield of any kind. Net operating profit for the first three quarters totalled $4.5 billion, averaging exactly $1.5 billion per quarter, and the company described the third-quarter figure as consistent with the previous quarter.

 

For holders the relevant figure is the buffer as a share of liabilities, 4.33%. It means that in the event of write-downs on non-cash-like assets, issuer equity can absorb losses equal to 4.33% of liabilities before holder principal is touched. How USDT works mechanically is what makes this so: a holder owns a redemption claim on the issuer, not direct title to any particular reserve asset.

 

 

The buffer rose from $4.11 billion on June 30 to $7.955 billion on September 30, an increase of $3.84 billion. Breaking it down, roughly $700 million came from gold appreciation and roughly $2.47 billion from bitcoin's rally, together $3.17 billion or 82.5% of the increase, with roughly $670 million from other movements.

 

The bitcoin line illustrates the point particularly clearly. Reserve bitcoin rose from $5.80 billion at the end of June to $8.37 billion at the end of September, an increase of $2.57 billion, while actual additions during the quarter were only about 1,130 coins. The reference price rose from $58,642 to $83,621.23, a gain of 42.6%. Almost the entire $2.57 billion increase therefore came from price rather than purchases. The September 30 line corresponds to roughly 100,060 bitcoin.

 

This creates a structural circularity. Part of the buffer's purpose is to serve as a first-loss layer when gold, bitcoin and other non-Treasury assets fall in value, yet this quarter's expansion came mainly from those same assets appreciating. When prices move the other way, the decline in carrying value reduces reserve assets and therefore reduces the buffer simultaneously, with both sides moving together, leaving the buffer thinnest exactly when it is most needed.

 

Price moves since the attestation date already allow this to be tested. Bitcoin currently trades near $77,446, 7.4% below the September 30 reference of $83,621.23. Applied to the roughly 100,060 coins disclosed, the line would be worth about $7.75 billion, roughly $620 million below the attested value. Holding everything else constant, the buffer would fall from $7.955 billion to about $7.33 billion. That is a calculation from the disclosed coin count and the current price rather than a company figure, and the actual position also depends on purchases, sales and movements in other assets during the period.

 

 

Setting the two quarters' line items side by side makes the direction of change more informative than the total.

 

 

The line items come from the two BDO reports. The four cash-like items fell from $140.64 billion to $139.58 billion, a decline of $1.06 billion, taking their share of total assets from 74.9% to 72.9%. The driver was overnight reverse repo falling $9.11 billion, more than offsetting the $7.88 billion rise in bills.

 

That substitution is a mild liquidity downgrade but a limited one. US Treasury bills carry a weighted average maturity under 90 days and overnight reverse repo converts to cash the following day, so both are highly liquid, differing mainly in the certainty and price risk of conversion. What warrants more attention is the absolute size of non-cash-like assets, $51.87 billion at the end of September or 27.1% of total assets, within which the $14.18 billion of secured loans and $6.00 billion of other investments have no disclosed detail on their realisation path.

 

Direct and indirect US Treasury bill holdings together total $132.4 billion, a figure above the $122.84 billion direct line in the table, with the difference coming from indirect exposure through money market funds and similar vehicles. Citing these two numbers requires naming which basis applies, since otherwise they produce different conclusions about the Treasury share.

 

 

Redemption capacity depends on assets convertible quickly at carrying value, not on the reserve total. On September 30 figures, cash-like assets of $139.58 billion against liabilities of $183.496 billion give coverage of 76.1%, slightly below the 76.6% at the end of June, leaving a gap of roughly $43.92 billion.

 

The meaning of that gap needs stating precisely. It does not indicate that Tether cannot meet redemptions, since overall asset coverage of liabilities is 104.33% and assets exceed liabilities. What it indicates is that if redemption demand in a short window exceeded 76.1% of liabilities, meeting the remainder would depend on selling gold, bitcoin, secured loans, other investments and public equities, assets whose realised prices under stress may fall below carrying value. The first 76.1% of redemptions is mechanically straightforward, and anything beyond introduces asset-sale price risk.

 

Historically, single-day stablecoin redemptions rarely approach a double-digit percentage of liabilities, so 76.1% cash-like coverage is ample in the large majority of scenarios. The genuine tail risk is a redemption wave coinciding with falling prices in the non-cash-like assets, because those two events are highly correlated in crypto: a market decline triggers redemptions, and the same decline depresses the realisable value of bitcoin and public equities. That is a correlation risk rather than a problem with the level of liquidity itself.

 

Worth noting is a capability the issuer holds outside the reserves, namely control over issued tokens. Tether said it froze nearly $1.57 billion of assets connected to the Binance security incident. That capability and reserve adequacy are separate matters: it affects the recoverability of stolen funds rather than the ability to meet ordinary redemptions, and the two should not be conflated.

 

 

A long-circulating claim needs updating here because it is no longer accurate.

 

This quarter's report is indeed an attestation rather than an audit. BDO performed it under agreed-upon procedures, confirming the state of reserve assets and liabilities at a single date, covering neither the full financial statements nor internal controls. Every figure above carries the date September 30, 2026, and movements since then are not included.

 

Tether has, however, completed one full financial statement audit. KPMG US issued an unqualified opinion on its 2025 financial year with a balance sheet date of December 31, 2025, published on August 14, 2026, covering the balance sheet, income statement, statement of changes in equity and cash flow statement, and examining underlying transactions and counterparties. That is where the difference from an attestation lies: an audit covers systems, valuations and internal controls, while an attestation confirms balances on one date.

 

Three qualifications belong alongside that audit. The opinion covers only the entity Tether International, S.A. de C.V., not the wider group that includes several British Virgin Islands entities and Tether Limited in Hong Kong. The full audit report is not publicly available, and what is public is the company's own summary. And the audit's reserve surplus of $6.814 billion differs from the BDO attestation figure for the same date by roughly $500 million, a gap that has not been explained.

 

The accurate statement is therefore that Tether has one annual audit of limited scope whose full text is not public, while quarterly disclosure remains attestation. That differs both from "never audited" and from "fully transparent." For holders the practical implication is that annual figures carry a higher level of assurance than quarterly ones, and the point-in-time nature of quarterly data requires readers to track asset price movements after the attestation date themselves. Issuer attestations and exchange proof of reserves are also two different mechanisms, and on verifying the latter, how to assess whether an exchange's funds are safe and how to check proof of reserves covers it separately, since neither substitutes for the other.

 

 

Placing the two issuers side by side shows how differently the profit from this business is distributed. The underlying activity is identical, holding reserve assets and earning the short-end rate, while the unit economics differ by an order of magnitude.

 

 

The gap originates in the distribution structure. Circle's agreement with Coinbase directs all reserve income on USDC held on Coinbase's platform to Coinbase, with Coinbase receiving 50% of residual income on USDC circulating elsewhere. The result is that $412.5 million of Circle's $667.7 million in second-quarter reserve income flowed out as distribution, transaction and other costs, and of the $618 million of incremental income from a 100 basis point rise, $315 million flows to distribution at the same time. Tether's public disclosure contains no revenue share of comparable magnitude.

 

A sharper comparison still: Tether's quarterly net operating profit per dollar of token, 0.8174%, equals 93.7% of Circle's quarterly reserve income per dollar of token, 0.8728%. What Tether retains as operating profit is close to everything Circle books as revenue before paying anyone.

 

Two points must accompany this comparison or it invites misreading. The first is basis: Tether's net operating profit and Circle's net income from continuing operations are differently defined, Tether does not publish a full quarterly income statement so the composition of its figure cannot be verified from the attestation, and Circle's number comes from financial reporting filed with regulators, so the levels of assurance differ. The second, and more important for holders, is that issuer profitability and holder protection are separate things. Neither token pays interest to holders, with reserve income accruing entirely to the issuer and its distribution partners, so the 13-fold gap is a fact about two companies rather than about two tokens. Judging the tokens themselves means looking at reserve composition, buffer depth and disclosure quality, not at how much the issuer earns. On the specific differences in reserve structure, regulatory path and fees, the USDT and USDC comparison works through them item by item. Readers wanting to check USDT's live scale and depth at any point can set the attestation date and the current moment side by side on the USDT price and market data page, which shows what happened between two disclosures better than reading the quarterly report alone. Stablecoin pairs on MEXC typically see short-lived shifts in trade composition around major disclosures.

 

 

The line in this report that deserves most attention is not $7.955 billion but the figure 82.5%. A buffer thickened mainly by appreciation in the very assets it exists to protect against is procyclical by construction: thicker when markets are good, thinner when they are bad, and the moment it is needed is the latter. Bitcoin alone already equals 105.2% of the buffer, gold 239.9%, and the two together 345.1%, meaning a combined 29.0% decline in both suffices to erase it entirely. That is not a remote scenario, since bitcoin has already fallen 7.4% since the attestation date, which on the disclosed coin count implies a buffer around $7.33 billion with everything else held constant.

 

What the market is most likely to misread is treating a near-doubling of the buffer as improved reserve quality. What actually happened was two components moving in opposite directions: the cash-like share fell from 74.9% to 72.9% with the absolute amount down $1.06 billion, while non-cash-like assets rose, mostly on price rather than purchases. The buffer's number got bigger and the assets behind it became more dependent on market prices. Both facts sit in the same report, and reading only one produces the opposite conclusion.

 

The metric genuinely worth tracking from here is cash-like coverage of liabilities rather than the absolute buffer. That ratio currently stands at 76.1%, half a percentage point below the end of June. It measures what scale of redemption can be met without selling any risk asset, which is the single figure most directly relevant to holders. Should it keep declining in later quarters while the non-cash-like share rises, then reserve structure would be weakening in substance even if the absolute buffer sets new highs on rising prices.

 

The cross-asset implication is that stablecoin issuers have become large long positions in short-end rates, an exposure running opposite to most crypto assets. The $132.4 billion of direct and indirect Treasury bill holdings, together with the fact that both Tether's and Circle's reserve income correlate positively with short rates, means that in the current tightening environment issuer profitability diverges from the risk appetite of the market they serve. The consequence is that rising rates simultaneously lift issuer revenue and depress crypto market activity, and the latter determines token scale. For holders this means the periods when issuer financials look strongest may be the periods when token circulation is under pressure, and the two metrics need reading separately.

 

 

 

It is reserve assets minus liabilities, measuring how much issuer equity sits ahead of holders' claims. As of September 30, 2026, reserve assets were $191.451 billion against liabilities of $183.496 billion, giving excess reserves of $7.955 billion, equal to 4.33% of liabilities, with assets covering liabilities at 104.33%. Because it ranks behind holders in any claim ordering, it functions as a first-loss layer.

 

 

It rose from $4.11 billion at the end of June to $7.955 billion at the end of September, an increase of $3.84 billion, of which roughly $700 million came from gold appreciation and roughly $2.47 billion from bitcoin's rally, together 82.5% of the increase. Reserve bitcoin grew by only about 1,130 coins during the quarter while the reference price rose from $58,642 to $83,621.23, so the increase came almost entirely from price rather than purchases.

 

 

As of September 30, Treasury bills were $122.84 billion, overnight reverse repo $9.52 billion, term reverse repo $7.18 billion and cash deposits and non-US bills $0.04 billion, together $139.58 billion of cash-like assets or 72.9% of total. Non-cash-like holdings comprise $19.08 billion of gold, $14.18 billion of secured loans, $8.37 billion of bitcoin, $6.00 billion of other investments, $4.18 billion of public equities and $0.06 billion of corporate bonds. Direct and indirect Treasury bill holdings together total $132.4 billion.

 

 

Cash-like assets cover 76.1% of liabilities, slightly below 76.6% at the end of June, leaving a gap of roughly $43.92 billion. This does not indicate an inability to pay, since overall asset coverage of liabilities is 104.33%. It indicates that the first 76.1% of redemptions is mechanically straightforward while anything beyond requires selling gold, bitcoin, secured loans and similar assets whose realised prices under stress may fall below carrying value. Bills carry a weighted average maturity under 90 days and overnight repo converts the next day.

 

 

An attestation is performed under agreed-upon procedures and confirms reserve and liability balances at a single date, covering neither the full financial statements nor internal controls. An audit covers the balance sheet, income statement, statement of changes in equity and cash flow statement, and examines underlying transactions, systems, valuations and internal controls. This quarter's third-quarter report is a BDO attestation dated September 30, 2026.

 

 

Once, with limited scope. KPMG US issued an unqualified opinion on its 2025 financial year with a balance sheet date of December 31, 2025, published on August 14, 2026. Three qualifications apply: the opinion covers only the entity Tether International, S.A. de C.V. and not the wider group including British Virgin Islands entities and Tether Limited in Hong Kong; the full report is not public and what is available is the company's summary; and the audit's reserve surplus of $6.814 billion differs by roughly $500 million from the attestation figure for the same date, a gap that has not been explained.

 

 

One is a flow and the other a stock, so no yield can be derived by comparing them. The $1.5 billion is third-quarter net operating profit, with $4.5 billion over the first three quarters, while $7.955 billion is the stock of buffer on September 30. Quarterly profit equals 18.9% of the buffer, meaning rebuilding it entirely at that earning rate would take slightly more than five quarters, which is not a return of any kind.

 

 

It does, and the relevant comparison is against the buffer rather than against total assets. Bitcoin at $8.37 billion equals 105.2% of the buffer and gold at $19.08 billion equals 239.9%, with the two together at 345.1%, meaning a combined 29.0% decline would erase the buffer entirely. Bitcoin has already fallen 7.4% since the attestation date, which on the disclosed coin count implies a buffer around $7.33 billion holding everything else constant.

 

 

That inference does not follow, because these are two separate matters. Tether's quarterly net operating profit per dollar of token was 0.8174% against Circle's 0.0630% on the same basis, a factor of 13.0, with the gap arising mainly because Circle pays 58.8% of revenue to partners as distribution and transaction costs. But neither token pays interest to holders, and reserve income accrues entirely to the issuer and its partners, so the gap is a fact about two companies. Judging the tokens means examining reserve composition, buffer depth and disclosure quality.

 

 

This article is for general information only and does not constitute investment advice, financial advice, legal advice, tax advice or any trading recommendation. Prices of crypto assets, equities and other related financial assets can move sharply, and the historical data, company disclosures and third-party statistics cited here cannot guarantee future results, while the attestation and audit figures referenced are point-in-time snapshots that will change with subsequent disclosures and asset price movements. Readers should conduct their own research and reach independent judgments based on their own financial circumstances, investment objectives and risk tolerance, consulting a qualified professional where appropriate. The MEXC Crypto Pulse team accepts no liability for any direct or indirect losses arising from use of the information in this article.

 

 

MEXC Editorial Team provides the latest platform updates, product launches, trading feature guides, and industry insights. With a focus on developments in the global crypto market, it delivers transparent information, professional data, and forward-looking analysis. Both beginners and experienced traders can find firsthand updates, in-depth observations, and official announcements related to the MEXC ecosystem.

 

 

 

Want the fastest access to MEXC's latest updates? Join our official Telegram group now!

MEXC Platform: Download App | Zero Fee Trading | Buy USDT Now

Join MEXC Community: X (Twitter) | Telegram | Discord

Get Insights: MEXC Learn | Terms of Service | MEXC Blog | MEXC News

Account Verification: Understand KYC | How to Complete KYC

External Content Platforms: Substack | Medium | Paragraph | LinkedIn | X(News)