The sUSD Wind-Down: Why an On-Chain, Over-Collateralized Dollar Still Lost Its Peg

Most of the synthetic dollars that broke in the last year broke in the dark. By the protocols' own disclosures and the reporting that followed, Stream's loss sat with an off-chain fund manager, Resolv's mint authority sat behind a single off-chain key, and Elixir's backing was exposed to Stream. In each case a holder could not see the thing that failed until the money was already gone. sUSD is the opposite case, and that is exactly why it is worth a careful read.

Synthetix's sUSD was about as transparent as a crypto dollar gets. It was backed by SNX staked on-chain, over-collateralized by design, with its collateral and its debt readable on Ethereum and Optimism the entire time. There was no custodian, no off-chain signer, no private reserve you had to take on faith. And in June 2026 it traded around twenty-five cents on the dollar, and its own governance voted to wind it down. If you could see all of the backing and it still failed, the obvious question is what you were supposed to be watching instead.

The answer is the part of a synthetic dollar that is not the reserves: the mechanism that defends the peg, and the governance that can change that mechanism. sUSD did not lose its backing to a thief. It lost the incentive that held its price to a dollar, and it lost it through a vote that anyone could read in advance. This is a teardown of how that happened, what it means for verifying any synthetic dollar, and how Kerne answers the same question for kUSD.

What just happened to sUSD

In June 2026, Synthetix governance moved to retire sUSD under SIP-423, a proposal introduced on June 12 by founder Kain Warwick and core contributor Benjamin Celermajer. Rather than try once more to restore the peg, the proposal winds the token down: it freezes the sUSD contract across Ethereum mainnet and Optimism, takes a holder snapshot, and pays holders back at face value in newly minted SNX, at a conversion of four SNX per one sUSD. That rate values SNX at a reference twenty-five cents and sUSD at its intended dollar. The compensation comes with strings: a one-year lock followed by a one-year linear vest, so holders are made whole on paper but not in liquid tokens for some time. Roughly forty million sUSD were still in circulation, about forty million dollars of face value to settle.

The price context is the reason a wind-down, rather than another rescue, was on the table. sUSD had drifted off its dollar through 2025 and never durably returned. It reached an all-time low near twenty-two and a half cents on June 23, 2026, about seventy-seven percent below par, and was trading around twenty-five cents, roughly three-quarters below its dollar, as governance moved to retire it. With the protocol now shifting its focus to perpetual futures on Ethereum mainnet, the legacy synthetic dollar was being closed rather than repaired. None of this is contested: it is Synthetix's own stated decision, made in public governance, and we report it as that and nothing more.

Why it depegged: a peg defense removed by a vote

Here is the part that matters for everyone else. sUSD's depeg is widely traced, in the protocol's own discussion and in independent analyses, to an earlier governance change rather than to any loss of collateral. The relevant change was SIP-420, which moved staker debt into a protocol-owned debt pool, a kind of delegated staking that let SNX stakers pool their debt and run at a lower collateralization ratio.

The mechanism is subtle and worth stating plainly, because it is the whole lesson. Under the older design, when sUSD traded below a dollar, individual stakers had a direct, personal reason to buy it back: each staker carried their own debt denominated in sUSD, and repaying that debt cheaply by buying discounted sUSD was profitable. That self-interest was the peg defense. Thousands of stakers quietly bidding sub-dollar sUSD is what pulled it back to par. When debt moved into a shared, protocol-owned pool, that individual incentive weakened. No single staker had the same reason to step in, and the self-correcting buyer of last resort that had defended the peg for years was no longer there in the same form. The collateral did not vanish. The reason anyone would defend the price did.

We are describing a reported and analyzed cause, not alleging fault. Governance changes are legitimate, they trade one set of properties for another, and Synthetix's own founder publicly took responsibility for the outcome and explored a basis-vault replacement before the protocol settled on a wind-down. The point is not that someone was careless. The point is that a property a holder was relying on, the thing that kept sUSD at a dollar, was an emergent feature of an incentive structure, and that structure was changed through ordinary governance that was fully visible in advance. The risk was legible. It just was not in the reserve table.

The inverse of the off-chain failures

Put sUSD next to the year's other breaks and a pattern sharpens. By their own disclosures, Stream, Resolv, and Elixir each failed where a holder could not see: an off-chain manager, a single off-chain mint key, an exposure routed through another protocol. apxUSD, by contrast, traded below a dollar in full view, its collateral a published, on-chain-legible basket, so the move was painful but never a surprise. We walked that one, with sources, in our apxUSD piece.

sUSD belongs with apxUSD on the legible side, and that is the uncomfortable part. Its reserves were visible the whole way down. A holder who only checked whether the backing existed would have seen a fully collateralized, on-chain dollar right up to the point it was trading at twenty-five cents. Reserve transparency, the thing proof-of-reserves is built to confirm, would have shown green. It was real, and it was not enough, because the failure was not in the reserves. It was in the mechanism that converts reserves into a stable price, and in the governance that changed that mechanism.

That is the upgrade to the verification thesis sUSD forces. Seeing the assets is necessary. It is not sufficient. A synthetic dollar is reserves plus a peg-defense mechanism plus a set of people who can change both. You have to be able to read all three.

Three things to verify, not just the reserves

Before depositing into any synthetic dollar, the reserve check is the first question, not the last. Add two more, both answerable from public sources.

1. Can you see the reserves? Are the backing assets readable on-chain, or do you stop at a dashboard figure you cannot re-derive? This is the question most coverage starts and ends with, and it is the one sUSD passed.
2. Can you see what defends the peg? What actually holds the price to a dollar: arbitrage against redeemable collateral, an incentivized buyer, a protocol backstop? Is that mechanism written into contracts you can read, or does it depend on a behavior that a parameter change can switch off? sUSD's defense was an incentive, and an incentive is a setting, not a guarantee.
3. Can you see who can change the rules, and how fast? Which address or multisig can alter the mint logic, the collateral ratio, or the incentive structure? Is it one key or a multisig? Is there a timelock between a decision and its effect, so a holder has time to exit? sUSD's change was legitimate and public, but the lesson holds for the cases that are not: the governance surface is part of the risk surface.

A token can pass the first question and fail the next two. That is precisely what happened here.

How Kerne answers the same three questions

kUSD has to survive its own checklist, so here is where it stands, including the parts that are still maturing.

The reserves. kUSD's backing lives on Base and is readable with raw on-chain calls: USDC held one for one in the Peg Stability Module, vault collateral, and a Hyperliquid hedge. The protocol publishes an hourly, cryptographically signed proof of reserves you can re-derive yourself rather than read off a dashboard, at /verify and /api/por.

The peg defense. kUSD holds its dollar through redeemable collateral rather than a discretionary incentive. The Peg Stability Module mints kUSD against USDC at one to one and redeems the other way, so the arbitrage that defends the price is a contract function, not a behavior that a parameter can quietly switch off. It is the kind of mechanism you can read instead of trust.

Who can change the rules. This is the question sUSD turns into the important one, so we answer it directly. The authority to administer kUSD's mint path is held by a two-of-three Gnosis Safe multisig, not a single key, and the mint privilege itself sits with on-chain contracts whose mint amount is derived from the contract's own USDC balance, not from an off-chain instruction. A one-transaction kill switch can halt minting if something looks wrong. Anyone can read who holds those roles, in three calls against Base, the same way we walked it in our Resolv mint-path teardown, and we assert those values continuously at /api/risk-status, rendered for humans at /risk.

We are not claiming kUSD has removed every trust boundary. It is pre-audit, at Genesis scale, its hedge runs on a single venue and is self-reported, and a per-day mint cap is in the contract source but not yet on the deployed minter. Those open gaps are listed, not hidden, at /legible, and kUSD's place in the field, including where it does not lead, is in the synthetic-dollar scorecard. The difference sUSD highlights is narrower and more honest than immunity: the mechanism that defends kUSD's peg, and the set of people who can change it, are things you can read before a bad day rather than reconstruct after one.

How to check any of this yourself

The skill this post is really teaching applies to anything you hold, not just kUSD. Here is where to run it.

• Verify any signed attestation in your browser. /verify is a free tool that runs on your machine, recovers the signer of any issuer's reserve attestation, rehashes the payload, and checks freshness. It proves a snapshot is authentic and fresh, not that an issuer is solvent, and it works on any signed attestation, not only ours.
• Commission an independent read of a counterparty. If you hold or lend against someone else's synthetic dollar, /counterparty-verification gives you a signed, independent read of that counterparty's public on-chain reserves and the authority surface around them, a proof you hold rather than their dashboard. One-off reviews run $5,000 to $15,000, scoped to your exposure.
• Get paged when a peg or a parameter moves. /monitoring is a standing watch on a peg, a reserve ratio, and the freshness of a feed, alerting your Discord or Telegram the moment a value crosses a line you set. From $99 a month. It tells you when something moves; it does not certify anything.

Conclusion

sUSD is not a cautionary tale about on-chain transparency failing. Its transparency worked exactly as advertised. It is a cautionary tale about what transparency was pointed at. A holder watching the reserves saw a fully backed dollar the whole way down, because the thing that broke was never in the reserves. It was the incentive that defended the price, and that incentive was a governance setting that a vote could change, in full public view, without anyone needing to touch the collateral.

So the question that sorts the field is not only whether you can see the backing. It is whether you can see the mechanism that holds the peg, and who can change it, and how fast. Run that on every synthetic dollar you hold, including ours. The reserves are where the checking starts. They are not where it ends.

Figures are as of June 30, 2026. Synthetix sUSD: the SIP-423 wind-down, the four-SNX-per-sUSD conversion at a reference $0.25 with a one-year lock and one-year vest, the freeze across Ethereum mainnet and Optimism, and the roughly 40 million sUSD outstanding are per The Defiant and Crypto Briefing; SIP-423 was introduced June 12, 2026 by Kain Warwick and Benjamin Celermajer and moved through Synthetix governance later that month. The all-time low near $0.225 on June 23, 2026 and the roughly $0.25 level are per public market trackers as reported by those outlets. The root-cause attribution to SIP-420's protocol-owned debt pool and the resulting loss of the individual buy-back incentive is per Synthetix's own SIP-420 and the analyses by Cointelegraph and crypto.news; it is presented as the reported and analyzed cause, not an allegation of fault, and the founder publicly accepted responsibility for the outcome. The earlier failures referenced for contrast, Stream, Resolv, and Elixir, are described per those protocols' own disclosures and the reporting that followed, and the Resolv case is walked with full sourcing in our Resolv mint-path teardown; apxUSD's sub-dollar trading and on-chain-legible collateral are covered with sources in our standalone apxUSD analysis. Kerne's own claims resolve to live endpoints: the hourly signed proof of reserves at /api/por/signed, the reserve breakdown at /api/por, and the live risk surface at /api/risk-status. Nothing here is investment or legal advice.