USDD: when the issuer pulled the Bitcoin

For much of its life USDD's published reserves included a large Bitcoin position — on the order of 12,000 BTC, worth roughly $726 million in August 2024. That holding was the reserve's single largest asset that had nothing to do with the TRON ecosystem: an exogenous, deeply liquid store of value that could hold its worth even if TRX did not.

In August 2024 that Bitcoin was removed from the listed collateral. On-chain analysts traced a large share of it moving to the HTX exchange, and the announcement came from Justin Sun's personal account rather than the TRON DAO Reserve, with no recorded governance vote on a decision of that size. What remained behind USDD was overwhelmingly TRX — reports cited roughly 98% — alongside a small USDT position.

Crucially, the token did not lose its peg in the process. USDD continued to trade near a dollar throughout. The event here is not a price collapse but a change in the quality and independence of the backing, and the governance questions that surrounded how that change was made.

The structural problem is that the remaining backing was correlated to the issuer. TRX is TRON's native token; its value rises and falls with confidence in the same ecosystem that issues USDD. A stablecoin backed mostly by its sponsor's own asset has no independent floor — in the precise scenario where you would most want hard backing, a loss of faith in the ecosystem, the collateral is falling for the very same reason.

This is the flaw that destroyed Terra, in a milder and non-algorithmic form. UST had no reserve at all, only a reflexive mint-burn relationship with LUNA, so a run hyperinflated the backing token without limit. USDD does hold real collateral and never relied on that doom loop, but removing the uncorrelated Bitcoin leg pushed it toward the same family of risk: the asset standing behind the dollar promise is wired to the issuer rather than independent of it.

Pharos treats issuer-correlated collateral as a structural fragility rather than a neutral design choice. Backing that moves with the sponsor's own token is weighted as lower-quality collateral in resilience scoring, because diversification away from the issuer is what gives a peg something to lean on when the issuer's ecosystem is the thing under stress.

The substance of the change was compounded by how it was made. A roughly $726 million reserve decision surfaced through an individual's social-media account, not through the entity nominally responsible for the reserve, and without a governance vote on record. For a project that markets itself around a decentralized reserve, that gap between the decentralization claim and the observed control was the sharper story.

It also validated an earlier warning. The rating firm Bluechip had already placed USDD at the bottom of its stability scale, citing precisely the TRX concentration and the thin transparency that the Bitcoin withdrawal then put on full display. A published transparency page is only as useful as the governance behind the assets it lists; here the page changed and the assets moved, with little in the way of binding process to explain or constrain it.

The durable lesson is that reserve transparency is two things, not one: the disclosure of what backs a coin, and the governance that controls whether that backing can change unilaterally. A dashboard that lists Bitcoin today tells you little if a single party can move it tomorrow without a vote.

In January 2025 USDD relaunched as USDD 2.0, a materially different system. The earlier model leaned on TRON DAO subsidies and hybrid-algorithmic mechanics; the new one is an overcollateralized Collateralized Debt Position protocol. Users lock TRX, sTRX, or USDT into publicly verifiable vaults and mint USDD against them under enforced minimum collateral ratios, with undercollateralized vaults cleared through permissionless liquidations and a Peg Stability Module offering 1:1 swaps against USDT and USDC.

This is a real improvement on the structure that drew the 2024 criticism. On-chain, auditable collateral with rule-based liquidations and no discretionary subsidy is a stronger foundation than an opaque reserve a single party can rebalance off-platform. By the protocol's own one-year review, collateral value stayed above circulating supply through 2025 and most vaults ran well above their minimum ratios.

What the redesign does not erase is the correlation concern. TRX and sTRX remain core collateral, so a sharp, ecosystem-wide drawdown in TRX would still stress the system through liquidations and vault health, even with the CDP guardrails working as intended. The mechanism is sturdier; the dependence on the issuer's own asset is reduced and governed, but not removed.

Collateral quality is about independence, not just amount. A high headline collateralization ratio means little if the collateral is the issuer's own ecosystem token, because that backing falls in the same crisis it is meant to cushion. Removing the one uncorrelated reserve asset can weaken a stablecoin structurally even when its price never moves — which is why this is read as a wounding, not a death.

Transparency is disclosure plus governance. Listing reserves on a page is necessary but not sufficient; what matters is whether a single party can move those reserves without a binding vote. The USDD episode is a clean example of a published reserve changing through an off-platform decision, and of why Pharos tracks the control over reserves alongside their composition.

A redesign can fix the mechanism without fully fixing the exposure. USDD 2.0's overcollateralized, on-chain CDP is a genuine step up from the model it replaced, but core collateral still skews toward TRX and sTRX. The structural improvement is real and worth crediting; the residual issuer-correlation is real too, and is the reason the durable signal here remains how much of a coin's backing is independent of the entity that issues it.