USDD — Asset Risk Assessment (Full)

Category: Stablecoin | Peg Mechanism: Hybrid (overcollateralized vaults + PSM + legacy TRX burn) | Issuer: TRON DAO Reserve (Justin Sun)

Note: This report covers USDD 2.0 (launched January 25, 2025), the current live version. USDD 1.0 (May 2022 - January 2025) is documented in the Historical Context section for background. The two versions have different architectures, different contracts, and different risk profiles — though the issuer and many structural concerns carry over.

Summary

USDD 2.0 is an overcollateralized stablecoin on TRON, forking MakerDAO’s smart contract architecture. Users deposit TRX, sTRX, or USDT into vaults to mint USDD. On Ethereum and BNB Chain, a Peg Stability Module (PSM) allows 1:1 swaps against USDT/USDC.

Total circulating supply is $1.558B across all chains (Tron: $1.328B, Ethereum: $124M, BNB: $111M), with total backing of $1.578B (101.3% headline CR). On-chain analysis shows ≈70.8% of collateral originates from HTX (Justin Sun’s exchange) via the Smart Allocator (≈$985M) and PSM (≈$132M). Independent user vaults contribute ≈$461M (29.6% independent CR). On-chain overlap verification confirmed no double-counting — the SA uses Spark on Ethereum (not Aave), and JustLend positions are separate from HTX’s. However, the admin multisig can access all user funds instantly (zero timelock) per the protocol’s own ChainSecurity audit — contradicting marketing claims of “no admin keys.”

Trajectory since the last review (Apr 1 → Jul 13, 2026): supply grew ≈16% ($1.344B → $1.558B), but the growth was funded by HTX Smart Allocator expansion, not independent user collateral (user vaults roughly flat at ≈$461M). As a result the headline surplus thinned from +14.8pp to +1.3pp (headline CR 114.8% → 101.3%), the independent/ex-HTX CR fell from 34.7% → 29.6%, and HTX’s share of backing rose from 69.8% → 70.8%. The “improving, HTX-share-declining” trend noted in the April review has reversed — USDD is now marginally more HTX-dependent with a thinner buffer. Peg held throughout (≈$0.999).

Live data: Backing Monitor (hourly on-chain updates)

⚠ Dashboard data-quality note (2026-07-13): the hourly backing snapshot at 2026-07-13T01:39Z returned a transient TRON-side Smart Allocator read of $0 (the smart_allocator.tron field), which collapsed the displayed headline CR to a false 48.55% and raised a spurious “CRITICAL — UNDERCOLLATERALIZED” flag. This is a data artifact, not a backing event: the immediately preceding hourly entry (00:40Z) read a clean CR 101.28% / backing $1,578M, USDD held ≈$0.999, and the supply/user-vault reads were unaffected. All figures in this report use the clean series, not the artifact snapshot. Flagged to the backing-monitor pipeline for a TRON-RPC read guard.

Contracts: Tron: TPYmHEhy5n8TCEfYGqW2rPxsghSfzghPDn | Ethereum: 0x0C10bF8FcB7Bf5412187A595ab97a3609160b5c6 | BNB Chain: 0xd17479997f34dd9156deef8f95a52d81d265be9c

Key Risk Notes (USDD 2.0 — Current)

Centralized control under Justin Sun — No functional DAO exists despite “TRON DAO Reserve” branding. The governance portal was removed. All major decisions made unilaterally. The admin multisig (unknown signers) has full system control with zero governance delay.
Admin can access user funds — ChainSecurity audit (Jan 2025) confirms the admin can “access the funds of users” via customExec() and execSpell() with no timelock. This directly contradicts USDD’s “no admin keys” marketing.
≈70.8% of collateral sourced from HTX — The Smart Allocator (≈$985M, 62%) and PSM (≈$132M, 8%) are both funded from HTX addresses. Independent user vaults contribute ≈$461M (30%). Trend reversed vs the April review: HTX share rose from 69.8% as 16% supply growth was funded by the SA, not user collateral. Heavily concentrated.
No double-counting detected (verified on-chain) — The SA deploys to Spark on Ethereum (not Aave) and has separate JustLend positions on Tron. HTX uses Aave on Ethereum. Different protocols = different positions. Verified by querying all 27 HTX PoR addresses live. Funds are same-source (HTX) but not same-position. See HTX Overlap Analysis.
Circular TRX backing — TRX vaults hold ≈$461M in user collateral. With the headline buffer now only +1.3pp, a ≈50% TRX drop pushes headline CR to ≈34% on the current mix — solvency depends almost entirely on the HTX-funded Smart Allocator holding.
Thin, concentrated liquidity — 24h volume ≈$4-5M. Concentrated on Sun-controlled venues (HTX, Poloniex, SunSwap).
PSM thinning structurally — total PSM reserves ≈$132M (down from $164M in April) as supply grew, so coverage fell to ≈8.5% of supply. Tron PSM ≈$48.5M (3.7% of Tron supply — critically thin). Ethereum PSM ≈$74.6M (60.1%). BNB PSM ≈$13.4M (12.1%). The acute -30%/24h Tron+Eth drain observed on Apr 1 did not recur, but Tron-side coverage has continued to erode.
Unsustainable yield subsidies — Launched at 20% APY, now 8-12%. Justin Sun: “fully subsidized” because “we have plenty of money.”
Terra/Anchor parallel — 61% of USDD 2.0 supply on JustLend at subsidized yield mirrors Anchor holding 75% of UST before collapse.
SEC fraud charges settled — Market manipulation, unregistered securities. $10M penalty (March 2026). Congressional analysis: 58% of illicit crypto finance on TRON.

USDD 2.0 Architecture (Current)

Peg Mechanism

USDD 2.0 operates differently across chains:

On TRON (primary issuance):

Overcollateralized vault model, forked from MakerDAO. Users deposit TRX, sTRX, or USDT as collateral to mint USDD.
Automated liquidations enforce solvency at per-vault minimum collateralization ratios.
Stability fees: 0.5% for TRX vaults, 1.0% for USDT and sTRX vaults.
Legacy: whitelisted institutions can also mint USDD by burning TRX through the authorized burning contract (retained from 1.0).

On Ethereum and BNB Chain:

Peg Stability Module (PSM): Fixed-rate 1:1 minting/redemption against USDT and USDC with zero fees and zero slippage.
This provides a hard price floor/ceiling on these chains — but only as deep as the PSM’s USDT/USDC reserves.

PSM reserves (on-chain, July 13, 2026):

Chain	PSM USDT	USDD Supply	PSM Coverage
Tron	≈$48.5M	$1.328B	3.7%
Ethereum	≈$74.6M	$124M	60.1%
BNB Chain	≈$13.4M	$111M	12.1%
Total	≈$132M	$1.558B	8.5%

PSM trend: Total PSM reserves have declined from $164M (Apr 1) to ≈$132M while supply grew ≈16%, so coverage fell from 12.2% to ≈8.5%. The mix shifted — Ethereum PSM rose (≈$58M → ≈$74.6M, now 60.1% of ETH supply) while Tron PSM eroded further (≈$70M → ≈$48.5M, just 3.7% of the $1.328B Tron float). The acute -30%/24h drain flagged on Apr 1 was a one-day event and did not repeat.

PSM Limitations (Why It Doesn’t Mitigate Core Risk)

The PSM is the strongest mechanical feature of USDD 2.0, but it does not meaningfully offset the HTX concentration and circular backing risks:

Scale mismatch: ≈$132M in PSM reserves vs $1.558B in outstanding USDD. The PSM covers only ≈8.5% of supply, and coverage has thinned as supply outgrew reserves.
Tron PSM is critically thin. Tron PSM has ≈$48.5M USDT covering $1.328B in supply (3.7%). Ethereum PSM at ≈$74.6M covers 60.1% of its smaller $124M supply.
Replenishment is opaque and discretionary. No documented mechanism exists for PSM refills. Presumably TRON DAO Reserve (Justin Sun) tops it up at their discretion — making it a discretionary backstop, not a mechanical guarantee.
Same-source risk. Both the PSM reserves and the Smart Allocator originate from HTX-controlled addresses. If HTX has a solvency issue, both the PSM and the collateral behind it are compromised simultaneously.
Death spiral drain sequence. If TRX crashes and Tron-side collateralization deteriorates, rational actors drain the PSMs for clean USDT. Ethereum PSM ($111M) gets drained first as the best exit. Once empty, the remaining ≈$1B+ in USDD has no stable exit.

sUSDD (yield-bearing wrapper):

Launched October 2025 on Ethereum and BNB Chain.
Accrues yield through increasing sUSDD-to-USDD redemption rate (similar to sDAI).
“Smart Allocator” deploys reserves into Aave and JustLend — introduces indirect protocol risk.
Current rates: 8% APY on Ethereum ($314.5M TVL), 6% on BNB Chain ($8.8M TVL).

USDD 2.0 Collateral — What’s Real

On-chain collateral breakdown (live, July 13, 2026):

Component	Value	% of Backing	Source	Independent?
Smart Allocator (HTX)	≈$985M	62.4%	HTX addresses	No
PSM Reserves (HTX)	≈$132M	8.4%	HTX addresses	No
TRX / sTRX / USDT Vaults (User)	≈$461M	29.2%	Open market	Yes
Total Backing	≈$1.578B

The Smart Allocator (HTX) figure is derived from the clean backing series (total backing × HTX share, less PSM); the raw hourly snapshot on 2026-07-13 had a corrupted TRON SA read and is not used (see data-quality note above).

Data source: On-chain VAT queries across Tron (TH5dhX...), Ethereum (0xff77...), and BNB Chain (0x41f1...). Live dashboard: Backing Monitor.

The collateral resides in MakerDAO-fork smart contracts (Vat), not in HTX wallets directly. The Smart Allocator deploys to Spark on Ethereum and JustLend on Tron (verified on-chain — not Aave as previously reported). This provides structural protection against third-party attacks. However, the admin multisig can access these funds at any time (see Admin Control).

Metric	Value	Trend (vs Apr 1)
Headline CR	101.3%	↓ from 114.8% (buffer thinned)
Independent CR (user vaults only)	29.6%	↓ from 34.7%
HTX share of backing	70.8%	↑ from 69.8% (more concentrated)
Surplus over supply	≈$20M (+1.3pp)	↓ from $199M (+14.8pp)
PSM reserves	≈$132M	↓ from $164M (coverage 8.5% vs 12.2%)

The April review’s “improving” thesis has reversed. Over ≈3.5 months supply grew ≈16% ($1.344B → $1.558B) but user-vault collateral stayed roughly flat (≈$461M), so the growth was carried by the HTX-funded Smart Allocator. The headline buffer collapsed from +14.8pp to +1.3pp, the independent (ex-HTX) CR fell to 29.6%, and HTX concentration ticked up. Peg held (≈$0.999) throughout, but the standalone equity cushion above par is now negligible.

Per-ilk detail (on-chain, as of April 1, 2026 — Tron VAT not re-queried this cycle; treat as structural shape, not current balances):

Ilk	Chain	Debt	Ceiling	Utilization	HTX?
SA001-A	Tron	$520M	$1.0B	52%	Yes
SA001-A	Ethereum	$318M	$620M	51%	Yes
TRX-A	Tron	$170M	$200M	85%	No
PSM-USDT-A	Ethereum	$111M	$500M	22%	Yes
PSM-USDT-A	Tron	$106M	$600M	18%	Yes
TRX-B	Tron	$80M	$200M	40%	No
TRX-C	Tron	$61M	$100M	61%	No
TRX-D	Tron	$17M	$50M	34%	No
PSM-USDT-A	BNB	$18.5M	$200M	9%	Yes
STRX-A	Tron	$8.9M	$50M	18%	No

HTX Overlap Analysis (Confirmed vs Suspected)

This section distinguishes between what is confirmed by on-chain evidence and what is suspected but unverifiable.

Data sources:

HTX PoR wallet addresses: github.com/huobiapi/Tool-Node.js-VerifyAddress (27 addresses across ETH, TRX, BTC)
Live on-chain verification: Backing Monitor HTX Overlap panel queries all 27 HTX PoR addresses hourly

Double-Counting: Resolved — No Overlap Detected

On-chain verification (April 1, 2026) confirmed that USDD Smart Allocator positions and HTX exchange positions are separate:

Chain	HTX Position	SA (USDD) Position	Overlap?
Ethereum	$276.5M in Aave aUSDT (1 address)	$0 in Aave — uses Spark instead	No — different protocols
Tron	$17.4M in JustLend (1 address)	$105.1M in JustLend (separate contract)	No — separate positions

The SA on Ethereum deploys to Spark (not Aave as previously reported based on Protos June 2025 tracing). This means HTX’s Aave positions and USDD’s Spark positions cannot be the same funds — they are on different lending protocols entirely. On Tron, both HTX and the SA use JustLend, but the positions are held in separate contracts with separate balances.

Verdict: Funds are same-source (HTX) but not same-position. No double-counting of lending protocol positions.

HTX PoR Addresses Still Active

All three previously flagged addresses remain in HTX’s PoR (verified live against all 27 addresses):

0x18709E89BD403F470088aBDAcEbE86CC60dda12e (Ethereum) — $276.5M in Aave aUSDT. This address funded the SA’s Ethereum deployment (Protos June 2025). The SA subsequently deployed to Spark, not Aave.
TDToUxX8sH4z6moQpK3ZLAN24eupu2ivA4 (Tron) — $17.4M in JustLend + $19.7M in USDD tokens. This address funded the SA’s Tron deployment. It also holds $19.7M in USDD directly — not included in HTX’s PoR (HTX excludes USDD from PoR).
TZ1SsapyhKNWaVLca6P2qgVzkHTdk6nkXa (Tron) — 118M TRX (≈$37M). The original USDD 1.0 double-counted address. Live balance has decreased from 1.064B TRX (March PoR CSV) to 118M TRX — approximately 946M TRX moved out, likely redistributed across other HTX wallets.

Additional finding: HTX addresses collectively hold 3.38B TRX (≈$1.065B) across 11 Tron addresses, plus $413.8M in stablecoins and $276.5M in Aave lending positions across 10 Ethereum addresses. All monitored live via the Backing Monitor.

Structural Risk Remains

Even without double-counting, the counterparty concentration risk is unchanged. One entity (Justin Sun) controls:

The collateral source (HTX)
The collateral contracts (admin multisig, zero timelock)
The primary lending deployment (JustLend, Sun-founded)
The primary trading venues (HTX, Poloniex, SunSwap)

A single event (regulatory action, hack, insolvency) compromises both sides simultaneously.

Circular Backing Risk

The heavy reliance on TRX as collateral creates a reflexive risk loop:

TRX price drops → reserve value falls → collateralization ratio declines
USDD confidence drops → sell pressure → USDD depegs
Depeg erodes confidence in TRON ecosystem → further TRX selling
Return to step 1

This is the same dynamic that destroyed UST/LUNA in May 2022 — the very event that occurred weeks after USDD first launched.

Terra/Anchor Parallel

This structure mirrors Terra/UST before its collapse:

Anchor held ≈75% of UST supply at ≈20% subsidized yield → JustLend holds ≈61% of USDD 2.0 supply at ≈20% subsidized yield
Both yields were explicitly subsidized by the issuer rather than generated organically
Both had circular backing (UST backed by LUNA; USDD backed by TRX)
Both had single points of control despite “decentralized” branding

USDD Supply by Chain (On-Chain, July 13, 2026)

Chain	Supply	% of Total	PSM Coverage
Tron	$1.328B	85.2%	3.7%
Ethereum	$124M	8.0%	60.1%
BNB Chain	$111M	7.1%	12.1%
Total	$1.558B		≈8.5%

Note: Tron supply grew from $1.091B (Apr 1) to $1.328B, driving essentially all of the ≈16% total-supply increase, while Ethereum and BNB floats were roughly flat. This is why Tron PSM coverage thinned further (to 3.7%) even though total PSM reserves held near their April level.

Issuer & Governance Risk

Justin Sun / TRON DAO Reserve

USDD is de facto controlled by Justin Sun. Despite “DAO” branding:

No functional DAO governance exists — the governance portal was removed
Only one governance vote has ever been held (May 2023, regarding burned TRX usage)
USDD 2.0 launch and all parameter changes decided unilaterally
Admin multisig signers are unknown

Justin Sun also controls HTX (formerly Huobi), Poloniex, and is a major investor in World Liberty Financial (Trump-linked). The concentration of control across the exchange, protocol, and stablecoin creates compounding counterparty risk.

Regulatory History

SEC charges (March 2023): Unregistered offer/sale of TRX and BTT, market manipulation ($31M in wash trading), undisclosed celebrity promoter payments. Settled March 2026 for $10M penalty.
Congressional scrutiny: Senator Warren and House members alleged preferential SEC treatment due to Sun’s $75M investment in Trump-linked crypto project. Congressional analysis cited 58% of illicit crypto finance occurring on TRON.
Dominica “authorized digital currency” status (October 2022) provides minimal practical regulatory protection.

Audit & Transparency

Smart Contract Audits

Five audits completed as of October 2025:

ChainSecurity — USDD v2 contracts on TRON (January 2025)
CertiK — USDD on Ethereum (September 2025), awarded AA score (87.5/100)
ChainSecurity — Ethereum and BNB Chain deployments (October 2025)
Two additional audits (details not fully public)

No critical vulnerabilities reported. However, the audit explicitly notes the economic model is out of scope, and that “there are many ways in which the governance can cause the system to break or lead to losses for its users.”

Admin Control vs “No Admin Keys” Claim (Audit Contradiction)

USDD marketing claims “no administrative keys” and “immutable contracts.” The ChainSecurity audit (January 2025) directly contradicts this. From the Roles & Trust Model (Section 2.2.10):

“The admin oversees the entire USDD V2 system and hence is fully trusted. The admin is able to trigger the execution of any privileged functions in the system through the DSPauseProxy (i.e. with customExec() and execSpell() defined in GovActionsProxy) and access the funds of users.”

The governance architecture, per the audit:

Admin Multisig → owns → GovActionsProxy → authority → DSPauseProxy → all system contracts

Key findings from the audit’s trust model:

The admin multisig (unknown signers) has full control over the entire system
customExec() and execSpell() allow the admin to execute arbitrary calls through DSPauseProxy
The admin can access user funds directly
Governance delay is set to 0 — meaning changes execute instantly with no timelock for users to react
Oracle signers are “fully trusted” and can manipulate prices within deviation bounds
The Flop operator can withdraw all auction collateral

This means: while collateral does reside in MakerDAO-fork smart contracts (not in HTX wallets directly), the admin multisig can move those funds at any time, instantly, without warning. The contracts provide structural protection against third-party attacks but no protection against the admin itself. The “no admin keys” marketing claim is false per the protocol’s own audit.

Reserve Attestations

None. No independent third-party reserve attestations from accounting firms exist. This is a significant gap compared to peers:

USDC has Deloitte attestations (monthly)
USDT has BDO attestations (quarterly)
USDD has only self-reported on-chain data from the protocol itself

Liquidity Analysis

Exchange Presence

Venue	Type	Notes
HTX (Huobi)	CEX	Sun-controlled — primary venue
Poloniex	CEX	Sun-controlled
KuCoin, Bybit, Gate.io	CEX	Independent but thin volume
SunSwap	DEX (Tron)	Sun-controlled
Uniswap V3	DEX (Ethereum)	Limited pools
PancakeSwap V2	DEX (BNB Chain)	Limited pools

Volume is extremely thin. The most active independent pair (USDD/TRX on KuCoin) shows only ≈$59K in daily volume. Liquidity is heavily concentrated on Sun-controlled platforms, which means exit liquidity in a crisis may be unreliable — the same entity controlling the stablecoin controls the primary trading venues.

Cross-Chain Considerations

Tron: Primary chain, deepest liquidity (such as it is) via SunSwap and JustLend
Ethereum: PSM provides 1:1 exit to USDT/USDC, but PSM reserves are limited (≈$74.6M)
BNB Chain: Smallest deployment (≈$111M supply), thinnest liquidity
Defunct bridges: Previously bridged to Fantom, Avalanche, Arbitrum, Optimism via Multichain (collapsed 2023) — those bridged versions are likely stranded

Scoring Rationale

Category	Score	Notes
Peg Mechanism	4.0	MakerDAO-fork vault model is a real structural improvement. PSM provides clean exit but coverage has thinned to ≈8.5% of supply (Tron side just 3.7%). TRX burn mechanism retained. System untested under severe market stress.
Backing	3.0	≈70.8% of collateral from one counterparty (HTX) — concentration ticked up vs April as supply growth was funded by the HTX Smart Allocator, not user vaults. Independent (ex-HTX) CR fell to 29.6% and the headline buffer is now only +1.3pp. On-chain verification confirmed no double-counting (SA uses Spark, HTX uses Aave — different protocols); collateral is real and in separate positions. No third-party attestations. Admin can access funds with zero delay. Score held at 3.0: the position-separation basis that lifted it off 2.5 still holds, but the “growing independent collateral” tailwind has reversed, so no further credit.
Liquidity	3.0	24h volume ≈$4-5M. Concentrated on Sun-controlled venues. PSM exit on Ethereum improved (60.1% of ETH supply), but Tron PSM is critically thin (3.7% of $1.328B) and eroding.
Issuer	2.0	De facto single-person control (Justin Sun). No functional DAO. SEC fraud charges (settled). Unknown admin multisig signers. “No admin keys” claim contradicted by own audit. Congressional allegations of TRON enabling illicit finance.
Overall	3.5	High risk, no longer improving — USDD 2.0’s architecture is a genuine mechanical improvement and on-chain verification still confirms no double-counting (positions are separate), which is what holds the score at 3.5. But the April “improving” thesis reversed this cycle: 16% supply growth ($1.344B → $1.558B) was carried by the HTX Smart Allocator, the independent CR fell to 29.6%, the headline buffer thinned to +1.3pp, and HTX concentration rose to 70.8%. The admin can still access all funds instantly and a TRX shock breaks solvency on the current thin margin. Would deteriorate toward 3.0 if the headline buffer erodes below par or HTX share climbs further; would improve only if independent collateral resumes growing and an admin timelock is enabled.

Chain-Specific Notes

Chain	Adjusted Overall	Notes
Tron	3.5 (default)	Primary chain, deepest (relative) liquidity, but fully within Sun-controlled ecosystem. PSM critically thin at 3.7% coverage.
Ethereum	4.0	PSM provides 1:1 exit to USDT/USDC at 60.1% coverage. Verified no overlap with HTX Aave positions.
BNB Chain	3.0	Smallest deployment, thinnest liquidity, but PSM provides 12.1% coverage.

Comparison to Peers

	USDD	FRAX (Legacy)	crvUSD	USDC
Market Cap	$1.558B	≈$153M	≈$60M	≈$52B
Collateral Ratio	101.3% headline / 29.6% independent	91.4% (under-collateralized)	>100% (algorithmic)	≈100% (fiat)
Reserve Verification	Self-reported; admin has full access	On-chain, auditable	On-chain, transparent	Deloitte attested
Governance	None (admin multisig, unknown signers)	DAO (active)	DAO (active)	Corporate (regulated)
Depeg History	Multiple under 1.0 (worst: 8%); untested under 2.0	Mild (≈$0.99)	Mild	One major (SVB)
Overall Score	3.5	2.5	5.0	9.0

USDD 2.0 scores higher than Legacy FRAX on headline collateralization, but FRAX’s risks are primarily mechanical (wind-down mode, frozen AMOs) rather than counterparty/trust risks. USDD’s risks are fundamentally about trust in a single individual and unverifiable claims — a qualitatively different and arguably more dangerous risk profile.

USDD 1.0 Historical Context (May 2022 – January 2025)

This section documents the deprecated USDD 1.0 for historical context. USDD 1.0 still has ≈$262M in circulating supply across chains (Tron: ≈28M, Ethereum: ≈124M, BSC: ≈111M) but is no longer the active system. The risks documented here informed the design of 2.0 but are distinct from current 2.0 risks.

USDD 1.0 Architecture

USDD 1.0 launched May 5, 2022 as an algorithmic stablecoin using a UST-style TRX burn/mint mechanism. The TRON DAO Reserve held collateral (TRX, BTC, USDT, USDC) in multi-sig wallets as an additional backstop.

USDD 1.0 Depeg History

Date	Trigger	Low Price	Duration
June 2022	3AC/Celsius/Voyager collapses	≈$0.96	Weeks
Nov 2022	FTX collapse; Alameda USDD selling	≈$0.985	Days
Dec 2022	Continued post-FTX pressure	≈$0.9695 (ATL)	Weeks
Mar 2023	SVB crisis, broad stablecoin contagion	≈$0.92	Days

USDD 1.0 HTX Double-Counting (Confirmed, January 2024)

Address TZ1SsapyhKNWaVLca6P2qgVzkHTdk6nkXa held ≈972 million TRX (≈$120M), nearly all staked. This address was simultaneously:

Listed as USDD 1.0 collateral by the TRON DAO Reserve
Included in HTX’s proof-of-reserves as exchange assets

The same tokens were counted twice — once as collateral backing USDD, once as reserves backing HTX exchange user deposits. First reported by Protos (January 2024), originally discovered by Twitter user @NoCryptFish cross-referencing HTX’s PoR snapshot (December 2023) against USDD’s reserve page. HTX never responded to press inquiries. No governance vote authorized storing USDD collateral at HTX.

The USDD 1.0 BTC reserves were also stored in wallet 1KVpuCfhftkzJ67ZUegaMuaYey7qni7pPj, which HTX claimed as their own. Those 12,000 BTC (≈$726M) were later removed from USDD reserves entirely (mid-2024) without a governance vote.

Bluechip (stablecoin rating agency, backed by MakerDAO’s Rune Christensen) rated USDD 1.0 an F and estimated effective collateralization at just 53% due to the HTX overlap. Their recommendation: do not use USDD.

USDD 1.0 Concentration

Chain	USDD 1.0 Supply	Concentration
Tron	≈28M	56% in single address
Ethereum	≈124M	≈99% in bridge contract
BSC	≈111M	≈99.5% in bridge contract

USDD 1.0 Whitelisted Minters (Defunct Partners)

Authorized minting institutions included several now-defunct or compromised entities:

Alameda Research — bankrupt, embezzled billions from FTX
Multichain — bridge hacked, CEO arrested, team lost web access
TPS Capital — connected to Three Arrows Capital, website down
FalconX — settled with CFTC for failure to register
Other partners (Mirana, Amber Group, Ankr, Wintermute) did not respond to Protos inquiries

Status Under USDD 2.0

The original double-counted address (TZ1Ss...) is not referenced in post-2.0 Protos investigations (June/September 2025). It is unclear whether the USDD 2.0 restructuring retired this address from USDD’s side. However, as of the March 2026 HTX PoR snapshot, this address is still listed in HTX’s proof-of-reserves with 1,064,078,456 TRX (≈$336M). HTX never removed it. Whether USDD still claims it as collateral is unconfirmed — but HTX is still counting it as exchange reserves.

Monitoring Recommendations

To properly monitor USDD 2.0, the following would be needed:

Tron chain support — Current backing monitor only covers EVM chains. Need TronGrid API integration to read reserve wallet balances and USDD supply on Tron.
Smart Allocator position tracing — Trace the Smart Allocator’s Aave and JustLend positions and cross-reference against HTX’s published PoR wallet addresses to determine whether positions overlap (confirming or ruling out double-counting under 2.0).
PSM depth monitoring — Track USDT/USDC reserves in Ethereum and BNB Chain PSMs. Alert if PSM reserves drop below meaningful exit thresholds.
TRX price sensitivity — Calculate “stress collateral ratio” at TRX -30%, -50%, -70% to quantify death spiral risk.
Admin multisig activity — Monitor the admin multisig for customExec() / execSpell() calls through GovActionsProxy, which could signal fund movements or parameter changes.
Cross-chain supply reconciliation — Sum USDD supply across Tron, Ethereum, and BNB Chain (both 1.0 and 2.0) to verify total supply matches claims.

Sources

USDD 2.0 Investigations (Current)

USDD 1.0 Investigations (Historical)

HTX Proof-of-Reserves

HTX PoR Wallet Addresses (CSV/XLSX, monthly) — March 2026 snapshot cross-referenced against USDD addresses
HTX PoR Verification Tool

General

Revision history

2026-07-13 — weekly refresh + backing-methodology reconciliation: supply $1.344B → $1.558B (+16%), backing $1.544B → $1.578B, headline CR 114.8% → 101.3%, independent/ex-HTX CR 34.7% → 29.6%, HTX share 69.8% → 70.8%, PSM reserves $164M → ≈$132M. The April “improving” thesis reversed — 16% growth was funded by the HTX Smart Allocator, not user vaults, so the headline buffer thinned to +1.3pp and concentration rose. Scores unchanged (peg 4.0, backing 3.0, liquidity 3.0, issuer 2.0, overall 3.5); rationale narrative flipped from “improving” to “no longer improving.” Added a dashboard data-quality note: the hourly snapshot at 2026-07-13T01:39Z had a transient TRON-side Smart Allocator read of $0 that produced a false 48.55% CR / “UNDERCOLLATERALIZED” flag — a fetch artifact, not a backing event; all figures use the clean series (≈$0.999 peg held).
2026-04-01 — verified position separation: on-chain overlap check confirmed no double-counting (SA on Spark, HTX on Aave); overall raised 3.0 → 3.5.

Live Backing Dashboard

Live data Open in new tab ↗

Open live dashboard