ART-E: Anonymous RWA Token Trading

• ART-E is a hybrid marketplace architecture that enables anonymous trading of RWA tokens (both ERC-20 and ERC-721/1155) while integrating off-chain KYC/AML processes.
• The system decouples deposits and withdrawals using an omnibus contract and secure off-chain channels, ensuring minimal on-chain exposure and enhanced privacy.
• ART-E offers competitive gas efficiency and lays the groundwork for future cryptographic enhancements such as zero-knowledge proofs to further secure anonymous transactions.

ARTeX is a hybrid on-chain/off-chain marketplace architecture designed to facilitate anonymous trading of Real-World Asset (RWA) tokens, encompassing both fungible (FT, e.g., ERC-20) and non-fungible tokens (NFT, e.g., ERC-721/1155). The system addresses the privacy and compliance challenges unique to RWA token transactions on transparent blockchains, introducing a workflow that decouples deposit and withdrawal events via an omnibus contract and off-chain KYC and communications infrastructure. ARTeX prioritizes practical privacy by organizational means while offering a blueprint for future cryptographic enhancements (Lee et al., 10 Oct 2025).

1. System Components and Operational Workflow

ARTeX comprises both on-chain and off-chain modules, integrating traditional KYC/AML backends with an omnibus smart contract architecture. The on-chain module centers on a smart contract (or a suite of contracts) deployed on Ethereum, which aggregates all deposited RWA tokens without directly linking user addresses to individual deposits or withdrawals. Standard token transfers (ERC-20, ERC-721, ERC-1155) interface with this contract.

Off-chain infrastructure includes:

• Web interfaces and API gateways for user interaction
• KYC database mapping real-world identities to internal user-IDs ({ID_s}, {ID_b})
• A secure channel for private key delivery (e.g., end-to-end-encrypted messaging or SFTP)
• Off-chain auction engine supporting standard mechanisms (English, Dutch, sealed-bid, etc.)

The transaction workflow is partitioned into three phases:

Phase	Key Actions	Anonymity Mechanism
Deposit	Seller logs in, submits token metadata, transfers token to omnibus contract, passes KYC/compliance review	All deposits via contract; user addresses never exposed
Matching	Buyers bid off-chain; auction engine selects winner; ARTeX generates ephemeral payment addresses	Payment split across contract-controlled addresses
Withdrawal	Buyer pays into omnibus; ARTeX disperses proceeds per seller's request; contract transfers token to a freshly created wallet for buyer	Fresh withdrawal address; private key delivered over secure off-chain channel

This workflow ensures only minimal on-chain activity is visible: the contract receives tokens from sellers, receives payments from buyers, releases proceeds to sellers, and delivers tokens to buyers, with all addresses contract-controlled.

2. Anonymity Guarantees and Informal Security Model

ARTeX does not deploy any on-chain zero-knowledge proof mechanisms (e.g., zk-SNARKs or Bulletproofs) in its initial implementation. Privacy is achieved through architectural and organizational means, with all sensitive mappings and credentials kept strictly off-chain.

The system defines an anonymity set $A = \{d_1, \ldots, d_N\}$, where $N$ is the number of deposits in an epoch. The privacy target is to ensure that, for any withdrawal $w$,

$$\Pr[\text{Adversary links } w \to d_i] = \frac{1}{|A|}$$

ARTeX models this with an indistinguishability game: the adversary selects two deposits $d_i, d_j \in A$; ARTeX randomly withdraws from one; the adversary guesses which. The theoretical bound for adversarial success is $1/2 + \epsilon$, where ideally $\epsilon = 0$ if all funds move exclusively through the omnibus contract, although side-channel information (e.g., timing, gas usage) may increase $\epsilon$ in practice (Lee et al., 10 Oct 2025).

Potential cryptographic extensions, not presently implemented, include:

• zk-SNARKs over Pedersen commitments for shielded deposit/withdrawal proofs
• Bulletproofs for confidential range proofs
• Merkle accumulators for withdrawn notes (à la Tornado Cash)

3. Token Handling and Linkage Resistance

ARTeX supports both FT and NFT modalities. For FT (e.g., ERC-20), token amounts are aggregated; for NFT, individual tokenIDs are managed directly in the omnibus contract. Settlement ensures that each withdrawal (of FT or NFT) goes to a newly generated address, which is created and delivered to the buyer off-chain.

Linkage resistance rests on the following practices:

• All on-chain operations employ contract-controlled multisig omnibuses, not revealable user addresses
• Buyers may split payments across multiple ephemeral contract addresses; sellers may similarly request proceeds to multiple addresses
• All withdrawals utilize fresh wallets generated per transaction and communicated only over private channels
• No public (or off-chain) disclosure of timing, order, or block identifiers tied to user activity

These measures mitigate address linkage and transaction graph analysis, with privacy fundamentally rooted in the decoupling of deposit and withdrawal flows.

4. Compliance, Security, and Disclosure Controls

ARTeX enforces comprehensive off-chain KYC/AML compliance before any deposit is permitted, and all on-chain addresses are whitelisted post-verification only. The identity linkage {ID_real ↔ deposit address ↔ on-chain deposit}, and similarly for buyers and withdrawal addresses, is stored solely off-chain in the ARTeX database.

Under legal request, ARTeX can selectively disclose these mappings from its database. No cryptographic selective-disclosure protocols (e.g., CL-signatures, Idemix) are presently used; all selective disclosure is manual and organizational. This enables regulatory compliance while maintaining transactional privacy within normal operations (Lee et al., 10 Oct 2025).

5. Performance Considerations

The ARTeX manuscript offers estimated on-chain gas costs based on observed norms:

• ERC-20 deposit: ≈ 65,000 gas
• ERC-721 deposit: ≈ 100,000 gas
• ERC-20 payout: ≈ 65,000 gas
• ERC-721 withdrawal: ≈ 100,000 gas

The total on-chain cost per auction is approximately 200k–300k gas, competitive with or more efficient than Tornado-style mixers (300k–400k gas). Off-chain performance is bounded by auction engine throughput and typical web infrastructure, with operational latency determined primarily by Ethereum block times (≈15 seconds). The paper does not present empirical benchmarking or scaling trials (Lee et al., 10 Oct 2025).

6. Comparison to Related Work and Open Challenges

ARTeX is contrasted with three main classes of RWA privacy projects:

• Front-end pseudonymity: hides names at the UI level only; addresses remain visible on-chain
• Metadata encryption: e.g., Aegis for NFT, which can obscure amounts at cost of trust in token identity
• Decoy/mixer protocols: high on-chain cost, limited anonymity, and frequent exchange blacklisting

ARTeX distinguishes itself by protocol-level separation of user identity from on-chain addresses, support for both FT and NFT, KYC-AML integration for legal compliance, and avoidance of high-risk mixed pools.

Identified open problems and avenues for research include:

1. On-chain trustlessness via replacing the operator trust with MPC or threshold-ZK constructs
2. Zero-knowledge enhancements for fully unlinkable deposit/withdrawal proofs (e.g., per-deposit zk-SNARKs)
3. Formal anonymity analysis, explicitly bounding adversarial advantage $\epsilon$ in the face of side-channel leakage
4. Gas efficiency improvements, potentially through batching, layer-2 rollups, or account abstraction
5. Integration of cryptographically selective disclosure (e.g., CL-signatures) for user-controlled proof of KYC without address revelation

A plausible implication is that a fully trustless and auditable ARTeX implementation will require substantive integration of zero-knowledge and threshold cryptographic constructions beyond the current architecture (Lee et al., 10 Oct 2025).