Liquid Network Deep Dive: How Blockstream's Federated Sidechain Works

The Liquid Network is a federated sidechain built on Bitcoin, operated by a consortium of exchanges, infrastructure providers, and financial institutions. Launched by Blockstream in 2018, Liquid enables faster settlement, confidential transactions, and native token issuance while anchoring security to Bitcoin's base layer. As of mid-2026, the network holds roughly $5 billion in total value locked across L-BTC, stablecoins, and tokenized securities.

This article examines Liquid's architecture in detail: how its federation secures funds, how Confidential Transactions hide amounts, how the peg mechanism works, and where Liquid fits alongside other Bitcoin Layer 2 solutions.

How Does Liquid's Federated Consensus Work?

Unlike Bitcoin's proof-of-work consensus, Liquid uses a model called Strong Federations. Fifteen functionaries operate custom-built hardware security modules (HSMs) that serve two roles: blocksigners, who produce and validate blocks, and watchmen, who manage the Bitcoin held in the federation's multisig wallet.

Blocks are produced in a round-robin schedule: each functionary takes a turn proposing a block every 60 seconds. The remaining functionaries validate the proposal and sign it. A block is finalized when at least 11 of the 15 functionaries have signed, providing Byzantine fault tolerance. This means the network continues operating even if up to 4 functionaries go offline: their rounds are simply skipped.

The result is deterministic 1-minute block times and 2-minute settlement finality (2 confirmations). Reorganizations beyond one block are structurally prevented by the consensus design, a significant advantage over Bitcoin's probabilistic finality where merchants typically wait for 6 confirmations (~60 minutes).

Federation vs. functionaries: Liquid distinguishes between federation members (87 organizations as of Q1 2026) and functionaries (the 15 nodes that actually sign blocks and manage the peg). Federation members participate in governance, but only functionaries run the consensus-critical infrastructure.

Federation membership and governance

The federation launched with 23 founding members in 2018, including Bitfinex, BitMEX, Bitso, and Xapo. It has since grown to 87 members across six continents. In late 2024, Blockstream open-sourced the functionary code, allowing anyone to audit the consensus implementation and potentially build their own Liquid-like network.

Dynamic Federations (DynaFed), a protocol upgrade deployed on the network, enables adding and removing functionaries without a hard fork. This mechanism is designed to allow the network to scale beyond its original 15 functionaries over time, though the 11-of-15 configuration remains in place as of mid-2026.

Emergency recovery

A 2-of-3 emergency backup key set, held by Blockstream in geographically distributed cold storage, activates via a timelock mechanism if the federation becomes non-functional. Upon each peg-in, a timelock of 4,032 blocks (~28 days) is set. If the federation fails to refresh these timelocks (indicating extended downtime), the emergency keys can recover the locked Bitcoin.

How Do Confidential Transactions Work on Liquid?

Every transaction on Liquid uses Confidential Transactions by default. Both the transaction amount and the asset type are cryptographically hidden from all observers, including federation functionaries. Sender and receiver addresses remain visible, as do transaction fees, but the economic substance of each transfer is private.

Pedersen commitments

Amounts are concealed using Pedersen commitments, a cryptographic scheme that replaces plaintext values with commitments that cannot be reversed to reveal the original amount. The critical property is homomorphism: commitments can be added together, allowing validators to confirm that transaction inputs equal outputs without knowing individual values. This proves no inflation occurred while revealing nothing about the actual amounts.

Range proofs and bulletproofs

Pedersen commitments alone have a problem: they could hide negative numbers, which would enable creating money from nothing. Range proofs solve this by providing a zero-knowledge proof that each hidden value falls within a valid range (0 to 2^64 satoshis) without revealing the value itself. Liquid uses bulletproofs, a more compact form of range proofs, to keep transaction sizes manageable.

Blinding keys

Confidential addresses on Liquid include a public blinding key. Only the recipient can decrypt the amount and asset type using their private blinding key. This key can optionally be shared with auditors or counterparties who need to verify transaction details, enabling selective disclosure for compliance without broadcasting data to the entire network.

Privacy vs. transparency tradeoff: Confidential Transactions on Liquid hide amounts from the public but allow selective disclosure via blinding keys. This design targets institutional use cases where traders need privacy from competitors but must still satisfy regulatory reporting requirements.

How Does the L-BTC Peg Work?

L-BTC is a 1:1 representation of Bitcoin on the Liquid sidechain. The peg mechanism converts BTC to L-BTC (peg-in) and L-BTC back to BTC (peg-out), with the federation managing the Bitcoin held in a multisig wallet on the main chain.

Peg-in: BTC to L-BTC

To peg in, a user sends BTC to a federation-controlled address generated by their Liquid client software. The transaction requires 102 confirmations on the Bitcoin blockchain (roughly 17 hours) before the equivalent L-BTC can be claimed. This high confirmation threshold protects against deep blockchain reorganizations. After the confirmations complete, the federation releases matching L-BTC within approximately 2 Liquid blocks (~2 minutes).

Peg-out: L-BTC to BTC

Peg-outs work in reverse: L-BTC is burned on the Liquid sidechain, and the federation releases an equivalent amount of BTC from its mainchain multisig wallet. Peg-outs require only 2 confirmations on Liquid and are processed by watchmen in batches, with typical processing times of 11 to 35 minutes.

An important limitation: general users cannot independently initiate a peg-out. The process requires a functionary or participant member to convert L-BTC back to BTC, using Peg-out Authorization Keys (PAKs) to prevent unauthorized withdrawals. This is a meaningful trust dependency that differentiates Liquid from trustless systems like Lightning.

What Assets Can Be Issued on Liquid?

Beyond L-BTC, Liquid supports permissionless issuance of new asset classes. Any participant can create tokens using the Issued Assets protocol, where each asset receives a unique 64-character hex identifier. All issued assets automatically benefit from the same Confidential Transaction features as L-BTC. In Q1 2026 alone, 18,857 new asset issuances were recorded on the network.

Stablecoins

Tether (USDT) is issued natively on Liquid, with approximately $97 million in circulation as of mid-2026. L-USDT benefits from 2-minute settlement and confidential amounts, though its supply on Liquid is small relative to USDT on Ethereum and Tron. Tether's Hadron tokenization platform has also integrated with Liquid, enabling real-time issuance with built-in KYC/KYB tools.

Tokenized securities and real-world assets

Liquid has become a significant platform for real-world asset tokenization. Two major issuance platforms operate on the network:

• Bitfinex Securities has surpassed $250 million in assets issued on Liquid, including tokenized bonds, equity, and the first regulated public offering of tokenized US Treasury bills on a Bitcoin sidechain
• STOKR crossed $1.5 billion in tokenized asset volume by end of 2025, with its flagship Blockstream Mining Note 2 (BMN2) valued at $790 million, making it one of the largest tokenized real-world assets globally

The network's TVL breakdown illustrates the shift toward tokenized assets: Mifiel Promissory Notes (~$2.1B), BMN2 (~$877M), USDT (~$97M), and USTBL tokenized Treasury bills (~$32M) make up the bulk of the $5 billion total.

How Does Liquid Compare to Other Bitcoin Layer 2s?

Liquid's federated model represents one point on a spectrum of Layer 2 trust assumptions. Each approach makes different tradeoffs between decentralization, performance, and functionality.

The key distinction is custody. Lightning and Spark preserve self-custody: users can always exit to Bitcoin L1 without anyone's permission. Liquid and Fedimint require trusting a federation to custody and return funds. This is a fundamental architectural difference, not merely a degree of decentralization.

Trust model nuances

Liquid's 11-of-15 threshold means that if 11 functionaries collude, they could theoretically steal the federation's Bitcoin. The mitigation is institutional: functionaries operate HSMs that auto-erase keys upon unauthorized physical access, and members span multiple jurisdictions. But the trust is placed in specific, known entities.

Spark uses a different approach: threshold signatures (FROST) where the security guarantee is 1-of-n. As long as one operator behaves honestly, funds remain secure, and users always retain the ability to exit unilaterally. The trust is moment-in-time: once a transfer completes and keys are deleted, operators cannot affect past transactions even if later compromised.

Lightning remains the only fully trustless option, but it introduces operational complexity: liquidity management, channel monitoring, and the requirement to be online to receive payments.

What Are Liquid's Primary Use Cases?

Inter-exchange settlement

Liquid's original use case remains its strongest: fast settlement between exchanges. Traders holding L-BTC in their own wallet can deposit to any supporting exchange within 2 minutes, execute trades, and withdraw to another exchange or back to self-custody. This is dramatically faster than the 60+ minutes typical for on-chain Bitcoin transfers between exchanges.

Privacy-preserving OTC trading

Confidential Transactions make Liquid particularly useful for large over-the-counter trades. When an institution moves $10 million in Bitcoin, the amount is invisible to blockchain observers, preventing competitors from front-running or acting on the information. Even federation functionaries cannot see transaction amounts or asset types.

Tokenized asset issuance

This has become Liquid's fastest-growing use case. The combination of Confidential Transactions (investor privacy), 2-minute settlement (capital efficiency), and native asset support (no smart contract deployment needed) makes it attractive for regulated securities issuance. Platforms like STOKR and Bitfinex Securities have issued bonds, equity, Treasury bill products, and mining notes on the network.

What Are the Criticisms of Liquid?

Liquid's federated design invites legitimate criticism, and understanding these limitations is essential for evaluating where it fits in the Bitcoin scaling landscape.

Permissioned validation

Only 15 functionaries validate blocks and manage the peg. This is a small, permissioned set compared to Bitcoin's thousands of independent miners and nodes. Users cannot run a functionary without being selected by the existing federation, making Liquid fundamentally different from permissionless networks.

Collusion risk

If 11 of the 15 functionaries colluded, they could steal the BTC backing all L-BTC in circulation. Blockstream points to HSM hardware safeguards, geographic distribution, and the reputational stakes of federation members as mitigations. The open-sourcing of functionary code in December 2024 improved transparency, but the fundamental trust assumption remains.

No unilateral exit

Unlike Lightning or Spark, Liquid users cannot exit to Bitcoin L1 without federation cooperation. If the federation collectively refuses to process peg-outs, users' only recourse is the emergency timelock mechanism, which requires Blockstream's 2-of-3 backup keys and roughly 28 days of federation downtime. This represents a meaningful custody tradeoff compared to statechain-based approaches that preserve unilateral exit.

Emergency key centralization

The 2-of-3 emergency backup keys are held by Blockstream in cold storage. If activated, Blockstream would control the recovery process. This creates a single organizational dependency in the most critical failure scenario.

Historical incidents

In 2021, a timelock bug was discovered where timelocks were occasionally being refreshed shortly after expiry rather than before, due to an inconsistency between HSM and server parameters. While patched without incident, it demonstrated the complexity of maintaining federation security and the potential for subtle implementation errors in critical infrastructure.

Recent Developments on Liquid

Liquid has seen significant technical evolution in 2025 and 2026, moving well beyond its original role as an inter-exchange settlement layer.

Simplicity smart contracts

In July 2025, after 8 years of development, the Simplicity smart contract language launched on Liquid mainnet. Simplicity is a purpose-built, non-Turing-complete language designed for formally verifiable contracts. It excludes recursion, unbounded loops, and mutable global state, meaning every execution path and fee cost can be proven in advance. A higher-level language (SimplicityHL) compiles down to raw Simplicity for developer ergonomics.

Post-quantum signatures

In March 2026, Blockstream deployed post-quantum signatures (SHRINCS) on Liquid mainnet using Simplicity smart contracts. These were the first post-quantum-signed transactions on a production Bitcoin sidechain securing real funds. SHRINCS signatures are as small as 324 bytes at NIST Level 1, significantly more compact than alternatives like ML-DSA-44 or SLH-DSA. This positions Liquid as a testbed for quantum-resistant cryptography before potential adoption on Bitcoin's base layer.

Network growth

Q1 2026 recorded 1,163,119 transactions, nearly 5x year-over-year growth and 72% above Q4 2025. December 2025 set the single-month record with 242,000 transactions. The federation added 7 new members in Q1 2026, bringing the total to 87, and peg activity showed 1,252 BTC in peg-ins against 1,144 BTC in peg-outs during the quarter.

Upcoming roadmap

Several features are in development or testing as of mid-2026:

• 0-conf settlement is in beta testing with federation members and the Boltz exchange
• Multi-asset fees (ELIP 204) would allow paying transaction fees in issued assets rather than L-BTC
• BitVM bridge research is exploring a 1-of-n bridge to reduce trust assumptions in the peg mechanism
• Taproot sweeps are targeted for 2027 network activation

When Should You Use Liquid vs. Other Layer 2s?

Liquid is best suited for specific use cases where its federation model is an acceptable tradeoff for the features it provides:

• Inter-exchange settlement where 2-minute finality and large-value privacy matter more than trustless custody
• Tokenized securities issuance where regulatory compliance, investor privacy, and Bitcoin-anchored settlement are priorities
• Large OTC trades where confidential amounts prevent information leakage to competitors
• Institutional custody where federation members are already trusted counterparties

For everyday payments, self-custodial wallets, and consumer applications, Lightning and Spark are generally better fits. Lightning provides fully trustless payment channels, while Spark offers self-custodial transfers without the complexity of channel management. Understanding the trust model of each Layer 2 is critical for choosing the right tool. For a broader comparison, see the Bitcoin Layer 2 comparison.

Different tools for different jobs: Liquid and Spark represent opposite ends of the federation spectrum. Liquid uses a fixed federation of known entities for privacy and token issuance, while Spark uses FROST threshold signatures for self-custodial payments with unilateral exit. Neither replaces the other: they serve fundamentally different use cases.

Developers interested in building self-custodial Bitcoin applications can explore the Spark SDK documentation for integration guides and API references. For a hands-on look at how Spark-powered wallets work in practice, see General Bread, a wallet built on Spark that demonstrates instant Bitcoin and stablecoin transfers. For deeper exploration of how these Layer 2 approaches compare at the protocol level, the Bitcoin scaling landscape overview provides additional context.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.