Cross-chain payment settlement is solved in principle. In practice, the gap between theoretical correctness and operational cost runs deep. The two dominant patterns, hash time-locked contracts (HTLC) and burn-and-mint bridges, occupy opposite corners of a cost-security spectrum. Neither dominates universally. The choice depends on acceptable trust assumptions, chain bandwidth, asset volatility, and failure recovery overhead. Understanding the tradeoffs is essential for payment infrastructure teams.
HTLC vs Burn-and-Mint
Hash time-locked contracts use cryptographic commitments and time-based release logic to enforce atomicity across chains without requiring trust in an intermediary. On chain A, Bob locks funds in a contract that releases only if a pre-image matching a known hash is revealed within a deadline. On chain B, Alice locks equivalent funds under the same hash. One party reveals the pre-image, both sides settle, or both time out and refund.
The security model is clean. No single oracle or relay can steal funds. Both sides either complete or revert. The contract logic is deterministic and verifiable on-chain. This self-custody finality is the primary advantage of HTLC.
Gas cost tells a different story. An HTLC deployment and settlement sequence costs approximately 450,000 to 600,000 gas across both chains for a single atomic swap. On Ethereum at current rates, that translates to 6-9 USD per transaction. On cheaper chains it remains material. The cost comes from storing the hash pre-image, checking timelocks, and executing conditional fund release.
Burn-and-mint bridges offer a simpler execution model. Funds are locked (or burned) on the source chain. A validator set or oracle publishes a signed proof that the lock occurred. The destination chain mints equivalent tokens or releases equivalent liquidity. Finality depends entirely on the validator set's honesty and economic incentives.
Gas consumption for burn-and-mint settlement is substantially lower. A token burn plus liquidity release runs 80,000 to 150,000 gas total. The relay overhead is amortized across many transfers. If 100 payments are batched in one settlement proof, per-transaction gas drops to 800-1500 gas. The cost difference from HTLC is a factor of 3 to 6 in favor of burn-and-mint.
The trade-off is trust. A compromised validator set or oracle can mint tokens without a corresponding lock. Burn-and-mint with a validator supermajority assumption works if validators are diverse, geographically distributed, and financially incentivized. If the validator set is concentrated or poorly incentivized, the risk is material. Some bridge protocols layer additional attestation requirements on top of burn-and-mint to reduce this risk.
When Each Approach Works
HTLC remains the preferred path when trust assumptions are unacceptable or impossible to meet. Settlement between chains with entirely independent validator sets benefits from HTLC's self-custody finality. Neither chain needs to trust the other's consensus. Both sides verify the other's lock condition before releasing funds.
Payments where the sender and receiver do not trust each other also prefer HTLC. A market maker using HTLC to swap stablecoins across chains knows that both sides will complete or both revert. No slippage from oracle manipulation. No counterparty risk from validator misconduct.
Large transfers where the percentage cost of HTLC overhead is small also justify the approach. A 10 million USDC transfer paying 9 USD in HTLC fees (0.00009%) is reasonable. A 100 USDC transfer paying 9 USD (9% fee) is not. At large transfer sizes, HTLC overhead becomes negligible.
Payments from a payer to a known recipient where settlement reversal is not operationally necessary can use burn-and-mint safely. A stablecoin payment from an exchange to a user's self-custody address across chains does not require mutual distrust verification. The exchange is trusted by the user to execute correctly.
High-volume, low-value transfers benefit from burn-and-mint's amortized costs. Payment networks settling millions of micropayments daily cannot absorb HTLC overhead per transaction. Batch settlement of HTLC is impractical.
Cross-chain atomic swaps of volatile assets benefit from HTLC's built-in timeout. If market conditions shift dramatically before both sides settle, either party can let the timelock expire and refund. Burn-and-mint settlements cannot be reversed without bridge governance intervention.
Decision Framework
Three variables determine the appropriate settlement path. They are trust environment, transaction size, and volume.
High-trust, low-volume, large-value scenarios favor HTLC. An example is market maker atomic swaps where both parties are sophisticated and want mutual assurance of completion.
High-trust, high-volume, small-value scenarios favor burn-and-mint batched. An example is payments within a stablecoin network where users trust the stablecoin issuer to maintain backing.
No-trust, any volume, any size scenarios require HTLC. An example is exchange-to-exchange cross-chain settlement with no third party intermediary.
Cost savings from burn-and-mint are real but not magical. A 3x to 6x improvement in settlement gas means replacing 9 USD per swap with 1.5 to 3 USD. At million-swap scale, that is 6 to 7.5 million USD per year in operational savings. Capital costs for maintaining cross-chain liquidity may or may not justify the engineering complexity of burn-and-mint systems.
The decision is not architectural philosophy but economic reality. Payment infrastructure teams must calculate the specific costs and risks for their use case and operating constraints.