Multisig and MPC Institutional Custody

Whoever holds a crypto private key can move the funds it controls, in full, irreversibly, with no bank to call and reverse the transfer. For an individual that's a feature. For a custodian holding client assets, a single key is a single point of catastrophic failure — one stolen laptop, one coerced employee, one bug, and everything is gone. Institutions solve this the same way boards solve any high-stakes decision: require more than one signature. Multisig and MPC are the two engineering approaches to making "more than one signature" actually happen on a blockchain.

Multisig: the rule lives on-chain

A multisig wallet is a smart contract (or, on Bitcoin, a special script) that holds funds and only releases them once a specified threshold of separate private keys has signed off — commonly written as m-of-n, such as 2-of-3 or 3-of-5. Each signer has their own ordinary private key; the wallet's logic, not any individual key, enforces the threshold. Anyone inspecting the blockchain can see that a transaction required, say, three valid signatures out of five authorized keys, because the multisig structure is a visible on-chain rule.

MPC: the rule lives in the math

Multi-party computation (MPC) custody splits a single private key mathematically into shares distributed among several parties, none of whom ever holds the complete key. To sign a transaction, the parties run a cryptographic protocol that jointly produces a valid signature without any party's share ever being combined into a whole key at any point, on any machine. The result on-chain looks like an ordinary single-signature transaction — the threshold logic is invisible to outside observers, unlike multisig's visibly-structured transaction.

Worked example

A custodian sets up 2-of-3 approval both ways, to compare. Under multisig, three separate keys exist; a withdrawal transaction must carry two distinct signatures, visible in the transaction data, and the blockchain's own validation rejects the transaction if only one signature is present. Under MPC, the same custodian splits one private key into three shares held by compliance, operations, and an offline backup; approving a withdrawal requires two of the three share-holders to participate in the signing protocol, producing a single ordinary-looking signature — cheaper in transaction fees (one signature instead of two-plus) and usable on any chain, even ones without native multisig scripting, but the 2-of-3 rule itself is enforced entirely by the custodian's internal software, not by the blockchain.

What this means in practice

Multisig's on-chain visibility is an auditability advantage — auditors and even clients can verify the threshold rule directly from the chain — but it costs more in fees (multiple signatures per transaction) and isn't supported identically on every chain. MPC is cheaper, chain-agnostic, and lets custodians change signing policy without moving funds to a new contract, but it pushes trust back onto the custodian's internal key-generation and signing software, which an outside observer cannot verify just by looking at the blockchain.