MPC vs Multisig: What's the Difference for Institutional Custody?
MPC and multisig solve the same problem, making sure no single key holder can move funds alone, but they solve it in different places. Multisig distributes control on the blockchain, requiring several complete keys to sign. MPC distributes control in the cryptography itself, splitting one key into shares so the complete key never exists anywhere. For institutional custody, that difference determines chain coverage, privacy, fees, and how governance actually gets enforced.
What is multisig?
Multisig (multi-signature) is an on-chain arrangement in which a wallet requires signatures from several complete private keys, for example 2 of 3, before a transaction is valid. Each signer holds a full, independent key, and the blockchain itself enforces the rule, through a script on Bitcoin or a smart contract on Ethereum and similar networks. The mechanism is battle-tested and transparent: anyone can verify the rules on-chain.
What is MPC?
Multi-party computation (MPC) splits a single private key into shares held by separate parties, and the parties jointly compute a signature without the full key ever being assembled in any one place, at generation, in storage, or at signing. The blockchain sees a standard, ordinary signature, indistinguishable from a single-key wallet. The threshold rule, for example 2 of 3 shares, is enforced by the cryptographic protocol itself.
How do MPC and multisig compare?
| Multisig | MPC | |
|---|---|---|
| Where the rule is enforced | On the blockchain, per chain | In the cryptographic protocol, independent of chain |
| Chain coverage | Depends on each chain's support; differs by network | Any chain using standard signatures, uniformly |
| What the chain sees | A visible multisig script or contract | A standard single signature |
| Governance privacy | Signer structure and policy visible on-chain | Nothing about the setup is revealed |
| Transaction fees | Higher: extra signatures or contract execution | Standard single-signature cost |
| Changing signers | On-chain change, often a new address and asset migration | Shares reissued or refreshed; address stays the same |
| Compromise recovery | Compromised key must be rotated on-chain | Proactive refresh re-randomises shares on schedule, invalidating anything previously stolen |
| Full key exposure | Each signer holds a complete key that can be individually stolen | The complete key never exists anywhere |
Where does multisig still make sense?
Multisig remains a reasonable choice when on-chain transparency is the point: a DAO treasury whose community needs to verify governance rules publicly, or a Bitcoin-only holding where the native multisig script is simple, proven, and sufficient. Its guarantees are enforced by the chain itself and are auditable by anyone, which for some governance models is exactly the property required.
Why do institutions increasingly choose MPC?
Because institutional requirements, multi-chain coverage, privacy, uniform policy, and recoverability, sit precisely where multisig is weakest. One MPC setup covers every supported chain with the same governance, instead of re-implementing (or discovering the absence of) multisig per network. Governance stays private rather than published on-chain. Signers can join, leave or be re-keyed without moving assets to a new address. And modern MPC protocols add properties multisig has no equivalent for: proactive refresh, which re-randomises shares on schedule so stolen material goes stale; identifiable abort, which cryptographically names a misbehaving participant; and signing fast enough, sub-second after preprocessing in the protocol family The Vault uses (MPC-CMP), that security no longer costs operational speed. Each share can additionally be sealed inside dedicated secure hardware, protecting it even from the administrators of the servers it runs on.
What should an institution check in an MPC implementation?
MPC shifts the trust from the blockchain to the implementation, so the implementation is what you diligence: whether the protocol has published, peer-reviewed security proofs, whether the code has been independently reviewed with findings remediated, and whether shares are protected by dedicated hardware rather than sitting in server memory. These questions are part of a broader framework we've laid out in How to Evaluate a Digital Asset Custody Provider.
If you're comparing custody approaches for your own institution, contact us. We're happy to answer your questions and book a demo for you.
Frequently asked questions
Is MPC safer than multisig?
They concentrate risk in different places: multisig in each signer's complete key and each chain's implementation, MPC in the protocol and its implementation. For institutions, MPC's advantage is that a complete key never exists to be stolen, and compromised shares can be invalidated by refresh.
Does MPC work on chains that don't support multisig?
Yes. Because MPC produces a standard signature, it works uniformly on any chain using standard signature schemes, regardless of whether that chain supports multisig natively.
Can you tell from the blockchain that a wallet uses MPC?
No. An MPC wallet's transactions carry ordinary single signatures, so the governance structure behind the wallet is not visible on-chain.
What is proactive refresh?
A scheduled re-randomisation of key shares that leaves the key and address unchanged while making any previously stolen share useless. It turns a share compromise from a standing threat into an expiring one.
Do institutions ever combine MPC and multisig?
Rarely in one wallet, but portfolios can mix models: for example, MPC for operational, multi-chain custody and a native multisig for a transparency-first Bitcoin reserve.