Lorenzo’s hybrid model — on-chain fundraising, off-chain execution, and on-chain settlement — exists to access strategies that blockchains cannot yet natively support. The hard problem is not execution itself, but proof. How does a user know that what happened off-chain actually occurred, and that the outcome reported on-chain reflects reality?
The answer is not a single cryptographic trick. It is a structured proof pipeline designed to translate off-chain financial activity into something blockchains can verify, enforce, and audit.
Why the Trust Gap Exists
Off-chain strategies operate in environments blockchains cannot directly observe: centralized exchanges, OTC desks, or private execution venues. Trades, balances, and profit calculations happen outside the chain’s line of sight. Without a bridge, users would be forced to rely on pure trust in the manager — a model Lorenzo explicitly avoids.
Instead, the protocol reduces this gap by requiring that every step of off-chain execution leaves behind a cryptographically verifiable trail.
Layer One: Signed Facts from the Source
The first layer of proof originates where execution actually happens. Trading activity occurs through permissioned accounts tied to institutional custody or exchange infrastructure.
These systems can produce digitally signed attestations that confirm objective facts: – Account balances at defined timestamps
– Executed trades with prices, quantities, and times
– Period-level profit and loss summaries
Each attestation is signed using keys controlled by the exchange or custodian. This makes the data tamper-evident. While trust in the institution still exists, the data can no longer be quietly altered without detection.
Layer Two: NAV Attestation as a Single Source of Truth
Raw trade data is not directly useful to smart contracts. What matters on-chain is the final Net Asset Value of the strategy.
A designated NAV oracle aggregates the signed off-chain attestations, performs accounting calculations, and produces a single NAV figure for the settlement period. This value is itself signed using the oracle’s cryptographic key.
At this point, responsibility becomes clear and bounded. The oracle is explicitly attesting: “Given the signed execution data, this is the correct NAV.”
Layer Three: Smart Contract Verification and Settlement
The final step happens entirely on-chain. The vault contract is programmed to accept NAV updates only from whitelisted oracle addresses.
When a settlement transaction is submitted, the contract verifies the oracle’s signature against its stored public key. If the signature is valid, the contract updates state: – User share values are recalculated
– Yield is finalized for the epoch
– Funds owed to the vault are reconciled
The transaction hash becomes the permanent, immutable record linking off-chain execution to on-chain settlement.
What the “Proof” Really Is
There is no illusion of full trustlessness here. The proof is a layered accountability system: – Exchanges or custodians sign execution facts
– Oracles sign derived financial truth
– Smart contracts enforce outcomes without discretion
Together, these layers convert traditional financial attestations into machine-verifiable actions.
How Trust Is Contained, Not Eliminated
Lorenzo does not pretend off-chain execution can be made trustless today. Instead, it contains trust by making it explicit, auditable, and economically accountable. Every actor signs their role. Every signature is visible. Every settlement is enforced by code.
Over time, this structure can evolve toward more decentralized oracle networks and redundancy. But even in its current form, it replaces blind trust with verifiable responsibility.
A short story to end:
I once explained this model to my friend 王浩 during a long walk after work. He paused and said something simple: “So it’s not about pretending the world is on-chain — it’s about making the off-chain world leave fingerprints.” That stuck with me. Lorenzo’s system doesn’t erase reality. It forces reality to sign its name.
@Lorenzo Protocol #lorenzoprotocol $BANK
