The story
Dividing an exposure into three assessable layers
A43 developed an architecture separating:
1. Asset risk – market behaviour, financial backing, volatility and relevant non-financial characteristics of the underlying asset. 2. Protocol risk – software and smart-contract exposure, governance, oracles, security controls, dependencies and historical incidents. 3. Pool risk – liquidity, concentration, positions, governance and risks inherited from assets and protocols.
The layers were designed to interact. A high-quality pool process does not remove exposure to an insecure protocol or unreliable collateral.
Combining data from blockchain and outside sources
The framework research explored indicators available from on-chain records, including liquidity and wallet concentration, alongside off-chain evidence such as audits, incident history, regulatory conditions and governance arrangements.
The accompanying risk taxonomy and early metrics material provided a structure for gathering evidence before deciding on mitigation or insurance treatment.
The technical and financial factors could vary significantly between use cases, so the work did not assume a single universal score would determine whether a position was safe.
From dividing cover to managing risk
The original Divide & Share Risks concept explored allocating different portions of an exposure to insurers or cover providers, with a possible Etherisc-based residual risk pool.
As the work progressed, A43 also developed the broader process of identifying risks, assessing significance, selecting responses, implementing controls and monitoring the result.
The surviving archive contains research, framework drafts, taxonomies and a roadmap for feasibility and proof-of-concept work.
Presenting the methodology
A43's founders presented the work at D1Conf in Amsterdam in April 2022, a conference devoted to blockchain insurance and risk.
The talk made the core design public: DeFi risk needs to be assessed in the context of the assets, protocols and financial products involved, rather than assigned a single generic blockchain-risk label.
The available evidence does not establish that the proposed risk pool launched or that a calibrated scoring engine entered production.