DeFi-Market-Risk-Report

DeFi Market Risk Report

Independent, evidence-based risk intelligence for DeFi lending markets.

Measure → Explain → Simulate → Let the user choose.

Product layers

v0.3 — Evidence-Backed Risk Methodology

The commercial beta report computes six dimensions from explicit factor rules and evidence:

Protocol / Asset / Liquidity / Oracle / Stablecoin / Chain

Risk Report #001 remains the Aave V3 Ethereum Core / USDC pilot.

v0.4 — Calibration & Stress Benchmark

v0.4 adds a protocol-neutral diagnostic layer:

Aave + Morpho + Compound + Spark
              ↓
     normalized market records
              ↓
      Stress Pressure Score
              ↓
 historical stress replays
              ↓
 calibration / ranking diagnostics

The benchmark score is deliberately separate from the commercial risk score. It exists to find ranking mistakes, threshold discontinuities and missing coverage before the methodology is scaled.

Live benchmark target: 20–50 markets.

• Aave: official Aave MCP.
• Morpho: official Morpho API.
• Compound: direct Ethereum Comet reads via ETH_RPC_URL.
• Spark: direct Ethereum ProtocolDataProvider reads via ETH_RPC_URL.

No RPC secret is committed to the repository.

Run

Requires Node.js 20+.

npm test
npm run report
npm run live:report
npm run benchmark

Without ETH_RPC_URL, the benchmark can pass as PASS_PARTIAL_RPC when Aave + Morpho provide at least 20 live markets and stress gates pass. With the RPC secret configured, Compound + Spark are included and the target becomes PASS_FULL.

See Risk Report #001, Calibration Benchmark v0.4, Risk Methodology v0.3, Commercial Beta, Verified Data Contract, and Product Boundary.

Commercial beta

Founding Beta price: €9.

A commercial beta report is informational decision support. It is not a probability of loss, credit rating, suitability assessment, or personalized recommendation.

Founding Beta validation

The first commercial test is intentionally small:

• price: €9;
• target: 10 qualified prospects;
• no mass outreach;
• no new feature work before objections are recorded.

See Founding Beta 10, Outreach Playbook, Offer, and Test Log.