Rewind

Phase 2. Blocked on Phase 1 + $420–985K raise.

Rewind is resolution-failure insurance. A policyholder buys protection on a specific PM contract; if that contract fails at resolution — oracle dispute, contract-text ambiguity, observable manipulation — Rewind pays out up to the policy limit.

Why it works

Resolution risk is inherent

Cong et al. 2025 formally proves the Oracle Trilemma: no protocol can simultaneously achieve decentralization, aggregate truthfulness, and scalability. This is an impossibility result, not a missing feature.

It is quantified

Adjudication Risk Index (Ma 2026): UMA voting-correctness ranges 0.6353–0.9995 over 747 days with measurable spikes tied to voting-power concentration. When error probability exceeds 1/3, no bond parameter can restore deterrence.
Semantic Risk Score (Sanjabi 2026): LLM-derived score predicts oracle disputes from contract text alone. Spearman ρ=0.157, p<0.001. Published with replication code. Tau = 0.10 SRS cutoff filters 59% of markets while doubling median volume.
Historical losses: $403M in oracle manipulation attacks in 2022 alone (Duley). MakerDAO Black Thursday: $8.32M liquidated at zero DAI. Compound malfunction: $89M bad debt. Pyth: 90% BTC/USD misreport.
Insider trading: $143M anomalous profit on Polymarket (Feb 2024 – Feb 2026) — Ofir & Mitts 2026. Flagged traders 69.9% win rate (p<0.001).
Enforcement gap: zero CFTC insider-trading cases on event contracts as of 2024 (Ashar 2025). CFTC budget $365M / 700 employees vs SEC $2.1B / 5,000.

It is currently unpriced

The market responds to resolution risk with liquidity flight, not risk-adjusted pricing. A 0.1 SRS increase reduces contract volume by ~60% (Sanjabi). Traders flee rather than demanding compensation — an Akerlof lemons dynamic and a classic insurance opportunity.

Precedent exists

Nexus Mutual and Etherisc operate DeFi insurance without derivatives classification. Rewind follows that pattern: infrastructure provider, not speculative instrument.

Pricing (v0, three layers)

premium = actuarial_base(event_type)
        + semantic_risk_surcharge(SRS)
        + concentration_surcharge(pool_exposure, correlation)
        − cs_discount(insured_counterparty)

actuarial_base — event-type hazard rate derived from UMA dispute history (Ma/Sanjabi). Starting estimate: 2–3% of notional, matching functionSPACE’s industry estimate of the unpriced risk premium.
semantic_risk_surcharge — SRS-derived scalar multiplier. Contracts above SRS τ=0.10 are either refused or surcharged heavily.
concentration_surcharge — classical insurance underwriting. Pool exposure concentrated on correlated events triggers surcharges until rebalanced.
cs_discount — high-Chrono-Score forecasters correlate empirically with well-specified markets (hypothesis to validate during Phase 1 data collection). Initial discount capped at base × 0.25.

Pool structure

Single pool, USDC-denominated in v0. Simplicity over capital efficiency.
Whitelisted markets — 10–20 to start, expanded as underwriting data accumulates.
Target loss ratio: 40%. 60% of premiums to protocol / reserve split.
Tranching (junior/senior) deferred to v0.2 — Slava’s fund-management experience applies here.

Build status

Spec: complete (CONCEPTUAL_MODEL.md §4c)
Team: existing Solidity dev from Phase 1 + quant consultant ($15–25K, part-time)
Audit: $50–80K (insurance contracts are higher-risk)
Legal: non-US entity + jurisdiction-specific opinion
Capital: $200–500K pool seed
Timeline: 14–20 weeks after Phase 1 has produced Polymarket CTF-integration operational experience

⏪REWINDPosition insurance

Position size (USDC)

$500

Recovery tier

Scenario: market resolves against you. You lost $500.

Citations

See resolution risk, ecosystem layers, institutional demand.