Economic intelligence for autonomous capital.

Markets generate activity faster than teams can establish economic truth.

Signals, wallet flows, funding, liquidations, and backtests are abundant. What is scarce is a coherent system that distinguishes trustworthy data from stale or inconsistent inputs, market state from alpha, statistical response from executable economics, and participant activity from persistent information.

A validation platform, not another signal feed.

• Market Truth: normalize books, trades, funding, OI, liquidations, wallet activity, freshness, and source provenance.
• Market State: generate causal liquidity, volatility, funding/leverage, and execution-confidence states independently from alpha.
• Wallet Intelligence: test peer-adjusted markout, persistence, specialization, metaorders, and participant heterogeneity.
• Validation: measure executable costs, MAE, false positives, tails, clustering, chronological OOS, and external-source agreement.
• Governance: keep research, paper, and live permissions separate with fail-closed defaults.

The platform already exists and is designed to reject weak evidence.

On 26 August 2026, the Market State Engine entered an observe-only canary on fresh production-research data. Liquidity, volatility, funding/leverage, and execution-confidence states are generated without changing trading behavior.

Carry remains a benchmark because realistic net economics are still negative. Wallet identity remains a confirmatory OOS research question, and external market-data adapters are used to challenge collector agreement rather than replace raw venue truth.

The moat hypothesis is measurement, not prediction theater.

• Participant-level historical data and wallet scoring, if incremental OOS value persists.
• Longitudinal Market State history tied to future MAE, liquidity, and execution-quality outcomes.
• Independent false-edge and strategy-validation datasets built with realistic cost and provenance controls.
• Cross-source truth reconciliation for funding, OI, liquidations, and market events.

Start with decisions institutions already need to make.

The first commercial wedge is not consumer trading. It is independent validation and risk intelligence for digital-asset teams: strategy review, market-state context, wallet research, data-quality scoring, and execution-risk workflows.

• Phase 1: validate Market State and wallet intelligence inside the existing research engine.
• Phase 2: run design-partner pilots around strategy validation and execution-risk intelligence.
• Phase 3: productize recurring software, data/API, and integration workflows.

The same economic-truth principle may extend to autonomous AI agents.

TheDuck Agents is an isolated read-only research track exploring agent balance sheets, treasury quality, compute capital, revenue, burn, and sustainability. It is not connected to market execution and does not include custody or treasury automation.

Prove usefulness before expanding scope.

• Show whether Market State materially improves future-risk separation and strategy false-positive rejection.
• Complete wallet-identity confirmatory OOS and external-provider benchmarking only if incremental value survives placebo tests.
• Validate external market truth against independent public datasets and document source disagreement.
• Run structured customer discovery and secure initial design partners around validation and risk workflows.
• Pilot read-only Agent Balance Sheet snapshots without expanding into custody, treasury execution, or token issuance.

Design partners, research partners, and pre-seed readiness.

We are looking for conversations with institutional digital-asset teams, quantitative researchers, fintech infrastructure operators, AI-agent builders, and investors who value measurable evidence before scale.