Skip to main content

Purpose

SigmaGrid exists to provide institutional-grade fundamentals for synthetic-equity perpetual markets.

The problem​

Synthetic-equity perps (SPY-PERP, QQQ-PERP, TSLA-PERP, etc.) trade 24/7, cross-venue, and increasingly agent-driven. But unlike listed equities, they typically trade without:

  • a stable institutional fair-value anchor
  • forward-looking volatility forecasts
  • regime classification (risk_on vs risk_off vs transitioning)
  • explicit event calendars and risk assessments
  • cross-venue spread detection

That absence creates structural issues:

  • Cross-venue mispricings persist longer than they should.
  • Directional signals are noisy and fragile.
  • Liquidation cascades cluster around macro events.
  • Funding patterns become unstable.
  • Risk controls for agent-driven execution remain weak.

The SigmaGrid approach​

SigmaGrid rebuilds the institutional fundamentals layer for synthetic-equity perps and exposes it as a machine-readable JSON API.

  • Every supported ticker receives fair value with confidence score and source (market_hours / futures_adjusted / pre_market_blend).
  • Per-venue premium-to-close in bps across Hyperliquid, Avantis, and Ostium.
  • Volatility forecasts (1h, 4h horizons).
  • Regime classification (risk_on / risk_off / transitioning) with VIX context.
  • Event risk (earnings proximity, historical avg move, implied-vs-historical assessment).
  • Cross-venue spread (cheapest/richest venue, max spread bps, arbitrage flag).
  • Funding rates, z-scores, anomaly flags, mean-reversion probability (Hyperliquid only).
  • Optional macro sensitivities (beta_macro, beta_yield, beta_dollar, beta_vol_index) in alpha-snapshot responses.

All fields are designed so execution engines, risk systems, and AI agents can plug them directly into routing, sizing, and hedging logic.

The goal is simple: make 24/7 synthetic-equity markets behave more like properly anchored institutional products, without forcing every desk to rebuild decades of fundamentals research.