Quantitative infrastructure
for the art market.
Decision-grade analytics for auction houses, lenders, allocators, and researchers.
Public auction records normalized into a research-grade panel, run through a cascade of econometric and probabilistic models. The pilot is opening to a small group of first users by invitation. Request access below.
Four audiences, one platform.
Auction houses & underwriters
Probabilistic guarantee pricing. Lot-level expected P&L, calibrated P(loss) ceilings, ensemble forecasts incorporating Hawkes intensity, MDN drift, and Bayesian level shifts.
guarantee calculator →Lenders & insurers
Liquidity tiers and time-to-sale estimates. Drawdown statistics, repeat-sales indices, value-at-risk with explicit normality caveats.
artist screener →Allocators, advisors & private collectors
Cross-asset benchmarking against S&P 500, gold, REITs, luxury equities. Portfolio concentration, correlation, and comparables.
portfolio context →Researchers
Public API. A research lab exposing every model with diagnostics, calibration plots, and slice metrics. Reproducible Python pipeline.
API & lab →Methodology & case studies.
Probabilistic Guarantee Pricing for Auction Houses
A mixture-density approach to pricing auction guarantees with calibrated downside risk. Backtested against the Macklowe, Newhouse, and Allen sales.
Repeat-Sales Indices for Art Segments
An open-method repeat-sales index family covering Blue-Chip Contemporary, Post-War Established, and Emerging segments — built using Bailey-Muth-Nourse on global auction records.
A Latent Factor Model for the Art Market
Asymptotic principal-components on the per-artist quarterly returns panel. Recovers the orthogonal factors driving the cross-section — equities have Fama-French / Barra; crypto has Liu-Tsyvinski-Wu; art now has its own.