Trade analysis and review at scale. Extract claims, verify them against independent data, and separate skill from luck.
A composite return stream tells you that a manager made money. It does not tell you whether any individual trade reflected genuine insight or happened to work. Allocators need to distinguish skill from luck at the single-trade level, not statistically in the aggregate, but structurally, by decomposing each idea into its thesis, timing, sizing, and factor exposure, then validating each against independent data.
Trade ideas, position rationale, and conviction signals are pulled from investor letters, transcripts, and communications in the Terrace graph.
Each claim is broken into verifiable components; thesis, catalyst, timing, sizing. Cross-referenced against SEC filings, earnings data, and market outcomes.
Returns are decomposed into systematic factor exposures and idiosyncratic alpha. What the market gave vs. what the manager earned.
Every claim is tested against independent sources. Trade analysis draws on a quantitative factor model for risk attribution, and a continuously updated fundamental data layer covering SEC filings, earnings, transcripts, and corporate actions across US public equities.
Multi-factor risk model decomposes returns into market, sector, size, value, momentum, and volatility exposures. Isolates what came from factor tilts vs. genuine selection.
SEC filings, earnings data, analyst estimates, and corporate actions. Continuously updated. Used to verify thesis claims and catalyst timelines against what actually happened.
Price history, volume, short interest, and options activity. Validates timing claims and measures realised P&L against the stated thesis window.
13F holdings, IAPD brochures, and regulatory filings. Cross-references stated positions with disclosed holdings to verify consistency.
Strata is the quantitative backbone of Terrace’s trade analysis. A multi-factor risk model built for portfolio-level attribution and security-level decomposition. Given a set of positions, Strata separates systematic factor returns from idiosyncratic alpha, answering how much of a manager’s P&L came from exposures anyone could replicate, and how much came from genuine security selection.