A Factor & Risk Model
You Can Query
Like an API
Barra/FactSet-adjacent quantitative factor exposures, betas and risk scores for ~10,000 US-listed companies - 13 years of point-in-time, survivorship-safe history, delivered through a licensed REST API.
Try live sample (no login) AAPL snapshot + screener slice, identical fields to the paid endpoint
Contact Data Team →What Is the Quantitative Factor & Risk Dataset?
A Standardized Factor & Risk Layer Over the US Market
Raw prices and filings are not a factor model. This dataset turns them into one: every one of ~10,000 US-listed companies carries cross-sectionally standardized factor exposures, risk scores, betas and market-structure metrics - the same kind of quantitative layer institutions buy from Barra or FactSet - computed daily and preserved as a point-in-time, survivorship-safe history back to 2013 Q1.
Two access patterns: a per-company endpoint for one name's live snapshot or full quarterly history, and a cross-sectional screener that filters and ranks the whole universe on any factor, risk or structure field. 90 documented fields, with units and a fixed column set on every row.
What's Included
90 Fields Across 10 Factor & Risk Families
Style Factors
10 fieldsCross-sectional standardized exposures: size, value, quality, growth, momentum, low-volatility and liquidity, plus a 0-100 composite factor score, a cross-sectional percentile rank and an index-weight percentile.
Factor Z-Scores
8 fieldsStandard-deviations-from-mean z-scores for market cap, 12-month momentum, 1-year volatility, beta, relative strength, liquidity, dollar volume and share volume.
Risk Scores
6 fieldsPercentile-ranked left-tail, stress, crash, trend-strength and concentration-impact scores, plus a market-level stress indicator - ranks, not probabilities.
Betas & Benchmark-Relative
9 fieldsBeta vs US-500, Technology, Nasdaq-Composite and Small-2000; correlation, relative strength, 1-year and 3-year alpha, and sector tracking error.
Market Structure
5 fieldsMarket cap, industry cap weight and equal weight, market-cap rank within industry, and cap-weighted return contribution.
Index Membership
6 fieldsUS-500, US-Tech-100 and US-Small-2000 membership flags and index weights, point-in-time as-of each period.
Ownership & Short Interest
5 fields13F institutional ownership, insider ownership, short interest as a percent of float, days-to-cover and free float (quarter-end as-of, lagged).
Liquidity
7 fieldsDaily turnover, dollar volume and rank, Amihud illiquidity, 50-day average daily dollar volume, a liquidity tier and a 0-100 liquidity score.
Technicals
17 fields1/3/6/12-month momentum, 30-day/90-day/1-year and downside volatility, RSI, distance from 50/200-day MAs, max drawdown, Sharpe, Sortino and information ratios.
Regime Labels
3 fieldsCategorical risk, volatility and momentum regime labels for each name and period.
Coverage & What Makes It Different
Point-in-Time, Survivorship-Safe, Never-Garbage
- Universe: ~10,000 US-listed companies, identified by SEC CIK, classified by CSIMarket, SIC and NAICS.
- History: quarterly point-in-time panel from 2013 Q1 (13 years) plus a daily live snapshot.
- Point-in-time: every row carries an as-of trading date and reflects only what was known then - no look-ahead, no restatement bias.
- Survivorship-safe: delisted, deregistered and merged companies stay in the historical panel, so backtests are not survivorship-biased.
- Never a garbage value: when a figure can't be computed reliably the field is served
nullwith a per-row "held" reason - the column set is identical on every row.
How to Access
Two REST Endpoints, One Licensed API Key
The dataset is delivered as JSON over two endpoints, secured with a licensed API key:
GET /api/v1/companies/{ticker}/quant-factors- one company's live snapshot (D) or quarterly history (Q).GET /api/v1/quant-factors/screener- cross-sectional filter and rank across the universe on any factor, risk or structure field.
Inspect the full field set, units and coverage before licensing, via the public metadata layer:
- Dataset catalog entry - families, field counts, coverage
- Schema · Field dictionary - names, types, units, descriptions
- OpenAPI spec · API docs
Access tiers and commercial/OEM licensing are arranged with the data team - see licensing or email data.info@csimarket.com.
Who Uses It
Built for Systematic & Data-Driven Workflows
Quant & Systematic Funds
Build and backtest factor and risk models on a survivorship-safe, point-in-time panel - exposures, z-scores, betas and risk scores already standardized cross-sectionally.
Fintech & Roboadvisors
Embed factor analytics, risk scores and screens directly from a REST API instead of building a factor pipeline from raw filings and prices.
Registered Advisors
Run cross-sectional screens, risk overlays and peer-relative factor profiles for portfolios and client reporting.
AI / LLM Agents
Ground models in structured, point-in-time factor data with a fixed, documented column set and machine-readable units - never a fabricated value.
Data Integrity
Methodology & Sourcing
Factors and risk scores are computed cross-sectionally each period over a daily market-structure model built from SEC filings and market data. Exposures are standardized as z-scores or percentile ranks within the period universe; scores in the 0-100 range are percentile ranks, not probabilities. Ownership and short-interest fields carry a quarter-end as-of lag (13F / FINRA). Unreliable rows are held field-by-field with a documented reason rather than dropped or guessed. See the methodology endpoint for definitions.
FAQ
Frequently Asked Questions
What is the CSIMarket Quantitative Factor & Risk Dataset?
It is a point-in-time, survivorship-safe dataset of 90 quantitative fields across 10 families - style-factor exposures, factor z-scores, risk scores, betas and alphas, market structure, index membership, ownership and short interest, liquidity, technicals and regime labels - for approximately 10,000 US-listed companies, with a daily live snapshot and quarterly history back to 2013 Q1, delivered via a licensed REST API.
What makes the dataset point-in-time and survivorship-safe?
Every row carries an as-of trading date and reflects only information available on that date - there is no look-ahead restatement. Companies that later delisted, deregistered or merged are retained in the historical panel, so backtests and factor research are not biased by survivorship.
Which factors and risk metrics are included?
Cross-sectional style factors (size, value, quality, growth, momentum, low-volatility, liquidity), a composite factor score with cross-sectional percentile rank, factor z-scores, percentile-ranked risk scores (tail, stress, crash, trend, concentration) and a market-stress indicator, betas/correlation/alpha vs US-500, Technology, Nasdaq-Composite and Small-2000 benchmarks, market-structure and index-membership metrics, 13F and insider ownership, short interest, liquidity and technicals.
How is the data accessed and licensed?
Through two REST endpoints - a per-company history endpoint (/companies/{ticker}/quant-factors) and a cross-sectional screener (/quant-factors/screener) - secured with a licensed API key. The public API metadata layer (discovery, schema, field dictionary, OpenAPI) lets you inspect the full field set, units and coverage before licensing. Licensing and access tiers are arranged with the data team.
How are unreliable or unavailable values handled?
The dataset never returns a fabricated number. When a value cannot be computed reliably (for example a contaminated price basis or an unresolved ownership figure), the affected field is returned as null with a per-row "held" reason, while the rest of the row is served normally - so a fixed, machine-readable column set is preserved on every row.
Who is the dataset built for?
Quantitative and systematic fund managers building factor and risk models, fintech and roboadvisor platforms embedding factor analytics, registered investment advisors running screens and risk overlays, and AI/LLM agents that need structured, point-in-time factor data rather than raw filings.
Ready to Build?
Start With the Public Schema, License the Data
Inspect the full field set free, then arrange a licensed key with the data team.