Factors and signals

Factor library

QuantGPT

Daily and monthly factors per stock, computed by QuantGPT: returns, value, quality, growth, risk, owners and shorts.

Use in my models: choose in My data All my data choices

What it is

Daily and monthly factors per stock, computed by QuantGPT: returns, value, quality, growth, risk, owners and shorts.

factors daily: 36 daily price and earnings factors: returns, RS rating, EPS and sales growth, SMA stack, bases, ATR, volume, the liquidity pass, the trend template, breakouts.
factors tier a daily: Risk and quality: beta, volatility, idiosyncratic volatility, max return, skew, Amihud, turnover, overnight and intraday returns, listing age, gross profitability, ROA, margins, asset turnover, SUE.
factors tier b daily: Valuation and holders: P/E and earnings-yield industry percentiles, F-Score, forward P/E and EPS growth, 13F concentration and holder counts, dividend yield, initiations, cuts and streak. From 1990.
factors extra monthly: Monthly valuation and flow: P/E, P/B, P/S, earnings, FCF and shareholder yield, accruals, asset growth, issuance, short volume ratio, put/call and options/stock ratios (stale), surprise.
factors extra 2 monthly: 25 more monthly factors: long momentum and reversal, consistency, seasonality, industry momentum, insider buyers and net, short z and days to cover, expected-earnings days, revision breadth, institutional change and holders, cash flow to assets, dividend yield, S&P 500 member, market cap.

Source and licence

Source: QuantGPT (computed from vendor prices and filings).

Licence: derived. This is our reading of the table; counsel decides.

Credit: the QuantGPT engine and warehouse (Data sources).

Coverage and gap years

49,590 names, 94,033,012 rows in factors daily, from Jan 3, 1950 to Oct 8, 2026.

Years with no rows: none between the first and the last.

Table	Rows	First	Last
factors daily	94,033,012	Jan 3, 1950	Oct 8, 2026
factors tier a daily	103,427,618	Jan 3, 1950	Oct 8, 2026
factors tier b daily	74,735,170	Jan 2, 1990	Oct 8, 2026
factors extra monthly	944,982	Apr 1, 1990	Oct 1, 2026
factors extra 2 monthly	4,646,510	Jan 1, 1962	Oct 1, 2026

Freshness

Last row Oct 8, 2026: 1 day behind on the day measured. Updates: every trading day; monthly factors at each month end.

Point in time

Keyed on the trading day, built from that day's prices and from fundamentals as of their filing date.

No version history: the factors are rebuilt from the tables under them.

Delisted names

Companies that later delisted are in from 1998; before that the history is survivors only, so early years lean toward the winners.

Known issues

Fundamentals factors inherit ACCURACY.md section 1: pe TTM uses the reported (adjusted) EPS, ps TTM uses gross bank revenue, and market cap uses stale share counts. (GAPS, 2. How a stock ranks: the factor profile (score 5.0))
Leave out putcall ratio and os ratio (stale since 2026-05-15). (GAPS, 2. How a stock ranks: the factor profile (score 5.0))
The universe is the engine's liquid names (pass liquidity), thin and survivor-heavy before 1990. (GAPS, 3. Market health: breadth and regime (score 4.0))
Omen and regime labels are research: show the numbers, no verdict. (GAPS, 3. Market health: breadth and regime (score 4.0))
Measured 2026-10-09: signals daily breadth pct above 200d, spy above 200d and regime risk on are null on every row, so the engine has no risk-on flag to show. 2027 counts the shares itself and shows SPY against its 200-day as a number. (GAPS, 3. Market health: breadth and regime (score 4.0))
48 splits since 2022 are unadjusted. (GAPS, 15. Dividend and split profile (score 3.0))
The far-future rows are scheduled coupons on baby bonds and preferreds: filter by is preferred and is note. (GAPS, 15. Dividend and split profile (score 3.0))

Where it is used

Engine tools that reach it (your AI, over MCP):

list_factors The factor vocabulary usable in run backtest filters and rank by: 74 factors the backtester can use (momentum, value, quality, growth, trend, size), and 28 more (beta, volatility, Amihud, SUE,...
run_backtest Backtest any criteria in seconds: up to 8 filters over the factor vocabulary, ranked by a numeric factor, top N equal weight (or weighting: market cap, inverse vol, factor, custom, with max weight),...
run_screen The CURRENT tickers matching a set of criteria against the latest data in the warehouse - the operational output of a backtest you validated.
query_warehouse Run read-only SQL (DuckDB dialect) directly against any research database.

Ask AI

Every count and date here was measured on Oct 9, 2026 from the QuantGPT engine's warehouse and is read from the catalogue file when the site is built. Metadata only: no vendor rows are shown.