StockAnalysis alternative · open Bayesian quant
StockAnalysis.com displays clean financial tables. Framler runs a Bayesian multi-factor engine on top of similar data, producing a 0-100 score with regime context and conformal prediction intervals. Tables are useful; a model that ranks them is more so.
One shows the numbers. The other tells you which numbers matter today.
StockAnalysis.com is excellent at presenting cleaned financial statements, ratios, and quarterly trends. It does not, however, tell you whether the data is bullish or bearish in the current regime. Framler runs a Bayesian factor model on similar inputs and produces a single 0-100 score with a verdict — bullish, mixed, bearish — plus a 90% conformal prediction interval.
A 30% revenue growth number means different things in a risk-on tech bull vs a risk-off contraction. Framler conditions every factor weight on the current market regime via a Bayesian Online Change Point Detection (BOCPD) model on SPY. A clean P/E in a transition regime gets weighted differently than the same P/E in a confirmed risk-on phase.
Framler ships a Mondrian conformal prediction interval on every score — a 90% coverage target, calibrated per bin against our own realised history rather than assumed. Where past reads in a bin landed far from the outcome, the interval is wider. Today those intervals span most of the scale, and we publish them at that width rather than dress them up. This kind of honest uncertainty is missing from a pure financial-tables product.
Yes. For pure data display — historical quarterly statements, ratios, comparison tables — StockAnalysis.com is excellent and we link to it from our methodology pages. Framler complements it with a quantitative score that ranks the same data through 13 academic factor families.
Our factor pipeline reads SEC EDGAR, ClinicalTrials.gov, options feeds, short-interest data and computes Bayesian regime posteriors for every ticker — 1,000+ names is the current quota-respecting limit. Coverage grows monthly as data sources scale.
Yes. Every factor, its academic source and the confluence-pattern definitions are documented at /methodology, with the structural invariants audited live at /coherence (9 invariants). The exact calibrated weights, thresholds and constants stay proprietary.
Look up any of 1,000+ tickers, see the full multi-factor breakdown, regime, prediction interval, and pattern flags. Free during early access.