Finbox alternative · open Bayesian model
Finbox is a fundamental-analysis platform — clean tables, DCF tools, scoring models you can build yourself. Framler is a single Bayesian engine with regime conditioning and conformal prediction intervals already built in.
One gives you the toolkit. The other ships the finished engine.
Finbox's strength is letting you build custom scoring models out of fundamental data. That's powerful but requires you to design the model, pick the weights, and validate it yourself. Framler ships one finished Bayesian engine — 13 factor families with regime-conditional weights and conformal prediction intervals — already calibrated and audited.
Framler conditions every factor weight on the BOCPD-detected market regime (risk-on / transition / risk-off). A momentum signal at z=2.0 means different things in a confirmed bull vs a transition phase. Finbox's scoring templates evaluate factors at fixed thresholds without regime conditioning.
Every Framler score ships with a 90% Mondrian conformal prediction interval — finite-sample coverage guarantee per regime and sector bin. Finbox's scoring outputs are point estimates without an explicit uncertainty quantification.
Yes — for DCF intrinsic value modelling and custom fundamental scoring templates, Finbox is excellent. Framler doesn't do native DCF; we focus on factors with published Information Coefficient evidence in academic literature (pure DCF historically has weaker forward-return IC than multi-factor composites).
Not in the UI — the composite is one finished, calibrated engine. Power users can read every weight at /methodology + /coherence and replicate the math externally if they want a different mix.
Framler's pipeline reads SEC EDGAR, ClinicalTrials.gov, options chains, short-interest, and computes Bayesian regime posteriors per ticker. 1,000+ names respects API quotas; coverage grows monthly. Finbox's 10k+ universe relies on lighter per-ticker depth.
Look up any of 1,000+ tickers and get the 13-factor composite + regime + conformal interval + pattern flags. Free during early access.