Zacks alternative · open methodology
Zacks Rank is a 1–5 ordinal score driven primarily by analyst earnings revisions — the actual weighting is proprietary. DeepVane is a 0–100 composite across twelve peer-reviewed academic factor families, with a Bayesian regime overlay and a confidence interval on every signal. The math is on a public methodology page; you don't have to take it on faith.
One ranks on analyst sentiment behind a paywall. The other ranks on twelve academic factors in the open.
Zacks earned its credibility from decades of running the Zacks Rank. DeepVane earns it differently: every formula is on /methodology, every factor cites a peer-reviewed paper, and 21 mathematical invariants run live at /diag/engine on every request. We let the math be auditable rather than trusted-because-old.
The Zacks Rank is anchored on analyst earnings estimate revisions. That's one academic factor (closest to PEAD / Bernard-Thomas 1989). DeepVane composites twelve: quality (Novy-Marx 2013), value (Fama-French), momentum (Jegadeesh-Titman 1993), insider flow (Seyhun 1986), NLP tone (Loughran-McDonald 2011), accruals (Sloan 1996), short interest, options flow, sector momentum, industry spillover, factor interaction, and PEAD itself. Diversifying across factors is what every academic study on factor decay shows is the right move.
A Zacks Rank #2 doesn't tell you how confident the model is. DeepVane ships a Mondrian conformal interval with every score: "78 ± 14 at 90% coverage on this regime/sector cohort." That width is what tells you whether to size big or stay light.
Zacks doesn't do this. DeepVane runs a Bayesian model on every Healthcare-sector ticker — Hay et al. 2014 base rate updated with log-likelihood-ratios from enrollment velocity, endpoint amendments, mechanism risk, and cash-burn runway. Phase 3 readouts move biotech stocks 30–50% overnight; knowing the predicted failure probability ahead of time is structural alpha for the sector.
Yes during the early-access window. All twelve factor scores, all eighteen confluence patterns, all Phase 3 failure predictions, and all conformal prediction intervals are publicly readable without a paywall. The Pro tier launches on 16 May 2026 once forward returns calibrate the model. Beta users stay free.
Two structural differences. (1) Continuous 0–100 vs ordinal #1–#5 — finer signal at the same compute cost. (2) Twelve factors with public weighting and academic citations vs analyst-revision-weighted proprietary formula. The Zacks Rank can be excellent (the long-term backtest in Mendenhall 2004 is real); DeepVane optimises for transparency and breadth of factor coverage.
Yes — earnings momentum is one of our twelve factors (PEAD, Bernard-Thomas 1989). It contributes to the composite alongside the other eleven. We don't isolate it as a standalone score because the academic literature on multi-factor combination consistently shows that single-factor ranks decay faster than composites.
The DeepVane factor pipeline reads SEC EDGAR 10-K / 20-F filings, ClinicalTrials.gov, PubMed, options data, and short-interest feeds for every ticker. Running this stack on 10,000 names exceeds free-tier API quotas. The 374 are curated for liquidity, factor coverage, and pharma/biotech depth. We expanded from 234 to 374 in one cycle and continue to grow as data quality allows.
Type any of 374 tickers and you get the 12-factor APEX breakdown, the regime context, every confluence pattern firing, and the conformal prediction interval. Compare against the Zacks Rank you're paying for — that's the cleanest A/B.