Skip to content
Salarium Classical Profile MarkSALARIUMAUTONOMOUS INVESTMENT RESEARCH
RankingsPortfolioSimulationResearchMethodologyArchitectureAbout
1.0.0-RC1/RELEASE CANDIDATE
Salarium Classical Profile MarkSALARIUMAUTONOMOUS INVESTMENT RESEARCH

Open-source systematic equity research: governed data, out-of-sample rankings, concentrated portfolio construction, and auditable risk decisions.

Research only. Not investment advice. No live order execution.

EXPLORE

RankingsPortfolioSimulationResearchMethodologyArchitectureAboutPerformanceExperimentsCandidatesDisclosures

EVIDENCE

Forward paper snapshotRelease snapshot20D release rankingsCandidate snapshot

PROJECT

GitHub repositoryModel cardRelease notes
COMMIT 7c730e84c409SNAPSHOT AUG 11, 2026, 11:14 PM UTC© 2026 NIALL GILLEN · MIT LICENSE

FORWARD PAPER / MARKET CLOSE

Ranked securities.Full context, no false precision.

Explore the top 25 names from the frozen 20D model's latest paper 499-security cross-section. Scores express relative conviction; they are not price targets, recommendations, or guaranteed expected returns.

SIGNAL DATE

Sep 29, 2026

PAPER / NO ORDERS

PUBLISHED RANKINGS

25

499-name cross-section

TOP MODEL SCORE

0.071448

relative cross-sectional output

AVERAGE 20D VOL

5.07%

recent single-name volatility

PAPER RISK STATE

63%EXPOSURE
99.8% feature coverage

SALARIUM EQUITY RANKING

The ranked cross-section.

Open any row for supported score, percentile, volatility, risk, selection-band, and provenance fields.

25 / 25 PUBLISHED SECURITIES
Risk stateSelection band
010.071448100.0%7.02%NeutralTop-10 selection band
020.05012199.8%5.07%NeutralTop-10 selection band
030.04396099.6%4.73%NeutralTop-10 selection band
040.04390799.4%4.01%NeutralTop-10 selection band
050.04373299.2%4.58%NeutralTop-10 selection band
060.04322399.0%6.20%NeutralTop-10 selection band
070.04321198.8%4.75%NeutralTop-10 selection band
080.04308698.6%4.97%NeutralTop-10 selection band
090.04305998.4%4.90%NeutralTop-10 selection band
100.04303698.2%5.15%NeutralTop-10 selection band
110.04301198.0%4.78%NeutralPersistence buffer band
120.04292297.8%5.29%NeutralPersistence buffer band
130.04287097.6%5.50%NeutralPersistence buffer band
140.04280197.4%5.34%NeutralPersistence buffer band
150.04259497.2%5.12%NeutralPersistence buffer band
160.04222197.0%5.09%NeutralOutside portfolio band
170.04215096.8%4.70%NeutralOutside portfolio band
180.04195396.6%4.58%NeutralOutside portfolio band
190.04194396.4%4.46%NeutralOutside portfolio band
200.04179596.2%6.11%NeutralOutside portfolio band
210.04179396.0%4.57%NeutralOutside portfolio band
220.04161995.8%4.92%NeutralOutside portfolio band
230.04158495.6%5.02%NeutralOutside portfolio band
240.04154595.4%4.70%NeutralOutside portfolio band
250.04142195.2%5.21%NeutralOutside portfolio band

IN PLAIN ENGLISH

A higher-ranked stock scored better than other stocks in the same governed Liquid-500 cross-section at this signal date. The paper feed scores fresh market-close data with frozen model weights; Top-10 selection, persistence, covariance, position limits, and exposure controls still operate before a name enters the paper portfolio.

Model score

A relative signal produced by the governed technical ranking model. Magnitudes are meaningful only inside the same model snapshot.

20-day volatility

A recent single-name risk estimate. It is not the portfolio forecast; the covariance engine evaluates how selected holdings interact.

Risk state

A macro/risk context flag used by downstream exposure controls. It does not overwrite the cross-sectional ranking.

FROM RANK TO PORTFOLIO

Rankings are an input—not the final product.

The release architecture uses a Top-10 selection, rank-15 persistence buffer, 60-day shrinkage covariance, and a 25% signal-aware blend before exposure control.

See constructionOpen candidate research

MODEL GOVERNED_TECHNICAL_HARDENED_HORIZON_20D · 976,668 TRAINING ROWS · SNAPSHOT GENERATED OCT 1, 2026, 1:26 AM UTC · RELEASE SHARPE REFERENCE 1.298

RESEARCH DISCLOSURE

What the numbers do—and do not—mean.

This ranking is a forward paper market-close feed, not live investment performance or an exchange-grade quote feed.
Rankings can change materially as prices, volatility, and features change. The frozen model is not retrained by the daily refresh.
A high score does not imply suitability, valuation support, catalyst certainty, or a buy instruction.
The release result is simulated and should not be interpreted as future performance evidence.