PatternRankPublic research ledger
Research note

The model had a weak month. Here is the complete sample.

Completed PatternRank outcomes through Jul 15, 2026, including the latest distribution, positive and negative rows, and full-sample context.
Typetrack record update
PublishedJul 24, 2026
SeriesResearch desk
StatusPublic
Open metricsTrack record
ExamplesSamples
UseResearch only

This Weekly Ledger is for self-directed US equity researchers who build after-close watchlists and want to examine a finite ranked set before doing their own chart, catalyst, liquidity, and risk work.

It is deliberately different from a highlight reel. Every completed ranking inside each reported window stays in the measurement set. Positive, negative, and flat outcomes are counted together, and recent weakness is not replaced with a more flattering historical period.

The result before the interpretation

The tables below separate three questions: what happened in the latest completed week, what the recent 30-day distribution looks like, and how that compares with the full measured sample. None of those views represents a live brokerage account.

7-Day Momentum

Latest completed window

From Jul 09, 2026 through Jul 15, 2026, the public ledger contains 41 completed rankings.

MeasureLatest 7 daysPrevious 7 days
Measured rankings4139
Positive outcomes1810
Win rate43.9%25.6%
Mean outcome-4.4%-7.1%
Median outcome-3.3%-7.9%

Recent distribution

The trailing 30-day window contains 174 completed rankings: 52 positive, 122 negative, and 0 flat.

Distribution measureTrailing 30 daysFull sample
Mean outcome-5.9%+3.9%
Median outcome-5.9%+0.0%
25th percentile-15.3%-6.9%
75th percentile+1.8%+9.9%
Win rate29.9%49.8%

The strongest completed row in the full sample was $WOLF on Jul 01, 2025 (+531.9%). The weakest was $CERO on Oct 24, 2025 (-96.5%). These are distribution extremes, not recommendations.

7-Day ETB

Latest completed window

From Jul 09, 2026 through Jul 15, 2026, the public ledger contains 14 completed rankings.

MeasureLatest 7 daysPrevious 7 days
Measured rankings1427
Positive outcomes78
Win rate50.0%29.6%
Mean outcome-0.0%-8.0%
Median outcome+0.8%-10.2%

Recent distribution

The trailing 30-day window contains 85 completed rankings: 30 positive, 55 negative, and 0 flat.

Distribution measureTrailing 30 daysFull sample
Mean outcome-3.9%+3.0%
Median outcome-5.2%+0.0%
25th percentile-15.4%-6.9%
75th percentile+4.8%+8.7%
Win rate35.3%49.4%

The strongest completed row in the full sample was $WOLF on Jul 01, 2025 (+531.9%). The weakest was $PEPG on Mar 25, 2026 (-63.3%). These are distribution extremes, not recommendations.

Why the distribution matters

A mean can be pulled upward by a small number of unusually large outcomes. A win rate can hide the size of both wins and losses. A cumulative research series can hide whether the most recent month is behaving differently from the longer record. Reading those measures together is more informative than turning any single field into a claim about what should happen next.

The 25th and 75th percentiles describe the middle half of completed outcomes. The median describes the middle ranking rather than the average ranking. The strongest and weakest rows show the tails that can distort summaries. PatternRank publishes all of them because an auditable process must remain visible when the current window is uncomfortable.

How to use the latest window

Treat the latest seven-day block as a monitoring instrument, not a verdict. A small completed set can move sharply when one observation enters or leaves the window. Compare its positive and negative counts with the previous block, then compare both with the 30-day quartiles and full-sample median. That sequence helps distinguish a short run of noisy outcomes from a broader distribution shift.

The same caution applies when the latest block improves. A stronger week does not erase a weak month, and a strong full-sample total does not answer whether current behavior has changed. The purpose of publishing overlapping windows is to keep those tensions visible instead of forcing them into one promotional number.

What this ledger does not measure

  • A personalized entry, exit, position size, or portfolio.
  • Fees, taxes, slippage, borrow availability at execution, or account constraints.
  • Whether a ranking was suitable for a particular researcher.
  • A guarantee that a historical distribution will persist.

The reported 7-day outcome is a consistent research observation attached to the original ranking. It is not the return a subscriber necessarily could or would have realized.

How to audit the process

Start with the public track record, then inspect completed sample rankings. The track record exposes trailing, year-to-date, and full-sample windows; the samples preserve the model, date, confidence context, and measured outcome on individual rows.

The live paid product provides the current after-close shortlist and broader model coverage. Completed evidence and the Weekly Ledger remain public. Get the next Weekly Ledger when a new completed window is published.


PatternRank is an educational analytics platform. Rankings are for research and informational use only and are not investment advice. Past performance does not guarantee future results.

Weekly Ledger

Get the next Weekly Ledger.

One email when the next edition is published. The complete analysis remains public.

PatternRank is an educational analytics platform. Rankings are for research and informational use only and are not investment advice. Past performance does not guarantee future results.