ANALOG LEDGER

Quantitative Equity Research

Four years of systematic research, every candidate on record.

Most trading research shows you the winners and quietly buries the rest. Ours can't - every candidate the model ranks, good or bad, is published the same week, locked in forever, and left untouched to prove itself in real time.

6.24M
Prices tracked
+520%
Total return
+127%
Average alpha
2.20
Sharpe (n=297)
8.16
Sortino
-14.2%
Max drawdown
Simulation

Simulated portfolio performance

Each week, the top 3 ranked candidates are bought at $500 per position. Each position exits at the 95% stop-loss, or once it first reaches the 150% take-profit target, a trailing stop arms and ratchets up with the position's peak weekly close, giving back at most 15% from that peak. If neither triggers within 30 weeks, it closes at whatever return it reached by then (a time exit). Hover to inspect any week.

Portfolio growth by year

Each calendar year runs as its own fresh $15,000 account starting January 1st, benchmarked against whichever major index performed best over those same dates.

$25,130+67.5%
$15,000 start · 35 weeks

Equity curve for 2026: $15,000 to $25,130, +67.5% over 35 weeks.

$14k$19k$25k
YearReturnS&P 500 (SPY)Nasdaq (QQQ)Russell 2000 (IWM)Dow (DIA)Bitcoin (BTC)Alpha
2026*+67.5%+10.8%+14.6%+13.2%+7.7%-13.3%+52.9%
2025+211.0%+15.5%+18.4%+11.3%+13.5%-14.7%+192.6%
2024+161.5%+23.8%+25.9%+9.9%+13.0%-+135.6%

* Year to date.

All years →
Methodology

How candidates get ranked

01

Build the weekly universe

Each week, every stock with a real technical trigger becomes a candidate.

02

Retrain from scratch

A machine learning model trains fresh against a rolling window of recent history — never a model carried over from a prior week.

03

Filter and rank

That fresh model scores the week’s candidates and ranks them by how likely each one is to break out.

04

Only publish real conviction

A candidate has to clear a rolling confidence bar before it earns one of the week’s picks — most weeks fill all 3 slots, some weeks fewer.

Read the full methodology →
Latest scan

Week of September 7, 2026

PAID MEMBERS
████#1
████████
PAID MEMBERS
████#2
████████
Track record

Recent matured outcomes

WeekCandidatesAvg peak return100%+ hits
Jun 15, 20262+25.9%0%
Jun 08, 20263+46.8%0%
Jun 01, 20263+12.5%0%
May 25, 20263+72.0%0%
May 18, 20263+370.7%100%
May 11, 20263+17.0%0%
Full track record →
Automated · Immutable · Auditable

No human edits, overrides, or hand-picks the model's output - the scan and ranking run without human input, and every result is cryptographically timestamped and permanently on the record. Four years of simulated results already show the edge: a cumulative return of +520%, independently auditable by anyone.

See how it's verified →

145 weeks of unedited research, freely searchable.

The full candidate archive - every ticker, every matured outcome - is open to anyone. Subscriber access unlocks the current week's ranked candidates before the 12-week maturation window closes.