Universe
We begin with global listed equities and remove names that appear stale, duplicated, delisted, acquired, suspended, or otherwise not suitable for a current public-market ranking.
How The Ranking Is Built
Signal Ledger starts with a broad public-equity universe, filters for investability, writes model-assisted research notes, then uses multiple judging passes to compress the field into a final top-100 list.
Research Layer
Each surviving company gets a compact investment thesis focused on forward valuation, upside torque, business inflection, capital allocation, balance-sheet quality, and the risk that the setup is a value trap.
Tournament
Candidates are grouped and compared in repeated model-judged rounds. The early passes emphasize breadth and consistency; the later passes use stronger reasoning and longer context.
Final Review
The final 100 receive a fresh high-intelligence thesis rewrite before the last stack-rank. The published list is the final ordering, not a price target service or a recommendation to trade.
What The List Is For
The ranking is designed as a research queue: a way to surface unusual small-cap setups that may deserve deeper human review. It can be wrong, stale, incomplete, or too optimistic. Use it as a starting point, not as investment advice.