Distress Radar
Ranks listed companies by adverse-event likelihood. Certified point-in-time AUC 0.7366 (all listed), 6 Nov 2026.
Economic intelligence, connected
BaselineWatch is an intelligence company. We connect public records and open-source intelligence — SEC filings, corporate events, government contracts, trade and regulatory data — with world events and supply shocks to research how companies and economies change. We publish only what survives independent verification.
Point-in-time data · Evidence-linked methodology · Independent certification
Research view
ILLUSTRATIVE ONLYFrom event to outcome
How a research question is framed before anything is tested.
SOURCE 01
Filing language change
SOURCE 02
Contract or regulatory event
OUTCOME TO TEST
Distress, delisting, recovery
Test design
Preregistered before the data is seen
Verdict
Certified, or sent to the graveyard
At a glance
As of 6 Nov 2026
0.7366
Distress Radar AUC, any adverse event
All listed firms · certified 6 Nov 2026
0.7571
AUC for bankruptcy
Firms priced at $1 or more · certified 6 Nov 2026
347,797
Company-quarters in the evaluation
Panel saved permanently and replicable
65
Entries in our graveyard index
Ideas we tested and killed or withdrew
AUC: 0.5 is a coin flip, 1.0 is perfect ranking. Lead time is not re-measured and is not claimed.
01 / Flagship
LIVE
Ranks listed companies by the likelihood of an adverse event — bankruptcy, performance delisting, going-concern stress — from point-in-time public records. One product: the Catalyst Atlas and dilution tracking are part of it.
| Outcome | Universe | AUC |
|---|---|---|
| Any adverse event | All listed | 0.7366 |
| Any adverse event | Price ≥ $1 | 0.7312 |
| Any adverse event | Liquid names | 0.7152 |
| Bankruptcy | Price ≥ $1 | 0.7571 |
| Performance delisting | Price ≥ $1 | 0.7671 |
Add-ons · buy online
Catalyst Atlas — weekly pack
How each catalyst type has behaved after the event. One-time purchase.
Dilution tracker
Alerts when a company files for new share issuance or dilutive financing.
Watchlist · risk ranking
PRODUCT PREVIEW · SAMPLE DATACompany A
CRITICAL
Company B
ELEVATED
Company C
ELEVATED
Company D
BASELINE
Why Company A is critical (score breakdown)
Delivery: dashboard · API · S3 · Snowflake share
02 / Marketplace
Only BaselineWatch’s own work. Open any card to see its testing history, who built it, who certified it and its sources.
Status badges describe where an item stands in our process. They are a status, not a rating or a recommendation.
Ranks listed companies by adverse-event likelihood. Certified point-in-time AUC 0.7366 (all listed), 6 Nov 2026.
How each catalyst type has behaved after the event, from measured price and volume history.
Alerts when a company files for new share issuance or dilutive financing.
How auditor-flagged risk language changes over time. Passed early tests; full preregistered battery pending. Not a claim.
Financial stress among hospital operators. Passed exploratory testing; confirmatory testing still ahead. No performance claim.
Point-in-time history of company event filings, mapped to a stable company identity.
The outcomes behind Distress Radar: bankruptcies, performance delistings and related events, by company and date.
Ten to fourteen pages on one urgent sector question, from primary government and filing data.
Withdrawn 4 Nov 2026: the factor window was misaligned. The reason and the re-measurement are public.
Also listed on Neudata and Eagle Alpha.
03 / Research standards
Every candidate faces one question: does it add reliable information that existing data does not already carry?
STEP 1
Pre-gate screen
Graveyard, sources, sample size, clear hypothesis.
STEP 2
Preregistration
The test is fixed before the data is seen.
STEP 3
Full test battery
Factor controls, placebos, out-of-sample.
STEP 4
Independent certification
Never certified by the team that built it.
STEP 5
Forward test, then production
Monitored live; scratch paper kept forever.
Point-in-time provenance
We separate when something happened from when it was public and when it entered our system.
Reproducible by design
Every published figure keeps its panel, script and checksums, so a buyer can replicate it.
Incremental value
Compared against established baselines, not a zero-effect strawman.
We publish our failures
When a result does not hold up, we withdraw it in public and say why.
04 / The graveyard
Most ideas fail. We keep a public record of every signal we killed or withdrew, with the reason — so nobody, including us, sells it again. Methods stay protected; verdicts are public.
65 entries in the graveyard index as of 6 Nov 2026.
Disclosure-drift alpha
WITHDRAWN · 4 NOV 2026
The published excess return used a misaligned factor window. Over the holding period it is not significant. The forensic-linguistics framework remains in use.
Executive-departure predictor
RETRACTED · JUL 2026
An audit found a sampling artifact. A forward holdout on new events remains open.
Linguistic freeze
KILLED · JUL 2026
Added no timing information beyond existing measures, and was not stable across eras.
05 / Who we help
Institutions
Funds and asset managers license served products and point-in-time datasets, with full Signal DNA for due diligence.
Discuss licensing →Data partners
Platforms and data vendors embed or resell independently verified BaselineWatch products.
Explore a partnership →Individual investors
Watchlists and add-ons priced for individuals — our founding mission is to open institutional-grade data to everyone.
See plans →06 / Special reports
Ten to fourteen pages on one urgent question, built from primary government and filing data and checked three times before delivery.
From $5,000 to $20,000 per report
Commission a reportReport titles shown are examples of the format, not published reports.
SPECIAL REPORT
Rare Earths: Who Is Exposed When Exports Tighten
BaselineWatch Research · Example
SPECIAL REPORT
Hospital Operators Under Pressure: A Filing-Level View
BaselineWatch Research · Example
07 / About
Disclosure intelligence is one core layer of what we do. Our research desk watches the world for the events that move companies — geopolitical shifts, supply shocks, export controls, regulatory changes, contract awards — and works backwards from the questions buyers actually need answered.
That is the foundation for our next frontier: event prediction — researching which world events lead to which corporate outcomes, tested to the same standard as everything else we publish.
From the founder
“Financial intelligence should be verifiable without trusting its author.”
BaselineWatch was founded by a principal who operates under the DBA name Sebastian — a real, directly reachable person. He works in a sensitive role in the public sector, which limits his public profile for now; he intends to take a fully public-facing role in the future.
Every published signal must survive a pre-registered validation battery. Failed signals are retired publicly — 65 recorded to date — and results are reproducible end to end.
BaselineWatch operates as BaselineWatch LLC, a Wyoming limited liability company (formed August 2026). All contracts and licensing agreements are executed by BaselineWatch LLC. Direct contact: baseline@baselinewatch.com
Licensing, partnerships, special reports or research access.