Five teams. Five questions.

The programme

Each team watches one question continuously, keeps score of its own work, and publishes under the same rules. Status lines are dated — this site never shows a reading a team did not produce.

01TEAM

Bottleneck Intelligence

Which small supplier does the whole AI build-out depend on?

Some parts of a chip can only be bought from one or two companies. When demand surges and supply cannot, those companies can grow extremely fast. This team tracks 13 candidates, scores each one out of 100, and only makes a call — a written prediction, recorded in advance — when a candidate scores 85 or more. So far, none has. That patience is deliberate.

Founding study complete13 candidates tracked

Status as of Oct 26, 2026

Read the founding study → How the team works →

02TEAM

Crash Watch

Is a crash or recession coming — and how much warning would we get?

US crash risk · AmberUS recession risk · Green-AmberEurozone · Red-AmberUK · Japan · China · Amber

Eight families of warning signs — jobs, lending, borrowing costs, share prices and more — checked every weekday across five economies and graded from Green (calm) to Red (danger). Today’s picture: high share prices and high long-term interest rates are the risk — not the bank-and-jobs pattern that came before 2008.

Baseline as of Oct 27, 2026

Read the founding study → How the team works →

03TEAM

Market Lens

What is the market thinking?

Twice a day, this team reads what CNBC and Bloomberg are saying and watches how prices actually react. Each lesson — “the market cared about X, not Y” — is written down, then re-checked every Friday and marked holding up, weakening, or retired. Nothing is quietly deleted. Its founding study is being prepared from the lessons gathered so far.

Midday + close editions, dailyLessons re-checked every Friday

Status as of Oct 28, 2026 · Founding study planned

How the team works →

04TEAM

Spring Loader

Which big stocks are coiling quietly before a jump?

Every weekday this team scans the S&P 500 for shares that have traded in a tight range for days — like a spring being compressed — and tracks what happens to each pick over the next five days. In October 2026 it found a mistake in its own maths: its price targets were set too far away. Fixing them lifted back-tested success from 4% to 41% — and the team published the mistake and the fix openly.

Daily weekday scanEvery pick scored after 5 days

Status as of Oct 28, 2026 · Founding study planned

How the team works →

05TEAM

Stock News Watch

What is really moving the four names under watch — fact or rumour?

Continuous monitoring of four names: company announcements, official filings, analyst rating changes, and unusual trading activity. Every item is labelled for what it is — confirmed or rumour — and shaky claims are flagged, not repeated as fact.

Four names under watchContinuous

Status as of Oct 28, 2026 · Continuous

How the team works →