This is the screen a portfolio manager opens at 7:15am. Everything below is clickable — start with a card, then look at the Belief map to see which unrelated positions rest on the same assumption.
Sample fund, sample positions. Not anyone's real book.
| Position | Side | Exposure | Thesis | What changed | Next catalyst | Action |
|---|
Sorted by decision urgency — not alphabetically, and not by weight. The first row is the one that needs a person. Two of the five positions carrying this morning's weakening assumption sit at the bottom of it, marked healthy.
Belief concentration, not sector exposure. A semiconductor designer, a power-equipment maker and a utility appear unrelated in every risk report you own. If this assumption breaks, they move together.
| Date | Position | What the firm believed | Confidence | Outcome |
|---|---|---|---|---|
| 2026-03-03 | NVDA | Hyperscaler capex stays above 20% through FY28 | 78% | weakening |
| 2025-11-14 | XYZ | Unit growth remains above 15% | 71% | wrong |
| 2025-08-02 | CEG | Power constraints ease through 2026 | 64% | correct |
| 2025-05-19 | KO | Consumer spending stays resilient | 83% | correct |
Nobody can reconstruct this later. It exists only because something was writing it down before the answer was known — which is why a competitor cannot backfill it, and neither can we.
Every calibration claim carries its sample size. A firm told it is wrong about something on three observations will either ignore the claim or wrongly believe it.
Onboarding starts from your positions and the research you have already written — not from typing three hundred theses.
Request access