Crene Logo
Review a thesisLive exampleMethodology
Crene
Platform
Review a thesisLive example
Data
Methodology
Company
About
Sign in
Explore
Review a thesisLive exampleAbout
Research
MethodologyData and calibrationAPI documentation
Legal
TermsPrivacySupport
© 2026 Crene, Inc.
Investment thesis tracking

How to track an investment thesis without losing the original view.

An investment thesis is not only a memo. It is a view about what must stay true for an investment decision to remain valid. Tracking the thesis means preserving that starting view, the assumptions underneath it, the evidence that changes those assumptions, and the decisions made as the view evolves.
Inspect a live record →Review a thesis →
Definition

Investment thesis tracking keeps the view and its changes in the same record.

A research memo captures a point in time. The investment decision usually lasts longer. New data arrives, assumptions move, models disagree, and the team revisits the view. If those changes live in separate notes, messages, spreadsheets, and meetings, the current thesis may be clear while the path that produced it becomes hard to reconstruct.

A durable thesis record therefore needs both the latest view and the preserved states that came before it. The purpose is not to freeze judgment. It is to make change explicit.

Starting view

What the team believes before the next round of evidence arrives.

Decision and horizon

The decision the thesis informs and the time horizon over which the view should be judged.

Assumptions and conditions

The facts, relationships, and observable conditions the thesis depends on.

Evidence and provenance

The new information that bears on those conditions, with enough source context to inspect it later.

Uncertainty and disagreement

Where independent models, analysts, or scenarios do not agree and therefore deserve review.

Review decision

Whether the team maintained, revised, replaced, or retired the view, with the rationale attached.

Version history

The prior states needed to reconstruct what changed without rewriting the original thesis.

Where it fits

A thesis tracker is not the same thing as a research repository.

Investment teams can use several systems at once. The distinction is the object each system is trying to preserve.

Research management system

Organizes the research corpus.

Notes, documents, meetings, search, collaboration, coverage, and workflow are the center of gravity. It answers where the team's research lives and how people retrieve it.

Thesis monitor

Watches selected conditions.

Alerts and monitoring focus on whether new filings, data, or news bear on stated assumptions. It answers what changed in the outside world that deserves attention.

Decision record

Preserves how the view changed.

The starting thesis, assumptions, evidence, disagreement, review decisions, and earlier versions stay connected. It answers what the team believed, what changed, and why the decision changed.

Workflow

A practical thesis review loop.

01

Define the view.

Write the thesis as a decision-relevant statement, name its horizon, and record the starting conviction or probability where appropriate.

02

Make the assumptions inspectable.

Break the view into the conditions, factors, or scenario paths that would strengthen, weaken, or overturn it.

03

Bring new evidence back to the record.

Review evidence against the saved assumptions rather than allowing the thesis to drift through scattered notes and memory.

04

Surface disagreement before the meeting.

Use model or analyst disagreement as a reason to inspect the underlying question, not as a substitute for judgment.

05

Record the decision and preserve the old state.

Keep the new view and the prior view together so the team can later explain what changed and why.

Crene

Crene treats the investment view as a governed record, not a disposable prompt.

Crene represents the question in the form that fits it: a Thesis for a directional view, a Factor for a continuous variable, or a Scenario for several coherent futures.

Independent model reads, model disagreement, evidence, and review conditions provide context for human review. Historical states remain preserved, and cleanly resolved forecast questions are evaluated separately through the public calibration layer.

Read the methodology →Inspect calibration and data →
Thesis, Factor, and Scenario records stay distinct rather than forcing every question into one representation.
Model disagreement is surfaced so a reviewer can inspect where independent assessments diverge.
Prior states are preserved so the current view does not overwrite the historical record.
Forecast calibration is disclosed as a measurement layer, separate from claims about investment performance.
Private evaluation workspaces are provided by invitation; the public live example can be inspected without an account.
Questions

Common questions about tracking an investment thesis.

What is investment thesis tracking software?
It is software that keeps an investment view connected to the assumptions, evidence, reviews, and decisions that change it over time. The useful unit is not only the latest note; it is the evolving record of the view.
How is thesis tracking different from a research management system?
A research management system is typically centered on capturing and finding research, notes, documents, and workflows. Thesis tracking is centered on the state of a specific investment view and how that state changes. The two can complement each other.
How often should an investment thesis be reviewed?
The cadence should follow the decision and the evidence. Some conditions change daily, while others are better reviewed around earnings, policy meetings, data releases, or scheduled portfolio reviews. A durable record should preserve both the cadence and the exceptional reviews triggered by material evidence.
Does Crene recommend what to buy or sell?
No. Crene is decision-record infrastructure. Model probabilities, disagreement, and evidence are inputs to research and review; they are not trade instructions or guarantees of an outcome.
How does Crene evaluate forecasts?
Cleanly resolved binary questions can be scored with Brier scores and calibration analysis against named resolution sources. Crene keeps that measurement layer separate from structural thesis, factor, and scenario records that may not resolve as one binary forecast.
Can teams use Crene data in their own models?
Crene exposes selected public read interfaces and structured records. Broader exports or private delivery can be arranged separately, so a team can discuss the schema and cadence appropriate for its research workflow.
Private review

Bring one live thesis.

Start with a real investment view and the decision it affects. Crene can structure the assumptions, preserve the starting state, and keep subsequent review decisions attached to the record.

Start a private review →See the public live example →