Methodology
How ThesisLoop research is made
Every ThesisLoop thesis is generated by our analysis pipelines from a company's own primary-source disclosures, with every insight cited back to the document it came from. This page explains exactly how that works — the pipelines, the inputs, how we score conviction, how often it updates, and where the limits are.
Last updated July 1, 2026
Important disclaimer
AI-generated research for informational and educational purposes only — not personalized investment advice, a stock recommendation, or a guarantee of returns. ThesisLoop is not a SEBI-registered investment adviser or research analyst. Do your own diligence and consult a registered professional before making any investment decision.
The five analysis pipelines
Each thesis runs five independent pipelines. They look for fundamentally different things, so we never blend them into a single black-box score — you see each pillar on its own.
Management Credibility
Backtests what management promised in earlier disclosures against what was actually delivered — a forward-looking track record of guidance, capital-allocation calls, and stated priorities versus outcomes.
Business Model
Assesses moat, unit economics, pricing power, and competitive position from the company's own reporting to judge how durable the earnings engine really is.
Future Growth
Identifies growth catalysts, addressable-market expansion, and reinvestment runway that management is pursuing, with each claim traced to the document it came from.
Risk
Surfaces regulatory, operational, concentration, and balance-sheet risks disclosed in filings and calls. This pillar is scored inverted — a higher risk score means more risk, not more quality.
Scenarios
Maps macro and industry events (rate moves, policy shifts, sector cycles) to a company-specific bull / base / bear read, so the thesis reflects the world it operates in — not just its own filings.
Primary-source inputs
We read what the company itself published. There is no anonymous opinion or scraped rumor in a thesis — every insight is cited back to one of these documents.
Annual reports
Audited financials, management discussion, and long-form disclosure — the backbone for business-model and risk analysis.
Concall transcripts
Quarterly earnings-call transcripts, where management commits to guidance and answers analyst questions — the primary evidence for credibility backtesting.
Investor presentations
Investor-day and quarterly decks that lay out strategy, segment detail, and growth targets in management's own framing.
AI-generated, honestly labeled
The analysis is produced end-to-end by AI reading primary-source documents — no human analyst writes or rewrites it, and we never present it as if one did. Quality control is structural: every published insight carries a citation to the exact document it came from, so any claim can be checked against the primary source in one click, and a pillar with no supporting documents is left unscored rather than guessed. We label our work as AI-generated because it is — and because being clear about that is part of the trust we're trying to earn.
The conviction-scoring rubric
Each pillar is scored 0–100 from the evidence found. Management, Business Model, and Future Growth are higher-is-better; Risk is inverted (higher = more risk). A blended conviction score averages the pillars, folding Risk in as (100 − risk) so the roll-up stays directionally honest.
Strong (70–100)
Evidence consistently supports the pillar. On the Risk pillar this scale inverts — 0–33 reads as Low risk.
Mixed (40–69)
Evidence is uneven or partially supportive; the thesis carries open questions. Risk reads 34–66 as Moderate.
Weak (0–39)
Evidence is thin or contradicts the pillar. Risk reads 67–100 as High risk.
Blended conviction reads as High (70–100), Moderate (40–69), or Low (0–39). A pillar with no supporting documents is left unscored rather than guessed.
Update cadence
A thesis is re-analyzed when new primary-source documents are published — a fresh annual report, a quarterly earnings call, or a new investor deck — or on demand. Every published report carries an “Analyzed” date so you always know how current the underlying evidence is. Older analysis is not silently presented as new.
Limitations
- AI outputs are probabilistic and can be incomplete, misread a source document, or reflect gaps in the underlying disclosure.
- Citations are provided so you can verify every claim against its source — but the presence of a citation does not guarantee the conclusion is correct or current.
- Coverage depends on what a company chooses to disclose. Sparse or delayed reporting produces a thinner thesis.
- Nothing here is a recommendation, price target, or suitability assessment. It is background research, not advice.
Common questions
Is ThesisLoop research written by a human or by AI?
Every thesis is AI-generated. ThesisLoop's analysis pipelines read a company's primary-source documents, extract findings, and cite each insight back to the exact document it came from. No human analyst writes or rewrites the output — which is why every claim carries a citation you can check yourself.
Is this investment advice?
No. ThesisLoop provides AI-assisted research for informational and educational purposes only. It is not personalized investment advice, and SIGNALSTACKS TECHNOLOGIES PRIVATE LTD is not a SEBI-registered investment adviser or research analyst.
Where does the underlying information come from?
From a company's own primary sources — annual reports, quarterly concall transcripts, and investor presentations. Every insight is cited back to the specific document it was drawn from so you can verify it.
How often is a thesis updated?
A thesis is re-analyzed when new primary-source documents are published — a new annual report, quarterly earnings call, or investor deck — or on demand. Each published report shows the date it was last analyzed.
Important disclaimer
AI-generated research for informational and educational purposes only — not personalized investment advice, a stock recommendation, or a guarantee of returns. ThesisLoop is not a SEBI-registered investment adviser or research analyst. Do your own diligence and consult a registered professional before making any investment decision.