Separating Facts, Opinions, and Market Narratives

Modern investors can encounter more information in one day than earlier investors encountered in weeks.

The challenge is not simply finding information. The challenge is determining what kind of information it is.

Nearly everything you encounter while researching a stock belongs to one of three core categories, summarized in the callout below.

Facts, forecasts, and interpretations can all be useful, but they should not be treated as interchangeable.

Within the three core categories, research material comes in more specific forms.

The Stock Research Information Map below lists the major types and what must be verified for each.

A fact should be tied to observable or documented evidence. Examples:

• Reported revenue
• Reported EPS
• Debt balance
• Filing date
• Announced acquisition
• Disclosed dividend

A number can be factually correct and still be misleading without:

• Reporting period
• Units
• Accounting definition
• Comparison period
• Source

Fact: The company reported $5.00 EPS last year.

Forecast: Analysts expect $6.00 EPS next year.

Interpretation: The company's earnings growth will remain strong.

These are three different statements: the first is documented, the second is an estimate, and the third is a conclusion drawn from evidence.

Companies communicate through:

• Filings
• Earnings releases
• Presentations
• Earnings calls
• Press releases
• Interviews

A company statement can itself be a documented fact—that management said something—while the substance of a forward-looking claim may still be uncertain.

Example: Management says a margin decline was caused by temporary freight costs.

Fact: Management gave that explanation.

Research question: Do later results support the claim that the pressure was temporary?

As covered in Lesson 7, guidance is a company forecast.

It should never be presented as though the future result has already occurred.

Analyst estimates are forecasts.

They may be informed by detailed research, but they remain estimates.

Distinguish between:

• Estimate
• Rating
• Price target
• Qualitative opinion

None of these is a historical company result.

Price targets reflect assumptions and methodologies that may differ among analysts.

A price target is not a guaranteed future stock price.

Journalism can contain, sometimes within the same article:

• Documented reporting
• Quotations
• Data
• Analysis
• Interpretation

Learn to distinguish these elements within the same article rather than treating the whole piece as one kind of information.

Headlines summarize complex events and may omit important context.

Read beyond the headline before using an article as research evidence.

Opinion and commentary may provide useful questions or alternative perspectives.

The researcher's job is to identify the evidence supporting the opinion.

Social media can quickly surface:

• Breaking information
• Filings
• Commentary
• Rumors
• Screenshots
• Interpretations

Speed does not establish reliability.

The callout below states a principle worth applying to every widely shared claim.

A knowledgeable analyst, executive, journalist, investor, or industry specialist may provide valuable insight.

But credentials do not convert a forecast or interpretation into a fact.

Confident language can sound persuasive.

The research question remains: what evidence supports the claim?

A rumor is information circulating without adequate verification.

This lesson deliberately uses fictional examples only; real company rumors should not be repeated as research material.

Reputable journalism may sometimes rely on anonymous sources.

Distinguish reported sourcing from independently verifiable company disclosures.

Do not automatically reject or accept information solely because a source is unnamed.

Screenshots can:

• Omit context
• Be outdated
• Be edited
• Show only part of a document

Whenever practical, locate the original document.

These choices can change how a chart appears:

• Start date
• End date
• Scale
• Benchmark
• Omitted periods

This lesson is about reading information critically, not about technical analysis.

Example: “Profits increased 100%.”

If profit rose from $1 million to $2 million, that is mathematically true. The absolute amount still matters.

Company-wide averages can hide:

• Segment differences
• Geographic differences
• Customer differences
• Product differences

Two events occurring together do not prove one caused the other.

Fictional example: A stock rises during the same week a company announces a new marketing campaign. The timing alone does not establish that the campaign caused the rise.

A stock falling after bad news does not prove every explanation offered for the decline is correct.

Likewise, a stock rising after an announcement does not prove every positive interpretation is correct.

Example: Revenue slows. Possible contributing factors:

• Weaker demand
• Currency
• Lost customer
• Product transition
• Pricing
• Supply constraint

Research must determine what the evidence supports.

A market narrative is a commonly repeated explanation or story about a company, industry, market, or economic development.

Narratives can contain facts, interpretations, assumptions, or combinations of these.

Narratives often begin with a real, documented fact—which is part of what makes them persuasive.

Example: Fact: Revenue growth slowed. Narrative: “The company's growth era is permanently over.”

The first statement may be documented. The second requires much more evidence.

The same company can move from a popular positive narrative to a negative one without the underlying business changing as rapidly as the commentary.

Track the business evidence separately from the story being told about it.

Do not scrutinize only negative narratives. Examples of positive claims requiring evidence:

• “This company cannot be disrupted.”
• “This product will dominate the market.”
• “Margins will keep expanding.”

Negative claims also require evidence. Examples:

• “The business is dying.”
• “Customers are leaving.”
• “The company cannot compete.”

Convert a broad narrative into specific claims that evidence can support or contradict, using the workflow below.

1. 1Narrative
2. 2Specific Claim
3. 3Evidence Required
4. 4Primary Source
5. 5Supporting / Contradictory Evidence
6. 6Unresolved Questions

The Market Narrative Evidence Test table below organizes each narrative claim, the evidence supporting it, the evidence contradicting it, and what remains unknown.

Confirmation bias means researchers may naturally notice information that supports what they already believe and discount information that challenges it.

This tendency affects everyone, including experienced investors.

If you believe a company is strong, deliberately search for evidence that could challenge that conclusion.

If you believe a company is deteriorating, deliberately search for evidence that could challenge that conclusion.

The purpose is not to argue with yourself endlessly. It is to test the research.

Different sources may have different incentives. Examples:

• Companies promoting their businesses
• Analysts publishing research
• Journalists attracting readers
• Commentators building audiences
• Investors discussing positions

Do not assume an incentive proves dishonesty or invalidates the information; it is context for evaluating it.

Promotional or sponsored material should be clearly distinguished from independent research.

Check whether content is paid placement before treating it as an independent perspective.

AI tools can assist with:

• Organizing information
• Explaining terminology
• Summarizing documents
• Generating research questions

But AI output can contain errors, outdated information, missing context, incorrect citations, or invented details. Important financial facts obtained through an AI system should be verified against reliable source material before they are used in an investment decision.

When the exact wording materially affects the research, check quotations against:

• Original transcript
• Filing
• Recording
• Official statement

Before relying on an important number, verify:

• Source
• Period
• Units
• Definition
• Whether historical or estimated
• Whether GAAP or non-GAAP where relevant

An accurate article from six months ago may no longer describe the current situation.

Research requires both reliability and recency.

After finding an older source, ask: what happened afterward? Examples:

• Later earnings
• Revised guidance
• Completed acquisition
• Regulatory decision
• New filing
• Debt refinancing

Credible sources can disagree. When they do, work through the process below rather than discarding whichever source you prefer.

1. 1Identify the disagreement
2. 2Check definitions
3. 3Check dates
4. 4Check reporting periods
5. 5Find original sources
6. 6Record what remains unresolved

Sometimes the correct research conclusion is the one stated in the callout below.

Classify each important claim into one of exactly four categories, explained in the framework below.

This framework deliberately uses no numerical scores, stars, letter grades, percentages, or rankings—it is a classification, not a scoring system.

The warning-signs box below lists research-process habits that deserve correction.

Stock Sage's fundamental qualification rules require Positive EPS in at least 4 of the previous 5 years and Total Assets greater than Total Liabilities in each of the previous 5 years.

These are objective first-stage historical qualification rules. They do not prove future growth, competitive strength, management quality, undervaluation, the accuracy of a market narrative, or investment suitability.

Stock Sage helps organize the starting universe and financial evidence. The researcher still must investigate why a stock is trading where it is and distinguish documented evidence from narratives and forecasts.

The workflow below organizes evidence classification into a repeatable sequence, followed by a checklist of the questions this lesson teaches you to answer.

1. 1Identify the claim.
2. 2Classify it as fact, forecast, interpretation, narrative, or unverified claim.
3. 3Identify the original source when practical.
4. 4Check the publication date.
5. 5Check the reporting period.
6. 6Verify important numbers.
7. 7Separate reported results from estimates.
8. 8Separate management statements from independently verified evidence.
9. 9Identify evidence supporting the claim.
10. 10Identify evidence contradicting the claim.
11. 11Check for subsequent events.
12. 12Record unresolved questions.
13. 13Avoid forcing certainty where evidence is incomplete.
14. 14Update the conclusion when new evidence appears.