Narrative Credibility Analysis: Testing Corporate Claims

Narrative credibility analysis is the structured assessment of whether the story a public company tells about itself is internally consistent, externally verifiable and borne out by later results. In practice, the method compares what management claims in earnings calls, SEC filings, press releases and proxy statements with what the record subsequently shows, and it records the difference as evidence rather than impression. The sections below set out the theory, the filing-level mechanics, the empirical findings that justify the effort, and a repeatable workflow for analysts, governance professionals and journalists.

TL;DR

• Credibility has two testable dimensions: coherence (the story does not contradict itself) and fidelity (the story matches verifiable facts). Both can be scored from public records.
• Credibility attaches to individual disclosures, not to a firm as a whole, so each claim needs its own evidence trail.
• Weakly codified channels, such as the "Other Events" item in Form 8-K, and footnotes are where engineered narratives are most likely to escape scrutiny.
• Dictionary-based sentiment scores are insufficient; tone dispersion, readability and claim-to-outcome tracking carry more information.
• A consistent method applied across issuers, with a logged evidence trail, is what makes the output auditable and comparable.

Table of Contents

• What narrative credibility analysis measures
• Coherence and fidelity: the two tests
• Credibility belongs to the disclosure, not the firm
• Where the narrative lives in the filings
• Why tone alone is not enough
• A six-step workflow for a credibility file
• Worked example: reading a single claim
• Red flags that merit escalation
• Perspective: why a consistent method matters
• Applying the method at the sector level
• Where Lacuna Index fits
• Limits of the approach
• FAQ
• Sources

What narrative credibility analysis measures

The phrase draws on two traditions. The first is narrative theory in communication research, which asks why some stories persuade and others fail. The second is disclosure research in accounting and finance, which asks why investors discount or accept management statements. Narrative credibility analysis for public companies sits at the intersection: it treats a corporate disclosure as a story with claims, causal links and promised outcomes, and it tests that story against the public record.

The object of study is not sentiment and not the reputation of the chief executive. It is the relationship between narrative (what is asserted), evidence (what is disclosed in audited or otherwise verifiable form) and delivery (what subsequently happens to revenue, margins, capital allocation, remediation timelines and other stated commitments). The distance between the first and the third is the aspiration-to-execution gap, and the quality of the second determines how confidently that gap can be measured.

This differs from the general source-credibility exercises that dominate many search results for the term, which evaluate an author's expertise or a publisher's reputation. Those checks are suitable for a reading list. A corporate disclosure is an interested party's statement on its own performance, and the relevant question is not who wrote it but whether it survives comparison with later facts.

Coherence and fidelity: the two tests

The most useful theoretical vocabulary comes from Walter Fisher's narrative paradigm. Credibility is assessed along two dimensions: narrative coherence, the internal consistency of a story whose events hang together without contradiction, and narrative fidelity, the degree to which the story accurately reflects the real world or shared values. A published empirical measure of narrative coherence and fidelity, built on Fisher's theory, reported satisfactory reliability for both scales (Cronbach alpha of .77 for coherence and .82 for fidelity) and accounted for up to 42% of the variance in the persuasiveness of written narratives. More recent work on assessing coherence and fidelity in non-fictional narratives extends the same logic to accounts that assign roles and values to the people who participate in them, which describes a corporate strategy narrative rather well.

Translated into corporate disclosure, the two tests operate as follows.

Coherence is cheaper to test, because it needs only the company's own documents. Fidelity is harder and more valuable, since it requires an external anchor. The strongest analyses report both and keep them separate: a narrative can be perfectly coherent and still lack fidelity, and that combination is precisely what engineered disclosure tends to produce.

Credibility belongs to the disclosure, not the firm

A frequent error is to assign a single credibility rating to a company and apply it to everything it says. A review of the disclosure-credibility literature, summarized in a paper on impression management and disclosure credibility of corporate narratives, draws on Mercer (2004) to make a distinction: investors evaluate credibility separately for each specific disclosure instead of relying on a monolithic view of the firm. The same source notes that bad news is often judged more credible than good news, because managers have fewer incentives to fabricate negative information.

Three methodological consequences follow.

1. Claims are the unit of analysis. Each quantified or verifiable assertion (a margin target, a remediation date, a customer-count figure) receives its own evidence trail and its own outcome.
2. Direction of the claim matters. Favorable claims about the future warrant more verification effort than unfavorable ones, since the incentive to embellish is asymmetric.
3. Aggregation comes last. A company-level score is defensible only if it is built from claim-level findings that a third party can inspect.

This is the logic behind validating corporate claims in public disclosures one assertion at a time, rather than forming a global impression from tone.

Where the narrative lives in the filings

Not every disclosure channel carries the same weight or receives the same regulatory attention. An analyst building a credibility file should know where narrative is concentrated and where it is least constrained.

MD&A: the designed home of management's story

The SEC's 2003 interpretive guidance on MD&A states that the purpose of the section is to give readers the information necessary to understand a company's financial condition, changes in financial condition and results of operations. Practitioners read the requirement as a mandate for management to confront disagreeable realities early: the diagnostic question is whether a company identifies emerging trends and uncertainties before they become probable, since that is where valuation misalignment tends to hide. Commissioner Elisse Walter's 2010 speech, Let the Story Shine Through, singled out Item 303 of Regulation S-K as a premier place for companies to communicate with their shareholders about performance and risk.

Footnotes: the quieter channel

The MD&A attracts most reader attention. A study of 10-K narratives over a 20-year period, comparing MD&A and footnote disclosures, found that investor reaction to the textual characteristics of the MD&A is much stronger and more timely than the reaction to footnotes. The practical implication for credibility work is that a gap between the MD&A story and the footnote detail is a place to look first. Information that sits in a footnote and contradicts a confident MD&A passage constitutes a coherence failure that the market may have been slow to price.

Form 8-K "Other Events": the lightly codified item

Mandatory disclosure rules differ sharply in how tightly they are specified. A working paper titled Sixty-Four Words of Guidance examines variation inside Form 8-K and reports that the residual "Other Events" item is governed by only 64 words of guidance. Items in this category show 70.5% lower odds of receiving an SEC comment letter than more tightly codified items. Optimistic tone in these channels predicts stronger announcement returns, but it often precedes price reversals and weaker operating performance. For an analyst, the implication is direct: optimistic language placed in the least supervised channel deserves the most verification. A five-step 8-K filing analysis is one way to operationalize that review.

Proxy statements and the board narrative

Proxy statements are the third pillar, because they link narrative to accountability: they describe what the board claims to oversee, how pay is tied to results and how strategy is characterized to owners. Comparing that board narrative with outcomes is a distinct exercise, covered in the guide to assessing board narrative versus outcomes, and it feeds directly into voting decisions, as described in the governance playbook for proxy decisions.

Why tone alone is not enough

Dictionary-based sentiment scoring, which counts positive and negative words, is the most common automated shortcut. It is also the weakest. It treats a document as a bag of words and ignores whether the optimism is supported. More informative features exist.

• Readability and complexity. Using a large panel of U.S. public firms from 1995 to 2017, a study of annual report readability and the cost of equity capital associates a move from the first to the third quartile of report complexity with a 54 basis point increase in the cost of equity capital, ceteris paribus. The effect is significantly more pronounced when the tone of the disclosure is perceived as negative or ambiguous.
• Tone dispersion. The same research stream points to the consistency of sentiment across a document, not just its average level. Uniformly upbeat language throughout a document that also reports deteriorating fundamentals is a pattern worth flagging, as is tone that tracks peers rather than the company's own circumstances.
• Specificity. Claims with a metric, a date and a definition can be tested. Claims without them cannot, and a narrative that relies mainly on the second kind has low auditability by construction.

The relationship between language and performance is covered more fully in the walkthrough on sentiment versus performance analysis, and the valuation consequences of deliberate framing are examined in the piece on narrative engineering in corporate reporting.

A six-step workflow for a credibility file

The following sequence converts the concepts above into a repeatable process. It requires no insider access and no proprietary data.

1. Define the claim inventory. Extract every verifiable claim from the last four earnings-call transcripts, the latest 10-K, recent 8-Ks and the most recent proxy statement. Record the speaker, the date, the metric and the stated horizon.
2. Normalize definitions. Note where the same term (for example "adjusted margin" or "recurring revenue") is defined differently across documents. Definitional drift is a coherence finding in itself.
3. Set the evidence anchor. For each claim, identify the audited or otherwise verifiable figure that would confirm or refute it, and the filing in which it will appear.
4. Track delivery. After the horizon passes, record whether the claim was met, missed, restated or quietly dropped. A dropped claim without explanation counts differently from a missed claim with explanation.
5. Test the channels. Compare the MD&A account with the footnotes, and compare press-release language with the 8-K item used to file it. Flag optimism concentrated in lightly codified channels.
6. Score and log. Convert findings into scores with a written rationale for each, so a reviewer can reproduce the result from the same documents.

The sixth step is the one most often skipped, and it is the one that separates analysis from opinion. A score without a log cannot be challenged, which means it cannot be trusted.

Worked example: reading a single claim

Consider a hypothetical issuer, constructed here only to illustrate the procedure and not drawn from any real company. On an earnings call, management states that a cost program will lift operating margin by 200 basis points within eighteen months. The claim inventory records the speaker, the date and the horizon.

The coherence check compares the call with the 10-K. If the 10-K MD&A describes the program but gives no margin figure, and the proxy ties no incentive metric to it, the claim is uncorroborated across documents. If the footnotes disclose rising restructuring charges that the call did not mention, coherence weakens further.

The fidelity check waits for the horizon. Eighteen months later, the filings show operating margin up only 60 basis points, with the difference explained by "market conditions." The finding is a partial delivery with an attributed shortfall, and the analyst records whether the 10-K risk factors had flagged those conditions beforehand. If they had not, the explanation is retrospective, and the claim's credibility is reduced accordingly.

The example shows why the unit is the claim: the same company might deliver fully on a separate dividend commitment, and that finding should not be averaged away before a reader sees it.

Red flags that merit escalation

Any single flag is a prompt for inquiry, not a verdict. Clusters across channels are more informative than isolated instances, and the guide to identifying high-risk narrative companies describes how oversight teams can prioritize them.

Perspective: why a consistent method matters

We approach narrative credibility as a measurement problem, and the principal obstacle is comparability. A careful analyst can build an excellent credibility file for one issuer, but the file is hard to compare with the next one if the criteria shifted between them. For that reason the Lacuna method standardizes its procedure across all covered companies, relies only on public records, and publishes explicit governance logs for every score, so that each finding can be traced to a document and a rule.

We also hold that the unasked question is as informative as the stated claim. Engineered disclosure rarely contains a false sentence; more often it contains an incomplete frame, in which a favorable metric is emphasized and the unfavorable comparator is never raised. Detecting that requires looking at what a disclosure omits relative to what the filing rules and the company's own earlier statements would lead a reader to expect. The classification of companies into archetypes such as earned, borrowed or undervalued follows from that view: narrative strength and delivery are measured separately, and valuation is read against both. The full procedure is set out in the methodology walkthrough.

— Glen

Applying the method at the sector level

Company-level work gains meaning from context. A margin claim that is ambitious in one industry is routine in another, and disclosure norms differ by sector. Sector benchmarks supply the baseline against which an individual issuer's narrative and delivery are judged. Lacuna publishes sector benchmark pages for Health Care, Consumer Staples, Industrials and Financials, with industry-level views such as Financial Services and Insurance. These sector benchmarks are open to all readers.

For journalists, the same discipline produces defensible story angles: a documented gap between claim and delivery, with dates and filings attached, is a stronger foundation than a characterization of tone. The 2026 guide to accountability story angles and the guide to annual report credibility and governance analysis cover that application. Complementary reading on the management side is available in the article on how public disclosures reveal management gaps.

Where Lacuna Index fits

Teams that need the claim-level evidence trail without building it from scratch can use Lacuna Index, which scores the gap between narrative and delivery for covered public companies using only public records. Detailed per-company analysis, including execution and narrative scores, is available through the per-company analysis service, and the Cohort Pulse and Sector Sweep plans are described on the pricing page. Sector benchmarks remain free to read, which allows a prospective user to inspect the framework before subscribing.

Limits of the approach

Public-record analysis has boundaries that should be stated plainly. It cannot observe private communications, and it cannot establish intent: a gap between narrative and delivery is consistent with optimism, poor forecasting or deliberate framing, and the record alone rarely distinguishes among them. Findings from the academic studies cited here are averages across many firms and periods; they describe tendencies, not the behavior of any one issuer. Two of them are working papers or abstracts that readers should consult directly before relying on specific figures. Finally, a credibility finding is not an investment recommendation. It is one input, to be weighed with valuation, balance-sheet strength and the analyst's own mandate.

FAQ

What do you mean by narrative credibility analysis?

It is a structured test of whether a company's story is internally consistent and matches verifiable facts. The analyst extracts claims from public documents, compares them across sources and then compares them with later reported results. The output is evidence-backed findings, not an impression of tone.

What is the difference between narrative coherence and narrative fidelity?

Coherence is internal consistency: the parts of the story agree with each other. Fidelity is external consistency: the story accurately reflects real-world facts or shared values. A measure built on Fisher's theory scores the two dimensions separately.

How do you measure credibility in corporate disclosures?

Measure it per claim, not per company. Record each verifiable assertion, identify the evidence that would confirm it, and track whether it was delivered, missed or dropped. Supplement this with readability, tone dispersion and a comparison of MD&A with footnotes.

Is bad news more credible than good news?

Research reviewed in the disclosure credibility literature suggests it often is, because managers have fewer incentives to invent unfavorable information. That asymmetry is why favorable forward-looking claims warrant the greater share of verification effort.

Sources

• Construction and test of an empirical measure for narrative coherence and fidelity (Communication Reports)
• Assessing coherence and fidelity
• Impression Management Motivations, Strategies and Disclosure Credibility of Corporate Narratives
• SEC Commission Guidance Regarding MD&A (2003)
• SEC Speech: Let the Story Shine Through (2010)
• The Information Content of 10-K Narratives: Comparing MD&A and Footnotes Disclosures (SSRN)
• Sixty-Four Words of Guidance: Disclosure Design and Narrative Credibility Within Form 8-K (SSRN)
• Annual report readability and the cost of equity capital (ScienceDirect)

Recommended

• 6 Step Forensic Test for Analysts to Verify Turnaround Credibility
• Annual Report Credibility Governance Analysis Guide
• Corporate Responsibility Claims Verification: 2026 Guide
• How to Validate Corporate Claims in Public Disclosures