Sentiment vs performance analysis, in our use of the term, compares a company's public narrative claims against its auditable delivery, measured from financial statements and operational disclosures already on the public record. The output is a quantified delivery gap, the distance between what management says and what the evidence shows, that institutional investors, governance professionals and journalists can use to prioritize engagement, reweight positions or sharpen reporting.
TL;DR:
- Begin with annual, quarterly, current, and proxy filings in EDGAR; earnings calls and presentations count as narrative unless matching figures enter filed disclosures.
- Give audited financial statements and independently verifiable operating measures more weight than filed releases, call transcripts, or promotional presentations when evidence conflicts.
- Flag KPI definition changes across periods, reconcile alternative financial figures to comparable GAAP measures, and cite each claim to a filing paragraph or transcript timestamp.
- Treat isolated optimistic framing as a watch list issue; recurring KPI drift across two or more periods warrants formal engagement or a proxy question.
- Compare gap trends with sector peers, then use persistent contradictions between claims and audited results to inform valuation, portfolio weight, or investigative coverage.
Table of Contents
- Defining the scope: what counts as narrative and what counts as delivery
- Why narrative-versus-delivery comparisons matter for fiduciaries and reporters
- Primary public-record evidence and the regulatory guardrails that govern it
- A repeatable checklist for comparing narrative and delivery
- Turning a measured gap into investor and governance action
- How to read a forensic narrative-versus-delivery report
- Why this discipline matters more than it appears to
- Scaling narrative-versus-delivery analysis with Lacuna Index
- FAQ
- Sources
Defining the scope: what counts as narrative and what counts as delivery
We draw a firm line between two categories of evidence. Narrative sentiment is the claim side: statements made by executives in forums where framing is a choice, not an obligation. Delivery is the measurement side: figures that have passed through an audit process or that can be independently verified against filed data.
- Narrative sources include management's discussion and analysis (MD&A), earnings call transcripts, press releases and proxy statement commentary.
- Delivery measures include GAAP financial statements, segment-level notes, audited figures and observable operational KPIs tied to those same notes.
- Social media sentiment, consumer confidence indices and trader psychology sit outside this scope entirely: they measure market mood, not corporate delivery, and conflating the two produces unusable comparisons.
This boundary matters because the forensic question we ask is narrower and more answerable than general sentiment: did the company do what it said, not how do people feel about it.
Why narrative-versus-delivery comparisons matter for fiduciaries and reporters
Engineered disclosures, selective emphasis on favorable metrics, quiet redefinition of a KPI, optimistic framing of a weak quarter, can mask underperformance long enough to distort valuation before audited results force a correction. For institutional investors, this gap is a mispricing signal. For governance professionals, it is an engagement trigger and a voting consideration ahead of compensation votes. For financial journalists, it is a sourcing method that replaces anecdote with filing citations.
Investors weigh narrative credibility against a company's historical earnings quality and the incentives present at the time of disclosure, according to academic research on narrative disclosure credibility. A firm with a track record of clean, well-reconciled earnings earns more benefit of the doubt in its narrative; one with a history of restatements or aggressive non-GAAP framing earns less. That asymmetry is itself decision-useful, and it is why a repeatable narrative risk analysis approach outperforms one-off reads of a single earnings call.
Primary public-record evidence and the regulatory guardrails that govern it
The evidentiary base for this work is narrow by design: filings and filed statements, not commentary about them. Treating oral claims and promotional materials as equivalent to filed disclosure is the most common error we see in informal analysis.
- Pull the 10-K, 10-Q, 8-K and proxy statement from EDGAR search and filing tools, which offer full-text search, Inline XBRL data and APIs for retrieving the exact sections cited in a claim.
- Treat earnings call statements and investor presentations as narrative evidence unless and until a matching figure appears in a filed 8-K or press release; the filed version controls.
- Check any non-GAAP metric against SEC staff guidance on non-GAAP financial measures, which requires a quantitative reconciliation to the most directly comparable GAAP measure.
- Read KPI disclosures against the Commission guidance on MD&A, which expects companies to define key variables and disclose when a calculation methodology has changed.
A few structural points carry weight beyond any single filing:
- Regulation G and Item 10(e) exist specifically to stop a non-GAAP figure from substituting for, rather than reconciling to, an audited one.
- PCAOB AS 1105 on audit evidence establishes that evidence from independent, directly inspected sources carries more weight than management assertion, and that professional skepticism is required when evaluating what management claims.
A repeatable checklist for comparing narrative and delivery
The following sequence is designed to produce findings that another analyst, or a regulator, could retrace line by line.
- Extract the claim. Record the exact wording of each explicit narrative statement, along with its venue (earnings call, MD&A, press release) and timestamp.
- Map to a comparable measure. Identify the GAAP or operational line item that most directly corresponds to the claim, and locate any required reconciliation table.
- Triangulate across sources. Compare the MD&A narrative, the call transcript and the press release for the same period; inconsistencies in timing or framing are themselves findings, a method detailed in our piece on spotting engineered disclosure patterns.
- Detect definitional drift. Flag any change in how a KPI is calculated period over period, since recurring methodological shifts are a stronger signal of narrative engineering than a single divergent quarter.
- Score the gap. Assign a simple RAG indicator (red, amber, green) based on the size and persistence of the divergence, then document the specific filing and paragraph that support each score. An execution score aggregates these individual gap scores across the claims reviewed for a given period, weighted by how central each claim was to the investment thesis management presented.
Practitioners who automate parts of this sequence, particularly the cross-period consistency check, can apply the tactics in our guide to monitoring disclosure consistency to flag drift as soon as a new filing lands. Reviewing an earnings call transcript line by line is also where many narrative claims first surface in full, a process walked through in this practical guide to earnings call analysis.
Pro Tip: Weight evidence by its position in the audit hierarchy: filed, audited notes outrank filed press releases, which outrank call transcripts, which outrank promotional slide decks.
Turning a measured gap into investor and governance action
A gap score only earns its keep once it changes a decision. Small, isolated gaps (a single quarter's KPI framing that proves optimistic) typically warrant a watch-list entry and a question at the next call. Medium gaps, recurring definitional drift across two or more periods, justify a formal engagement letter or a proxy question. Large gaps, where audited results contradict a sustained narrative theme, support a valuation discount, an underweight decision or an investigative story, the kind of translation from finding to publishable angle covered in our guide to accountability story angles.
- Convert the gap size into a percentile against sector peers rather than an absolute number, since baseline disclosure quality varies by industry.
- Use the gap trend, not a single period's reading, to decide whether an issue is improving, stable or worsening.
- Classify the company by the relationship between its gap score and its valuation: companies that deliver in line with a modest narrative sit in one archetype, companies whose premium valuation rests on an unproven narrative sit in another, and companies delivering more than their narrative suggests sit in a third.
This archetype step is what lets a research desk triage coverage instead of reading every filing with equal intensity.
How to read a forensic narrative-versus-delivery report
A report built on this methodology should let a reader retrace every score to its source. Evidence citations sit next to each claim and point directly to a filing, a paragraph or a transcript timestamp, not to a summary of one. Execution scores are built exclusively from public-records inputs, filed financials, reconciliation tables, transcript text, so two analysts working from the same filings should reach comparable conclusions. Archetype assignments (how a company's execution score relates to its valuation) sit at the top of the report as a quick orientation, with the underlying claim-by-claim evidence available beneath it for anyone who wants to verify the classification rather than accept it. A methodology walkthrough and sector benchmarks typically accompany the per-company findings so a reader can check how a given score compares to its peer group before acting on it.
Why this discipline matters more than it appears to
Public-records-first forensic checking does something that broader sentiment reads cannot: it forces every claim back to a citation. That constraint changes behavior on both sides. Analysts stop treating a confident tone on a call as information, and management teams, aware that their definitions and reconciliations will be checked against the next filing, have less room to redefine a KPI quietly between quarters.
The real value is not catching a single bad disclosure. It is the routine: checked claims, cited evidence, repeated every quarter, which closes the information gap between issuer and reader faster than waiting for a restatement to do it. That is a discipline worth building into any research process, not a one-time audit.
— Glen
Scaling narrative-versus-delivery analysis with Lacuna Index
Running this checklist manually across a coverage list of any size is slow, which is the gap our platform is built to close. We produce audit-traceable forensic reports that score the distance between what a company claims and what its public filings show, using only public records, with every score linked back to supporting evidence.
- Per-company analysis delivers a full execution score and archetype classification for a single issuer, built from the same claim-to-filing mapping described above, detailed on our benchmarks page.
- Cohort Pulse applies that scoring across a defined group of companies for comparative monitoring.
- Sector Sweep extends it across an entire sector for relative benchmarking.
Pricing and current plan details for Cohort Pulse and Sector Sweep sit on our pricing page, and our full methodology walkthrough explains the scoring logic behind every report before you commit to a subscription.
FAQ
What is the difference between narrative sentiment and performance delivery?
Narrative sentiment refers to claims executives make in forums like earnings calls, MD&A sections and press releases, where framing is discretionary. Performance delivery refers to audited financial results and operational KPIs that can be verified against filed statements, independent of how confidently they were described.
Which public filings should I check first in a forensic comparison?
Start with the 10-K and 10-Q for audited figures, then the 8-K and proxy statement for interim disclosures and governance detail, all searchable through EDGAR. Cross-check any non-GAAP figure against its required reconciliation before treating it as comparable to a GAAP line.
How is an execution score calculated?
An execution score aggregates the sizes and persistence of gaps found between specific narrative claims and their corresponding audited or filed measures over a period, weighted by how central each claim was to the company's stated thesis. We calculate ours using only public-records inputs, with each component traceable to a specific filing citation.
Why exclude social media sentiment from this type of analysis?
Social media and consumer sentiment measure market mood and trader psychology, which are different signals requiring different methods than a forensic comparison of management claims to audited results. Mixing the two produces findings that cannot be traced back to a verifiable filing, which breaks the audit trail this methodology depends on.
What does a large narrative-delivery gap typically indicate?
A large, persistent gap between what a company claims and what its audited results show often signals that a valuation premium rests on an unproven narrative rather than demonstrated performance. Academic work on narrative disclosure credibility finds that investors already discount such claims when a company's historical earnings quality is weak, which this kind of gap scoring makes explicit and measurable.
Sources
- Commission guidance on management's discussion and analysis (MD&A)
- PCAOB AS 1105 — Audit evidence
- EDGAR search and filing tools
- Academic research on narrative disclosure credibility