BCI Structural Integrity Protocol

Standardizing structural diagnostics and data governance for capital markets — the operating specification, revised as calibration work produces new results.

Summary of Changes (v2.0 → v4.0)

• Retired the Decision Binding Interface; replaced with historical association language under an explicit, standing prohibition on point-estimate valuation instruction (§0)
• Retired the Confidence-Band / DRG Tier-1/2/3 mechanism; consolidated all data-reliability grading onto a single DRG-A/B/C/D standard (§2)
• Formalized the Comparable Company Selection Standard, now the shared basis for both the MT and ES calculations (§2.2)
• Revised the ES formula from a period-over-period ratio to a level-based ROIC-ratio formulation, resolving a known zero-crossing instability (§3.4)
• Resolved the κ / n notation conflict; κ is now confined to internal use and does not appear in external materials (§3.3)
• Renamed Category C from prediction-adjacent language to Structural Forward Scenario, and prohibited predictive framing across all three categories (§6)
• Added Section 9, Known Limitations and Calibration Status, as a standing disclosure reviewed at every protocol revision

This document serves as the operating specification for BCI Lab under BSIP v4.0. It sets out the standardized procedures for data filtration, epistemic grading, variable definition, output calibration, and delivery taxonomy, and states plainly where that calibration is not yet complete. The objective is for every BCI structural reading to function as a rigorously bounded, jurisdictionally insulated capital-markets tool: one whose limits are specified as clearly as its methods.

Protocol Scope Statement. Current protocol coverage spans luxury (hard assets, leather goods and apparel, beauty and fragrance), performance apparel, consumer technology, automotive, and select PE/M&A transaction contexts, with sector-specific calibration tables at varying stages of completion (Section 9). The protocol is designed to extend into a cross-sector structural-diagnostic standard for intangible-asset risk. Sector coverage expands only as each sector clears the comparable-company and evidentiary thresholds set out in Section 2 — coverage is a function of what has actually been validated, not of ambition.

BCI’s external rating uses three tiers — R1 / R3 / R5 (>8.5 / 6.5–8.0 / <6.0) — with two internal-only transitional monitoring bands. All boundary values are current working assumptions pending calibration; see Section 9.2.

Across historical samples, assets falling into different tier ranges have exhibited the following directional association patterns in subsequent observation windows. These descriptions are drawn from real cases in BCI Lab’s internal M&A / impairment backtest registry and are offered to institutional readers as one reference input among several for governance discussion and risk identification, not as a forecast.

• R1-range assets. In the historical sample, structural-stress signals appear infrequently; where a stress signal does appear, the interval before financial confirmation has typically been longer.
• Transitional-band assets. In the historical sample, these assets are mid-transition and directionally unresolved. Elevated observation frequency is recommended; a determinate structural judgment on an asset in this range is not recommended.
• R5-range assets. In the historical sample, an observable interval exists between a structural-stress signal and subsequent financial confirmation (impairment, profit warning, governance change, or similar). Across the cases on record, that interval has mostly fallen within a 6–27 month range, varying materially by asset type; negative confirmations have generally arrived faster than positive ones (see the backtest registry’s methodology notes).

No statement in this section constitutes, or may be used as, advice or instruction regarding the terminal growth rate, cost of equity, valuation discount factor, or any other parameter of any specific asset’s valuation model. Converting the structural-historical associations described in this section into a specific valuation-model parameter assumption is a matter for the report user’s own professional judgment; BCI does not provide that mapping and accepts no responsibility for a user’s own application of one. Any scenario comparison or governance-option description appearing in a BCI report template (for example, “if Path A is taken, the reading moves to…”) that gives a specific, point-estimate score trajectory must be revised in the next template update to a directional, non-point-estimate statement — this is now a standing template rule, not a case-by-case editorial judgment.

The lag between structural reality and financial reporting routinely distorts capital markets: a brand can be losing pricing power for a year or more before that loss appears in a P&L. BCI Lab treats marketing metrics, superficial social engagement, and short-term Gross Merchandise Value spikes as Structural Noise: signal that tracks the business cycle but not the structural variables this protocol measures.

Three filtration rules separate the asset’s underlying structural behavior from that noise:

1.1 Decoupling Revenue from Sovereignty

Top-line growth is frequently achieved by liquidating long-term scarcity — expanding distribution, deepening discount cadence, or both. BCI isolates revenue generated by structural compounding (the TSⁿ term) from revenue generated by extractive distribution (the ES⁻¹ term). A growing top line is not, on its own, evidence of a healthy structure.

1.2 Proxy Translation

Qualitative phenomena are translated into the specific financial proxies defined in Section 3, not into looser stand-ins. “Brand heat” is not a BCI input; gross margin relative to sector median (MT), SG&A relative to revenue (PL), secondary-market residual value (TS), and ROIC relative to peer median (ES) are. Where a qualitative observable is blended with a financial proxy — as with known cases where BCI’s default PL proxy understates genuine brand friction — the blend and its components are disclosed in the report (Section 3.2, Section 9.1).

1.3 Temporal Neutralization

Inputs are smoothed to remove seasonal anomalies before entering the calculation. A single reporting period is not sufficient to confirm a structural shift; see the review cadence set out in the Comparable Company Selection Standard, Section 2.2.

BCI Lab does not claim uniform certainty across all inputs. Every reading is built from data of varying reliability, and this protocol requires that reliability be graded, disclosed alongside the reading it supports, and never silently averaged away.

2.1 The DRG Scale

Circuit breakers.

• Data-Missing Circuit Breaker: any dimension lacking DRG-A or DRG-B data is marked N/A. It is never filled by estimation or averaging.
• Logic-Conflict Circuit Breaker: a reading in which, for example, MT rises while TS falls sharply, or any comparable internal contradiction, triggers mandatory manual review before publication.

Where DRG-A/B data is unavailable for a given dimension, a structured internal observation matrix (twenty audit points, five per variable, scored 0–10) may be used as a DRG-C-grade substitute input. This substitute score is never merged with, or presented as equivalent to, DRG-A data; the report must state clearly that the reading’s ceiling grade for that dimension is DRG-C.

Source-convergence practice. Where two independent data sources are available for the same input, convergence within a 5–10% variance band is treated as a positive reliability signal and may be noted in the report. This is a documented best practice, not a formal gate on any DRG grade.

2.2 Comparable Company Selection Standard

Any calculation requiring a sector median or peer median — currently the MT base ratio and the ES peer-median-ROIC term — draws its comparable set under this standard. Comparable sets are not assembled case-by-case by individual analysts.

Minimum sample size. Fewer than six genuinely comparable companies: no sector median or peer median is computed. Each company’s absolute reading is reported with its DRG grade instead; no relative benchmark is generated.

Inclusion criteria (all three required):

1. The relevant product category represents ≥70% of the company’s total revenue, or the company discloses an independent segment report that allows the category to be isolated.
2. Distribution-channel model (direct / wholesale) is held as consistent as reasonably achievable across the set. Where it cannot be, the constituent is flagged “channel model not fully comparable,” triggering the PL channel-structure correction protocol (Section 9.1).
3. Data for the evaluation period meets DRG-A or DRG-B grade.

Exclusion criteria. A candidate may be excluded on business-mix-contamination grounds (for example, a company whose reported gross margin is distorted by an unrelated commodity super-cycle). Every exclusion is logged with a one-line stated reason and retained alongside the comparable set; silent exclusion is not permitted.

Review cadence. Each comparable set carries a “last reviewed” date. Default cadence is quarterly, ahead of earnings season. A material change in a constituent’s business structure or channel model triggers an immediate review regardless of the standing cadence.

Relationship to the DRG circuit breaker. This standard is a precondition to Section 2.1’s data-missing circuit breaker: a comparable set that fails the minimum-sample or inclusion criteria is treated as equivalent to unavailable DRG-A/B data for that dimension, and handled under the same rule: marked N/A, never filled by averaging.

The minimum-sample-size figure (six) is a current working threshold, not a backtested optimum. Changing it requires the formal version-change process, not an ad hoc adjustment.

3.1 MT — Meaning Tension

Meaning Tension measures the degree to which a brand can command price for symbolic meaning, rather than for material cost.

Mathematical nature: scalar, range [0, ∞).

Primary proxy (DRG-A):

Secondary proxy (DRG-B): the ratio of search volume for a brand’s name alone versus the brand’s name paired with “discount” or “resale” terminology, with seasonal peaks (e.g., major shopping holidays) excluded.

3.2 PL — Perceptual Legibility

Perceptual Legibility measures how easily a brand’s value proposition can be read, understood, and replicated by the broader market.

Canonical direction: higher PL means lower cognitive friction, easier acquisition, and greater commoditization: an unfavorable direction for an asset whose pricing power depends on scarcity.

Mathematical nature: coefficient, range [0, 1].

Primary proxy (DRG-A):

Secondary proxy (DRG-B): global store/city distribution density, SKU count; semantic entropy (the rate at which a brand’s keyword coupling with “value” or “dupe” terminology is rising).

BCI’s default PL proxy does not, on its own, distinguish genuine brand-driven pricing discipline from a self-retail / direct-distribution business model that structurally depresses SG&A/Revenue independent of brand strength. This is an active, disclosed limitation, not a resolved one.

3.3 TS — Time Structure

Time Structure measures an asset’s resistance to value decay over time: whether a premium compounds or dissipates.

Mathematical nature: the exponent n in the master formula, range [0.5, 2.0]. This exponent has been public since the framework’s initial publication (approximately one year) and is unchanged in this revision.

Notation. TS’s own residual-value decay-curve fit uses a separate internal parameter, κ (kappa), distinct from the public exponent n. κ is used only in internal calculation and does not appear in external materials.

Primary proxy (DRG-B+):

Secondary proxy (DRG-A): inventory write-down reserve ÷ total inventory. A lower ratio indicates a stronger TS reading.

Decay model: TSt = TS0 × e−λt, where λ is an entropy coefficient; λ increases when an asset lacks structural maintenance, and κ is revised downward once a defined threshold is crossed.

Working n-Regime Classification

Current working assumption; in-band interpolation not yet defined — see Section 9.2.

3.4 ES — Energy State → Capital Efficiency Multiplier

Energy State measures whether a structural system is growing through internal nourishment or being sustained through external capital extraction.

Mathematical nature: dimensionless multiplier. A nourishing reading is >1.0; an extractive reading is <1.0. No fixed theoretical ceiling; the floor is strictly greater than zero — the composite formula is multiplicative, and an ES reading of zero or below is not a permitted output under any calculation path.

Primary proxy (DRG-A/B):

The first term (ROIC ratio) is the structural fix relative to the formula’s prior version. ROIC is a level, not a period-over-period ratio of changes, and crosses zero far less often than a delta-based ratio does. Peer median ROIC is computed under the Comparable Company Selection Standard, Section 2.2.

The second term has two canonical forms. The form used must be disclosed alongside any published reading, and the two forms are not mixed without a stated reason:

• Form A (default): organic-traffic share of total customer acquisition, used directly. Naturally bounded [0, 1], monotonic in the intended direction, with no division-by-near-zero risk.
• Form B (used where the relative organic-versus-paid emphasis needs to be preserved explicitly): log(1 + organic share ÷ paid share). Compresses the raw ratio so that near-zero paid spend — precisely the profile of the assets this framework most wants to score well — does not send the term toward infinity.

ES⁻¹ discipline (protocol-wide, no exceptions). The master formula’s denominator always uses ES⁻¹. “ES” is never written bare in a denominator position, in this document or in any BCI report. Working assumption ranges — nourishing 1.2–2.5, extractive 0.4–0.8 — are current working assumptions, not final calibrated values, and convert accordingly once inverted into ES⁻¹.

Mandatory error-source disclosure (may not be omitted from any published ES reading):

1. Macro credit-cycle contraction can suppress ROIC comparability independent of an asset’s own structural health.
2. Acquisition-structure data (organic vs. paid share) is subject to cross-platform and cross-source methodology differences; this component’s DRG grade should be adjusted downward accordingly.

Reassessment Trigger (mandatory reset rule). Either of the following immediately triggers a forced re-basing of an asset’s ES reading: (a) a major reputational or public-relations crisis with observable, measurable impact on core-audience demand; (b) a structural, not short-term, supply-chain disruption. Pending completion of the reassessment, the asset’s ES reading defaults to below 1.0 (an extractive-state assumption); the pre-event reading is not carried forward during this window.

Secondary proxy (DRG-A, retained): Operating Cash Flow ÷ Net Income — a measure of earnings quality / cash content, distinct from the primary proxy’s capital-efficiency concept. The two are not summed or substituted for one another.

Retired: ES⁻¹ = ΔMarketing Expense ÷ ΔNOPAT. This period-over-period formulation was found unstable across multiple 2026 validation cases and is replaced by the level-based ROIC-ratio formulation above.

3.5 The Master Formula

Notation discipline. Two rules apply protocol-wide, without exception:

• κ ≠ n. n is the sole public master-formula exponent. κ is TS’s internal decay-fit parameter and is never used in, or confused with, the public n.
• Friction or consumption is always written ES⁻¹, never bare ES. This single-symbol, single-meaning rule applies to every symbol in this protocol going forward; a symbol is not reused for a second quantity, even where the second quantity is a simple reciprocal of the first.

Cumulative reading (point-in-time trend tracking). For quarterly or annual trend reporting, discrete point-in-time BCI readings are accumulated by trapezoidal approximation across the observation window. This is not an independent formula; it is the point-in-time formula summed over time.

Category C forward-scenario formula:

The “−1” term treats MT×TSⁿ÷(PL×ES⁻¹) = 1 as structural equilibrium, so the integral captures the cumulative deviation from equilibrium over the observation window — conceptually analogous to cumulative abnormal return in event-study methodology. It is the canonical formula for Category C (Structural Forward Scenario) reporting.

These are built on top of the four core variables in Section 3. They do not replace the core variables and are not part of the canonical definitions in that section.

SIR — Symbolic Insulation Ratio

Measures whether MT is large enough to buffer against the cognitive erosion PL imposes. Where MT is insufficient relative to PL, an asset is understood to enter commodity drift. A specific numeric breach threshold for this transition is not yet empirically validated against the backtest registry; treat SIR as a directional supplement to MT and PL individually, not as a metric with its own confirmed break-point. Because SIR is calculated directly from PL, it inherits PL’s channel-structure limitation (Section 9.1) in full; a fix to the PL proxy resolves SIR automatically and does not require separate remediation.

BDV — BCI Dilution Velocity

Fully defined; no outstanding methodological issues.

BDI — Structural Duration Index

Conceptually defined as governed jointly by the stability of TSⁿ and the consumption rate of ES⁻¹. No executable formula for BDI currently exists. No numeric BDI value may be published for any asset until a formula is finalized and documented in a future version of this protocol.

t½ — Structural Half-Life

The rate at which an asset’s structural premium erodes. The concept is retained. This metric may not be presented, directly or by implication, as an input to any specific valuation-model parameter, including, without limitation, a terminal growth rate assumption in a discounted cash flow model. This restriction follows the same discipline set out in Section 0 and applies with equal force. Any report-template section describing itself as a “Terminal Value Sensitivity Analysis” using t½ is deprecated pending legal review (Section 7.3) and is not to be used in new reports under this protocol version.

Canonical output scale: 0.0–10.0.

All boundary values above are current working assumptions pending calibration; see Section 9.2 for the method that will be used to derive final values and the current state of that work.

Relationship to the Public Protocol Annex. R1/R3/R5 are the universal, cross-sector external labels. The Public Protocol Annex (v1.0, unchanged by this release) publishes a sector-specific numeric implementation of this same tier system for the luxury hard-asset sector: 9.0+ / 8.5–8.9 / 8.0–8.4 / 7.5–7.9 / <7.5. These bands sit higher than the cross-sector boundary because luxury hard assets, as a population, skew structurally high, consistent with the cross-sector standardization logic in Section 2.2. Other sectors require their own sector-specific tables before their scores can be read against a sector-calibrated line; current sector tables (technology, automotive, athletic apparel, PE/M&A) are at varying stages of completeness (Section 9.4).

Sector-specific calibration. Where a sector’s structural relationships genuinely differ from the default calibration — for example, a sector in which higher PL does not straightforwardly dilute MT — the adjustment is made to the sector-specific normalization curve that translates a raw ratio into a standardized reading, never by altering the shape of the master formula itself. This mechanism is governed by the Sector-Specific Calibration Governance Protocol (a separate document). As of this version, no sector has cleared that protocol’s adoption threshold; every sector currently uses the default calibration curve.

BCI Lab’s outputs are standardized to three report categories, calibrated to the depth of the underlying question. BCI Lab does not offer subjective brand consulting; it offers structural diagnostics calibrated to a defined category.

• Category A | Structural Snapshot. The asset’s current structural state: a static, point-in-time reading. Core output: BCI Score, per-variable DRG grades, current tier.
• Category B | Structural Attribution. What happened, structurally, to produce the asset’s current state: an explanation of the past, not a statement about the future. Core output: variable-level decomposition (MT/PL/TS/ES), historical trajectory matrix, relevant governance events.
• Category C | Structural Forward Scenario. A directional projection of an asset’s structural trajectory, built from its current and historical state. This category does not generate return forecasts, price targets, or point-estimate outcomes; it applies the Section 3.5 trend formula to produce a scenario-based, directional reading. This category is not described, internally or externally, as prediction. Any report language implying predictive certainty is inconsistent with this protocol and with the Independence & Conflict Policy.

Longitudinal consistency. Every asset under BCI coverage carries a persistent Asset ID. Subsequent Category A/B/C reports on that asset must reconcile with prior BCI readings unless explicitly invalidated by a Reassessment Trigger event (Section 3.4). A variance exceeding the documented working assumption ranges requires formal annotation and structural justification in the report. This treats BCI as a time-series diagnostic system, not a set of isolated analyses.

7.1 Core Declarations

Every BCI Lab output is subject to the following four declarations, which constitute the single canonical legal language for the BCI system and are not restated differently across documents:

1. Non-Valuation Declaration. BCI outputs do not constitute, and are not to be construed as, fair value opinions, market price indications, or asset pricing recommendations of any kind.
2. Non-Culpability Declaration. No structural reading constitutes an evaluation of any management team’s competence, motive, or compliance.
3. Governance-Use Limitation. BCI reports serve solely as inputs to governance discussion, risk foresight, and capital-efficiency review.
4. Non-Backtestable-Returns Declaration. The structural formulas in this protocol describe dynamic relationships among variables. They do not possess backtestable-return properties, are not used to generate investment-return forecasts, and do not constitute a source of alpha.

Clarification on Declaration 4. This declaration prohibits presenting BCI as a signal for generating excess returns: that use case belongs to investment research, carries different regulatory exposure, and remains one BCI Lab prohibits without exception. It does not prohibit validating the methodology’s own descriptive accuracy: confirming whether an R5 reading, for instance, actually preceded discount-depth acceleration, SKU-level markdown frequency, or secondary-market price decline in a given case is a test of descriptive accuracy, not a return forecast. The two activities are not in tension and may proceed in parallel.

Pre-publication word check. Before any report naming and scoring a specific real asset is published, the draft is checked for “valuation,” “buy/sell,” “should invest,” “expected return,” and comparable terms. Any occurrence triggers a rewrite, repeated until only structural-state description remains.

7.2 Business Model: Mode B

BCI Lab’s sole commercial model is subscription-based or fixed-fee institutional research access, purchased by parties on the buy side of a transaction or governance decision: private equity and strategic acquirers during diligence, institutional subscribers, credit and risk analysts. BCI Lab does not accept compensation from the entity being assessed, in any form, and does not adjust a calibration, a reading, or a publication timeline in exchange for compensation from a covered entity. This is the operating model, not a policy under periodic review, and is consistent with the Independence & Conflict Policy.

Two standing safeguards apply to every engagement:

• Quiet Period. Where a paid diligence engagement on a specific asset is active, public rating commentary on that same asset is suspended for the duration of the engagement.
• Standard Disclosure Line. Every report carries a fixed disclosure that BCI Lab may have a commercial relationship with a party related to the transaction under review, without naming that party.

Full independence provisions — draft-review boundaries, response windows, data-retention periods, and the anti-weaponization clause — are set out in the Independence & Conflict Policy, which governs and is not restated in full here.

7.3 Coverage of Non-Consenting Entities

BCI Lab publishes structural observations about entities that have not commissioned, and have not consented to, that coverage. This is standard practice across sell-side equity research, credit analysis, and financial journalism, and is not, on its own, disqualifying. It carries a risk profile distinct from the independence questions addressed in Section 7.2: the question is not whether BCI Lab is captured by the entity it covers, but whether unsolicited critical coverage of a specific, named company exposes BCI Lab to a claim of commercial disparagement. Established research houses carry this risk with the support of standing outside counsel, errors-and-omissions insurance, and a body of case law distinguishing fact-based structural observation from subjective disparagement. As of this protocol version, BCI Lab has not yet completed a dedicated legal review of this practice. That review is pending and is tracked as an open governance item; it is not resolved by this section’s language.

7.4 Jurisdictional Limitation Clause

The analytical frameworks, protocols, and intellectual property contained within BCI Lab’s outputs are governed strictly by the legal framework of the Hong Kong Special Administrative Region.

Every BCI output is structured for machine ingestion, quantitative-model interoperability, and AI-based retrieval. As of v4.0, the standard field set is:

• Asset ID — persistent across all reports on that asset
• Report Category — A / B / C, per Section 6
• BCI Score — point-in-time composite, where fully computed, or Partial Reading, where one or more variables remain DRG-N/A (Section 9)
• Output Tier — R1 / R3 / R5, with an internal-only R2/R4 flag where applicable
• MT, PL, TS, ES — raw values and, where a validated cross-asset normalization exists, normalized values
• n — applied value or working regime label (e.g., “Regime: (1.1, 1.5]; in-band value not yet interpolated, see Section 9.2”)
• Per-variable DRG grade (A/B/C/D) — disclosed individually, never averaged into a single blended figure
• Known Limitation Flags — e.g., “PL: channel-structure-affected, see Section 9.1”
• Protocol Version — this document’s version number
• Timestamp — data cut-off date, distinct from publication date

Retired fields. The Confidence Band (± X.XX) and the DRG Tier-1/2/3 fields, used under the document formerly circulated as BCI Audit Protocol v2.0 Documentation, are retired as of this version and must not appear in any report published under BSIP v4.0 or later. See Section 2 and Section 9.6 for the reason.

A methodology that only describes what it can already do is not a complete methodology. This section is reviewed and updated at every protocol revision.

9.1 Perceptual Legibility and Channel Structure

BCI’s default PL proxy, SG&A ÷ Revenue, does not distinguish between two structurally different causes of a low ratio: genuine brand-driven pricing discipline, where fewer channel intermediaries are required because the brand itself commands the terms of distribution, and a self-retail or direct-distribution business model, where fewer intermediaries are required because the company owns its retail footprint, independent of brand strength. Internal validation completed in July 2026 confirmed this produces a systematically inflated PL reading, and therefore an inflated composite score, for issuers with a predominantly self-retail distribution model.

Interim treatment. Any report on a self-retail-heavy issuer carries a standard channel-structure disclosure and treats the resulting PL and composite readings as directionally informative, not as precise measurements.

Path to resolution. Four demand-side cross-validation indicators — discount depth and frequency, order backlog / waitlist duration, price elasticity derived from real historical repricing events, and organic / direct-navigation traffic share — have been identified as candidates not distorted by channel structure. These are being layered in as DRG-B/C cross-checks, not replacements, ahead of a permanent fix to the primary proxy.

9.2 Composite Calibration: Current Status

The 0–10 output scale and the R1/R3/R5 boundary values (Section 5) are provisional working assumptions, not values derived from a completed distribution of real composite readings. Deriving that distribution requires cases with all four variables — MT, PL, TS, and ES — independently computed from real data. As of this protocol version, that has not yet been completed for any case in the internal backtest registry. Three cases (Porsche AG; Mandarin Oriental; Arc’teryx / Amer Sports) have complete MT and PL computations; none yet has a completed TS or ES computation.

The planned calibration method is set out here so the gap is a known destination, not an open-ended one:

• Raw-ratio-to-0–10 conversion: percentile-rank mapping across completed cases, chosen over a Z-score approach because the current and near-term sample size is too small to support a distributional (normality) assumption.
• R1/R3/R5 boundaries: to be set by outcome separation — the boundary value that best separates cases with a known structural-confirmation outcome from cases with a known healthy-or-improving outcome, drawn from the backtest registry — rather than by a fixed percentile cut.
• n in-band interpolation: the three-regime classification (Section 3.3) is in active use; the rule for selecting a specific value within a regime is not yet defined and will not be defined until TS computation across enough cases makes the interpolation meaningful.

9.3 Sector Weighting

The governance process for sector-specific calibration adjustment is fully documented (Sector-Specific Calibration Governance Protocol) and requires written justification, cross-validated evidence from real cases, and a documented before/after comparison before any sector-specific adjustment moves from analyst judgment to standing policy. As of this version, zero sectors have cleared that process; all sectors use the default calibration curve.

9.4 Comparable-Company Coverage Gaps

The automotive comparable set currently stands at three companies, below the six-company minimum set in Section 2.2. MT relative-ratio reporting for automotive issuers is suspended (absolute gross-margin readings with DRG grade may still be reported) until the set is expanded.

9.5 Energy State: Acquisition-Structure Term, Sector Applicability

The organic-versus-paid acquisition-structure adjustment term (Section 3.4) was validated using ROIC and acquisition-channel data typical of direct-to-consumer and digitally-native issuers. For traditional manufacturing issuers — automotive tested specifically — organic/paid traffic-split data is generally not disclosed, and internal marketing-spend line items are typically bundled with distribution and logistics costs in a way that would double-count against the PL proxy if used as a substitute. Where this data is unavailable, the acquisition-structure term is held at a disclosed placeholder value of 1.0 (no adjustment) rather than estimated; this is stated explicitly on any affected reading, not silently assumed.

9.6 Retirement of the Confidence-Band / DRG-Tier Mechanism

A separate three-tier data-reliability system, using a ± confidence band (±0.15 / ±0.25 / ±0.50) rather than the DRG-A/B/C/D weighting system in Section 2, was introduced under the document formerly circulated as BCI Audit Protocol v2.0 Documentation and carried forward into a subsequent methodology-disclosure update. That mechanism produced statistically styled outputs — most visibly, a published 96% confidence interval on one Category B report — without a corresponding sampling or estimation process to support them. This version retires the confidence-band mechanism in full. It is not replaced with a narrower version of the same idea: per Section 8, no report published under this protocol may carry a confidence interval or ± band until a genuine estimation process, backed by a real sample of fully computed cases, exists to support one.

This document does not assert its own precedence. The authoritative record of every publicly released BCI Lab methodology document — including this protocol, the White Paper, the Independence & Conflict Policy, and the Public Protocol Annex — together with each document’s version number, status, and relationship to every other document, is maintained in the Methodology Version Register, published and updated independently of this protocol text.

Citation Standard. When referencing this protocol, cite as: “BCI Structural Integrity Protocol (BSIP v4.0), BCI Lab, 2026.” When referencing the master formula or the canonical variable definitions specifically, cite the companion White Paper: “The Physics of Pricing Power (v4.0), BCI Lab, 2026,” which is the designated authoritative source for those definitions.