Independence & Conflict of Interest Policy

How BCI Lab enforces independence from the entities it evaluates, stated plainly, including what is still being built.

Summary of Changes (v1.2 → v1.3)

• Replaced the statement that BCI Lab's sole commercial model is buy-side research access with a two-part model, adding the Self-Assessment Engagement (§I)
• Stated expressly that the prohibition of equity, revenue-sharing, success-based, and other performance-linked compensation is unconditional and applies to every engagement (§I)
• Added Section I-A: definition, fee terms, firewall, quiet period and disclosure, no solicitation, conflicts between engagements, use of the report, records and oversight, and annual aggregate disclosure (§I-A, new)
• Extended the data-retention standard to Self-Assessment records (§V)
• Recorded acceptance of a Self-Assessment Engagement as an independence-safeguard determination presently resting with the Founding Partner, and added the first aggregate disclosure to the events recorded in the Version Register (§VIII)
• Extended the analytical-weaponization test to engagements sought to obtain a favorable reading for use in marketing or financing (§IX)
• Conforming edits to title, version metadata, structured-data FAQ, and footer references

This document sets out BCI Lab's independence commitments and the safeguards that enforce them. Where this document and any other BCI Lab publication describe the same commitment differently, this document governs.

BCI Lab operates as an independent research institution. It is not a registered investment adviser, a credit rating agency, a broker-dealer, a marketing consultancy, or a public relations entity.

BCI Lab does not accept equity participation, revenue-sharing arrangements, success-based compensation, contingent fees, or any other performance-linked compensation from an entity subject to its structural diagnostics. BCI Lab holds no equity or derivative exposure in any entity it evaluates. No compensation is linked to a valuation outcome, a capital raise, stock performance, or the reputational consequence of a published finding. This commitment is unconditional. It applies without exception to every engagement, including a Self-Assessment Engagement, and it cannot be waived by written disclosure.

BCI Lab's commercial model has two parts. The first 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, and credit and risk analysts. The second is the Self-Assessment Engagement, a fixed-fee, confidential engagement commissioned by an entity to assess its own structural condition, which is defined and constrained in Section I-A. BCI Lab has no other commercial model. Neither part is under periodic reconsideration; changing either would require a public amendment to this document, recorded in the Methodology Version Register. Version 1.3 is that amendment for the second part.

The only fee-bearing relationship BCI Lab has with an entity that it also evaluates is a Self-Assessment Engagement, governed by Section I-A. Any other commercial engagement touching a covered entity, including a buy-side engagement in which the covered entity is the target, is disclosed internally and reviewed under the arrangements set out in Section VIII.

An entity that pays to see its own structural condition is a different party from an entity that pays to influence how it is seen. The arrangement in which the assessed party pays is the one most historically associated with conflicts of interest in rating and assessment work. This Section therefore defines it narrowly, states what it cannot do, and records where its safeguards are not yet independently reviewed.

I-A.1 Definition. A Self-Assessment Engagement is a fixed-fee engagement in which an entity commissions a confidential structural diagnostic of its own brand or asset, as a governance and risk-monitoring input for its board and management and, in complete and unaltered form, for its investors, lenders, and advisers. BCI Lab currently offers it as Structure Readout, including a one-time baseline assessment. It is not a buy-side engagement, a credit rating, a certification, or an advisory mandate. It produces no public reading, and Section VI, which governs responses to published reports, does not apply to it.

I-A.2 What the fee buys. The client pays a fixed fee for method and time. The fee is agreed before work begins, does not vary with any reading, and is not linked to any outcome, including a capital raise, a transaction, a valuation, or the client's later use of the report. It is not refundable or reducible on the basis of what a reading says. The client cannot purchase a conclusion: Section III applies in full, so the client may verify the facts it has supplied and may not influence variable weighting, request alteration of a model output, or negotiate a structural conclusion. Such a request is declined and logged under Section V. BCI Lab applies the same methodology and data-reliability standard it applies to every other entity, issues one reading per engagement, and does not re-run an engagement to obtain a different result. It does not provide implementation advice, and it accepts no advisory mandate, sponsorship arrangement, equity participation, or contingent compensation from the same entity.

I-A.3 Firewall. The report and the client's non-public data are confidential to the client and are used only for that engagement. They do not enter, inform, or influence any BCI Lab public reading, diagnostic, case-registry entry, or other research output about the client, its competitors, or any other entity, and they are not carried into any other engagement. Anonymized reuse is not permitted. Purchasing, declining, or ending a Self-Assessment Engagement has no bearing on whether, when, or how BCI Lab publishes about the client, and no published finding is adjusted, delayed, or withheld because an entity is or has been a client.

I-A.4 Quiet period and disclosure. While a Self-Assessment Engagement is active, BCI Lab does not publish commentary, diagnostics, or updated readings on the client. If BCI Lab later publishes on the client, or on an asset the client owns, the existence and dates of the engagement are disclosed in that publication, including where the engagement was ended early. The quiet period is never a reason to withhold or delay a published finding: if BCI Lab concludes that a public finding on the client is required while an engagement is active, it ends the engagement first, consistent with Section IV.

I-A.5 No solicitation, no linkage. BCI Lab does not solicit a Self-Assessment Engagement from an entity on which it has published a diagnostic. Where such an entity approaches BCI Lab, BCI Lab accepts the engagement only if the client agrees that its existence will be disclosed by addendum on each published diagnostic of that entity for as long as the engagement is active, and in any later publication. No public reading is offered, withheld, or timed as an inducement to buy or renew a Self-Assessment Engagement, and BCI Lab does not represent that a published reading can be changed by purchasing one.

I-A.6 Conflicts between engagements. BCI Lab does not accept a Self-Assessment Engagement from an entity that is the subject of an active buy-side engagement, and does not accept a buy-side engagement on an entity that is an active Self-Assessment client. Where both requests arise, BCI Lab declines the later-in-time engagement, disclosing only that a conflict prevents acceptance.

I-A.7 Use of the report. A Self-Assessment report is not a certification, rating, endorsement, or third-party attestation, and may not be described as one or used in advertising, marketing, or promotion of the client, its securities, or its financing. The complete report, unaltered and with its disclaimers and data-reliability grades, may be shared under confidentiality with the client's board, shareholders, existing and prospective investors, lenders, and professional advisers. Excerpts may not be published or distributed. BCI Lab may act under Section IX against a misuse. Every Self-Assessment report carries the standard disclaimer language of the Legal Disclaimer & Risk Governance Framework.

I-A.8 Records and oversight. Every acceptance and refusal of a Self-Assessment Engagement, and every request declined under I-A.2, is recorded under the audit-log standard set out in Section V. The engagement record, including the delivered report, is retained on the terms stated there, measured from delivery, and retention confers no right to use client data for any other purpose. The decision to accept an engagement is an independence-safeguard determination that currently rests with the Founding Partner, which means no independent party yet reviews it. This document states that gap rather than describe a control that does not exist. Once the Governance Committee described in Section VIII is seated, it will review the acceptance criteria and a sample of engagement records, and its doing so will be recorded in the Methodology Version Register at the point it becomes real.

I-A.9 Aggregate disclosure. At least once a year, BCI Lab publishes the number of Self-Assessment Engagements active in the period and the share of total revenue they represented, without naming clients except where I-A.4 or I-A.5 requires disclosure. The first such statement will be published within twelve months of the first Self-Assessment Engagement accepted under this Policy, and will be referenced from the Methodology Version Register.

BCI Lab is building toward a personal conflict-of-interest standard governing individual analysts: a prohibition on holding a financial position in a covered entity while an engagement on it is active, and a recusal requirement where an analyst has a pre-existing personal or professional relationship with a covered entity.

That standard has not yet been formalized into a standing policy. This document states that plainly rather than assert a control that does not yet exist in written form. Until it is formalized, a determination of this kind is made directly by BCI Lab's Founding Partner on a case-by-case basis, and each instance is retained under the audit-log standard set out in Section V.

A covered entity may be invited to verify factual accuracy: financial data, public disclosures, and operational facts. A covered entity may not use that review to influence variable weighting, request alteration of a model output, negotiate a structural conclusion, or condition the timing of publication on reputational concern.

A draft report is not subject to narrative negotiation. Language may be clarified for factual precision. It is not softened for reputational accommodation.

Final authority over the timing, structure, and content of a published reading currently rests with BCI Lab's Founding Partner, pending the formation of the Governance Committee described in Section VIII.

BCI Lab does not suppress a structural finding because of commercial pressure, reputational sensitivity, or a client relationship. Model parameters are applied consistently across entities within the same category classification, and any sector-specific calibration adjustment is disclosed under the Sector-Specific Calibration Governance Protocol before it is used in a published reading.

Every report discloses its data cut-off date, the protocol version under which it was produced, and a revision-log identifier. A post-publication revision requires a documented rationale and a corresponding version update, recorded in the Methodology Version Register.

All analytical inputs, model outputs, calibration logs, and published diagnostics are retained for a minimum of three years from the date of publication, to support independent verification and internal audit. Retained materials include data-extraction timestamps, sampling summaries, model parameter configurations, per-variable data-reliability grades, and revision records.

Retained materials are stored with controlled access. Modification of an archived record requires logged authorization, and deletion before the retention period expires is prohibited absent a legal obligation to do so.

Records of a Self-Assessment Engagement, including the delivered report, are retained on the same terms, measured from the date of delivery, and remain confidential to the client as set out in Section I-A.

A covered entity may submit a written response within fourteen business days of a report's publication. A response may address a factual inaccuracy or a publicly verifiable data discrepancy. A response may not challenge the model's structure, a variable's definition, or the calibration logic behind a weighting.

Where BCI Lab judges a response warranted, it is published as an addendum noting that a formal response exists. BCI Lab retains sole discretion over the addendum's format and placement, and a response does not retroactively alter the original reading.

BCI Lab's outputs are structural diagnostic observations, issued under the Category A, B, or C taxonomy set out in the BCI Structural Integrity Protocol. They do not constitute a credit rating, a fairness opinion, an investment recommendation, legal advice, or an accounting determination, and they are not a source of investment-return forecasts. BCI Lab assumes no fiduciary responsibility for a decision made in response to its publications, and no structural reading constitutes an evaluation of any management team's competence, motive, or compliance.

Methodological changes and independence-safeguard determinations currently rest with BCI Lab's Founding Partner. BCI Lab is building toward an independent, multi-member Governance Committee with standing authority to review conflict-of-interest determinations, approve material calibration changes, and investigate a suspected breach of independence, without requiring the Founding Partner's sign-off. The acceptance or refusal of a Self-Assessment Engagement (Section I-A) is an independence-safeguard determination of this kind, and is among those the Committee is intended to review.

That committee has not yet been formed. This document states that plainly rather than describe a body that does not yet exist. Until it is seated, a suspected breach of independence, a data-integrity concern, or an allegation of publication interference is reviewed directly by the Founding Partner, and the outcome of that review is retained under the same audit-log standard set out in Section V.

Separately from that governance function, BCI Lab may consult a practitioner with direct, ongoing experience inside a covered sector, or an academic researcher with relevant subject-matter expertise, on an ad hoc basis, to check the factual and contextual accuracy of a specific analysis before it is published. This consultation is advisory only. It informs BCI Lab's own independent judgment; it carries no authority over a reading's conclusions, tier, or methodology, and it is not a substitute for the Governance Committee described above. Where a consulted practitioner has a current or recent affiliation with the entity under analysis, that relationship is disclosed internally and handled with the same confidentiality standard applied to any other source.

This Policy is reviewed whenever a material governance gap is identified, and in full at each material revision to the BCI Structural Integrity Protocol. The formation, membership, and first review cycle of the Governance Committee, the eventual formalization of the personal conflict-of-interest standard described in Section II, and the first aggregate disclosure under Section I-A are each recorded in the public Methodology Version Register at the point they become real, not announced in advance of being true.

BCI Lab does not accept an engagement whose evident purpose is to produce a negative reading of a competitor, a litigation target, or an activist campaign subject, rather than an independent structural assessment. Where an engagement's framing suggests the client has predetermined the conclusion it is seeking, BCI Lab declines the engagement rather than proceed and disclose the conflict afterward. The same test applies in the other direction: BCI Lab declines an engagement whose evident purpose is to obtain a favorable reading, or a document for use in marketing or financing, rather than an independent structural assessment.

BCI Lab prohibits the use of its published research for market manipulation, a coordinated reputational attack, or selective quotation built to support a misleading narrative, and reserves the right to issue a clarifying statement where its findings are materially misrepresented.

A structural reading reflects the asset's structural state as of its data cut-off date. A subsequent macroeconomic shock, regulatory change, or comparable external event can alter that state without triggering an immediate revision.

The absence of a near-term financial consequence does not invalidate a structural observation. Structural deterioration and its financial confirmation are frequently separated by a real interval, documented case by case in BCI Lab's backtest registry; a reading is not wrong merely because the market has not yet caught up to it.

The core variable definitions, the master formula, and the dimensional framework underlying BCI Lab's diagnostics are published openly, in the BCI Structural Integrity Protocol, the White Paper, and the BCI Lexicon, and are made available for reference and citation. What remains proprietary is the specific weighting architecture, internal scoring coefficients, and calibration data behind an individual reading. Unauthorized commercial replication of BCI Lab's scoring engine, or reverse engineering undertaken to reproduce it for a competing commercial service, is prohibited.

This Policy, and BCI Lab's frameworks and reports generally, are governed exclusively by the laws of the Hong Kong Special Administrative Region.

This Policy is amended only through the version-controlled process that governs the BCI Structural Integrity Protocol. A material change to a commitment in this document is documented, dated, and preserved in the Methodology Version Register alongside the version it supersedes. No prior commitment is quietly dropped.