BCI Lab Institutional Guide

What BCI Lab provides to investors, boards and their advisers; the standards it holds itself to; and the terms on which it engages.

1. BCI Lab is an independent research institution that measures the structural durability of pricing power in brand-dependent assets. Its work addresses a question that valuation, credit ratings and commercial due diligence are not designed to answer: whether the pricing power behind an asset’s financial performance is intact, and how it has moved.
2. It serves investors, lenders, and the boards and advisers who act for them, and provides research access to academic, policy and media users.
3. Its outputs are structural readings. Every input is graded for reliability, every report states its protocol version and data cut-off date, and data that cannot be graded to the required standard is reported as not assessable rather than estimated.
4. Its outputs are not credit ratings, valuations, audit or assurance, or investment advice. They are not benchmarks, and they are not to be referenced in any financial instrument or contract.
5. BCI Lab is paid for research access and fixed-scope diagnostic services. It does not sell sponsored content, advertising, endorsements or paid coverage of any brand, and no fee is contingent on a reading.
6. Its evidence is a public registry of dated cases. The registry shows that an interval exists between structural change and financial confirmation. It does not show that BCI Lab would have signaled any case in advance, and this Guide makes no such claim.

The standards that bind BCI Lab are set out in Sections 1, 7, 8 and 9. The services are described in Sections 3 to 6.

1.1 What BCI Lab Is

BCI Lab is an independent research institution based in Hong Kong. It measures the structural durability of pricing power in brand-dependent assets: whether the premium an asset commands is compounding, holding or being spent down. Its framework, the BCI Structural Integrity Protocol (BSIP), defines four structural variables, a data reliability standard, and the terms on which readings are issued. The framework is published in full and is version-controlled.

BCI Lab’s readings are prepared for investors, lenders, and the boards and advisers who act for them. They are inputs to a decision-maker’s own judgment, and they are written to be cited, examined and challenged.

1.2 What BCI Lab Does Not Do

• BCI Lab does not issue credit ratings, ESG ratings, valuations, fairness opinions or price targets.
• It does not provide investment advice, and it does not recommend that any person buy, sell or hold any security or interest.
• It does not provide audit, review or other assurance services.
• It does not forecast prices, returns or financial results.
• It does not sell sponsored content, advertising, endorsements or paid coverage of any brand, and it declines requests for them.
• It does not accept fees that are contingent on a reading, on the completion of a transaction or financing, or on any publication outcome.

1.3 Principles of Engagement

• Every input to a reading is graded for reliability, and the grade is disclosed with the reading.
• Where data of the required grade is unavailable, the affected element is reported as not assessable. It is not estimated.
• Each report states the protocol version under which it was prepared and its data cut-off date.
• A reading describes structural condition at its cut-off date. Forward scenarios are directional, conditional on stated assumptions, and carry no point estimates.
• Conclusions follow from the data and the published method. They are not negotiated with a client, a covered entity or any other party.
• Limitations are disclosed, including those the framework has not yet resolved.

1.4 Terms Used

• BSIP: the BCI Structural Integrity Protocol.
• Policy: the Independence & Conflict of Interest Policy.
• Client: a party that commissions an engagement or holds a subscription or license.
• Covered entity: an entity on which BCI Lab publishes a diagnostic, or which is the subject of a report in an engagement.
• Reading: a structural reading of an asset issued under BSIP. A Partial Reading reports each variable that can be computed and identifies each that cannot.

Brand-dependent assets are underwritten, financed and monitored largely on financial indicators. Those indicators record outcomes. The condition that produces them, an asset’s capacity to command a price premium, changes first and is recognized later.

Three features of current practice account for the interval.

1. Accounting records goodwill and acquired intangible assets, including brand-related value, at the acquisition date and reviews them at intervals, and on triggers, set by the applicable standard. The lag between structural change and accounting recognition is a recurring subject of investor commentary and standard-setting.
2. Management’s account of an asset during diligence or monitoring is necessarily partial, and consolidated reporting can conceal divergent conditions among the brands of a group.
3. The instruments available to an investment committee or a lender address adjacent questions. Valuation addresses what an asset is worth at a date. Credit analysis addresses an obligor’s capacity to meet its obligations. Commercial due diligence addresses market position and outlook. None is designed to establish whether the pricing power on which an asset’s financial performance depends is structurally intact, or how it has moved.

The practical consequence is that the assumptions on which price, leverage and capital allocation rest are commonly tested principally against the lagging record. BCI Lab’s work is directed to that interval. It does not replace valuation, credit analysis or due diligence. It adds a structural reading that those processes do not otherwise contain.

BSIP defines four structural variables. Each is tied to observable financial or market evidence, and each is reported with a stated data reliability grade.

The variables are related by a published master formula (BSIP Section 3.5).

The formula describes a structural relationship among the variables. It is not a valuation formula, and it is not used to forecast.

3.1 Data Reliability Grading

Every input carries one of four grades. Grade A denotes audited filings, official statistics and exchange data. Grade B denotes structured third-party observational data. Grade C denotes unstructured semantic data, such as news and social media sentiment. Grade D denotes anecdotal evidence, which is excluded from calculation. Grades are reported for each variable and are never averaged into a single figure.

3.2 Comparable Sets and Entity Integrity

Relative measures are computed only against a comparable set that meets the Comparable Company Selection Standard in BSIP, including its minimum sample size. Where no such set exists, the reading is reported in absolute terms. BCI Lab does not attribute segment or group data to an individual brand: a reading is reported for the entity or segment to which the underlying data relate.

3.3 Disclosed Limitations

Known limitations are disclosed with each reading and are tracked in BSIP Section 9. Where a proxy is known to be distorted by an asset’s business model, for example where a direct-to-consumer distribution model affects the perceptual legibility proxy, the reading carries a limitation flag and is not presented as a precise measurement.

3.4 Coverage

Sector coverage is stated in the BSIP Protocol Scope Statement. Coverage extends to a sector only after it clears the comparable-company and evidentiary thresholds set out in BSIP.

BCI Lab’s outputs follow the three report categories defined in BSIP.

4.1 Supporting Materials

• Information request list. A list scoped to the asset and the engagement, identifying the evidence required to grade each variable. It has no fixed length. It is limited to what the reading requires, and it is settled at scoping.
• Management question set. A core set of 20 questions, five for each structural variable, extended as the asset requires. It is designed to be put to management by a deal team, a board or diligence advisers.
• Interpretive session. A structured session with the client’s decision-makers on the reading, its data reliability grades and its limitations, followed by a period of 14 calendar days for written queries.
• Monitoring. Periodic re-readings of a defined set of assets, identifying movement since the prior reading, any Reassessment Trigger event, and the status of an Active Forward Watch List of defined conditions.

Every output identifies its protocol version and data cut-off date and follows the machine-readable field set defined in BSIP Section 8.

4.2 Research Access

Research is available on a subscription basis. Research Briefs are monthly thematic studies of a single sector or question. The Pattern Library and Case Registry is a structured library of recurring structural patterns and the dated cases that support each. Both are available as individual subscriptions and as team or institutional licenses.

BCI Lab’s work enters an institution’s process at points where a decision depends on the durability of pricing power. The table is illustrative and not exhaustive. Engagements are scoped to the decision in hand.

Where the client owns the asset assessed, as in holding-period governance, exit preparation or an owner’s own impairment review, the engagement is a Self-Assessment Engagement and is subject to Section 8.3.

BCI Lab engages on four bases. Scope and fees are agreed in writing before work begins, and no fee is contingent on a reading.

6.1 Basis of Pricing

Fees for each engagement, and terms for each team or institutional license, are set by reference to the depth and breadth of the research required, the complexity of the asset and of the questions to be answered, and the scope agreed. Any fee or price that BCI Lab indicates, in this Guide or elsewhere, is a reference baseline and is not a fixed price. Published prices may change over time, and the price stated at the point of subscription, or in the engagement letter, applies.

Once a fee is agreed in writing, it is fixed for the agreed scope. No fee varies with the content of a reading. Fees are quoted in the currency agreed at scoping and are exclusive of applicable taxes.

7.1 The Registry

BCI Lab maintains a backtest registry of dated cases. Each entry sets out a chronology drawn from public filings, announcements, regulatory and court documents, and reporting by established media; the financial confirmation event used; and a data reliability grade. The index is published with the Pattern Library at byamanda.net/reports/bci-pattern-library. The case library forms part of the Pattern Library and Case Registry subscription.

7.2 Findings as at September 20, 2026

As at September 20, 2026, the registry contained 15 cases, of which 8 allow a signal-to-confirmation interval to be calculated. For the 6 deteriorating cases, financial confirmation followed the first publicly observable structural stress by 6 to 18 months. For the 2 strengthening cases, the interval was 20 to 27 months. Across all 8, it ranged from 6 to 27 months. Current figures are published in the registry index.

7.3 What the Evidence Shows and Does Not Show

• The intervals show that, in these cases, financial confirmation followed the first publicly observable structural stress. They do not show that BCI Lab would have identified that stress at the time or reached the same reading, and they do not show that the interval will recur.
• The signal date in each case is fixed retrospectively from public disclosure. The registry is not a record of forecasts, and BCI Lab makes no claim as to predictive accuracy.
• The samples are small, and the sample of strengthening cases is smaller still. The figures are descriptive and should not be treated as statistical estimates.
• The event that constitutes financial confirmation differs among cases, and includes profit warnings, guidance reductions, dividend cuts, impairments and changes in management. The event used for each case is identified in the registry.
• Cases were assembled after the events they describe, and their selection is constrained by the availability of reliable public data. The registry is not a random sample.

7.4 Calibration Status

The status of composite calibration, tier boundaries and sector-specific calibration is stated in BSIP Section 9. A reading is issued as a composite only where that status supports one. Otherwise it is issued as a Partial Reading, and no tier is assigned.

7.5 Correction of the Record

Corrections to BCI Lab’s own earlier statements, including statements of evidence, are recorded in the Methodology Version Register. Retired mechanisms and superseded documents are retained there and are not removed. Where a misclassification, data error or methodological defect is identified in a published reading, BCI Lab corrects the record, discloses the affected output, and annotates the version history.

8.1 Commercial Model

BCI Lab’s revenue derives from two sources: research access and fixed-scope diagnostic services purchased by parties on the buy side of a transaction, financing or governance decision; and, within the limits set out in Section 8.3, confidential fixed-fee Self-Assessment Engagements. BCI Lab has no other commercial model. The Policy governs the commercial model and prevails over this Guide where the two differ.

8.2 Compensation

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 diagnostics. It 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 and cannot be waived by written disclosure.

BCI Lab does not sell sponsored content, advertising, endorsements or paid coverage of any brand. Clients and subscribers pay for research access or fixed-scope diagnostic services, and no fee is contingent on a reading or on any publication outcome.

8.3 Self-Assessment Engagements

One category of engagement is commissioned by the assessed party, or by the owner of the assessed asset: the Self-Assessment Engagement, currently offered as Structure Readout. It is confidential and fixed-fee, and it produces no public reading. Because the party assessed is also the party paying, it is the arrangement most historically associated with conflicts of interest in assessment work. BCI Lab therefore defines it narrowly, in Section I-A of the Policy, and summarizes the safeguards here.

• The fee is fixed before work begins, does not vary with any reading, and buys no influence over a conclusion. The client may verify the facts it has supplied. It may not influence weighting, request alteration of an output, or negotiate a conclusion.
• The report and the client’s non-public data are confidential to the client and never enter BCI Lab’s public research, including in anonymized form.
• BCI Lab does not publish on the client while the engagement is active. If it later publishes on the client, the engagement is disclosed. If a finding on the client is required while an engagement is active, BCI Lab ends the engagement rather than delay the finding.
• BCI Lab does not solicit a Self-Assessment Engagement from an entity on which it has published a diagnostic. Where such an entity engages BCI Lab, the engagement is disclosed by addendum on each published diagnostic of that entity for as long as it is active.
• 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 active Self-Assessment client. The later request is declined.
• The report is not a certification, rating, endorsement or attestation, and may not be used in marketing or in the promotion of the client, its securities or its financing. The complete, unaltered report may be shared under confidentiality with the client’s board, shareholders, investors, lenders and professional advisers.
• BCI Lab publishes, at least annually, the number of Self-Assessment Engagements active in the period, without naming clients except where the Policy requires disclosure.

8.4 Buy-Side Engagements

Where a paid diligence engagement on a specific asset is active, BCI Lab suspends public commentary on that asset for the duration of the engagement. Every report carries a standard disclosure that BCI Lab may have a commercial relationship with a party related to the transaction under review, without naming that party.

8.5 Draft Review and Response

A covered entity may be invited to verify factual accuracy. It may not influence variable weighting, request alteration of a model output, negotiate a structural conclusion, or condition the timing of publication. For published reports, a covered entity may submit a written response within fourteen business days addressing a factual inaccuracy or a publicly verifiable data discrepancy. BCI Lab may publish the response as an addendum. A response does not alter the original reading.

8.6 Oversight

The Policy states which controls exist and which are still being formed. As at the date of this Guide, the independent Governance Committee described in the Policy has not been seated, and independence determinations rest with the Founding Partner. A written personal conflict-of-interest standard for individual analysts is in formation. Each will be recorded in the Methodology Version Register when it exists. A report of a suspected breach of independence, a data-integrity concern or an allegation of publication interference may be made through the channels in Section 12. Each such report is reviewed by the Founding Partner, and the outcome is retained under the audit-log standard.

9.1 Status of BCI Lab and Its Outputs

BCI Lab does not operate as a credit rating agency, investment adviser, broker-dealer, valuation firm or statutory auditor. Its outputs are not credit ratings within the meaning of Regulation (EC) No 1060/2009 or ratings of a nationally recognized statistical rating organization. They are not ESG ratings, valuations or fairness opinions, and they are not audit, review or other assurance.

9.2 Investment Recommendations

BCI Lab’s outputs do not recommend or suggest an investment strategy, explicitly or implicitly, in respect of any financial instrument or issuer, and they are not prepared for that purpose. They are not a substitute for a recipient’s own analysis or for professional advice. The structural formulas in BSIP describe relationships among variables. They do not have backtestable-return properties, they are not used to generate return forecasts, and they are not a source of excess return.

9.3 Benchmarks and Reference Use

Readings are not benchmarks. They are not intended to be used as a reference for the value of any financial instrument or for determining any amount payable under a covenant, facility or other contract, and BCI Lab does not consent to that use.

9.4 Information

BCI Lab’s published research is built exclusively on public financial disclosures and lawfully obtained third-party data. Where an engagement involves non-public information supplied by the client, the engagement letter allocates responsibility for the client’s authority to provide it, and BCI Lab handles it under the confidentiality terms of the engagement.

9.5 Reliance and Decision Responsibility

Decisions remain the responsibility of the client’s decision-makers. Reliance on a BCI Lab report does not transfer fiduciary responsibility to BCI Lab. A party that receives a report from a client receives it on the terms stated in the report, including its disclaimers.

9.6 Audience and Jurisdiction

The services described in this Guide are directed at institutional and professional users. Users in the European Union, the United Kingdom, the United States and elsewhere are responsible for ensuring that their use of BCI Lab’s outputs complies with the laws and regulations that apply to them. Limitation of liability is set out in the Legal Disclaimer & Risk Governance Framework. BCI Lab’s frameworks and reports are governed by the laws of the Hong Kong Special Administrative Region, and each engagement is also governed by the terms of its engagement letter. Nothing in this Guide is directed at any person in a jurisdiction where its distribution would be unlawful.

9.7 Amendment, Precedence and Local Law

BCI Lab may revise this Guide from time to time. This Guide does not form part of any contract unless an engagement letter or the terms of a subscription incorporate it. Where this Guide and an engagement letter differ, the engagement letter prevails as to that engagement, save that no engagement letter varies a commitment in the Policy. BCI Lab may vary, suspend or decline a service where necessary to comply with applicable law, sanctions or regulatory requirements. If a provision of this Guide is unenforceable in a jurisdiction, or inconsistent with mandatory law there, it applies to the extent permitted and the remaining provisions are unaffected. Nothing in this Guide excludes or limits any liability that cannot be excluded or limited under applicable law.

10.1 Process

1. Inquiry and conflict check. BCI Lab checks an inquiry against active engagements and published coverage before any confidential information is exchanged.
2. Scoping. Scope, evidence required, timetable, fee and permitted use are agreed by reference to the decision the client is taking.
3. Engagement letter. The letter records scope, fee, timetable, confidentiality, permitted use, data handling and conflicts provisions.
4. Information. BCI Lab issues an information request list scoped to the asset. Elements for which information of the required grade is not provided are reported as not assessable.
5. Fact verification. The client may verify the accuracy of the information it has supplied, and BCI Lab may clarify language for factual precision. Conclusions are not open to negotiation.
6. Delivery. Each report states its protocol version, data cut-off date, revision-log identifier, per-variable data reliability grades and limitation flags.
7. Interpretive session and queries. The client may submit written queries during the 14 calendar days following delivery.
8. Records. Inputs, outputs, parameter configurations and revision records are retained for at least three years.

10.2 Confidentiality

The identity of a client, the scope of an engagement and the content of a report are confidential, subject to legal obligation and to the disclosure provisions of the Policy summarized in Section 8. Retained materials are stored with controlled access, and modification of an archived record requires logged authorization.

10.3 Permitted Use

A report is provided for the client’s internal decision-making. A complete, unaltered report may be shared, under confidentiality, with the client’s investment committee, board, financing sources, co-investors and professional advisers. Excerpts may not be published or distributed. A report may not be described as a rating, certification, valuation or forecast, and may not be cited in offering documents, fund marketing or other promotional material without BCI Lab’s prior written consent.

10.4 Citation

Published research may be cited in accordance with the Citation Standard in BSIP Section 10. A reading supplied under an engagement may be cited in the client’s internal decision papers by report identifier, protocol version and data cut-off date, together with its data reliability grades.

11.1 What Distinguishes BCI Lab’s Work

The following features distinguish BCI Lab’s work from adjacent services. Each can be verified against the documents listed in Section 14.

• It addresses one question, the structural durability of pricing power, and offers no valuation, credit rating or investment advice.
• Its definitions, master formula, data reliability standard and known limitations are published and version-controlled.
• Its compensation prohibitions and conflicts provisions are stated in advance, and they bind BCI Lab’s own engagements.
• Its findings rest on a dated registry of sourced cases, and the limits of that evidence are stated beside it.
• Corrections to its own statements are made in public and retained.

11.2 The Institution

BCI Lab was founded in 2025 by Amanda Zhang, its Founding Partner. She trained in economics and accounting, worked within investment institutions, spent a decade in fashion media, and founded and operated the luxury fragrance house UMFD for four years. The framework reflects that background. It requires each structural conclusion to be expressible as a financial signal that can be audited, and it draws on first-hand observation of how distribution decisions and discounting can draw down a brand’s pricing power well before the effect appears in reported results.

BCI Lab applies to its own work the disciplines that institutional users expect of established research and assurance providers: written methodology under version control, disclosure of protocol version and data cut-off in every report, graded data, a conflicts policy with stated prohibitions, retention of records, and public disclosure of limitations and corrections.

BCI Lab is based in Hong Kong SAR. Its scope of work is set out in the Institutional Charter, which includes the boundary that its diagnostics are not designed for an entity with less than 24 months of operating history.

Inquiries may be made through the Contact page of byamanda.net, under which this Guide is published, or through any of the channels listed in the website footer. An inquiry should identify the organization, the decision to which it relates, the asset or assets concerned and the timetable. It should not contain material non-public information.

BCI Lab confirms whether it is able to act after a conflict check and before any confidential information is exchanged. It does not solicit engagements from entities on which it has published a diagnostic.

Is BCI Lab a credit rating agency?

No. BCI Lab does not issue credit ratings. Its outputs are not credit ratings within the meaning of Regulation (EC) No 1060/2009, and they are not ratings of a nationally recognized statistical rating organization. See Section 9.

Does BCI Lab make investment recommendations?

No. Its outputs do not recommend or suggest an investment strategy in respect of any financial instrument or issuer, and BCI Lab does not provide investment advice. A reading is one input to a recipient’s own analysis.

Can an entity that BCI Lab covers pay to influence its reading?

No. No payment can influence a reading, a conclusion or the timing of a publication, and performance-linked compensation is refused without exception. The one fee-bearing relationship BCI Lab can have with an entity it also evaluates is a confidential, fixed-fee Self-Assessment Engagement, which is separated from BCI Lab’s public research. See Section 8.3.

Can a reading be referenced in a loan agreement, a covenant or fund marketing?

No. Readings are not benchmarks and may not be used as a reference for the value of any instrument or for any amount payable. A report may not be cited in offering documents, fund marketing or other promotional material without BCI Lab’s prior written consent. See Sections 9.3 and 10.3.

What information does BCI Lab rely on?

Published research rests on public financial disclosures and lawfully obtained third-party data. In an engagement, the client may supply non-public information under the confidentiality terms of the engagement letter. Elements for which information of the required grade is unavailable are reported as not assessable and are not estimated.

Are the fees and prices indicated by BCI Lab fixed?

They are reference baselines and are not fixed prices. Fees are set for each engagement by reference to the depth and breadth of the research required, the complexity of the asset and of the questions to be answered, and the scope agreed. Once a fee is agreed in writing, it is fixed for that scope and does not vary with the content of a reading. See Section 6.1.

How reliable are the readings?

Reliability is reported. Each variable carries a data reliability grade, and a composite reading is issued only where BSIP supports one; otherwise the deliverable is a Partial Reading. The evidence base, and its limits, are described in Section 7.

Does BCI Lab hold positions in the companies it evaluates?

BCI Lab holds no equity or derivative exposure in any entity it evaluates. A written personal conflict-of-interest standard for individual analysts is in formation. Until it is formalized, such determinations are made by the Founding Partner and recorded in the audit log.

How long does BCI Lab retain the records behind a reading?

For at least three years from publication or delivery. Records include data-extraction timestamps, sampling summaries, model parameter configurations, per-variable data reliability grades and revision records.

Which law governs BCI Lab’s frameworks and reports?

The laws of the Hong Kong Special Administrative Region. Each engagement is also governed by the terms of its engagement letter.

How should BCI Lab be cited in an investment committee or board paper?

Published research may be cited in accordance with the Citation Standard in BSIP Section 10. A reading supplied under an engagement may be cited in the client’s internal decision papers by report identifier, protocol version and data cut-off date, together with its data reliability grades. See Section 10.4.

• BCI Structural Integrity Protocol (BSIP): definitions, master formula, data reliability standard, output tiers and known limitations.
• White Paper: the framework’s concepts and governing philosophy.
• Institutional Charter: mandate, scope boundary and standard of output.
• Independence & Conflict of Interest Policy: commercial model, conflicts, oversight and disclosure.
• Legal Disclaimer & Risk Governance Framework: standard disclaimer language accompanying a report.
• The BCI Protocol: the governance protocols beneath BSIP.
• Registry index: dated cases, published with the Pattern Library.
• Methodology Version Register: the current version and status of each document.

Links point to the version current at the date of this Guide. The Methodology Version Register records any later version. This Guide is reviewed whenever a governing document that it summarizes is revised.