TB

Risk Disclosures

Last updated: March 2026

TrancheBook provides AI-generated analysis and intelligence for private markets. The following disclosures describe material risks associated with private market investments and the use of AI-generated analysis. You should read these disclosures carefully before using the Platform.

Important Notice

TrancheBook is an information and analysis platform. We do not provide investment advice, and nothing on this Platform constitutes a recommendation to buy, sell, or hold any security. Praevex Ltd does not owe any fiduciary duty to users of the Platform. Consult qualified professional advisers before making any investment decision.

1. Investment Risk

Investments in private markets carry significant risks that differ from public market investments. You should be aware of the following:

1.1 Illiquidity Risk

Private market securities are inherently illiquid. There is no public exchange on which these securities trade. Secondary market transactions may take months to complete, may require GP consent, and may not be available at all. You should be prepared to hold investments for their full duration and may not be able to sell when you wish to do so.

1.2 Loss of Capital

Private market investments may result in partial or total loss of invested capital. Past performance of any fund, company, or strategy does not guarantee future results. The value of private investments can decline substantially and without warning.

1.3 Valuation Uncertainty

Private market assets are not marked to market on a continuous basis. Valuations are typically reported quarterly by fund managers, may be based on subjective methodologies, and may not reflect the price at which a transaction would actually occur. Net Asset Values (NAVs) reported by GPs may be stale, optimistic, or based on assumptions that do not reflect current market conditions.

1.4 Limited Information

Private companies and funds are not subject to the same disclosure requirements as public companies. Information may be limited, delayed, or unavailable. You may not have access to the full picture when evaluating an investment opportunity.

1.5 Concentration Risk

Private market portfolios often involve concentrated positions. Individual fund or company failures can have an outsized impact on overall portfolio performance. Diversification does not guarantee returns or eliminate the risk of loss.

2. AI Analysis Risk

Cole, our AI analysis engine, generates scores, projections, IC packages, and recommendations. These outputs carry specific risks:

2.1 Model Error

Cole's analysis is generated by large language models and proprietary algorithms. These models can and do produce errors. Factual claims in AI-generated content may be incorrect. Financial projections are model outputs based on assumptions that may not hold. Scores and ratings reflect algorithmic assessment, not ground truth.

2.2 Hallucination Risk

Large language models can generate plausible-sounding but factually incorrect content. While we implement extensive validation and grounding techniques, Cole may occasionally produce statements that appear authoritative but are not supported by underlying data. Always verify critical facts independently.

2.3 Projection Limitations

Financial projections, including revenue forecasts, exit multiples, and IRR estimates, are model-generated scenarios based on historical patterns and assumptions. They are not predictions of future outcomes. Actual results may differ materially from projected values. Sensitivity analyses show how projections change under different assumptions but do not capture all possible outcomes.

2.4 Not Professional Due Diligence

Cole's analysis is not a substitute for professional due diligence conducted by qualified attorneys, accountants, and investment professionals. IC packages are designed to supplement, not replace, your own analysis process.

3. Data Quality Risk

The quality of AI-generated analysis depends on the quality of underlying data. TrancheBook uses a Data Quality Level (DQL) framework to communicate data reliability:

  • DQL 5 (Highest): Verified data from audited financial statements, regulatory filings, or direct GP reporting. Multiple independent sources confirm key figures.
  • DQL 4: Data from reliable institutional sources such as SEC EDGAR filings, investor letters, or established data providers. Some figures may be unaudited.
  • DQL 3: Data from credible but unverified sources including press releases, news reports, and industry databases. Figures may be estimates or self-reported.
  • DQL 2: Data inferred from indirect sources, comparable analysis, or model estimation. Significant uncertainty around key figures.
  • DQL 1 (Lowest): Sparse or unreliable data. Analysis is largely model-driven with limited factual grounding. Treat all outputs at this level with heightened scepticism.

Each analysis and IC package displays its DQL rating. Lower DQL ratings indicate higher uncertainty in the underlying data and, consequently, higher uncertainty in the analysis outputs.

4. Third-Party Data Risk

Cole aggregates data from multiple third-party sources, including:

Praevex does not guarantee the accuracy, completeness, or timeliness of third-party data. Errors in source data can propagate through analysis outputs. We make reasonable efforts to validate third-party data but cannot verify every data point.

5. No Fiduciary Duty

Praevex Ltd is not a registered investment adviser, broker-dealer, or fiduciary. We do not owe you any fiduciary duty, duty of care, or duty of loyalty in connection with your use of the Platform. The relationship between you and Praevex is that of a service provider and subscriber, governed by our Terms of Service. You are solely responsible for evaluating the merits and risks of any investment decision. We strongly recommend consulting qualified legal, financial, and tax professionals before acting on any information obtained through the Platform.

Questions

If you have questions about these risk disclosures, contact us at compliance@tranchebook.com.