Research Integrity Policy

Every MCH investment conclusion should be capable of being challenged, traced and reproduced.

The classification, governance attributes, reproducibility, labelling and revision rules below give effect to this principle.

MCH research distinguishes three kinds of statement, so that a reader can tell which kind of statement supports a conclusion.

FACT — An objectively verifiable claim supported by identifiable evidence or source material.

INFERENCE — An analytical conclusion derived from identified facts, models or assumptions.

SPECULATION — A plausible proposition with material uncertainty or insufficient direct evidence.

Where appropriate, material research should identify:

• publication timestamp
• information cut-off
• source provenance
• methodology or model version
• material assumptions
• key risks
• probability or scenario framework
• valuation framework
• falsification criteria
• kill switch, where used
• conflicts and holdings
• revision history

Where commercially and legally possible, MCH should retain enough internal provenance to reproduce material quantitative outputs.

This does not require public disclosure of proprietary code, licensed datasets or confidential intellectual property.

Performance outputs must carry one of the following explicit labels:

• Backtest
• Simulated
• Paper
• Live
• Audited actual

Different categories must not be merged into one apparent historical-performance series without clear disclosure. The classification and disclosure rules for performance outputs are set out in the Performance & Backtesting Policy.

Where a claim materially drives a conclusion, MCH prefers more than one independent source where practical.

Material historical conclusions must not be silently overwritten. Corrections and amendments to published research are handled under the Research Corrections & Amendments Policy.

Holdings and conflicts that bear on a publication are disclosed under the Conflicts of Interest & Securities Ownership Policy.

Questions about methodology, data or a specific publication may be sent to research@mchadvisoryservices.com.