What counts as investment advice

Sonar treats published output as educational research, not trade instruction. Its research-to-publishing rules prohibit specific trade recommendations and permit method descriptions, tests, and analytical frameworks instead. The overfitting audit and Deflated Sharpe Ratio materials show a process centered on research validation rather than actionable trade calls. SEC/FCA regulatory definitions are not referenced in the current material, so that broader legal claim cannot be established here.

The line between educational research and investment advice matters because the format and specificity of a publication can change its regulatory character. Sonar’s publishing framework is designed to stay on the educational side of that line by focusing on methods, validation, and research hygiene rather than specific trade calls.

Sonar’s own research-to-publishing rules state that published research must present a method, test, or analytical framework, and must not instruct readers to enter a named position or execute a specific trade. In practice, that means a piece can explain how a factor is constructed, how a signal is evaluated, or how robustness is assessed, but it should stop short of telling a reader to buy, sell, short, or allocate to a particular instrument at a particular time. The stated purpose is to keep output in the category of research education rather than individualized or actionable recommendation content.[1]

That distinction is reinforced by the rest of Sonar’s research stack. The backtest overfitting audit emphasizes whether a result is statistically credible, whether multiple testing has inflated apparent significance, and whether a strategy’s observed edge survives audit checks. Its role is methodological: it evaluates the integrity of a research claim, not whether a reader should take a live position.[2] Likewise, the glossary entry on the Deflated Sharpe Ratio explains a statistical adjustment used to account for selection effects and multiple trials when interpreting Sharpe ratios. Again, the focus is analytical interpretation, not a directive to transact.[3]

Together, these sources support a practical compliance boundary:

• Permissible educational content: explaining a signal-generation method, defining evaluation metrics, showing how to test robustness, or teaching readers how overfitting can distort backtests.[1][2][3]
• Prohibited recommendation-style content: naming a current instrument and telling the reader to take a position, set an allocation, or act on the signal as a trade instruction.[1]

Illustratively, a compliant methodological statement would look like this: a publisher explains that a cross-sectional momentum model ranks assets on trailing returns, then evaluates the ranking rule with out-of-sample tests and a deflated Sharpe framework.[1][3] A non-compliant trade-style statement, under Sonar’s own rules, would be a publication that says a reader should buy or short a specific named asset now because the model currently signals it.[1]

The audit process also helps explain why publishers stay on one side of the boundary. A workflow centered on overfitting checks and statistical deflation pushes authors toward generalizable research claims: how the method was formed, what tests were run, what assumptions were made, and how uncertainty was handled.[2][3] That is structurally different from publishing a directional call. The former is about research process and evidentiary quality; the latter is about prompting action.

Regulatory definitions from bodies such as the SEC or FCA are not present in the current material, so external regulatory definitions cannot be cited.

The defensible conclusion is narrow but clear: Sonar’s publishing process is built to categorize output as methodological education, and its rules expressly prohibit specific trade recommendations. That is why a publisher of quantitative research would confine itself to teaching methods, validation standards, and statistical interpretation rather than naming trades for readers to execute.[1][2][3]

Drafted with AI assistance from cited sources. Reviewed and approved by Sonar Sciences Quant & Research Team.