An analyst develops a quantitative trading strategy and backtests it exclusively on five years of in-sample historical data, achieving a simulated Sharpe ratio of 2.5. She does not perform any out-of-sample testing or walk-forward validation, nor does she examine whether the strategy's parameters were overfit to the historical data. She then presents the strategy to clients as a "proven, reliable approach" for their portfolios. Which Standard has she most likely violated?
- AStandard V(B), Communication with Clients, because she failed to update clients periodically on strategy performance
- BStandard I(A), Knowledge of the Law, because backtesting without disclosure may violate securities regulations
- CStandard I(C), Misrepresentation, solely because the word "proven" overstates certainty
- DStandard V(A), Diligence and Reasonable Basis, because she lacks adequate basis for the investment recommendation
Show answer & explanationAnswer & explanation
Correct answer: D. Standard V(A), Diligence and Reasonable Basis, because she lacks adequate basis for the investment recommendation
Standard V(A) requires members to have a reasonable and adequate basis for investment recommendations, supported by appropriate research and investigation. Relying solely on in-sample backtested results without out-of-sample validation or overfitting checks does not constitute a reasonable basis, especially before presenting the strategy as "proven" to clients — this compounds a V(A) violation, though the misrepresentation is a secondary consequence of the underlying diligence failure.
Why the other options are wrong
- A. Communication with clients pertains to ongoing reporting, not the initial basis for the recommendation.
- B. There is no indication of a legal/regulatory violation regarding backtesting disclosure requirements.
- C. While the language is a concern, the root cause and best-fit violation is inadequate research basis under V(A).
Diligence and Reasonable Basis (V(A))
Members must exercise diligence and have a reasonable and adequate basis, supported by appropriate research, for any investment analysis, recommendation, or action.
- In-sample backtests alone are insufficient; out-of-sample/walk-forward testing is needed
- Overfitting risk must be assessed before relying on quantitative models
- Reasonable basis requirement applies regardless of how strong backtested statistics appear
Memory trick: "A shiny backtest isn't proof — test it out-of-sample for the roof."