Standard V: Investment Analysis, Recommendations, and Actions

V(A) Diligence and reasonable basis

Members must exercise diligence, independence and thoroughness and have a reasonable and adequate basis, supported by appropriate research, for any analysis, recommendation or action. How much research is enough depends on the investment philosophy, the member's role and the resources available, but all pertinent issues must be covered.

A basis is built from a balance of sources: macroeconomic conditions, the company's history, the industry and business cycle, a fund's fees and management, a securitisation's asset quality, model outputs and their limits, and sensible peer comparisons. A sound process cannot remove downside risk: a well-researched investment that later fails does not prove a violation.

Secondary research (by others in your firm) and third-party research (outside the firm) may be used if you make reasonable efforts to confirm they are sound. Check the assumptions, the rigour of the analysis, how current it is, and its objectivity. Blogs and social media need more scrutiny than established research houses. You may rely in good faith on your firm's approved vendors unless you have reason to doubt them; the firm should review approved providers regularly.

Group research: a team report reflects the consensus. If you disagree but believe the process was sound, independent and had a reasonable basis, you do not need to remove your name. You may ask to have it removed.

Models and external managers

Users of quantitative models need not master every technical detail, but must understand the parameters, assumptions and limitations, and how the output feeds into decisions. Test model output before relying on it, and make sure scenarios include a broad range, including severe negative outcomes beyond the historical data. Leaving out downside scenarios can misstate an investment's true value.

People who build models face a higher standard than those who use them: they must understand the technical workings, test thoroughly before release, consider whether the data cover both good and bad market cycles, and keep the model updated. Components tested separately must also be tested as a whole.

Selecting an external adviser or subadviser requires the same diligence as selecting a security. Do not pick on fees alone, on returns alone, on friendship or on the data a manager submitted without verification.

Recommended procedures for V(A): a policy that research must have a reasonable basis, reviewed by a supervisory analyst or committee before release; written due diligence guidance; measurable criteria for research quality and accuracy; minimum scenario testing for models; criteria and review frequency for outside data providers; and standardised criteria for evaluating external advisers.

V(B) Communication with Clients and Prospective Clients

Clients can only make informed decisions if they understand what you do for them. Members must disclose the nature of the services they provide and information about the costs the client pays. Best practice is to set both out in writing at the start. Exact amounts are not required (an asset-based fee cannot be known in advance), and you need not disclose what the services cost the firm. Costs charged by affiliates and third parties that are part of the service must be included.

These disclosures are ongoing: when services or costs change, including the method used to calculate fees, tell affected clients promptly, even if the change lowers fees. Client sophistication is no excuse for not disclosing, though the detail can be tailored. If firm-produced disclosures are inadequate, supplement them, raise it with the firm, and dissociate if it cannot be fixed. The duty applies to members who face clients.

Disclose the basic format and general principles of the investment process: how you analyse, select and build portfolios. You need not reveal every model detail, but clients must understand the logic. Material changes must be disclosed promptly: a new valuation model, a style or capitalisation shift, moving stock selection from analysts to a committee, adding external managers, or a material error in the process.

Disclose significant risks and limitations, judged on what was known at the time (ex ante). Leverage is generally a significant risk. Limitations include liquidity (can the position be sold quickly at reasonable cost?) and capacity (the asset size beyond which returns suffer). A risk you could not have known about is not a V(B) breach, though it may point to weak diligence under V(A).

Use judgment to include the important factors so readers can follow and challenge your reasoning; you may limit a report's scope if you say so. If recommendations come in capsule form (a list), tell clients more detail is available. Always distinguish fact from opinion: forecasts, estimated reserves and expected dividends are opinions. For model-based work, state the assumptions and limits; outputs are estimates, not certainties.

V(C) Record Retention

Members must develop and maintain records supporting their analyses, recommendations, actions and client communications, including reviews that did not lead to a trade. Examples: notes from company meetings, company releases, model inputs and outputs, risk analyses, external-adviser selection criteria, notes from IPS review meetings and outside research. Electronic and social-media formats (e-mails, texts, blog and social posts) count.

Records created for the employer are firm property. A member who changes jobs may not take them, even copies, without consent, and may not reuse old recommendations at the new firm without recreating the support from public sources or the company itself, not from memory or old files.

Follow local regulation and firm policy on how long to keep records; doing so satisfies V(C). If neither applies, CFA Institute recommends at least seven years. Keeping records is generally the firm's job, but individuals must archive the notes that support their work.