Boutique Asset Manager: A 'GIPS-Aligned' Presentation and a Personal Trade
You are a senior research analyst at a EUR 2 billion boutique asset manager. In January you prepare the annual performance presentation that marketing describes as GIPS-aligned, although the firm has never claimed compliance. The CIO asks you to (a) remove a EUR 120 million segregated account that terminated in November from the three-year composite, which lifts the headline return from 6.1% to 6.9%, because the client left over fees, not performance; (b) display a backtest of the model portfolio beside the actual composite with equal prominence and no label; and (c) label a 40% cash position that a client imposed during a transition as non-discretionary so that account can be excluded. Separately, you cover a mid-cap retailer in which you personally own shares. On a call, a supplier mentions the retailer is pulling forward orders. You publish a positive note three days later and your spouse's account buys the stock the day before publication.
1. Assess the three presentation requests against performance presentation and misrepresentation standards.
Read the worked answer
(a) Removing the terminated account - not permissible. A portfolio that was discretionary and managed within the composite's strategy must remain in the composite for the periods it was managed, including the final period up to termination. The reason for termination is irrelevant: performance presentation standards require that composite results not be distorted by survivorship, and the removal of a single account that drags a three-year return from 6.9% to 6.1% is textbook survivorship bias. The only legitimate removals are accounts that were never discretionary, or a documented, consistently applied change in composite definition - which requires restatement of all periods and disclosure, not a selective edit before year-end.
(b) Backtest beside actual results - not as proposed. Hypothetical or model performance is not actual performance and cannot be presented with equal prominence or without a label. If shown at all it must be clearly labelled as hypothetical, accompanied by the assumptions and limitations, and kept separate from the actual composite. Presenting a backtest beside a real track record with no label is a misrepresentation of what the firm has achieved - the exact failure the misrepresentation standard is written to catch, regardless of whether the firm claims GIPS compliance.
(c) The 40% cash restriction - not on these facts. A client-imposed restriction can make a portfolio non-discretionary if it significantly hinders implementation of the strategy, but the determination must be documented, applied consistently to all similar accounts, and made when the restriction arises - not retroactively to remove an account already in the composite for three years. A 40% cash constraint during a transition, with the manager retaining 60% discretion, is more likely a restricted portfolio that belongs in a separate composite with disclosure than an excluded one.
Common thread: each request is a change to the presentation that improves the number without changing the underlying management. That is the test; all three fail it.
2. Address the personal holding, the supplier call and your spouse's purchase.
Read the worked answer
The personal holding. Owning shares in a company you cover is a conflict of interest. It must be disclosed to your employer and, in most firms, disclosed in the research report; many policies prohibit holding covered securities outright. Standard on conflicts requires disclosure; the standard on independence and objectivity requires that your recommendation not be influenced by the position. You should have pre-cleared the holding; if you did not, that is a separate compliance breach.
The supplier call. Combining a supplier's comment with public information is classic mosaic analysis and is legitimate - if the information is not material nonpublic information. The test: is the pull-forward information material, and is it nonpublic in the sense that the retailer has not disclosed it and the supplier is breaching a confidence? A single supplier's comment on order timing is normally a piece of the mosaic; but if it is specific enough to move the stock and is clearly confidential, publishing on it is a problem. Document the call, the source's capacity to disclose, and what public information corroborates the conclusion.
The spouse's purchase the day before publication. This is the clearest failure. Whether or not the information was material and nonpublic, buying ahead of your own published recommendation uses the firm's research as a personal trading signal - a priority-of-transactions breach, and a front-running of clients who will read the note. Beneficial ownership through a spouse does not create distance; if you benefit, the trade is yours for policy purposes.
What to do now: disclose the holding and the spouse's trade to compliance immediately, place the security on the firm's restricted list, refrain from publishing or voting until the conflict is cleared by someone independent, and expect the spouse's position to be liquidated or held in a monitored account. Do not publish first and disclose later; sequencing is evidence.
3. What controls and disclosures would prevent a recurrence, and what would you say to the CIO?
Read the worked answer
To the CIO, directly: the presentation requests are declined as framed, and the basis for declining should be documented in the file, not just in a conversation. Offer the workable alternatives - keep the terminated account, label the backtest as hypothetical and separate, and place the restricted account in a disclosed separate composite.
Presentation controls:
- A written composite definition, approved and dated, with a change log. Any change to definitions is prospective, disclosed, and restated where required.
- A policy that terminated portfolios are included for all periods managed, with an automatic reconciliation of accounts in and out of each composite.
- A separate, labelled section for hypothetical performance with assumptions, or none at all.
- Annual independent verification, or at minimum a second-reviewer sign-off by someone who does not report to marketing.
- Marketing copy that uses terms like GIPS-aligned only if legal and compliance have approved the specific wording; if the firm is not compliant, it should say so plainly.
Personal trading controls:
- Pre-clearance for the analyst and household accounts in all covered and watch-list names.
- A blackout window around publication - no trading from the moment a recommendation is being formed until it is published, and for a defined period after.
- Quarterly attestation of holdings, including spouse and dependent accounts, with a beneficial-ownership definition that leaves no gaps.
- An outside-activity and outside-directorship register.
The underlying judgement. None of these controls is exotic and all were available. The real failure is cultural: a CIO who asks for a 6.9% instead of a 6.1% number is telling the analyst that the presentation serves marketing, not clients. The analyst's obligation is to say no in writing, escalate to compliance, and keep the evidence - because when the regulator or a departing client asks how the composite was constructed, 'the CIO told me to' is not a defence, and a file note is.