The Asset Manager Code, Conduct in Practice & the Advice Perimeter

Branch: Ethics, Standards & GIPS (Finance-Plus ring) · Node ES1.03 · ~16 focused hours · Mastery gate ≥ 85%

Four hundred people get a message at 9:05 on a Tuesday morning. Buy this cement stock, target ₹640, stop ₹470. It arrives free of charge. Four weeks later the same message goes out with a payment link under it, and 46 of those 400 pay ₹2,000 a quarter for the version that arrives first.

The analysis did not change. The legal position changed completely, in India and in the United States both, and it changed at the first rupee.

That first rupee is one of three lines drawn here. The second runs through a phone call. A chief financial officer tells a fund analyst, nine days before results, that the quarter is coming in much better than the street expects, and the analyst has about ninety seconds to work out whether he is now holding unpublished price sensitive information, what that forbids, and what abstaining will cost his fund. In the worked case below the answer is ₹3.54 crore of foregone gain, which is exactly why the reasoning has to be automatic rather than negotiated.

The third line runs between two rulebooks that resemble each other and are not the same object. The CFA Institute Code of Ethics and Standards of Professional Conduct bind a person. The CFA Institute Asset Manager Code of Professional Conduct binds a firm. A firm can adopt the second, comply with every provision of it, and still employ someone breaching the first that morning. Both of the worked examples built on that pairing below are ordinary situations rather than paradoxes, and the gap between them is the reason two documents exist.

There is a fourth thing here that is not a line but a habit. Ethics teaching in this program so far has been diagnostic: read a vignette, name the Standard, state the verdict, point at the deciding fact. That method grades well and it does something real, which is to convert a feeling into a citation. It does not tell you what to do at 11pm on a Thursday when the investment committee has already voted, the raising credit riding on your silence is 2.14 times your base salary, and the person asking you to sign is the person who hired you. A diagnosis is retrospective and a decision is not. Section 9 supplies the second thing, and is careful to say how it differs from the first.

One honest limit before any of it. Nothing here is legal advice, no rule number below is safe to rely on without opening the primary text, and the Indian conduct regime in particular has been amended repeatedly since 2019: the trading-plan cooling-off period, the research analyst regime, and the rules on associating with unregistered promoters all moved between 2024 and 2025. Every citation carries a verify flag naming the source and the date it was checked. Treat those flags as part of the content rather than decoration on it, because a conduct rule you cite from memory is a conduct rule you have not checked.

The work divides into roughly five hours on the core content read with the Asset Manager Code open beside it, three hours on the worked examples, three and a half on the practice set, two and a half on the mini-project, and two on the flashcards and the mastery check.

Learning objectives

You can:

1. State what the Asset Manager Code binds and what the Code and Standards bind, name the subject of each, and explain why a firm's adoption of the one discharges no individual obligation under the other.
2. Recite the six general principles of the Asset Manager Code and state precisely what a claim of compliance commits a firm to: the whole Code rather than a selection, a notification to CFA Institute, and no verification of any kind by anybody.
3. Apply each of the Code's six sections to a firm-level fact pattern, and say for any given failure which section it lands in and what evidence would settle it.
4. Run the four-stage ethical decision framework (identify, consider, decide and act, reflect) on a live situation, name the situational influences at work, quantify the incentive where it can be quantified, and distinguish that framework from the four-step vignette diagnosis it sits beside.
5. Test information against the unpublished price sensitive information definition limb by limb under SEBI's insider-trading regulations, decide whether a person is connected, and compute the trading-window and continual-disclosure consequences from a dated calendar.
6. Work the United States mirror: Rule 10b-5 under the classical and misappropriation theories, tippee liability and the personal-benefit test, Regulation FD's promptness clock, Form 4 timing, and the cooling-off arithmetic of a Rule 10b5-1 plan and its amendment.
7. Place mosaic theory legally beside both regimes, state what an analyst may assemble, quantify an assembled estimate against consensus, and name the one fact that would move the work across the line.
8. List the operative obligations of a person who publishes research for others in India: holdings disclosure on every report, the trading blackout around publication, the prohibition on trading against a live recommendation, certification, and the mandatory disclosure block, and write that block from a blank page.
9. Classify any activity on the advice perimeter for both markets, from a private memo to a discretionary account, and say for each what it triggers under the SEBI research analyst and investment adviser regimes and under the United States Investment Advisers Act, including the publisher's exclusion and where that exclusion stops.
10. Explain the finfluencer regime: what triggers the perimeter, what a SEBI-regulated entity may not do with an unregistered promoter, and the shape enforcement takes when it arrives.

Prerequisites & connections

Builds on. ES1.01 supplies the Code of Ethics, all seven Standards and the four-step vignette method, and is a hard prerequisite rather than a soft one: everything below that says "Standard VI(A)" or "Standard III(D)" assumes you can already state the rule and find the deciding fact. ES1.02 supplies GIPS in full and the three-regulator architecture, including the Investment Advisers Act registration bands, the SEC Marketing Rule, the SEBI Investment Adviser fee-cap machinery and the FCA's Consumer Duty. M0.02 supplies the market structure the conduct rules sit on top of: who the referees are, what a listed company must publish and when. M8.04 supplies the scuttlebutt craft whose legal boundary Section 12 draws. AA1.03 supplies performance measurement and the composite arithmetic, which Section 7 uses without re-deriving.

Feeds forward. PW1.01 owns the private-client investment policy statement, and the instruction there to stage a client's sale around insider-trading-window compliance is the rule Section 10 finally teaches. M10.04 owns the long-run practice audit, and the perimeter tree of Section 14 is what turns its "not investment advice" guardrail into a decision a learner can actually make. The Phase 8 and Phase 10 memos are the raw material for the mini-project. Anyone who goes on to publish anything, for money or otherwise, meets Sections 13 to 15 again the first week they do it.

Six modules own material that borders this one, and none of it is rebuilt here. ES1.01 owns the seven Standards and the vignette method, including the 2024 updates to the Standards themselves; the Standards are cited below by number and never re-taught. ES1.02 owns GIPS, verification, the SEC and FCA architecture and the SEBI registration categories as categories; what follows takes the operative conduct obligations that sit inside those categories and leaves the architecture alone. AA1.03 owns performance measurement, attribution and the composite arithmetic; Section 7 states only what the Asset Manager Code demands of a firm's performance and valuation, and hands the arithmetic back. M0.02 owns market structure and the referees. M8.04 owns mosaic theory as an analytical craft, and Section 12 attaches statute names to a boundary that module already draws correctly. PW1.01 owns the private-client policy statement and its constraints. Cross-links below point at those owners rather than duplicating them.

Core teaching content

3. Section A, loyalty to clients

Three provisions, and they compress into one instruction with two failure modes attached.

The instruction is that client interest comes before the Manager's own. That is the same substance as Standard III(A) at the individual level, arriving one storey up: the Standard asks what a person did, and Section A asks what the firm's arrangements make likely. A revenue-sharing agreement with a custodian, an affiliated broker that receives order flow, a distribution arm paid on gross sales, a research budget funded by client commissions and spent on the marketing team: none of those is a decision anybody made this morning, and each of them bends outcomes every morning.

The first failure mode is confidentiality. Client information stays confidential within the scope of the relationship, and the duty is the firm's as well as the individual's, which means it survives staff turnover and system migrations. A firm that lets a departing team walk out with a client list has breached Section A whatever the individuals did, because the system permitted it.

The second is the arrangement that impairs loyalty before anyone acts on it. The Code's language is about refusing relationships and gifts that could reasonably be expected to affect independence, objectivity or loyalty. The tense matters. A firm does not wait to see whether the arrangement bent a decision; it declines the arrangement.

Where this bites hardest in practice is the affiliate. Consider a firm running discretionary mandates whose parent also owns a broking arm. The mandate documents disclose that affiliated execution may be used. Disclosure is necessary and it is not sufficient, because Section A asks whether the arrangement itself could reasonably be expected to impair loyalty, and Section C then asks separately whether best execution was actually sought. Two provisions, two tests, and passing the disclosure test tells you nothing about the execution one. Worked example 7 puts numbers on the second.

4. Section B, investment process and actions

Six provisions covering how investment decisions get made, and the three that carry the most weight in practice are reasonable basis, mandate consistency and the prohibition on manipulation.

Reasonable basis is the firm-level twin of Standard V(A). At the individual level the question is whether this analyst had adequate support for this recommendation. At the firm level it is whether the process reliably produces adequate support: whether there is a research file, whether models are reviewed by someone other than their author, whether a decision taken on a Monday can be reconstructed on the following Friday. A firm with a brilliant analyst and no file has a Section B problem that will outlive the analyst.

Mandate consistency splits into two duties that are easy to collapse into one. For a pooled fund or a mandate, the Manager takes only actions consistent with the stated objectives and constraints, and separately discloses enough that investors can judge any proposed change of style or strategy. Style drift is therefore two failures rather than one: the drift itself, and the silence about it. A fund whose factsheet says large-cap value and whose portfolio has quietly become mid-cap growth has broken the first even if every individual purchase was well researched.

For separate accounts the Code adds a suitability duty with an ordering: understand the client's objectives, risk tolerance, horizon, liquidity needs, constraints and circumstances first, then determine that an investment is suitable. The ordering is the content. Suitability determined after the trade is a justification rather than a determination, and the private-client policy statement module owns the craft of writing the document that makes the determination possible.

Manipulation is the provision that looks least likely to matter and matters more than it looks. The prohibition covers practices designed to distort prices or inflate volume with intent to mislead, which covers the obvious cases and also the ones firms talk themselves into: marking the close on a valuation date, trading between affiliated accounts to establish a price, and placing orders with no intention of execution. Intent is the element, and intent is proved from the pattern rather than from a confession. The Indian and American statutory versions of the same idea sit in SEBI's prohibition of fraudulent and unfair trade practices regulations and in Rule 10b-5, and a firm that satisfies Section B on this point is usually well clear of both.

5. Section C, trading

Five provisions, and this is where the firm-level Code and the two national insider-trading regimes below meet.

The first is the prohibition on acting, or causing others to act, on material nonpublic information that could affect the value of a publicly traded investment. As a firm-level duty this is a systems requirement: information barriers, a restricted list, a watch list, a wall-crossing procedure with records, and training that has actually happened. Sections 10 and 11 supply the legal content of what counts as such information in India and the United States.

The second is the priority of client transactions over the Manager's own, and it is the provision most often satisfied on paper and defeated in practice. A pre-clearance system that inspects trades will not see a director's fee, a private ownership stake, a spouse's employment, or a promise of future business. Worked example 3 is a system that ran 1,462 pre-clearances in a year, flagged nothing, and missed a conflict worth 7.86 times the analyst's salary, because the conflict was never a trade.

The third governs soft commissions. Client brokerage may buy investment-related products and services that help the investment decision-making process, and may not buy the running of the firm. A firm paying for its order-management system out of client commissions is on the wrong side of that line, and the disclosure obligation in Section F then requires the amount, the goods received and how they benefit the client.

The fourth is best execution, and it is a total-cost test rather than a commission test. Commission is visible, quoted and easy to negotiate; implementation shortfall against the arrival price is invisible, unquoted and usually larger. A broker charging 0.8 cents a share against another charging 2.5 cents looks 68 percent cheaper and can be 16.8 basis points more expensive once the shortfall is counted, which on $2,000,000 of turnover is $3,359.58. Worked example 7 runs that arithmetic in full.

The fifth is fair allocation, and it needs a written policy that survives a partial fill. An order for 90,000 shares that fills 54,000 has to be split by a rule chosen in advance. Pro rata gives 27,000, 18,000 and 9,000 against orders of 45,000, 30,000 and 15,000. First come, first served gives 45,000, 9,000 and nothing, and if the stock then rises $2.19 the smallest account has been handed $19,710 less than the pro-rata rule would have given it. Neither rule is dishonest. Choosing between them after the fill is.

6. Section D, risk management, compliance and support

Seven provisions, and they are the least glamorous part of the Code and the part that most often decides whether anything else in it works.

A Manager develops and maintains policies and procedures that make its activities comply with the Code and with applicable law. It appoints a compliance officer with actual responsibility for administering those policies and for investigating complaints about the firm or its people. Both provisions are about designation: an unnamed responsibility is nobody's.

Portfolio information given to clients must be accurate and complete, and the Manager arranges independent third-party confirmation or review of it. That provision is the one Worked example 4 turns on, and its practical form is a custodian statement, an administrator's books, or an independent valuer, depending on what the assets are.

Records are kept for an appropriate period in an accessible format, which in most jurisdictions means the local regulatory retention period is a floor rather than a target. Staff and technology must be sufficient for the work actually being done, which is a real constraint on a firm that has grown its assets faster than its middle office. A business-continuity plan must exist and address both disaster recovery and periodic market disruption. And a firmwide risk process must identify, measure and manage the firm's risk exposures, including their sources and nature.

The connecting idea across all seven is that a firm is a machine for producing decisions, and the Code is asking whether the machine has the parts. A compliance officer who is also the chief investment officer, a risk report nobody reads, a continuity plan written in 2019 and never tested: each of those satisfies a sentence and fails the provision.

7. Section E, performance and valuation

Two provisions, and between them they carry more money than any other pair in the Code.

Performance information presented to clients must be fair, accurate, relevant, timely and complete, and a Manager must not misrepresent the performance of any portfolio or of the firm. That is the firm-level statement of the same duty Standard III(D) places on the individual, and the GIPS standard is the elaborated, auditable version of it. GIPS itself, its composite construction, its required disclosures and the meaning of verification belong to the GIPS and regulation module; the attribution and composite arithmetic belongs to the performance module. What the Asset Manager Code adds is that the duty binds whether or not the firm claims GIPS compliance at all. A firm outside GIPS is not outside Section E.

The second provision is valuation, and it is the one that turns into cash. Holdings are valued at fair market prices, and where no independent third-party quotation is readily available, the Manager applies good-faith methods to determine fair value. The phrase doing the work is "independent third-party". A manager's own mark on its own unlisted position, feeding its own fee, is not an independent quotation however carefully it was derived.

Put numbers on it. A fund reports a net asset value of ₹340 crore including one unlisted position carried at its ₹48 crore cost twenty months after the round that set that cost. An independent valuer marks the position at 0.62 times cost, or ₹29.76 crore. The overstatement is ₹18.24 crore, which is 5.36 percent of reported net asset value. On a 2.00 percent management fee that is ₹36.48 lakh a year of overcharge. On invested capital of ₹300 crore with an 8 percent hurdle and 20 percent carry, the reported ₹340 crore clears the ₹324 crore hurdle and earns ₹3.20 crore of performance fee, while the corrected ₹321.76 crore clears nothing and earns zero. Total gap for one year: ₹3.5648 crore.

The seam to remember is that the valuation provision and the performance provision fail together. An overstated mark produces an overstated return, which produces an overstated composite, which produces a marketing document that is false in a way no arithmetic check on the composite will ever catch, because the arithmetic is correct and the input is not.

8. Section F, disclosures

Four provisions, of which the fourth carries a long list, and the whole section is organised around four questions: what, to whom, when and in what language.

What gets disclosed is every material fact about the Manager, its personnel, the investments and the process. The Code's own enumeration is long, and it groups into three families. Conflicts come first: those arising from broker relationships, from other client accounts, and from fee structures, together with any regulatory or disciplinary action against the firm or its people. Economics come second: fees and costs with the methodology for computing them, and any soft or bundled commissions with the goods received and the benefit to the client. Mechanics come third: the investment process itself with its lock-ups, strategies, risk factors and any use of derivatives or leverage, plus performance, valuation methods, shareholder voting policy, trade allocation policy, the results of any review or audit, significant personnel or organisational change, and the risk-management process.

To whom is clients and prospective clients, which includes the second group deliberately. A disclosure made only after the money arrives is a disclosure made to somebody who has already decided.

When is ongoing and timely, which rules out the annual-report-only pattern. A firm that changes its trade-allocation policy in March and discloses it in the following February's report has satisfied the letter of a document and none of the provision.

In what language is the requirement people ignore. Disclosures must be truthful, accurate, complete and understandable, in a format that communicates effectively. A thirty-page offering document with the leverage limit on page 24 inside a defined term is not an effective format, and the Code says so without needing a regulator to agree.

One test makes the section operational. Take any fact about your firm and ask whether a prospective client who learned it after investing would feel they had been told. If the honest answer is no, Section F already required it.

{
  "skill": "Auditing a firm's disclosure set against the Asset Manager Code and drafting the gap list",
  "use": "Have an assistant extract every disclosed item from a fund's offering document, factsheet and website into a structured table, then map that table against the Code's Section F list and produce a candidate gap list that you check provision by provision against the primary text yourself.",
  "tools": ["Chat assistants with document handling - Claude, ChatGPT, Gemini", "The Asset Manager Code text from rpc.cfainstitute.org, open beside the output", "A spreadsheet for the provision-by-provision mapping"],
  "prompt": "Here is a fund's offering document, its latest factsheet and the text of its website disclosures. Extract into one table every disclosure the fund actually makes, with the exact source document and page or URL for each. Do not judge completeness yet. Then, in a second table, list each item and quote the sentence that discloses it. Where a topic is mentioned but no specific fact is given (for example leverage is named but no limit is stated), put it in a third table headed 'named but not disclosed'.",
  "verify": "Never accept the mapping. Open the Code's Section F yourself and walk its provisions in order, because an assistant that has been told to find gaps will find gaps, including ones that are not there, and an assistant that has been told a document is compliant will agree. Trace every extracted quote back to the page it came from; document AI drops rows out of dense offering documents silently. Match the effective date of the document against the period you are auditing. Recompute any fee or performance figure the disclosure states. Cite document and page for each item, and record what you could NOT find as a finding rather than as an absence.",
  "diy": "You must be able to state Section F's four provisions and the substance of its enumerated list from memory before an extraction is worth anything, because the gap you are looking for is a topic that is missing entirely, and nothing in the source document will point at it. Reading a disclosure set for what is not there is the gated skill; the assistant only speeds up reading what is.",
  "market": "US"
}

Common mistakes & how experts think differently

1. Treating the firm's adoption of the Asset Manager Code as personal cover. The mistake is tempting because the compliance manual is thick and somebody else wrote it, so the individual reasons that conduct has been outsourced. An expert reads a compliant manual as evidence about systems and as evidence about nothing else, and keeps asking the individual-level question separately, because the two rulebooks have different subjects and neither discharges the other.

1. Believing a claim of Asset Manager Code compliance has been checked by somebody. The claim is self-made, notified rather than approved, and expressly unverified by CFA Institute. The word "certified" attached to it is always wrong. An expert treats the claim as a representation whose truth is a question of fact, and reads the firm's actual policies before believing it.

1. Applying the trading window to the wrong person. The window closes for the company's designated persons. Outside analysts are bound by the possession rule and by their own firm's code, which is a different prohibition with a different trigger. The error runs both ways: analysts who think they are free during an open window and analysts who think they are frozen during a closed one are both misreading who Schedule B binds.

1. Testing materiality and forgetting the other limb. A fact can be highly material and entirely published, in which case it is not unpublished price sensitive information at all. The generally-available test in Regulation 2(1)(e) asks whether the information is accessible to the public on a non-discriminatory basis, not whether the analyst happened to know it. Experts run both limbs, in order, before doing anything else.

1. Confusing Regulation FD with an insider-trading rule. Regulation FD binds the issuer and imposes no duty on the analyst who receives the disclosure. What the analyst risks is tippee liability under Rule 10b-5, which requires a breach by the tipper for a personal benefit. Two rules, two defendants, two element sets. Firms that conflate them end up with a compliance policy that protects nobody.

1. Assuming the classical theory reaches everyone. A buy-side analyst is not an insider of the issuer and owes its shareholders nothing, so Chiarella does not reach him. O'Hagan does, whenever the information's source is a principal he owes a duty to, and his own employer is the commonest such source. The expert's habit is to ask "whose information was this, and what did I owe them?" rather than "was this material and nonpublic?"

1. Amending a 10b5-1 plan as though it were an edit. A change to amount, price or timing is treated as a termination and a new adoption, and the cooling-off clock restarts. In the worked case the amendment costs 102 days. The tempting reading is that a small change is a small change; the rule does not measure size.

1. Reading the disclosure block as boilerplate rather than as the deliverable. A research report's disclosures are the part a regulator reads first, and the two ownership tests behave differently: beneficial ownership is disclosed at one percent and above, financial interest at any size. Experts draft the block before the report, because writing it early surfaces the conflicts while there is still time to decline the coverage.

1. Thinking the perimeter has a revenue threshold. Consideration is a switch. The first rupee does all the legal work, and the annual total does none of it beyond sizing the eventual disgorgement. People wait until the income "gets serious" and then discover that seriousness was never the test.

1. Using investor education as a shelter for a return claim. The education carve-out survives only where there is no recommendation and no claim about return or performance. A screenshot of a profit and loss statement is a performance claim in every relevant sense, and calling the surrounding text educational does not change what the screenshot says.

Worked examples

No figure below has been retyped from another source. Each one is computed, and the Verify block at the end of this section restates every one at full precision.

Worked example 1: The call nine days before results (India, ₹)

Bhavnagar Speciality Chemicals Ltd (synthetic) has a September quarter end. Its board approves and publishes Q2 results at 16:40 IST on 9 November 2026. On 31 October 2026, nine days before that, the company's chief financial officer telephones an analyst at an Indian asset manager and says, unprompted, that the quarter is coming in much better than the street expects. The fund already holds 8,20,000 shares at an average cost of ₹1,104. The analyst was about to add 2,00,000 shares at the prevailing ₹1,286. Consensus for the quarter is ₹412 crore of revenue at a 17.2 percent EBITDA margin. The actual numbers, when they land, are ₹468 crore and 21.4 percent, and the stock closes the next session at ₹1,463.

Solution. Step 1, run the definition, limb by limb. Is the information unpublished? Yes: a directional statement about an unannounced quarter, made in a private call, is not accessible to the public on a non-discriminatory basis, which is the test in Regulation 2(1)(e). Is it price sensitive? The ex-post evidence is decisive and the ex-ante reasoning has to stand on its own: a 13.59 percent revenue beat and a 4.2 percentage-point margin beat are the kind of surprise that moves a mid-cap, and in the event the stock moved 13.76 percent on publication. Both limbs hold, so this is unpublished price sensitive information under Regulation 2(1)(n).

Step 2, is the analyst a connected person? Regulation 2(1)(d) reaches anyone associated with the company in the previous six months in any capacity giving access to such information, and expressly includes association by frequent communication with the company's officers. An analyst who speaks to this chief financial officer four times a quarter is inside that limb. The consequence is evidentiary rather than substantive: the presumption in Regulation 4 puts the burden on him to show the trade was not motivated by the information.

Step 3, does the trading window bind him? No, and this is the step that is most often got wrong. Schedule B's window closes for the company's designated persons, from the end of the quarter until 48 hours after results are declared. For this calendar that is 1 October 2026 until 16:40 on 11 November 2026, which is 42 days. The analyst is not a designated person of the issuer. What binds him is Regulation 4's possession rule and his own firm's code of conduct maintained under Regulation 9.

Step 4, price the abstention. Buying 2,00,000 shares at ₹1,286 and holding to ₹1,463 would have produced ₹3,54,00,000, or ₹3.54 crore. The existing 8,20,000-share position gains ₹14.514 crore on publication either way, and no rule requires the fund to sell what it already owns. The forbidden act is the incremental purchase while in possession, and its cost is ₹3.54 crore of foregone gain.

Step 5, what actually has to happen. Log the call, including what was said and when, in the firm's records. Escalate to the compliance officer the same day. Place the security on the restricted list until 48 hours after publication. Do not trade, and do not pass the information on, because Regulation 3 prohibits communication except for a legitimate purpose and a trading idea is not one.

Step 6, the second call, for contrast. Two weeks later the same officer mentions that input costs have fallen. The benchmark input has moved from ₹214 to ₹174.2 per kilogram over the quarter, a fall of 18.60 percent, published weekly by a trade body. This fails the first limb: the information is generally available. No abstention is required and no restricted-list entry is warranted, though logging the call still is.

Read it. The ₹3.54 crore is not a penalty and nobody imposes it. It is the ordinary price of a rule that binds on possession rather than on use, and an analyst who has not priced it in advance is an analyst who will negotiate with himself at the moment it arrives. Every regulation number in this example carries a verify flag: check Regulations 2(1)(d), 2(1)(e), 2(1)(n), 3, 4 and 9 and Schedule B at sebi.gov.in, checked 7 September 2026.

Worked example 2: A 10b5-1 plan, its amendment, and the Regulation FD clock (US, $)

The chief financial officer of Calder Ridge Semiconductor Inc. (synthetic), a Section 16 officer, adopts a Rule 10b5-1 plan on 21 April 2026. The company's scheduled earnings release is 11 May 2026, twenty days later. The plan's own first scheduled sale is 1 June 2026. On 3 August 2026 the officer modifies the plan's limit price. The company's third-quarter results are disclosed on 10 November 2026.

Solution. Step 1, the cooling-off arithmetic at adoption. For a director or Section 16 officer the period is the later of two arms, capped. Arm one is 90 days after adoption: 20 July 2026. Arm two is two business days after disclosure of the results for the quarter in which the plan was adopted: 13 May 2026. The later is 20 July 2026. The cap is 120 days after adoption, 19 August 2026, which is not reached. The first permitted trade is therefore 20 July 2026, and the plan's own 1 June date is 49 days too early.

Step 2, the amendment. A change to amount, price or timing is treated as a termination and a fresh adoption. Arm one is now 90 days after 3 August, which is 1 November 2026. Arm two is two business days after the 10 November results, which is 13 November 2026 rather than 12 November, because 11 November is a federal holiday and business days skip it. The later is 13 November 2026; the cap of 1 December 2026 is not reached. The modification costs 102 days.

Step 3, the same plan for a non-officer. An employee who is neither a director nor a Section 16 officer has a flat 30-day cooling-off, so an identical plan adopted 21 April 2026 could first trade on 21 May 2026. Same document, same company, different clock, because the rule keys on the role.

Step 4, Regulation FD from the company's side. On Friday 15 May 2026 at 16:45 ET an investor-relations officer non-intentionally gives a segment margin figure to two analysts. Public disclosure is required promptly, which is defined as the later of 24 hours or the commencement of the next day's trading on the New York Stock Exchange. Twenty-four hours is 16:45 on Saturday 16 May. The next trading session begins at 09:30 on Monday 18 May. The later is Monday 09:30, giving 64.75 hours. Move the same slip to 18:05 on a Thursday and the 24-hour arm wins, giving exactly 24 hours.

Step 5, which theory reaches the analyst. Suppose an analyst at Two Rivers Capital (synthetic) buys 45,000 shares at $38.10 on the strength of a figure passed to him and the stock trades at $46.85 after the release, a gain of $393,750. The classical theory does not reach him: he is not an insider of Calder Ridge and owes its shareholders no duty. Tippee liability under Dirks reaches him only if the tipper breached a duty for a personal benefit and he knew or should have known. The misappropriation theory reaches him directly if the source of the figure was a principal he owed a duty to, his own employer being the commonest case. Under Section 21A the civil penalty ceiling is three times the profit gained, or $1,181,250.

Verify. Rule 10b5-1(c)(1) as amended in December 2022 and effective 27 February 2023, Regulation FD at 17 CFR 243.100 to 243.103, Section 16(a) with its two-business-day Form 4 deadline, and Section 21A, all at sec.gov, checked 7 September 2026.

Read it. The two clocks in this example fail in opposite directions. The 10b5-1 clock punishes people who think an amendment is a small thing, and the Regulation FD clock rewards anyone who has worked out which arm wins before the phone call happens rather than after it.

Worked example 3: The firm complies and the analyst does not (US, $)

Hollis & Vane Asset Management (synthetic) claims compliance with the CFA Institute Asset Manager Code. Its policy set satisfies Section D on inspection: a named compliance officer with an independent reporting line, a personal-account dealing system requiring pre-clearance of every trade, annual attestations, seven-year record retention, and a tested continuity plan. In the last year the system pre-cleared 1,462 personal trades and flagged none.

A senior analyst holds a 6.5 percent stake in Tenpenny Fluidics (synthetic), a private company last priced at $26.0 million. Tenpenny's revenue is $18.4 million, of which $11.2 million comes from a single customer: Marrow Industrial Inc., a listed company the analyst covers and has rated a buy for two years. He has never traded Marrow personally and has disclosed nothing about Tenpenny, because the attestation form asks about securities transactions.

Solution. Step 1, the firm's position. Nothing in the fact pattern breaches Section D. The policies exist, the officer is appointed, the records are kept, the system runs. A reviewer auditing the manual would find it compliant, which is exactly the problem: the manual is compliant and the firm is exposed.

Step 2, the individual's position. Standard VI(A) requires full and fair disclosure of all matters that could reasonably be expected to impair independence and objectivity, and it is the broadest of the conflict Standards. An ownership interest whose value depends on the commercial health of a company he rates is squarely inside it. The stake is worth 6.5 percent of $26.0 million, or $1,690,000, which is 7.86 times his $215,000 base salary. Tenpenny takes 60.87 percent of its revenue from the company he is recommending. Standard VI(A) is violated, and the deciding fact is the undisclosed ownership rather than any trade.

Step 3, the Code provisions that fail with it. Section A.3 asks a firm to refuse business relationships that could reasonably be expected to affect loyalty, and Section F requires disclosure of conflicts arising from relationships and other interests. The firm's claim of compliance is therefore false as applied, not because the manual is wrong, but because the manual has no question that would ever surface this.

Step 4, the systems fix. The gap is that a personal-account dealing screen inspects trades. Closing it requires an outside-interests declaration covering private holdings, directorships, advisory roles, and immediate-family employment, refreshed annually and on change, cross-referenced against the covered-company list. That control would have fired on day one.

Read it. A screen that runs 1,462 times and flags nothing is not evidence of a clean population. It is evidence about what the screen looks at, and the conflict here was never a trade at any point.

Worked example 4: The analyst is clean and the firm is not (India, ₹)

Sarayu Alternative Managers (synthetic) runs a Category II alternative investment fund reporting a net asset value of ₹340 crore against ₹300 crore of invested capital, a reported one-year return of 13.33 percent. Included in that net asset value is a position in Rohtak Cold Chain (synthetic), unlisted, carried at its ₹48 crore cost twenty months after the round that set the price. Fees are 2.00 percent on net asset value, with 20 percent carry over an 8 percent hurdle. An independent valuer engaged for a different purpose marks the position at 0.62 times cost.

A junior analyst notices that the carrying value has not moved in twenty months, writes a memo saying so, sends it to the compliance officer and the investment committee, and is overruled on a documented basis. He makes no public statement, signs nothing, and does not trade.

Solution. Step 1, the valuation gap. At 0.62 times cost the position is worth ₹29.76 crore, so net asset value is overstated by ₹18.24 crore, which is 5.36 percent of the reported figure. Corrected net asset value is ₹321.76 crore.

Step 2, the fee consequence. Management fee at 2.00 percent on the ₹18.24 crore of overstatement is ₹36.48 lakh a year. The hurdle requires ₹324.00 crore. The reported ₹340 crore clears it by ₹16 crore and earns 20 percent of that, ₹3.20 crore. The corrected ₹321.76 crore, a 7.25 percent return, clears nothing and earns zero. The total one-year gap is ₹3.5648 crore, of which 89.77 percent is the performance fee.

Step 3, the Code provisions. Section E.2 requires fair-market prices and, where no independent third-party quotation is available, good-faith fair-value methods. Cost twenty months after a round is neither a market price nor a method. Section D.3 requires the Manager to arrange independent third-party confirmation or review of the portfolio information given to clients. Section E.1 then fails downstream, because the performance figure derived from the mark is not accurate. Three provisions, one root cause.

Step 4, the individual's position. Up to the point of publication the analyst has done what the Standards require: he identified the problem, escalated it in writing, and created a contemporaneous record. Standard V(A)'s reasonable-basis duty is satisfied by the memo. If the firm then publishes the reported figure to investors and he participates in producing or distributing it, Standard I(A)'s dissociation requirement bites and III(D) becomes live for him personally. His clean position is real and it is time-limited.

Step 5, the seam. The composite arithmetic, the recomputation of net from gross, and the GIPS provisions that formalise all of this belong to the performance and regulation modules. What the Asset Manager Code adds is that the duty applies whether or not this firm claims GIPS compliance, and this one does not.

Read it. The ₹3.5648 crore does not require anybody to be dishonest. It requires only that nobody re-marks a position that nothing forces them to re-mark, and that is why the Code puts the obligation on the firm's arrangements rather than on anyone's intent.

Worked example 5: Six activities on the perimeter (India ₹ and US $)

A learner who has been writing memos for two years is considering six things. Classify each in both markets.

Solution.

Activity 1, a memo written for her own file and shown to nobody. India: outside every regime, because there is no publication and no client. United States: outside the Advisers Act, because there is no compensation and no advice given to another person. No action needed beyond the habit of dating it.

Activity 2, an unpaid memo emailed to two friends with a target price. India: the investment adviser definition requires consideration and there is none. The research analyst definition turns on publication rather than payment, so the safety here comes from the communication being genuinely private rather than from its being free. United States: the three-prong test fails at compensation. Both answers change the moment anybody pays anything, including in kind.

Activity 3, a free public blog post with a target price. India: this is a recommendation and a price target published to the public, which is the substance of a research report; treat the research analyst regime as live and verify current guidance. United States: likely within the publisher's exclusion, if the blog is a genuine publication of general and regular circulation and the advice is impersonal.

Activity 4, a paid subscription newsletter, 180 subscribers at ₹499 a month. Annual receipts are ₹10,77,840. India: consideration plus recommendations puts this in registration territory. United States: the publisher's exclusion can still apply notwithstanding payment, which is the holding in Lowe, provided the publication stays bona fide, regular and impersonal. Note the asymmetry: the same activity is more likely to require registration in India than in the United States, and a writer with readers in both places has to satisfy the stricter one.

Activity 5, a free group of 400 with a paid premium tier at ₹2,000 a quarter taken up by 46 members. Premium receipts are ₹3,68,000 a year, 11.5 percent of the group is paying, and 354 members receive the same calls for nothing. India: both regimes are live. The premium tier supplies consideration, and answering individual members' questions about their own holdings supplies personalisation. United States: the exclusion fails once the advice becomes person-specific, and it fails one answered question at a time rather than all at once.

Activity 6, a discretionary account for a cousin under a power of attorney, with no fee. India: discretionary management of another person's assets is portfolio-management territory, and the family exemption is narrow and fact-specific. United States: this is advisory activity, and holding the authority to trade raises custody questions on top of it. The absence of a fee helps with one prong of one test and does nothing about discretion.

Read it. The three switches do all the work: consideration, personalisation, discretion. Activity 4 flips one, activity 5 flips two, activity 6 flips two different ones, and the reason activity 3 is the hardest cell in the table is that it flips none of the three in the United States and still looks like a research report in India.

Worked example 6: The disclosure block and the blackout (India, ₹)

A registered research analyst intends to publish initiating coverage on Nandurbar Agri Foods (synthetic) on 18 March 2026, with a BUY rating and a twelve-month target of ₹640 against a market price of ₹512. She holds 62,000 shares. The company has 55,00,000 shares outstanding. She has received no compensation from the company and has never been an officer, director or employee of it, and her firm does no investment banking.

Solution. Step 1, the ownership arithmetic. 62,000 of 55,00,000 shares is 1.127273 percent, which is above the one percent beneficial-ownership threshold in Regulation 19. Both the beneficial-ownership disclosure and the financial-interest disclosure are therefore required, and the second would have been required at any size. The holding is worth ₹3,17,44,000 at ₹512, and the target implies 25.00 percent upside, or a further ₹79.36 lakh on her own position.

Step 2, the blackout. Regulation 16 forbids dealing in the security within thirty days before and five days after publication. For an 18 March 2026 report that is 16 February 2026 to 23 March 2026, 36 calendar days inclusive. She may not add to or reduce the position inside that window, and separately may not trade contrary to the recommendation while it stands.

Step 3, the block itself. Nine items have to be addressed, and each is addressed even when the answer is nothing.

DISCLOSURES (SEBI (Research Analysts) Regulations, 2014, Regulation 19)
1. Registration: [name], SEBI Research Analyst, registration no. [INH...].
   Certification: NISM-Series-XV Research Analyst.
2. Financial interest of the analyst and relatives in the subject company:
   YES. The analyst holds 62,000 equity shares of Nandurbar Agri Foods Ltd.
   No relative of the analyst holds any interest.
3. Actual or beneficial ownership of 1% or more of the subject company's
   securities at the end of the month immediately preceding publication:
   YES. 62,000 shares = 1.13% of 55,00,000 shares outstanding, as at
   28 February 2026.
4. Any other material conflict of interest at the time of publication: NONE.
5. Compensation received from the subject company in the past 12 months:
   NONE, of any kind.
6. Public offering: the analyst and associates have NOT managed or
   co-managed any public offering of the subject company's securities.
7. Investment banking, merchant banking or brokerage compensation from the
   subject company in the past 12 months: NONE.
8. The analyst has NOT been an officer, director or employee of the
   subject company.
9. Market making: the analyst and associates have NOT engaged in market
   making in the subject company's securities.

Recommendation definitions: BUY = expected total return above 15% over
12 months; HOLD = -5% to +15%; SELL = below -5%.
Valuation basis and risks: [state the method and the three risks].
Trading restriction: the analyst has not dealt in this security between
16 February 2026 and 23 March 2026, and will not deal contrary to this
recommendation while it stands.
This report is for information and does not consider the objectives,
financial situation or needs of any specific person.

Step 4, what the block does not fix. A disclosed conflict is still a conflict. A 1.13 percent holding in a company she is initiating coverage on with a 25 percent target is a fact her readers can now weigh, and weighing it may lead them to discount the report entirely. Disclosure transfers the judgement to the reader; it does not neutralise the interest.

Verify. Regulations 16 and 19 and the certification requirement at sebi.gov.in, checked 7 September 2026. The research analyst regime was substantially amended with effect from 2025, and the numbering, the registration mechanics and several operative conditions should be confirmed against the current consolidated text before this block is used for anything real.

Worked example 7: Section C in numbers (US, $)

A manager places a single order for 90,000 shares across three accounts that want 45,000, 30,000 and 15,000. The market gives 54,000 at an average of $27.41. Two weeks later the stock is $29.60. The manager also has a choice of brokers: A charges 2.5 cents a share and has historically delivered an 18 basis-point implementation shortfall against the arrival price; B charges 0.8 cents and delivers 41 basis points.

Solution. Step 1, allocate pro rata. The fill ratio is 54,000 divided by 90,000, or 0.600000. Allocations are 27,000, 18,000 and 9,000, and they sum to the fill exactly.

Step 2, allocate first come, first served. Taking the accounts in order of arrival gives 45,000 to the first, 9,000 to the second and nothing to the third.

Step 3, price the difference. At $2.19 of subsequent appreciation, the pro-rata rule moves $39,420 of gain away from the largest account and hands $19,710 to each of the other two. Neither rule steals; the difference is entirely distributional, and both are defensible in advance. Choosing after the fill is what Section C.5 forbids, and it is why the policy has to be written down before there is a fill to allocate.

Step 4, best execution as total cost. At $27.41 a share, A's 2.5-cent commission is 9.1208 basis points and B's 0.8-cent commission is 2.9186. Total cost is 18 plus 9.1208, or 27.1208 basis points for A, against 41 plus 2.9186, or 43.9186 for B. B is 16.7979 basis points more expensive despite charging 32 percent of A's commission. On $2,000,000 of turnover that is $3,359.58.

Step 5, the cross-check. The two halves of this example test the same idea from opposite sides. Allocation asks whether the rule was fixed before the outcome was known. Execution asks whether the number being optimised is the number that leaves the client's account. In both cases the visible figure, the order size and the commission, is the one that misleads.

Read it. Best execution is a duty about total cost and a firm that reports commission savings to its clients is reporting the smaller half of the number. What would break this analysis is a shortfall estimate drawn from a different order profile: 18 basis points on a liquid mid-cap is not 18 basis points on a small-cap block, and a firm using one broker's historic average across every order size has not done the work.

Verify block

Every printed figure above, restated at full precision.

Practice set

Work each problem fully before reading its solution. A calculator is assumed. Date arithmetic must be exact; monetary and percentage answers within ±1 percent score as correct unless the problem asks for an exact figure. Every regulation cited in a solution should be treated as a pointer to a primary source rather than as the source.

P1 (guided). Two codes, one firm. Meghna Capital (synthetic) publishes on its website that it "is certified compliant with the CFA Institute Asset Manager Code, Sections A through E." Name every distinct thing wrong with that sentence, and say what a client should ask for next.

Solution. Three things are wrong. First, "certified" describes something that does not exist: CFA Institute does not verify, audit or endorse any claim of compliance, and the required claim language says so. Second, "Sections A through E" is a partial adoption, and the Code is adopted in its entirety or not claimed at all. Third, the claim as written is a misrepresentation of the firm's compliance status, which makes Standard I(C) live for any individual who repeats it. The client should ask for the firm's actual claim-of-compliance statement in its required form, ask whether CFA Institute has been notified, and then ask for the policies implementing Section F, since a firm careless about the wording of its claim is unlikely to be careful about its disclosure set.

P2 (guided). The window, from a calendar. A listed Indian company has a December quarter end and publishes results at 11:20 on 6 February 2027. State when the trading window for designated persons closes and reopens, and how many days it was shut. Then say whether an outside analyst at an asset management company is bound by that window.

Solution. The window closes at the end of the quarter, 31 December 2026, so the closed period runs from 1 January 2027. It reopens 48 hours after the results are made public, at 11:20 on 8 February 2027. Counting 1 January to 8 February 2027 inclusive gives 39 days. The outside analyst is not a designated person of the issuer and the Schedule B window does not bind him. He is bound by Regulation 4's prohibition on trading while in possession of unpublished price sensitive information, and by his own firm's code of conduct maintained under Regulation 9.

P3 (guided). Three facts, two limbs. Decide for each whether it is unpublished price sensitive information. (a) A chief executive tells an analyst privately that the board will recommend a bonus issue next month. (b) An analyst calculates from published customs data that a company's export volumes rose 22 percent. (c) A plant manager mentions that a line was down four days, in a quarter where that line makes about 1 percent of output.

Solution. (a) Yes. A change in capital structure sits in the deemed list, the communication was private, and a bonus issue is likely to affect the price. Both limbs hold. (b) No. It fails the first limb: customs data is accessible to the public on a non-discriminatory basis, so the underlying information is generally available. The analyst's conclusion is his own work and is the mosaic. (c) No, on the second limb. Four days of downtime on a line contributing about 1 percent of output is not likely to materially affect the price. Record it, because the answer would change if the line were 30 percent of output or the downtime were four weeks.

P4. Aggregation to the disclosure threshold (India, ₹). A designated person of a listed company sells 150 shares at ₹2,180 on 8 January 2027, 210 at ₹2,260 on 12 February 2027, and 120 at ₹2,410 on 3 March 2027. State whether Regulation 7(2) is triggered, on which trade, by how much, and what has to happen next.

Solution. The three trades are worth ₹3,27,000, ₹4,74,600 and ₹2,89,200, a running total of ₹3,27,000, then ₹8,01,600, then ₹10,90,800. All three fall in the January to March calendar quarter, so they aggregate. The threshold of ₹10,00,000 is crossed on the third trade, which itself is only ₹2,89,200; the shortfall before it was ₹1,98,400. Within two trading days of that third transaction the designated person must disclose to the company, and the company must notify the exchange within two trading days of receiving the disclosure. The disclosure covers the trades, not merely the excess of ₹90,800.

P5. Cooling-off, two roles (US, $). A Section 16 officer adopts a 10b5-1 plan on 2 February 2027. The company discloses its quarterly results on 24 February 2027. Compute the first date on which the plan may trade. Then compute it for a non-officer employee adopting the same plan on the same day.

Solution. Arm one is 90 days after adoption: 3 May 2027. Arm two is two business days after the results: 26 February 2027. The later is 3 May 2027, and the 120-day cap of 2 June 2027 is not reached, so the first permitted trade is 3 May 2027. For a person other than a director, Section 16 officer or the issuer, the cooling-off period is a flat 30 days, giving 4 March 2027. The gap between the two answers is 60 days and it is created entirely by the filer's role.

P6. The promptness clock (US). An investor-relations officer non-intentionally discloses a material figure to two analysts at 18:05 ET on Thursday 10 June 2027. State the Regulation FD deadline and the number of hours available, and say how the answer would change if the slip had happened at 16:45 on a Friday.

Solution. Arm one is 24 hours later, 18:05 on Friday 11 June 2027. Arm two is the commencement of the next day's trading on the New York Stock Exchange, 09:30 on Friday 11 June 2027. The later is 18:05 on Friday, giving exactly 24.00 hours. A Friday 16:45 slip pushes arm two to Monday's open at 09:30, which beats the Saturday 16:45 twenty-four-hour mark, giving 64.75 hours. The rule is the same; only the calendar changes, which is why an issuer's disclosure policy names the deadline computation rather than a number of hours.

P7. Which theory, if any (US). Classify each of four fact patterns. (a) A company's controller buys shares knowing of an unannounced acquisition. (b) A lawyer at a firm advising the acquirer buys shares in the target. (c) An analyst infers the acquisition from public filings and buys. (d) An analyst is told by a friend at the acquirer's investment bank, who tells him because they are friends.

Solution. (a) Classical theory. The controller is an insider of the issuer and owes its shareholders a duty of trust and confidence. (b) Misappropriation theory, which is O'Hagan on its facts: he owes no duty to the target's shareholders and deceives his own firm and its client by using their confidential information to trade. (c) Neither. This is Chiarella's outcome and the mosaic: no duty, no deception, no liability, however material the inference. (d) Tippee liability under Dirks, with the personal benefit supplied by the gift itself under Salman, provided the analyst knew or should have known the tipper was breaching a duty.

P8. The mosaic in numbers (India). Channel checks across three dealer buckets, weighted by share of national revenue, give: 0.44 weight reporting sell-through up 11.5 percent, 0.33 weight up 7.5 percent, and 0.23 weight up 2.0 percent. Consensus revenue growth is 6.0 percent. Compute the weighted estimate and the gap, compare it with the unweighted average, and state the one fact that would end the mosaic.

Solution. The weights sum to 1.00. The weighted estimate is 0.44 times 11.5 plus 0.33 times 7.5 plus 0.23 times 2.0, which is 5.06 plus 2.475 plus 0.46, or 7.995 percent. The gap to consensus is 1.995 percentage points. The unweighted average of the three bucket figures is 7.00 percent, so weighting adds 0.995 points and is where the analysis lives. The fact that would end the mosaic is a company officer supplying the actual quarterly revenue or sell-through figure: that single tile is material and unpublished by itself, and possession of it stops the analyst trading until publication, regardless of how good the estimate already was.

P9. Perimeter, five activities. Classify each for India and the United States. (a) A free weekly newsletter with no recommendations, only sector commentary. (b) The same newsletter with a monthly stock pick, still free. (c) The same newsletter, paid, at ₹299 a month. (d) A paid one-to-one call in which the writer reviews a subscriber's portfolio. (e) A free Telegram channel with a tip jar.

Solution. (a) Neither regime, in either market: general commentary on market trends is outside the research report definition, and there is no compensation for the United States test. (b) A recommendation published to the public. India: the research analyst regime should be treated as live, and verified. United States: likely inside the publisher's exclusion if genuine, regular and impersonal. (c) India: consideration plus recommendations, so registration territory. United States: the exclusion can still apply notwithstanding payment, per Lowe. (d) Both markets: this is personalised advice for consideration, so India's investment adviser regime and the Advisers Act both engage, and the publisher's exclusion fails on impersonality. (e) The tip jar supplies consideration, unreliably and after the fact, which is the worst of both positions: the writer has the compensation prong satisfied for whoever paid and no clarity about who did. Treat a tip jar as payment.

P10. Audit a disclosure block (India). A draft research report carries five items: registration number, financial interest of the analyst, compensation from the subject company, market making, and the recommendation definitions. Nine are required by Regulation 19. Name the missing items and compute the completeness percentage.

Solution. Four of the nine are missing outright: beneficial ownership of one percent or more at the end of the preceding month; any other material conflict; whether the analyst or an associate managed or co-managed a public offering; and whether the analyst has been an officer, director or employee. The investment banking, merchant banking or brokerage compensation item is absent as well, and some drafts fold it into the general compensation line when it belongs on a line of its own. Five are therefore missing, and only four of the draft's five entries count toward the nine, because the recommendation definitions are worth having and are not one of the required disclosures. Completeness is 4 of 9, or 44.4444 percent. The fix states every item explicitly, including the ones whose answer is none, because a silent item is indistinguishable from an overlooked one.

P11. Allocation and execution (US, $). An order for 100,000 shares across accounts of 60,000, 24,000 and 16,000 fills 40,000 at $154.20. Compute the pro-rata allocation. Then compare broker A at 3.0 cents a share with a 12 basis-point shortfall against broker B at 1.1 cents with 33 basis points, and price the difference on $6,168,000 of turnover.

Solution. The fill ratio is 0.400000, giving 24,000, 9,600 and 6,400, which sum to 40,000. At $154.20, A's commission is 1.9455 basis points and B's is 0.7134. Total cost is 13.9455 basis points for A and 33.7134 for B, a gap of 19.7678 in A's favour. On $6,168,000 that is $12,192.80. B's commission is 36.7 percent of A's and B is the more expensive broker by a wide margin, which is the standard result whenever a low-touch venue is used for an order it cannot absorb.

P12. Valuation and fee (India, ₹). A fund reports ₹512 crore of net asset value including a ₹96 crore unlisted position carried at cost. An independent valuer marks it at 0.55 times cost. The management fee is 1.75 percent of net asset value. Compute the overstatement in rupees and as a percentage of net asset value, and the annual fee overcharge. Then name the two Code provisions engaged.

Solution. The position is worth ₹52.8 crore at the independent mark, so net asset value is overstated by ₹43.2 crore, which is 8.4375 percent. The management fee overcharge is 1.75 percent of ₹43.2 crore, or ₹75.6 lakh a year. Section E.2 is engaged because carrying an unlisted position at cost is neither a fair market price nor a good-faith fair-value method, and Section D.3 is engaged because the firm has not arranged independent third-party confirmation of the portfolio information it gives clients. Section E.1 fails downstream, because the return computed from the mark is not accurate.

P13. Situational influence, priced. A relationship manager on an ₹18,00,000 base is asked to sign off on a suitability assessment for a ₹25 crore mandate that carries a 1.25 percent fee. The client's documented risk tolerance does not support the product. Compute the fee at stake and its multiple of the manager's base, then run the four-stage framework and list four alternative actions between "sign" and "resign".

Solution. The fee is 1.25 percent of ₹25 crore, or ₹31,25,000, which is 1.7361 times the manager's base salary. Identify: the client is the primary stakeholder, with the firm, colleagues, other clients in the same product and the market beyond them; the conflict is that the manager's own standing depends on the mandate. Consider: the situational influences are the incentive just priced, loyalty to the colleague who sourced the mandate, and obedience to whoever is asking; guidance is available from compliance and from outside counsel; alternatives are more than two. Decide and act, with a contemporaneous record. Reflect afterwards on whether the firm produces this situation repeatedly. Four alternatives between the extremes: request the instruction in writing; escalate to compliance in writing with the documented tolerance attached; propose a different product that fits the documented profile; recuse yourself from this specific assessment while continuing in the role. Standard III(C) is the individual rule and Section B.6 of the Code is the firm-level one, and both are broken by a signature.

Applied mini-project

Take a memo you have already written and put it through the perimeter. Budget about two and a half hours. The deliverable is one existing memo, amended, plus a one-page conduct note. Everything you assert about a rule carries a verify flag naming the primary source and the date you checked it.

Start by choosing the memo, in about ten minutes. Pick one you actually wrote in Phase 8 or Phase 10, on a real company, with a view in it. A memo with no recommendation and no target is the wrong choice, because the exercise turns on the fact that you said something.

Then build the disclosure block, in about thirty-five minutes. Address every item Regulation 19 requires, including every item whose honest answer is none. State your own holding in the subject company in shares and as a percentage of shares outstanding, computed from the latest shareholding disclosure, with the date of that disclosure. State any holding by an immediate relative. State the recommendation definitions you are using, with the return bands, and state your valuation basis and three risks. If you have never held the security and never will, the block still gets written, because the block is the artefact being graded.

Next, classify the activity on the perimeter twice, in about thirty minutes. Run the memo through the three switches for India and for the United States. For India, name the regime it engages and the specific regulation, or state that it engages none and say which switch it fails. For the United States, apply the three-prong test in Section 202(a)(11) and then the publisher's exclusion, naming which of the exclusion's three edges is closest to your activity. Where the two markets disagree, say so and say which answer binds you.

Then the UPSI and Regulation FD reasoning, in about thirty-five minutes. List every non-public conversation, call, meeting or channel check behind the memo, with a date. For each, run the two limbs of the unpublished price sensitive information test in writing, and say whether the source was the issuer or somebody acting on its behalf. Where it was, add the Regulation FD analysis from the issuer's side: was this a selective disclosure, was it intentional, and what deadline followed. If there were no such conversations, say so explicitly and say what your sources were instead, because a documented negative is a finding.

The conduct note takes about thirty minutes. One page, at most 500 words, answering four questions. Which regime binds this activity today, and under which rule. What would have to change for it to become registrable, naming the switch that would flip. What you would have to stop doing tomorrow if it flipped this evening. And one sentence, in the first person, stating plainly what you are not qualified to do. That last sentence is graded and it is not a formality; a note that cannot name a limit has not been written honestly.

Close with a verify pass, in about ten minutes. Go back through every rule you cited and confirm each against the primary source, recording the source and the date beside the citation. Anything you could not confirm is marked unverified rather than deleted, because an unverified citation you have flagged is useful and one you have quietly removed is not.

The rubric runs to 12 points, and the pass mark is 10 with no zero on any row. Self-score against the tests rather than against how the work reads.

Any row whose test fails scores zero for that row however well the memo is written. A conduct note that reads beautifully and cites a regulation number nobody checked is worth less than a scruffy one with a date beside every rule, because the first will be repeated with confidence and the second will not.

Reading & resources

Core, do these.

• CFA Institute, Asset Manager Code of Professional Conduct, current edition, free at rpc.cfainstitute.org. Short enough to read in one sitting. Read the six general principles, then Sections A to F, then the guidance on claiming compliance. [Free] [Beginner]
• CFA Institute, Standards of Practice Handbook, current edition. The individual-level rulebook the firm-level Code sits beside. Read the guidance on Standards II(A), VI(A) and VI(B) alongside Sections C and F here. [Free/Paid] [Intermediate]
• SEBI (Prohibition of Insider Trading) Regulations, 2015, as amended, at sebi.gov.in. Read the definitions in Regulation 2, Regulations 3, 4, 5, 7 and 9, and Schedules A, B and C. The schedules are where the operative codes of conduct live and they are routinely skipped. [Free] [Intermediate]
• SEBI (Research Analysts) Regulations, 2014, as amended, at sebi.gov.in. Read the definitions, the registration provisions, the limitations on trading, and the disclosure requirements, and read the current FAQ set alongside them, because the 2025 amendments changed the mechanics. [Free] [Intermediate]
• 17 CFR 243, Regulation FD, at sec.gov, with the SEC's adopting release. Twelve pages of rule and a great deal of useful commentary, including the treatment of mosaic theory. [Free] [Intermediate]
• SEC, Fact Sheet: Rule 10b5-1 Insider Trading Arrangements and Related Disclosure (Release 33-11138, 2022), at sec.gov. The clearest short statement of the cooling-off periods, the certification and the Item 408 disclosure. [Free] [Beginner]

Going deeper.

• Chiarella v. United States, 445 U.S. 222 (1980); Dirks v. SEC, 463 U.S. 646 (1983); United States v. O'Hagan, 521 U.S. 642 (1997); Salman v. United States, 580 U.S. 39 (2016). The four cases that built the American law of insider trading out of the word fraud. Read them in date order; each one is intelligible without a law degree. [Free] [Advanced]
• Lowe v. SEC, 472 U.S. 181 (1985). The publisher's exclusion, and the reason the American perimeter sits where it does. [Free] [Advanced]
• SEBI's circulars on the trading-window closure framework and the freezing of designated persons' permanent account numbers at security level, at sebi.gov.in, including the April 2025 extension to immediate relatives. Short, operational, and the clearest illustration of a rule being turned into a control. [Free] [Intermediate]
• SEBI's 2024 amendments to the intermediaries framework on association with unregistered persons, with the October 2024 circular and the January 2025 clarifications, at sebi.gov.in. [Free] [Intermediate]
• Investment Advisers Act of 1940, Section 202(a)(11), at sec.gov. Three prongs and five exclusions, in about a page. [Free] [Intermediate]
• The Financial Services Register (FCA) and SEBI's list of registered intermediaries. Both are searchable and both answer the only question that matters about a person offering advice, which is whether they are registered to give it. [Free] [Beginner]

One exercise beats all of the above. Pick one listed company you follow and reconstruct its last four quarters of trading-window closed periods from its own results filings. Then pull its disclosed insider transactions for the same period and see how many sit inside a window you computed. If any do, you have either mis-dated a results announcement or found something worth understanding, and both outcomes are worth the forty minutes.

Teach it back & journal

Teach it back this way. Explain to a friend who runs a 400-person stock-tips group, and who has just been offered ₹40,000 a month by a broker to add a referral link, exactly what has changed about his position. Use the three switches rather than the regulations: consideration, personalisation, discretion. Then explain the second problem, which is that the broker is now asking him to be the unregistered half of a prohibited association, and that the broker has a compliance department and he does not. Do not tell him what to do. Tell him what each option costs, including the one where he registers, and make sure he can repeat back to you the difference between what triggers the perimeter and what triggers the association rule. If he ends the conversation asking "how much of what I have already collected would I have to give back", you have taught the enforcement shape correctly, because disgorgement runs on gross receipts rather than on profits.

For the journal, take the most recent memo you wrote with a view in it and answer four questions in writing. Who saw it, and did any of them pay you anything, in money or in kind or in access? What were your sources, and was any of them the company or somebody acting for it? If one of them was, run the two limbs of the unpublished price sensitive information test on what you were told, in writing, and say honestly whether you ran that test at the time or are running it now for the first time. And what is the one thing you would have to stop doing tomorrow if a subscription button appeared under that memo tonight?

Then write a paragraph on something harder. The ethical decision framework asks you to name the situational influences acting on you, and the honest version of that exercise is uncomfortable, because the influences that work are the ones that do not feel like influences. Identify the person whose approval you would least like to lose, and the number, if there is one, that would be at stake in a decision you might have to make in the next year. Put both on paper. The purpose is not to resolve anything today; it is that a pressure you have named in advance is a pressure you will recognise when it arrives, and a pressure you have never named will simply feel like a good reason.

This module's flashcards and mastery quiz are wired into the app: see the node's Quiz and Reviews.