CAIA® integrated portfolios: who should make each decision?

On this page

In an integrated portfolio, give each investment decision to a body that can make it competently and in time, while leaving the board able to set objectives and check results. The candidate’s task is to match authority, expertise, reporting and escalation to the decision at hand. A sound allocation proposal can still be poorly governed if nobody can say who may approve it or challenge it.

Start with the decision, not the organisation chart

Investment governance is the set of rules that turns a portfolio objective into authorised action. A board, investment committee, internal team and external manager can all contribute, but their roles are different. CAIA Association’s discussion of institutional portfolio objectives describes a total portfolio approach in which the board retains responsibility for defining the objective while investment responsibility is largely delegated to the team. That is a useful model for analysis, not a universal prescription for every asset owner.

First identify the actual choice. Setting a long-term risk objective, permitting a new illiquid strategy, selecting a specific manager, executing a rebalance and accepting a valuation are separate decisions. They need not sit at the same level. A board may approve the portfolio’s risk limits, a committee may permit an alternatives sleeve, staff may select managers within a mandate and an external agent may implement a trade. The precise split must be in the applicable policy and delegation documents. A job title alone does not establish authority.

The first diagnostic question is whether the stated portfolio objective can be implemented with the organisation’s real decision timetable. Private investments can require decisions on capital calls, follow-on commitments, liquidity and manager changes at times that do not match a committee calendar. Public-market adjustments can be faster still. A policy that demands prior approval for every operational move may preserve formal control yet leave the institution unable to act within its own agreed limits. Speed matters only when accompanied by records and a clear route for exceptions.

Map authority, information and accountability together

A practical governance map records who proposes, who checks, who approves, who executes, who receives the result and who can stop or reverse an action. Candidates should look for a broken link in that chain. A committee cannot assess a manager decision if it receives only a final performance number months later. A staff member cannot be held fairly responsible for an allocation the board repeatedly overrides. An external agent cannot be monitored on a mandate whose boundaries were never written down.

Authority: Identify the document and limit that allow each decision, including any amount, concentration or liquidity threshold.
Information: Specify what the approver sees before acting and what the overseer receives afterwards; delayed information can make oversight ceremonial.
Accountability: Define the measure and review interval against which the decision maker is assessed, and identify who handles exceptions.

CAIA Association’s implementation discussion explains that the level of detail in an alternatives mandate affects the balance between allocator control and manager discretion. It also stresses monitoring, feedback and alignment with the investor’s overall goals. In an exam case, that means the candidate should ask whether constraints are detailed enough to prevent an unintended exposure but broad enough to let the delegated team use its expertise. A long list of limits can be as unhelpful as no limits if it leaves no workable decision space.

An illustrative governance case

The following fictional example uses invented round numbers only to show the decision process. An endowment has a £400 million portfolio, including a £40 million private-markets allocation authorised by its board. Its policy permits commitments in that sleeve, but the committee has reserved approval of each individual commitment. The committee meets every three months. A manager asks for a £4 million commitment within the existing sleeve and requires an answer before the next scheduled meeting. The internal team has completed its diligence. A quick calculation shows that the proposal is one tenth of the private-markets allocation and one hundredth of the total portfolio. Those ratios describe size; they do not prove that the proposal is suitable.

The authority problem is clear: staff have information and time to decide, but the written policy gives the committee the decision. Staff must not treat a favourable diligence report as implied approval. The committee could convene under an authorised process, or the board or committee could revise the delegation for future commitments. A revised mandate might let staff approve qualifying commitments within a specified size and concentration cap, subject to documented diligence, conflict checks and prompt committee reporting. Larger or unusual proposals would be escalated. The example does not assert that any particular cap is an industry standard.

Suppose the same fictional team later proposes a £7 million follow-on commitment that would push exposure to one manager beyond its written concentration limit. Even if staff had delegated authority for ordinary commitments, this is now an exception. The correct path is to escalate before execution, disclose the effect on portfolio concentration and liquidity, and seek the approval required by the mandate. The issue is not simply that £7 million exceeds £4 million; it is that a stated boundary would be crossed.

What delegation changes, and what it does not

Delegation can improve timeliness and bring specialist skill closer to the decision. It can also add a fee layer, reduce direct visibility and make it harder to tell whether poor outcomes came from policy, selection or implementation. Retaining every decision centrally may look cautious but can create bottlenecks and leave a small board making specialist judgments without the relevant evidence. A case answer should identify the particular trade-off instead of ranking models from best to worst.

An outsourced investment function illustrates the distinction. The owner can retain the portfolio objective, risk budget and reporting requirements while giving an agent discretion within an investment mandate. That agent’s discretion is real, but it does not remove the owner’s need to monitor mandate compliance and evaluate performance. Conversely, hiring an internal team does not automatically produce oversight. The board still needs clear exception reports and a way to challenge valuations and conflicts.

For alternatives, add the lifecycle of the investment to the governance map. A private fund commitment can lead to later capital calls and distribution decisions, while reported valuations may arrive with a lag. Ask who monitors unfunded commitments and liquidity, who reviews marks and who receives information about a breach. A manager-selected valuation without independent challenge may be a control weakness. The response should propose a review channel and evidence, not merely say that valuation should be “independent”.

How to answer a CAIA case under exam pressure

Read the objective and constraints first. Then mark every decision verb: sets, approves, recommends, selects, executes, reports and reviews. If a case says the board set a policy but the CIO acted outside it, state the conflict plainly. If the committee approved a strategy but cannot meet before a time-sensitive decision, describe the bottleneck without assuming staff already have authority. If an external agent makes decisions, trace what the owner still controls and what information it receives. This order prevents a generic essay about good governance.

A strong recommendation is conditional. For example, “delegate routine manager selection within the agreed sleeve, require a written diligence record and conflict assessment, and escalate proposals that breach concentration or liquidity limits.” The reason is that routine decisions then fit the available timetable while the owner retains control over exceptions. The candidate should also explain how the board will learn whether the arrangement works: timely reporting, breach logs, exposure summaries and periodic review of the mandate.

Do not substitute a cost calculation for a governance answer. An external manager’s asset-based fee can be compared with the fixed cost of staff, but the cheaper arrangement may still lack needed expertise, speed or transparency. Equally, more delegation is not automatically a control failure. The failure is an undocumented or unmonitored transfer of decision power. CAIA Association’s discussion of portfolio objectives helps separate strategic ownership from implementation authority.

For further preparation, the CAIA programme overview provides the curriculum context, while Omni’s methodology explains how its learning approach is organised. Use the governance case as a repeatable method: identify a decision, locate its written authority, test the information flow and propose a bounded correction. That is more useful than memorising a hierarchy of titles.