CAIA® valuation controls: who can challenge a manager’s mark?

On this page

A valuation policy is credible when it explains where each price comes from, who can challenge it, and what happens when evidence conflicts. For a hard-to-price asset, the decisive question is whether someone with genuine authority and some distance from portfolio management can test the mark. A third-party quote helps, but an unchecked override can defeat that safeguard.

Why the source of a price matters

A reported fund return depends on the values assigned to the assets at the measurement date. A frequently traded security may have a usable market quotation. A privately negotiated loan may instead require a model and assumptions about future cash flows, credit quality and a market participant’s required return. The second estimate can be reasonable without being uniquely observable. The candidate’s task is to assess the procedure that produces and reviews it, rather than treat any model output as a fact.

The SEC staff’s discussion of private assets explains why individually negotiated private credit often lacks readily available quoted prices and needs significant judgement. That observation does not establish a particular hedge fund’s value. It explains why evidence about inputs, timing, responsibility and challenge becomes more important as observable prices become scarcer.

Start with a position inventory, grouped by how it is valued. For each group ask whether the policy chooses a source in advance, whether the source is appropriate for the instrument, and how stale or conflicting observations are identified. A document that merely says assets are valued fairly leaves the most consequential operational choices unstated. A detailed procedure for liquid holdings does not fill a gap for an illiquid sleeve.

Follow the price through the process

Imagine a pricing vendor sends a quote directly to operations. Operations compares it with the policy’s other permitted evidence and records an exception when the difference exceeds the fund’s stated threshold. An authorised reviewer then resolves the exception before the administrator calculates net asset value. That sequence creates records an examiner or investor can inspect: the original observation, any challenge, the reason for the decision and the person who approved it.

Now imagine the portfolio manager sees the quote first and replaces it with a higher model value without leaving the original quote or an explanation. The final figure might still be defensible, but the process gives the person whose performance is being measured control over the measurement. The central defect is the missing challenge trail. Neither a vendor’s name nor the existence of a committee cures that defect by itself.

The SEC has highlighted methodology, override controls and prompt review by people outside portfolio management as issues for alternative fund valuation policies. Its framework for registered funds also stresses testing methodologies, evaluating pricing services, documenting decisions and separating duties. Those are useful control principles for this exam explanation; they are not a claim that every private fund is directly subject to the same registered-fund rule.

A worked valuation-control review

Consider an invented fund with an illustrative net asset value of £100 million. Exchange-traded holdings account for £55 million, broker-quoted instruments for £25 million, and privately negotiated loans for £20 million. The policy names an external feed for the first group and independent quotes received by operations for the second. For the loans it says only that the portfolio manager will provide a quarterly fair-value estimate. No reviewer, override record or escalation path is specified.

The arithmetic matters because the loans make up one fifth of the stated net asset value: £20 million divided by £100 million equals 20%. If the manager’s aggregate loan mark were £2 million too high in this simplified scenario, reported net asset value would also be £2 million too high before other adjustments, equivalent to 2% of the originally reported total. The example is invented to show exposure, not an estimate of real-world mispricing. It does not prove the marks are wrong.

What should the candidate conclude? The two more observable sleeves have named sourcing procedures, while the loan sleeve has concentrated judgement and no stated independent challenge. Ask for a documented loan valuation method, the underlying cash-flow and credit inputs, an independent check suited to the assets, and records of exceptions. Also ask who approves a departure from the selected source and whether that person can reject the manager’s proposal. The appropriate assessment is a material control gap requiring investigation, not an automatic calculation of a fair price from the information supplied.

A useful sensitivity check changes one assumption at a time. If a loan model is sensitive to recovery assumptions, compare the manager’s selected recovery with other supportable values and show how the position value moves. If a broker quote is stale, establish its date and whether an intervening credit event makes it unreliable. A control review becomes stronger when it tests the evidence that would have changed the mark, not merely whether a completed form exists.

Independence is a property of the decision

Independent pricing does not mean that the reviewer must always disagree with the manager. It means the reviewer receives or obtains relevant evidence without filtering by the portfolio manager, has a documented role in deciding exceptions, and can require a different value when the evidence warrants it. An external provider can still rely heavily on inputs supplied by the manager. Its engagement terms, assumptions and response to challenges therefore need scrutiny.

Ask to see the policy, a recent pricing file, an exception register and meeting records for disputed assets. Match the people named in the policy to the people who acted in the file. Was the price transmitted to operations or the administrator directly? Did a named reviewer examine a stale quote? If the manager changed a vendor price, was the original value retained, with reasons and approval? A record of a sensible challenge is stronger evidence of a working control than an organisational chart alone.

A committee can coordinate these decisions, but its title is not proof of independence. Examine whether its members report to portfolio management, whether they can obtain alternative evidence, and who has the final say. Minutes that record only approval of all supplied values tell less than minutes explaining a disagreement and its resolution. Equally, the absence of a past override is not proof of failure; the reviewer should test whether the process could actually challenge a mark when necessary.

Treat special situations separately

A side pocket can separate a difficult-to-sell asset from the fund’s ordinary dealing pool. The key valuation questions then extend to who decides to designate the asset, what value is used at the transfer, how it is reviewed while held, and how investors’ interests are treated under the fund’s own documents. Do not assume all side pockets have identical subscription or redemption terms. Read the specific terms before drawing a conclusion about who participates in a later gain or loss.

Similarly, an administrator’s involvement can mean several different things. It may calculate net asset value from manager-supplied marks, independently verify selected prices, or perform a broader pricing function. The label “independent administrator” does not reveal which role it performs. Ask what it prices itself, what exceptions it escalates, and whether the manager can overrule it. These questions focus on the decision rights behind the reported number.

How to use this in a CAIA answer

Build a short chain of reasoning: identify the hard-to-price exposure; identify the chosen evidence; identify who can alter the mark; then state what document would confirm the control operated. Separate valuation uncertainty from governance failure. A range of defensible model values is possible for an illiquid loan even under a strong process. Conversely, a plausible reported return cannot demonstrate that the pricing control was sound.

For revision, practise with invented small portfolios rather than memorising a list of job titles. Change one fact, such as direct delivery of broker quotes or an unrestricted manager override, and explain how the conclusion changes. The CAIA program page gives the relevant programme context; the methodology page explains Omni’s study approach. Keep the exam answer focused on evidence, authority and the specific exposure at risk.