Questions to Ask a Chief Investment Officer
For trustees, analysts and clients sitting down with a chief investment officer at an endowment, foundation, pension scheme or asset manager. Twenty questions on what the portfolio is for, allocation and liquidity, total fees, how managers get hired and fired, and who can overrule the CIO.
The questions
Open any question for the note
What are you responsible for here, and what does the board decide instead of you?
Why ask it
The split between a CIO's discretion and the board's reserved powers decides whether you are talking to a decision maker or an adviser. It also tells you which of your remaining questions should be directed at someone else.
What is this portfolio actually for: what spending or liability does it have to fund?
Why ask it
A portfolio supporting a fixed spending rate, a pension promise or a family's income defines risk differently in each case. If the answer is framed purely as beating a benchmark, the mandate and the real purpose may have drifted apart.
What is the allocation today, and how far is it from your policy targets?
Why ask it
This is a factual question a CIO should answer from memory. Drift is not automatically wrong, but the reason for it is the useful part: a deliberate tilt, an illiquid holding they cannot trim, or nobody rebalancing.
What are you not allowed to invest in, and does that constraint cost you anything?
Why ask it
Exclusions come from statute, donor agreements, boards or beliefs, and each narrows the opportunity set. A CIO who can put a rough number on what a constraint costs has examined it. One who says the constraints make no difference has not measured.
How much of the book could you turn into cash inside a week?
Why ask it
Liquidity is where portfolios fail in a crisis rather than in a spreadsheet. You want a figure in days, with redemption terms and gates behind it. Treating a fund with quarterly windows as liquid is exactly what produces forced selling later.
What is the deepest drawdown this portfolio should expect, and what happens operationally when it arrives?
Why ask it
You are asking for a number with a plan attached. Someone who has run the scenario knows whether the spending policy still works at that level and who has to be told what. Answers about long horizons skip the operational half of the question.
What are you paying all in, including carry and any underlying fund fees?
Why ask it
Total cost is usually well above the headline management fee once carry, layered funds and transaction costs are counted. A reluctance to produce a single total, or a figure covering only the top layer, tells you how closely fees are being tracked.
Over the last five years, how much of the return came from allocation and how much from selection?
Why ask it
Attribution separates skill from luck. Most institutional returns come overwhelmingly from allocation, so a CIO who credits selection should be able to show the split. Vagueness usually means nobody has decomposed the performance at all.
How do you value the private holdings you cannot mark against a screen?
Why ask it
Private marks are stale and generally supplied by the manager who benefits from them. Listen for whether they adjust independently, lag public comparables, or simply accept what arrives. It also determines whether the reported allocation is real.
What has to be true for your approach to work, and what would tell you it is not?
Why ask it
Asking for the falsification condition is the quickest way to distinguish a considered strategy from a set of habits. If nothing could disprove it, weak results will always be attributed to the environment rather than the process.
Which position would be hardest to sell in a bad month?
Why ask it
The illiquid or emotionally protected holding is where the real risk sits. Sometimes it is a lock-up; sometimes it is a legacy position nobody wants to be the one to exit. Either matters more than the volatility figures in the report.
When did you last terminate a manager, and what actually triggered it?
Why ask it
A CIO who has never fired anyone has either a remarkable record or an unwillingness to act. The trigger is the informative part: departures of key staff, style drift or a broken process are better reasons than a run of poor numbers.
How do you tell a manager having a bad run from a manager who is broken?
Why ask it
Good answers describe criteria set in advance: the people, the process, asset growth, whether the losses came from where the manager said the risk was. Deciding after the fact reliably produces selling at the bottom, which is what most investors do.
What was the last thing you were wrong about, and how long did it take you to say so?
Why ask it
Watch how the answer is built. Naming a specific call, roughly when it was made and how long it was held is evidence of a working feedback loop. An answer about market conditions rather than a decision is a polished way of declining to answer.
What does the investment committee add, and what does it slow down?
Why ask it
Honest replies contain both: a check on impulse, and a two-month delay on anything time sensitive. A CIO who says the committee only adds value is managing you, and one who says it only obstructs has a governance problem you should know about.
Who can overrule you, and when did that last happen?
Why ask it
Formal authority and real authority often differ, particularly where a chair, a donor or a founder holds sway. The specific example matters, because the pattern of overrides shows which decisions are genuinely the CIO's to make.
How is your team paid, and what behaviour does that encourage?
Why ask it
Pay structures drive conduct more than stated philosophy does. Annual bonuses measured against a benchmark quietly reward hugging the index and chasing the last twelve months, whatever the stated horizon says.
Who on your team is allowed to tell you that you are wrong, and do they do it?
Why ask it
You are asking about dissent, which is what most often protects a portfolio from its own leader. Look for a named person and a real instance. A team where challenge is only described as welcome in principle usually has very little of it.
What do you do differently from most people in your seat?
Why ask it
This produces either something concrete about structure, horizon, fee negotiation or direct deals, or language that could come from anyone. A generic answer is not disqualifying, but it argues for judging the results against a cheap index alternative.
If we spoke again in three years and this had gone badly, what would the story be?
Why ask it
Asking for the failure narrative up front gets past optimism. Credible answers name the actual vulnerability: too concentrated, too illiquid, too dependent on one manager, or a spending rate the portfolio cannot sustain. Someone who cannot picture the bad version has not stress tested it.
Getting a useful hour with a CIO
Practical guidance for the conversation itself
Before the meeting
Read the policy statement, not just the performance page
The investment policy statement contains the targets, ranges, benchmarks and constraints the CIO is actually working against. Nearly every question above becomes sharper once you know what they have already committed to in writing.
Look up what is already disclosed
Endowments, foundations and public schemes publish more than people expect: annual reports, filings, sometimes manager lists. Asking about a number they have already published wastes the slot; asking why it moved does not.
Know which conversation this is
A trustee is testing governance and whether risk is understood. An analyst is testing process. A prospective client is testing whether the pitch survives contact with detail. The same twenty questions serve all three, but the follow-ups differ.
How to ask
- 1Ask for the number first and the reasoning second. Reasoning offered without figures is difficult to check and easy to produce.
- 2When you get a framework, ask what it has ruled out in practice. A risk limit that has never stopped anything is decoration.
- 3Follow every claim of skill with a request for attribution. It is a normal question and the answer is usually already prepared internally.
- 4Ask about the last time rather than the general policy. Policies are written to be shown; the last time is not.
- 5Let a pause sit after an answer about a loss or a termination. The second half of that answer tends to be the candid half.
- 6Ask who else you should speak to. A CIO who offers the risk lead and an analyst is confident the account holds up from below.
Answers worth weight, and answers to discount
Worth weight
- Figures given from memory, with the assumption behind each one stated.
- A decision they got wrong, described without repackaging it as a learning experience.
- Clear labelling of what is measured, what is estimated and what is a manager's own mark.
- Explicit trade-offs: what a given return target costs in liquidity or in drawdown risk.
To discount
- Long-term horizon used as the answer to a question about a specific loss.
- Risk discussed only as a governance process rather than as an exposure with a size.
- Performance credited to insight and shortfalls credited to the market.
- Fees described as competitive without a total being produced.
- Alternatives justified by diversification while the underlying exposure is plainly equity risk with a lock-up.
If you are a trustee
- Your job is to test whether the risk is understood and appropriate, not to have a better market view than the CIO.
- Ask for the same risk figures in the same format at every meeting. Changing presentation is how deterioration gets absorbed unnoticed.
- Insist on a liquidity schedule alongside the allocation, showing what can be realised in a week, a quarter and a year.
- Ask what the spending policy does in a bad year, and get the answer before the bad year.
- Ask once a year what the whole arrangement costs in cash terms, including consultants and custody, not only as a percentage.
- Note who does the talking. If the consultant answers the questions put to the CIO, that is worth raising separately.
If you are being pitched
A CIO presenting to prospective clients is selling, so the useful questions are the ones a pitch deck does not answer: what the total cost is, what the worst historical drawdown of this specific strategy was, how much of the assets belong to the largest client, and who exactly would manage your money if the person in front of you left. Ask for the track record of the strategy as run today rather than the composite, and ask what has been closed or written off along the way.