An investment policy statement is a document the family writes in calm markets and opens in a crisis. While markets rise it looks like paperwork: why commit to paper what everyone at the table already agrees on. Its value shows up the moment the principal calls the office asking "are we selling everything?" — and a pre-written answer preserves more capital than a decision taken in that same minute. The size of that effect has not been measured directly: the quantitative estimates in the literature cover the contribution of an adviser in general, not the presence of an IPS as such.
An investment policy statement (IPS) is a set of decisions taken in advance. CFA Institute describes its function plainly: the document supplies "an objective course of action" during market disruption, "when emotional or instinctive responses might otherwise motivate less prudent actions." Its second purpose is allocative in the legal sense: the IPS draws the line between what the family owns as a decision and what the family office or an external manager owns as an execution duty. Without that line every loss becomes a personal conflict and every gain becomes grounds to renegotiate the mandate.
The gap is wide. In the UBS Global Family Office Report 2026, 60% of family offices run an investment committee, fewer than half have a formal governance framework with board-level oversight, and only 28% maintain risk management processes outside the investment function. The sample averages roughly USD 1.3 billion under management — families with ample resources to formalise governance.
Anatomy: what a working IPS actually contains
CFA Institute organises the document around four pillars: scope and purpose (which assets it governs at all), governance (who sets policy, who executes, who monitors), objectives and risk, and risk management. Translated into family terms:
A hierarchy of objectives. The first is almost always preservation of purchasing power after tax, costs and inflation. The second is the spending policy: how much the family may withdraw each year and under what formula. The endowment canon is useful here — Yale applies a 5.25% target spending rate with a smoothing rule that blends 80% of the prior year's spending, adjusted for inflation, with 20% of the target rate applied to the endowment's year-end market value from two years earlier; the result is further bounded by a 6.5% cap and a 4.0% floor. The purpose is intergenerational equity and, equally, insulating the household budget from following quotations up and down. The third objective is liquidity: how many years of spending are covered by assets sellable within a week.
Strategic allocation and ranges. Target weights per asset class plus permitted corridors. A corridor of plus or minus fifteen percentage points constrains nothing — it merely authorises any decision after the fact.
Rebalancing policy. This matters more than manager selection and is the most under-written line in most documents. Vanguard's research states the objective without ambiguity: rebalancing exists to minimise risk relative to the target allocation, not to maximise returns, with annual or semiannual monitoring against roughly 5% thresholds as a workable balance between risk control and cost. Calendar rules are easier to administer; threshold rules catch stress more precisely. Pick one and write it down before the portfolio drifts.
Constraints. Concentration in the legacy asset — the family's operating business — is usually the portfolio's dominant risk and almost never appears in the allocation table. Sector and ethical exclusions, a leverage prohibition or hard limit, and derivative restrictions belong in the same section.
Currency policy. A family's base currency follows from the structure of its future liabilities. The hedging policy follows from that, and must be written down in advance.
An illiquidity budget. The distinction that matters is between capital commitments and actual capital calls. Work by Dimmock, Wang and Yang (NBER) shows what this distinction costs: in a crisis illiquidity rises, holdings deviate materially from target weights, and capital calls arrive precisely when secondary-market discounts are widest — so the call is met with borrowed money. The IPS should cap unfunded commitments and size the reserve behind them. For scale: in the UBS 2025 report alternatives accounted for 44% of the average actual allocation (2024 data), with private equity at 21%. The 2026 report puts alternatives lower, at roughly 42% — a figure reported in secondary coverage, since UBS's public materials do not carry the allocation table.
Benchmarks. An honest benchmark is one the manager can both beat and lose to: a blended index reflecting the actual allocation, plus a separate absolute reference of "inflation plus X" tied to the purchasing-power objective.
Valuation of illiquids. For positions in funds and direct deals, a reference to the IPEV Valuation Guidelines (current edition December 2025) removes the argument about whose number is right: the fair-value methodology is fixed in advance.
Reporting. What, to whom, how often, and who signs it. The principal's report and the committee's report are different documents with different granularity.
| IPS section | Question it closes | Typical failure |
|---|---|---|
| Objectives and spending policy | How much can be spent without eroding capital | The objective is stated as "10% return" |
| Allocation and ranges | Where the risk boundaries run | Ranges so wide they forbid nothing |
| Rebalancing | Who must sell the winners, and when | No rule exists at all |
| Liquidity and commitments | What pays capital calls in a bad year | Calls budgeted, unfunded commitments ignored |
| Constraints | What is off limits even by unanimity | The family business is not treated as a position |
| Benchmarks and valuation | How to separate skill from market | The benchmark is chosen after the fact |
| Authority and reporting | Who may decide and to whom they answer | Delegation is oral |
The investment committee owns policy
The standard failure mode is a committee that spends quarter after quarter reviewing individual transactions. Deal review belongs to execution. The committee owns policy: it approves the IPS and amendments to it, authorises departures from the ranges, appoints and dismisses managers, and accepts reporting.
A workable composition: the principal (or two representatives of different generations), an independent member with investment experience and no economic tie to the service providers, and the office CIO or an external adviser. Quorum and frequency are fixed in writing — at least four meetings a year, plus a procedure for convening an extraordinary session when a rebalancing threshold is breached. Every meeting's minutes record the alternatives considered, the reasoning, the votes and any dissent — a bare "discussed and agreed" falls short.
Minutes are fiduciary insurance. The prudent investor rule assesses the decision process — compliance "is determined in light of the facts and circumstances existing at the time of a trustee's decision or action and not by hindsight." To any future adversary, an undocumented process is indistinguishable from no process. Conflict disclosure is a standing agenda item: ERISA requires fiduciaries to avoid conflicts of interest and not to transact in the interest of related parties, and that logic transfers cleanly to a private committee.
Where the IPS meets the family charter
A family charter speaks about values, roles, entry of heirs and succession. The IPS speaks about money. They meet at two points: who may amend the investment policy (typically a qualified majority of the committee plus consent of the principal or family council), and how a generational disagreement about risk is resolved. The second is settled by architecture: different horizons go into different pools with their own IPS documents instead of one compromise portfolio.
The fiduciary layer: why trusts and foundations need it more
Once assets sit in a trust, the IPS stops being an internal document and becomes evidence. The Uniform Prudent Investor Act requires each investment to be evaluated "in the context of the trust portfolio as a whole and as a part of an overall investment strategy having risk and return objectives reasonably suited to the trust"; it mandates diversification unless special circumstances make non-diversification better suited to the trust's purposes; and it permits delegation provided the trustee exercises reasonable care in selecting the agent, establishing the scope and terms of the delegation, and periodically reviewing the agent's actions. ERISA states a parallel duty — diversify so as to minimise the risk of large losses, and follow the governing documents.
The practical consequence: a trustee is judged against a standard of conduct — the process by which the decision was reached. An approved IPS both protects him (he followed the policy) and binds him (he cannot depart from it silently). This is why a concentrated holding in the family business must be expressly permitted in the IPS or the trust deed — otherwise it reads as a breach of the duty to diversify. A letter of wishes is no substitute: it is non-binding and expresses preference, whereas the IPS sets testable parameters. Any power of a protector to approve changes to investment policy is drafted separately.
Selecting and replacing managers
Investment due diligence and operational due diligence are two different processes, and the second cannot be delegated to whoever is selling the first. Capco's 2003 study of more than one hundred failed hedge funds found that 54% had identifiable operational issues, and that half of all failures were attributable to operational risk alone. ODD examines the administrator, custodian and auditor, valuation policy, segregation of duties, key-person exposure and the cyber perimeter — everything that is not the investment idea.
Replacement criteria are written in advance and expressed in process terms: departure of a key person, strategy drift without notification, a change of auditor or administrator, deviation from the stated mandate, refusal to report in ILPA format. The ILPA Principles rest on three tenets — alignment of interest, governance and transparency — and are backed by templates (Reporting Template, Capital Call & Distribution Template, DDQ) that an IPS can cite directly as the required reporting standard. For a mandate handed to an external asset manager, those references are the only way to make manager comparison meaningful.
Recurring failures
The IPS was drafted by a consultant and never read by the family, so at the first shock it does nothing: the signatory does not remember what he agreed to. Ranges are so wide they constrain nothing. There is no rebalancing rule, so "do we sell the winners" is relitigated every time. Concentration in the family business is not treated as a portfolio position. And the most expensive omission: no procedure for death or incapacity of the principal — who may sign instructions in that window, who convenes the committee, what happens to outstanding commitments.
What must be signed before the first trade
The minimum set: the IPS itself, with objectives, allocation, ranges, a rebalancing rule and currency policy; the committee's terms of reference, with composition, quorum and an authority matrix; a cap on unfunded commitments; a valuation policy for illiquids referencing IPEV; reporting format and frequency; and a procedure for incapacity of the principal. All of it fits into twelve to fifteen pages if the text sticks to mechanics.
Q/A
Which sections make an IPS operational rather than aspirational?
Record its scope, objectives and acceptable risk, liquidity and horizon, strategic allocation and ranges, constraints, rebalancing rule, benchmarks, roles and delegations, reporting and review procedure. Separate capital pools with different liabilities may need separate policy statements.
Who may amend the IPS or approve a departure from it?
The authority matrix should state who proposes a change, who checks conflicts, which body votes, the required quorum and who records the reasons. A manager should not expand its own mandate; the governing instrument and applicable law take priority over the IPS.
How should a rebalancing rule be written so that it works in a crisis?
Specify target weights and bands, monitoring frequency, the exact trigger, the responsible executor, permitted liquidity sources and an exception process. The rule is designed to restore risk to the approved allocation; it does not promise higher returns and should not be invented after markets move.
Does the prudent-investor rule always require immediate diversification?
There is no universal answer without the governing law and trust instrument. The model UPIA assesses each investment in the context of the whole portfolio and generally requires diversification unless special circumstances make concentration better suited to the trust’s purposes. Any exception needs contemporaneous reasons.
What should the IPS provide for the principal’s death or incapacity?
Name the temporary signatory, the body that can convene urgently, authority to meet capital calls and maintain hedges, limits on new discretionary risk and the route back to the ordinary mandate. These powers must align with the trust, corporate documents and powers of attorney.