Before Copying the Endowment Model: A Spending-First Review Protocol for Family Offices

Transfer the question, not the portfolio

A prestigious institution’s investment approach can become a shorthand for sophistication. Yet an endowment and a family office may have different obligations, contribution patterns, governance arrangements and capacities to withstand interruption. A useful research translation begins by asking which problem the published model solved and whether the receiving institution has the same problem. The presence of familiar asset-class labels is insufficient evidence of comparability.

Arnold, Earl, Farizo and North examine endowment asset allocations using monthly data from 1997–2023 and mean-variance optimization (10.1057/s41260-023-00346-9). The study reports variation in optimized allocations across subperiods and addresses reporting biases in hedge-fund data. Its findings are historical and conditional on data and assumptions. They cannot establish a future allocation or a guaranteed outcome for a family office.

Define obligations and protected decisions

The proposed protocol begins with a statement of what the assets must support, who may revise those purposes and which obligations are contractual rather than discretionary. The family and its qualified advisers should identify relevant entity boundaries and constraints. This exercise deliberately precedes discussion of attractive asset categories. It creates an accountable basis for asking whether a proposed strategy can serve the institution’s purposes.

A spending record distinguishes timing, uncertainty and authority. A recurring commitment, an anticipated gift and a contingent obligation should retain their different evidentiary status. The protocol does not supply a standard spending rate. It requires the institution to preserve the reason for its own assumptions and to make a change in those assumptions visible to the people authorized to respond.

Expose the model’s conditions

A proposal that relies on optimization should disclose the observation period, asset proxies, estimation method, constraints and treatment of incomplete data. Reviewers should be able to explain why the selected inputs are relevant and how the conclusion changes under reasonable alternatives. A single efficient-frontier illustration is not a substitute for that record. Historical estimates should remain labeled as historical estimates.

The study’s discussion of hedge-fund reporting biases is particularly useful as a discipline for interpretation. The committee should ask whether its data exclude failed or nonreporting vehicles, whether an index resembles an accessible investment and whether returns are comparable after relevant costs and restrictions. These are diligence questions prompted by the research. Resolving them requires evidence about the specific proposal, not a generic adjustment copied from the article.

Test feasibility separately from attractiveness

A feasibility review asks whether the institution could perform its obligations under adverse but plausible conditions before debating comparative attractiveness. The office should document which resources can actually be accessed, the conditions that limit access and the people required to act. A hypothetical sale or distribution should not be treated as certain simply because a planning model includes it. Each scenario should retain the supporting evidence and the unresolved uncertainty.

The committee can then compare proposals using the same set of institutionally approved conditions. Its decision minute should state which model outputs informed the decision, which considerations came from professional judgment and which uncertainties remained. If a model’s favored portfolio is infeasible under the family’s approved constraints, the constraint should be addressed explicitly. It should not disappear through an undocumented change to the spreadsheet.

Create an empirical test of the protocol

To evaluate the protocol, researchers could give independent reviewers historical or synthetic cases with the same investment data but different spending and authority conditions. Reviewers would use either a conventional presentation or the spending-first record. Outcomes could include recognition of infeasible assumptions, clarity about authority, consistency of recorded reasons and preparation burden. The evaluation should be specified before results are collected.

Improved documentation would not by itself establish superior investment performance. The immediate test is whether decision makers can explain the relationship between obligations, evidence and judgment. If the protocol encourages ritual completion while hiding uncertainty, it fails its purpose. A family office should be able to learn from endowment research while retaining a clear account of why its own institutional responsibilities require a particular decision.

Source, contribution and limits

Source: Endowment asset allocations: insights and strategies. Tom Arnold; John H. Earl; Joseph Farizo; David North. Journal of Asset Management. 2024-05-20. https://doi.org/10.1057/s41260-023-00346-9. The original work is licensed CC BY 4.0. Original authors remain responsible for their original contribution; they are not represented as authors or endorsers of this companion.

Educational process guidance only. No recommended allocation, spending rate, security, fund or expected return is provided; the proposed protocol has not been evaluated.

This is an AI-assisted editorial research analysis prepared for the Family Office Private Wealth Journal at Dr. Freedom Cheteni’s direction. Its operational proposals have not been tested in a new empirical study. It is not an independently peer-reviewed research article. The named editorial authority retains responsibility for decisions and corrections.

Implementation record