Strategic asset allocation for family offices

Why institutional models are the wrong starting point

The endowment model, as popularised by Yale and Harvard in the 1990s and 2000s, offered a seductive prescription: reduce listed equities and fixed income, rotate toward private equity, real assets, and hedge funds, and harvest illiquidity and complexity premia over long horizons. Pension funds followed a different but equally systematic path, anchoring allocations to liability-matching fixed income and targeting actuarially determined return thresholds. Both frameworks were built for specific institutional mandates: tax-exempt status, perpetual or quasi-perpetual horizons, governance by committee, and no single controlling beneficiary.

Family offices share some of these characteristics and none of them completely. A single-family office (SFO) managing assets for a founder who sold a technology business in their mid-fifties faces a radically different constraint set: significant embedded capital gains in rollover equity, an immediate income need to fund lifestyle and philanthropy, children with divergent risk appetites and time horizons, and a tax domicile that makes every allocation decision a post-tax calculation. Applying the Yale model to this family produces the wrong answer at nearly every margin.

The efficient frontier shifts materially once you introduce taxes, concentration, and a named beneficiary. Optimising for pre-tax Sharpe ratios is the right answer to the wrong question.

The after-tax objective function

The single most consequential adjustment a family office can make to its allocation framework is to replace pre-tax expected return with after-tax expected return as the primary objective. This sounds obvious, yet most commercially produced asset allocation studies, manager pitchbooks, and asset-class return forecasts present pre-tax figures. A family domiciled in a high-tax jurisdiction such as California, the United Kingdom, or France faces marginal rates on income and short-term gains that can reach 45-54%, depending on the composition of returns. Even a moderate 150 basis point difference in annual tax drag, compounded over 30 years, translates to a final portfolio value roughly 36% lower, all else equal.

Tax efficiency therefore becomes a first-order allocation input, not an afterthought. This has direct implications for asset-class selection within each bucket. In the equity sleeve, low-turnover index strategies and direct indexing structures that allow systematic loss harvesting can meaningfully reduce realised gain recognition relative to actively managed funds with high portfolio turnover. In fixed income, municipal bonds may be preferable to equivalent-yielding taxable bonds for families in the highest brackets, even though their headline yield is lower. In alternatives, the difference between an evergreen vehicle and a traditional drawdown fund involves not just liquidity mechanics but also the timing and character of gain recognition.

Locating assets across entities and jurisdictions

Beyond asset-class selection, asset location, meaning the deliberate placement of different return streams within different holding entities, is a powerful lever. Interest-generating fixed income held inside a tax-deferred retirement account or a life insurance wrapper compounds without annual friction. High-growth equity with minimal current income can sit in taxable accounts where unrealised appreciation builds without immediate cost. Philanthropic capital destined for charitable giving is often most efficiently held in a donor-advised fund or private foundation, where appreciated assets can be contributed and the embedded gain extinguished. Each of these location decisions requires a clear view of the family's overall entity structure, a task complicated by FATCA and CRS disclosure requirements that now make offshore opacity both legally and reputationally untenable.

Concentrated positions as a distinct asset class

Among UHNW families, concentrated positions, typically in a founder's operating business, a legacy real estate portfolio, or inherited stock in a single company, frequently represent 40-70% of gross wealth at the outset of a family office mandate. Standard mean-variance optimisation ignores this reality entirely; it assumes a clean slate and diversified starting position. In practice, the concentrated position is the dominant risk factor, and every other allocation decision must be understood relative to it.

The appropriate framework treats the concentrated holding as a separate sleeve with its own return expectation, volatility estimate, correlation to other assets, and disposition timeline. A private business generating 18-20% returns on equity but carrying single-company operating risk is not equivalent to a diversified private equity fund with similar headline returns. The idiosyncratic risk is far higher, the liquidity is essentially zero, and the family's human capital is typically also tied to the same entity, compounding the concentration. Advisors who present a diversified portfolio alongside a concentrated business as though the two are independent are understating total family risk by a large margin.

Disposition strategies for concentrated positions must balance tax efficiency, family control preferences, and market timing. Structured equity collars, exchange funds, and instalment sales are among the tools used to manage this exposure, each with distinct tax and liquidity trade-offs. The key governance principle is that the disposition strategy should be encoded in the investment policy statement (IPS) with explicit targets: reduce the concentrated holding to below 20% of net worth within seven years, for example. Without that specificity, the position persists indefinitely, driven by inertia and tax aversion.

Liquidity as a strategic allocation, not a residual

Institutional investors often treat liquidity as a buffer, a cash or near-cash sleeve that absorbs unexpected outflows. Family offices need a more deliberate framework, because their liquidity demands are more varied and less predictable than those of a pension fund with actuarially stable payments.

A proper liquidity architecture begins with a liability schedule. This document maps expected cash outflows across three time horizons: operating needs over the next 12 months (family living expenses, payroll, tax payments, charitable commitments), medium-term obligations over the next one to five years (real estate transactions, private capital commitments, potential business acquisitions or estate-planning transactions), and long-term contingent obligations beyond five years (estate taxes at the death of the first and second generation, large philanthropic pledges, potential liquidity events for family members). Each category carries a different probability weight and calls for a different liquidity instrument.

Liquidity planning is not about holding enough cash. It is about matching the duration and reliability of liquid instruments to the timing and certainty of known obligations.

The illiquidity premium available in private equity, private credit, and infrastructure is real, but it is estimated in many studies at roughly 150-300 basis points annually over comparable public market equivalents, net of fees. Capturing that premium requires the family to commit capital it will not need for seven to twelve years. Families that over-allocate to illiquid strategies without stress-testing their liability schedule often find themselves unable to meet capital calls in a downturn, forced to sell public assets at depressed prices, or locked out of opportunistic investments precisely when they are most attractive.

Generational time horizons and their implications

One genuine structural advantage family offices hold over many institutional investors is the potential for a multigenerational time horizon. A first-generation founder in their fifties, together with their children and grandchildren, may have an effective investment horizon of 60-80 years. This is longer than most pension fund liabilities and comparable only to endowments and sovereign wealth funds. In theory, it justifies a higher structural allocation to illiquid assets and equities relative to the conventional wisdom for individual investors.

The qualification is governance. A long time horizon is only useful if the family has the investment policy discipline to hold through volatility, the legal structure to prevent premature liquidation driven by family disputes or estate processes, and the next generation's genuine alignment with the strategy. A dynasty trust in a jurisdiction such as South Dakota, Nevada, or Jersey can provide structural continuity, insulating long-duration assets from forced-liquidation events. However, legal structure without family governance, meaning regular family council meetings, documented investment principles, and next-generation financial education, is insufficient. Many family offices find that the human governance problem is harder to solve than the technical asset allocation problem.

Building a family-specific investment policy statement

The investment policy statement is the document that translates family circumstances into portfolio guardrails. For a family office, it must go further than the standard IPS used in institutional mandates. At a minimum, it should specify: the net-worth return target expressed on an after-tax, after-fee, after-inflation basis; the maximum permitted allocation to a single name, sector, or illiquid strategy; the liquidity reserve floor; the family's tax domicile and any relevant regulatory obligations under FATCA, CRS, or AIFMD; the criteria for rebalancing and for reviewing the concentrated position; and the governance process for overriding strategic allocations.

The IPS should also address BEPS Pillar Two implications where relevant. Families with operating businesses or investment holding structures in multiple jurisdictions now face a 15% global minimum tax on profits in low-tax entities. This changes the calculus for certain offshore holding structures that were previously used partly for tax deferral on investment returns. Advisors who have not revisited their clients' holding-vehicle architecture since Pillar Two came into force in major jurisdictions from 2024 onward are leaving a material risk unaddressed.

Rebalancing discipline and behavioural constraints

Rebalancing in a family office context is more complicated than in an institutional setting, primarily because realising gains to rebalance triggers immediate tax events. A threshold-based rebalancing policy, where rebalancing is triggered only when an asset class drifts more than a defined percentage from its target, reduces unnecessary turnover relative to calendar-based approaches. More sophisticated approaches use new capital contributions, trust distributions, or charitable transfers to rebalance without realising embedded gains. The IPS should specify which method takes priority and who has decision-making authority when family members disagree about the timing or method.

Behavioural discipline is a separate but related challenge. Research across institutional and family investor populations consistently shows that rebalancing is abandoned most often precisely when it is most necessary, during sharp drawdowns when the instinct to reduce equity exposure is strongest. A governance structure that separates the investment committee (which sets policy) from the family principal (who may be psychologically exposed to market volatility) provides a useful check. The investment policy statement, once ratified, should require a supermajority of the family council to amend during a period of market stress, creating a structural cooling-off mechanism.

Translating principles into a practical allocation framework

Assembling these considerations into a working allocation framework requires mapping the family's specific circumstances against five dimensions: tax status and domicile, concentration and disposition timeline, liquidity liability schedule, generational horizon and governance quality, and return objectives on an after-tax basis. Each dimension will push the efficient frontier in a different direction for different families. A family with a fully diversified liquid balance sheet, a 60-year time horizon, robust governance, and a low-tax domicile can reasonably allocate 40-50% to illiquid strategies and should accept meaningful equity volatility. A family still holding a 60% concentrated position, with near-term estate tax obligations and a contested succession, needs a far more liquid and conservative posture until those constraints are resolved.

The practical output is not a single model portfolio but a tiered architecture: a liquidity sleeve sized precisely to the liability schedule, a diversified core of public equity and investment-grade fixed income that can be rebalanced without illiquidity constraints, and an opportunistic long-duration sleeve for private assets calibrated to the surplus capital the family can genuinely lock up. Within each sleeve, after-tax return is the sorting criterion. Fees matter in direct proportion to time horizon: at 200-300 basis points of annual fees across two layers of a fund-of-funds structure, a family paying those fees over 30 years surrenders a compounding cost that a well-run SFO with a senior investment team costing 30-40 basis points on assets can avoid, freeing that differential to compound in the family's favour. The goal is not to replicate what Yale built. It is to build what this family, with its particular history, obligations, and ambitions, actually needs.