Curriculum Part 1 Chapters 1–6

Asset Allocation & Institutional Investors

The chapter-depth version of Topics 4 and 6: the allocation process itself, tactical allocation and its costs, risk budgeting and parity, factor investing, and each owner type in the detail the textbook gives it.

The asset allocation process

Why it dominates. Classic studies of pension plans found that strategic asset allocation explains roughly 91–94% of the variance of fund returns, with security selection and market timing accounting for the small remainder. Note the precise claim — it is variance explained, not the level of return, and the distinction is examinable. For endowments the picture differs: research on endowment returns attributes around 74% to strategic allocation, with roughly 15% from market timing and 8% from security selection, i.e. a much larger role for dynamic allocation and manager selection.

Five steps. (1) Identify the asset owner's objectives and constraints; (2) form capital market assumptions — expected returns, volatilities and correlations; (3) determine the strategic asset allocation; (4) implement it through vehicles, managers and rebalancing rules; (5) monitor, review and rebalance, feeding results back into the assumptions.

Objectives divide into return objectives (required versus desired, stated in real or nominal terms and after fees) and risk objectives (ability versus willingness to bear risk — where ability is a function of horizon, liquidity needs and the size of the asset base relative to the liability). Constraints are the standard five: liquidity, legal and regulatory, tax, time horizon, and unique circumstances.

Implementation decisions the curriculum expects you to weigh: active versus passive within each class, internal versus external management, the number of managers, the vehicle (commingled fund, separate account, co-investment), currency policy, and the rebalancing rule — calendar versus threshold bands, with wider bands trading tracking error against transaction cost. For portfolios containing illiquid assets, rebalancing must be executed through the liquid sleeve, because drawdown funds cannot be traded.

Exam trap
Mean-variance optimisation is an input-sensitive exercise: expected returns matter far more than variances or correlations, and small changes in them produce extreme weights. The failure mode is error maximisation — the optimiser concentrates in whatever asset's return you happened to overestimate.

Tactical asset allocation & its costs

The fundamental law of active management frames the whole discussion: risk-adjusted value added rises with the manager's skill (the information coefficient), the square root of the number of independent decisions (breadth), and the transfer coefficient — how completely the manager can express views given constraints. Applied to TAA, the implication is that breadth is the scarce resource: an allocator making a handful of large asset-class calls a year needs an implausibly high hit rate to justify the activity.

Why TAA is expensive in alternatives specifically. Illiquid positions cannot be traded quickly or cheaply, so tactical views must usually be expressed synthetically. The textbook enumerates the costs of moving between managers:

Forgone loss carryforward
Redeeming from a manager who is below the high-water mark surrenders the right to earn back those losses free of incentive fees. The investor pays incentive fees to the new manager on gains that would have been fee-free at the old one — a real, quantifiable cost of firing a manager after a drawdown.
Dormant cash
Capital sitting uninvested between redemption and the new manager's subscription date earns nothing while the target exposure is missed.
Opportunity losses
Notice periods, lock-ups and gates mean the exit cannot be timed; the investor is out of the market on the market's schedule, not their own.
Slippage and market impact
Liquidating one portfolio and building another moves prices, and in capacity-constrained strategies the investor bears that cost directly.

Keys to a sound TAA process. The core requirement is a model that forecasts returns across asset classes out of sample. Three characteristics of sound model development: economically meaningful signals rather than data-mined ones; a parsimonious specification, since adding variables always improves in-sample fit; and stability, meaning the estimated relationship does not shift materially with small changes in the data. Fundamental models are usually linear regressions producing conditional expectations (for example, conditioning next year's equity return on today's dividend yield); technical models supplement them with price-based signals. In practice, TAA in an alternatives portfolio is best implemented with futures overlays rather than by moving capital between managers.

Extensions to mean-variance

Every extension answers a specific violation of the base model's assumptions, and the exam rewards naming the violation before the fix.

Higher moments
Returns are not normal. Incorporate skewness and kurtosis preferences, or optimise mean-CVaR instead of mean-variance — the latter is a linear program and therefore tractable, while mean-VaR is not convex.
Estimation error
Resampled efficiency averages optimal portfolios across simulated input draws; shrinkage pulls noisy estimates toward a structured target; constraints bound the damage crudely but effectively.
Black–Litterman
Reverse-optimise market weights to recover equilibrium implied returns, then blend in the investor's views weighted by stated confidence. Produces intuitive, diversified portfolios and makes the source of every tilt explicit.
Liquidity and horizon
Single-period models ignore the multi-period funding path. Model commitments and cash flows, not just weights.
Stale pricing
Unsmooth appraisal-based series before optimising, or the model will over-allocate to private assets.

Risk budgeting, risk parity & factor investing

Risk budgeting allocates a limited quantity of risk rather than capital. The machinery is marginal and component risk: a position's marginal contribution times its weight gives its component contribution, and components sum to total portfolio risk — which is precisely why component measures, not marginal ones, are used for budgeting and attribution. A capital-weighted 60/40 portfolio typically shows equities contributing around 90% of total risk, which is the observation that motivates the next idea.

Risk parity equalises risk contributions across assets. Because low-volatility assets then carry large capital weights, reaching a target return requires leverage — so risk parity embeds an assumption that leverage is available, cheap and stable, and its worst outcomes occur when funding costs rise and equity and bond risk premia compress together. The theoretical case rests on maximising diversification when Sharpe ratios are similar and correlations low; the practical case rests on avoiding the estimation of expected returns entirely, since the method needs only a covariance matrix.

Factor investing allocates to underlying risk drivers — equity, rates, credit, inflation, liquidity, and style factors such as value, momentum, carry and quality — rather than to asset-class labels. The argument is that asset classes are bundles of factors, so a portfolio that looks diversified by label may be concentrated by driver: private equity and public equity share the growth factor, and private credit and high yield share the credit factor. Implementation questions: are the factors investable and cheaply accessible, is the factor premium compensation for risk or a behavioural anomaly, and does the factor survive transaction costs and crowding? The honest caveat for essays is that factor definitions are not standardised, so two providers' "value" portfolios can differ substantially.

The endowment model

The model. A perpetual horizon, an equity-oriented portfolio, heavy allocation to illiquid alternatives, extensive use of external active managers, and a spending rule that smooths the operating budget. The intellectual case: a perpetual investor should be paid for bearing liquidity risk that shorter-horizon investors cannot, and inefficient private markets reward manager selection more than public markets do.

Six advantages that may explain large endowments' outperformance — and the reason the model does not transfer automatically: an aggressive asset allocation; effective investment manager research; a first-mover advantage into new asset classes; access to a network of talented alumni; genuine acceptance of liquidity risk; and sophisticated investment staff with capable board oversight. An institution lacking these should not expect the returns by copying the allocation alone.

Four risks of the model, each a standard essay prompt:

Spending and inflation
Tension between the spending rate, portfolio risk, and preserving real corpus. Higher spending demands higher return, which demands more risk — and the relevant inflation for a university (salaries, healthcare, construction) has historically run above headline CPI.
Liquidity
The 2008 experience: unfunded commitments plus a collapsing liquid book plus an operating budget dependent on distributions. Liquidity must be budgeted explicitly against the worst plausible call-and-distribution path.
Rebalancing and TAA
Aggressive contrarian rebalancing added real value at the largest endowments, but it is psychologically hard and constrained by illiquidity — you cannot rebalance into a crisis with capital you have already committed elsewhere.
Tail risk
A large drawdown when systemic risk spikes and correlations converge. Mitigations — cash reserves, lower equity beta, trend-following, explicit tail hedges — all carry a cost that must be compared per unit of drawdown avoided.

Spending rules. A simple rule spends a fixed percentage of a trailing average market value; a smoothing rule blends last year's spending inflated by the price level with the target rate applied to market value, weighted to trade budget stability against corpus preservation. Higher weight on prior spending gives a smoother budget and more drift in the real value of the corpus. US private foundations additionally face a statutory payout requirement of about 5% of assets, which converts a policy choice into a constraint.

Pension fund portfolio management

Defined benefit. The sponsor promises a formula-based benefit and bears investment and longevity risk. The objective is the surplus, so risk is the volatility of assets minus liabilities, not of assets alone — an all-equity portfolio and an all-cash portfolio can both be high surplus risk. Liability-driven investing splits the fund into a hedging portfolio matched on duration and inflation sensitivity and a return-seeking portfolio, with a de-risking glidepath that shifts weight to hedging assets as the funded ratio improves. Risk tolerance depends on the funded ratio, the sponsor's covenant and its correlation with the plan's assets, plan maturity (active versus retired members), and the regulatory and accounting discount-rate regime.

Defined contribution. No plan-level liability; the member bears investment, contribution and longevity risk, and the objective is an adequate replacement ratio. Design questions dominate: default fund and target-date glidepaths, the number of options offered (choice overload is a documented problem), fee levels under litigation scrutiny, and the operational barriers to including alternatives — daily valuation and liquidity, and member comprehension. Contrast the two plan types on who bears risk, portability, accounting treatment, and the ability to hold illiquid assets.

Governmental social security plans are typically pay-as-you-go rather than funded, so their solvency depends on demographics — the dependency ratio — rather than on investment returns, and reserve funds exist to smooth that transition. Annuities convert accumulated assets into lifetime income and are the natural decumulation answer to longevity risk: know the trade-offs between immediate and deferred, fixed and variable, and inflation-linked forms, and the reasons for the annuity puzzle — cost, loss of bequest and flexibility, credit exposure to the insurer, and behavioural reluctance to hand over a lump sum.

Sovereign wealth funds

Sources of sovereign wealth: commodity export revenues (principally hydrocarbons), persistent trade surpluses and accumulated foreign-exchange reserves, and fiscal surpluses or privatisation proceeds. The source shapes the risk: a commodity-funded SWF's inflows are correlated with commodity prices, so holding commodity-linked assets doubles the country's exposure.

Stabilization funds
Insulate the government budget and the economy from commodity price and revenue volatility. Short effective horizon, high liquidity, low risk — the portfolio looks like a reserve portfolio, because withdrawals arrive exactly when markets are weak.
Reserve (investment) funds
Seek higher returns on excess foreign-exchange reserves, reducing the carrying cost of holding reserves. Moderate risk, still liquidity-conscious.
Savings funds
Convert finite resource wealth into a permanent financial endowment for future generations. The longest horizon of any institutional investor, hence the largest allocations to private markets, direct investment and co-investment.
Development funds
Fund domestic economic and infrastructure development. The examinable tension: a domestic development mandate is by construction concentrated in the home economy, which is already the source of the wealth — so the fund is not diversifying the sovereign balance sheet.

Governance and political risk are the distinguishing risks. Political withdrawal risk means the money can be called on for fiscal purposes regardless of the investment horizon; recipient-country scrutiny can restrict access to strategic sectors; and transparency expectations are set by the Santiago Principles, a voluntary framework covering legal and institutional structure, a clear policy purpose, independence of investment decisions from short-term political interference, disclosure, and risk management. A well-governed SWF separates the owner (the state) from the investment decision-maker, publishes its objective and reports against it.

The family office model

What they are. Single-family offices serve one family and are effectively private investment and administrative firms; multi-family offices serve several and resemble a boutique wealth manager. Beyond investment they typically provide tax and estate planning, philanthropy, reporting and consolidation, concierge and lifestyle services, next-generation education, and family governance.

Goals by generation is a distinctive framing: the wealth creator's generation is often concentrated in the operating business with high risk tolerance and a growth objective; the second generation diversifies and professionalises, and objectives shift toward preservation and income; by the third and later generations the family is larger and more dispersed, objectives conflict across branches, and governance and liquidity for distributions dominate. Portfolio advice that ignores which generation is asking will be wrong.

Macroeconomic exposures
Family wealth usually carries a concentrated exposure — an operating business, a sector, a currency, a country. The investment portfolio should be constructed to offset that exposure, not to replicate it. This is a total-balance-sheet argument, and it is the most commonly examined family-office idea.
Income taxes
The family office is the one owner type whose objective function is genuinely after-tax return. Consequences: preference for deferral and long-term capital gains over ordinary income, sensitivity to turnover, tax-loss harvesting, entity and domicile structuring, and the fact that a strategy's pre-tax alpha can vanish after tax.
Lifestyle assets
Homes, aircraft, yachts, art and collectibles. They consume liquidity through carrying costs, generate little or no income, and are often illiquid and emotionally held — but they belong on the balance sheet and in the liquidity plan.
Governance
Family constitution or charter, a family council separate from the investment committee, defined decision rights, conflict-resolution and succession mechanisms, and education of the next generation. Governance failure, not investment failure, is the main documented cause of wealth dissipation across generations.
Philanthropy & impact
Charitable vehicles with their own payout and governance rules, and impact investing where the family's operating expertise and network can supply genuine additionality.

Ten competitive advantages, in the textbook's framing, cluster into four ideas worth being able to argue: an unusually long and flexible horizon with no external mandate or peer benchmark; speed and simplicity of decision-making (a small committee, no consultant chain); access to proprietary deal flow through the family's business network and reputation; and the freedom to be genuinely contrarian, concentrated and opportunistic — including in direct deals and control positions — because there is no career risk from tracking error.

Confusion pairs

Variance explained vs return explained
Strategic allocation explains ~90% of the variance of pension returns — not 90% of the return level. Endowment studies find a much lower figure (~74%) with bigger roles for timing and selection.
Ability vs willingness to bear risk
Objective capacity from horizon, liquidity and asset size vs the committee's stated tolerance. Where they conflict, the lower governs and the gap is addressed through education.
Risk budgeting vs risk parity
Allocating a chosen amount of risk across exposures vs the specific rule that equalises risk contributions and typically requires leverage.
Marginal vs component contribution
Sensitivity to a small change vs contribution × weight, which sums to total risk. Only the second is used for budgets.
Forgone loss carryforward
The incentive-fee cost of firing a manager below their high-water mark — a cost of the switch, not of the original loss.
Stabilization vs savings SWF
Short horizon, liquid, buffers the budget vs multi-generational, growth-seeking, heavy in private markets.
Single vs multi-family office
One family, fully bespoke, high fixed cost vs several families sharing infrastructure, closer to a boutique manager with its own conflicts.
Pre-tax vs after-tax alpha
The only owner where the second is the objective is the family office — high-turnover strategies can be additive pre-tax and destructive after tax.

Practice

Six multiple-choice questions in exam style, with the reasoning — not just the letter.

1. Studies finding that strategic asset allocation explains over 90% of pension fund results are most accurately described as explaining:

A. The level of returns
B. The variance of returns over time
C. The cross-sectional dispersion of manager skill
D. After-fee outperformance
B. The classic result concerns variance explained, and it is routinely misquoted as being about return levels. Endowment-specific research finds a materially lower figure, with larger contributions from market timing and manager selection.

2. An investor redeems from a hedge fund currently 15% below its high-water mark and reallocates to a new manager. The distinctive cost incurred is:

A. A higher management fee
B. The forgone loss carryforward — future gains now bear incentive fees that would have been free
C. Increased tracking error
D. A redemption penalty
B. Below the high-water mark the old manager must recover the loss before charging incentive fees; the new manager charges from the first dollar of gain. Dormant cash, opportunity losses and slippage are the other three costs of switching.

3. The most fundamental practical assumption embedded in a risk parity allocation is:

A. That expected returns can be forecast accurately
B. That leverage is available, affordable and stable
C. That correlations are exactly zero
D. That all assets have equal Sharpe ratios ex post
B. Equalising risk contributions loads capital into low-volatility assets, so leverage is required to reach a return target. Notably the method avoids forecasting expected returns entirely — which is its main attraction.

4. A commodity-exporting nation establishes a fund to insulate its annual budget from oil price swings. The appropriate portfolio is:

A. Long-horizon and heavily weighted to private equity
B. Short-horizon, liquid and low-risk, avoiding assets correlated with oil
C. Concentrated in domestic infrastructure
D. Equity-dominated to maximise intergenerational wealth
B. That is a stabilization fund: withdrawals coincide with weak oil prices and typically weak markets, so it must be liquid and uncorrelated with the revenue source. A is a savings fund; C is a development fund.

5. A family's wealth derives from a private manufacturing business in one country. The best portfolio guidance is:

A. Overweight the same sector, where the family has expertise
B. Construct the financial portfolio to offset the concentrated operating exposure on a total-balance-sheet basis
C. Match a standard institutional 60/40 policy
D. Maximise pre-tax expected return
B. The operating business is the dominant risk on the family balance sheet; the liquid portfolio's job is diversification against it. And for a family office the objective is after-tax, not pre-tax, return.

6. A DB plan invests entirely in short-term cash instruments while its liabilities have a 15-year duration. Its surplus risk is:

A. Zero, because cash has no volatility
B. High, because the duration mismatch leaves the surplus exposed to falling rates
C. Equal to the volatility of the cash portfolio
D. Undefined without an equity allocation
B. Surplus risk is the volatility of assets minus liabilities. Falling rates raise the present value of the liability while cash does not respond, so the funded ratio deteriorates — the "risk-free" asset is the liability-matching bond, not cash.

Constructed-response practice

Write these under time. Each outline is the shape the rubric rewards, not a model answer to memorise.

Prompt A (15 minutes). A $400m university endowment with a 4.5% spending rate proposes to adopt the allocation of a $30bn peer. Advise the board.

Outline:
  • Distinguish the allocation from the capabilities that made it work: list the six advantages (aggressive allocation, manager research, first-mover, alumni network, liquidity-risk acceptance, staff and board sophistication) and assess which this endowment has.
  • Governance budget: a private programme needs staff, diligence capacity and decision speed; without them, use funds of funds or an OCIO and accept the fee layer.
  • Liquidity: model the worst plausible call-and-distribution path against the spending requirement; size the liquid reserve and set policy ranges.
  • Spending and inflation: relate the spending rate to the required return and to institution-specific inflation; discuss the smoothing rule's weight.
  • Tail risk: name the mitigations and their cost per unit of drawdown avoided.
  • Recommend a phased path with vintage pacing and explicit review triggers rather than immediate adoption.

Prompt B (12 minutes). A committee wants to add tactical asset allocation across its alternatives portfolio. Explain why this is harder than in a public portfolio and describe how you would implement it if the committee insists.

Outline:
  • Frame with the fundamental law: few independent decisions means low breadth, so a very high information coefficient is required; constraints also lower the transfer coefficient.
  • Enumerate the switching costs: forgone loss carryforward, dormant cash, opportunity losses from notice periods and gates, slippage and market impact.
  • Note that illiquid positions cannot be traded on a tactical horizon at all.
  • Implementation: express tactical views through liquid futures and overlays rather than by moving capital between managers; keep the strategic private programme on its pacing schedule.
  • Model discipline: economically meaningful signals, parsimony, stability, and out-of-sample validation; pre-agree the size of tactical deviations and the review process.
← 15 · Ethics 10 · Private Equity →