lev-os Constructing Portfolio Allocations
When To Use
- Building a strategic asset allocation (SAA) for a new portfolio or investment policy statement (IPS)
- Constructing a tactical asset allocation (TAA) overlay to tilt exposures relative to a policy benchmark
- Optimizing an existing portfolio's risk-return profile under updated capital market assumptions
- Creating model portfolios for client risk-profile tiers (e.g., conservative, moderate, aggressive)
- Evaluating allocation trade-offs when adding alternative asset classes or illiquid holdings
Inputs To Gather
- Investment objective and time horizon — total return target, income requirement, liability-matching need, or spending-rate constraint
- Risk parameters — maximum drawdown tolerance, volatility budget, tracking-error limit versus benchmark, and any VaR/CVaR constraints
- Eligible asset classes — equities (domestic/international/EM), fixed income (duration/credit tiers), real assets, alternatives (PE, hedge funds, real estate, infrastructure), cash equivalents
- Capital market assumptions (CMAs) — expected returns, standard deviations, and correlation matrix for each asset class; source and vintage of CMAs [VERIFY: confirm CMA provider and date]
- Constraints — regulatory limits (e.g., ERISA prudent-investor, insurance statutory caps), liquidity minimums, ESG/SRI exclusions, concentration caps per asset class or issuer, currency-hedging policy
- Current portfolio (if rebalancing) — existing holdings, unrealized gain/loss positions, and transaction-cost estimates
- Benchmark — policy benchmark or composite index for performance attribution
Workflow
Set the allocation framework
- Choose methodology: mean-variance optimization (MVO), Black-Litterman, risk-parity, minimum-variance, or liability-driven investing (LDI)
- For MVO: build the efficient frontier from CMAs; identify the tangency portfolio and the minimum-variance portfolio
- For Black-Litterman: establish equilibrium returns from market-cap weights, then incorporate the manager's active views with confidence levels
- For risk-parity: equalize risk contribution across asset classes using marginal risk decomposition
Apply constraints
- Encode upper/lower bounds per asset class (e.g., alternatives ≤ 20%, domestic equity 30–60%)
- Layer in liquidity requirement: ensure sufficient allocation to daily-liquid instruments to meet redemption or spending needs
- Integrate ESG screens or exclusion lists if mandated by IPS
- Apply regulatory floors/caps [VERIFY: jurisdiction-specific statutory allocation limits for insurance, pension, or sovereign wealth mandates]
Run optimization and scenario analysis
- Generate optimal allocation at target return or target risk level
- Run sensitivity analysis: stress-test output against ±1–2σ shifts in key CMAs (equity risk premium, credit spreads, inflation)
- Compare efficient frontier portfolios at multiple risk points to give decision-makers a menu
- Evaluate corner solutions — if optimizer pushes any asset class to a bound, document why and whether the constraint should be revisited
Construct model portfolio tiers (if applicable)
- Map allocations to 3–5 risk-profile tiers aligned with client suitability questionnaires
- Ensure each tier shows monotonically increasing equity/growth exposure and expected volatility
- Assign implementation vehicles (index funds, ETFs, active managers, direct holdings) per sleeve
Document rebalancing and governance rules
- Define rebalancing triggers: calendar-based (quarterly/annual) vs. threshold-based (±5% drift bands)
- Specify TAA authority: allowable active tilts, maximum deviation from SAA, and approval process
- Note tax-aware rebalancing considerations for taxable accounts (harvest losses, avoid short-term gains)
Output
- Allocation summary table — target weights per asset class with permissible ranges
- Efficient frontier chart — plotted portfolios with the selected allocation highlighted
- Risk decomposition — contribution to total portfolio risk by asset class (marginal and percentage)
- Scenario/stress-test results — portfolio return and drawdown under bull, base, and bear CMAs
- Model portfolio tiers (if multi-tier) — side-by-side allocation grids with expected return, volatility, Sharpe ratio, and max drawdown estimate per tier
- Methodology narrative — rationale for framework choice, key assumptions, and known limitations
- Rebalancing policy summary — triggers, bands, and governance authority
Quality Checks
- Weights sum to 100% across all asset classes in every tier
- No allocation breaches stated upper/lower bounds or regulatory caps
- Expected portfolio return and risk metrics are arithmetically consistent with CMAs and weights
- Correlation and volatility inputs match the stated CMA source and vintage — flag stale data (>12 months old) with [VERIFY]
- Liquidity profile supports stated spending/redemption needs without forced selling of illiquid sleeves
- Sharpe ratio and risk-contribution metrics are reasonable relative to historical ranges for similar portfolios
- Tax-lot and transaction-cost implications are noted for rebalancing recommendations in taxable accounts
- All assumptions and data sources are explicitly cited; no inferred figures presented as confirmed