Asset Allocation - Strategic, Tactical and Regime-Based
By EC Assets Research Team, Portfolio Construction · Published
Asset Allocation — Asset allocation is the decision of how to split a portfolio across asset classes (equities, bonds, alternatives, cash) to achieve a stated return-and-risk objective. Academic studies attribute roughly 90% of long-run portfolio variance to the allocation decision, dwarfing the impact of security selection.
Definition
Asset allocation is the most consequential single decision an institutional investor makes. Brinson, Hood and Beebower's 1986 study and its many successors established that approximately 90% of the variance of long-horizon portfolio returns comes from the asset-class mix, with security selection and market timing accounting for the residual. The number has been debated, but the directional message - allocation dominates - has survived four decades of scrutiny.
A practical allocation framework operates on three nested time horizons:
| Horizon | Decision | Driven by |
|---|---|---|
| Strategic (SAA) | 10–30 year target weights | Long-run risk premia, liability profile, mandate |
| Tactical (TAA) | 6–24 month tilts vs SAA | Cycle, valuation, sentiment, regime |
| Dynamic / regime | Conditional rules within TAA | Macro state, volatility regime, drawdown bands |
The three are not alternatives; they are layers. A pension fund sets SAA in its investment policy statement, applies TAA tilts within risk bands, and may apply dynamic overlays based on volatility or macro regime indicators.
Strategic Asset Allocation (SAA)
SAA is the long-run policy mix that the investor would hold if they had no view of near-term market conditions. It is set against the investor's investment objectives - typically a return target, a risk tolerance, and a liability profile if applicable.
Three common SAA archetypes:
| Archetype | Equities | Bonds | Alternatives | Cash |
|---|---|---|---|---|
| Conservative (retiree) | 30% | 55% | 10% | 5% |
| Balanced (60/40) | 55% | 35% | 8% | 2% |
| Aggressive (endowment) | 35% | 15% | 45% | 5% |
| Risk parity | varies | varies | varies | varies |
The academic foundation is Markowitz's mean-variance optimisation: given expected returns, volatilities and correlations, identify the portfolio that maximises expected return for a given risk level. In practice, allocators use Markowitz outputs as a sanity check rather than a prescription, because forward-looking inputs (especially expected returns) are notoriously unstable.
Tactical Asset Allocation (TAA)
TAA is the discretion to tilt away from the SAA based on shorter-horizon signals. A typical TAA framework defines:
- A policy band around each SAA weight (e.g. equities target 55% ± 10%)
- A signal set combining valuation, momentum, sentiment and macro
- A sizing rule mapping signal strength to position size within bands
- A drawdown trigger that forces de-risking on portfolio losses beyond a threshold
Empirically, well-disciplined TAA adds 50–150 basis points of annual return over pure SAA over long cycles, with most of the value coming from avoiding the worst 5% of regimes rather than catching the best 5%.
Regime-Based Allocation
A regime-based approach explicitly classifies the macro/market environment into discrete states and varies the allocation accordingly. A common four-regime framework links growth and inflation:
| Regime | Growth | Inflation | Favoured assets |
|---|---|---|---|
| Goldilocks | Up | Stable | Equities, credit |
| Reflation | Up | Up | Commodities, real assets, value equity |
| Stagflation | Down | Up | Inflation-linked bonds, gold, defensive equity |
| Recession | Down | Down | Long-duration government bonds, cash, quality equity |
The attraction is intellectual coherence: positioning explicitly reflects the macro thesis. The execution risk is regime classification - the framework only adds value if you can identify the regime quickly enough to position before the regime is already priced in.
Worked Example - 60/40 versus Regime-Tilted
Consider a 60/40 strategic mix versus a regime-aware tilt over a hypothetical decade with the following regime pattern:
| Period | Regime | 60/40 (annualised) | Regime-tilt (annualised) |
|---|---|---|---|
| Yr 1–3 | Goldilocks | +9.5% | +12.0% (overweight equity) |
| Yr 4–5 | Reflation | +4.5% | +8.0% (overweight commodities) |
| Yr 6–7 | Stagflation | −2.0% | +5.0% (gold + linkers) |
| Yr 8–9 | Recession | −5.5% | +6.5% (long bonds) |
| Yr 10 | Recovery | +12.0% | +10.0% (catch-up) |
| Total | +4.5% | +8.3% |
The regime-tilt approach outperforms in this stylised example by ~380 bps per year, but the example assumes correct regime classification. In practice, allocators who get one regime wrong (e.g. positioning for stagflation when the regime is actually recession) can lose more than the entire premium they otherwise would have earned.
Common Allocation Pitfalls
| Pitfall | What it looks like | Fix |
|---|---|---|
| Recency bias | Overweighting whatever has performed best recently | Use 10+ year input estimates; avoid 3-year backward-looking inputs |
| Home bias | Domestic equity at 70%+ of equity sleeve regardless of mandate | Benchmark to a global cap-weighted index |
| Risk-budget myopia | Sizing positions by capital weight rather than risk contribution | Risk-parity or volatility-targeted construction |
| Liquidity blindness | Allocating to illiquid assets without modelling redemption needs | Liquidity tier-mapping; cash-flow forecasting |
| Rebalancing drag | Trading too often through fees and slippage | Define rebalancing bands, not calendar triggers |
Why Allocation Matters
For any portfolio above a few million dollars, the asset-allocation decision determines roughly 90% of the long-run variance of outcomes. Security selection within asset classes - even very skilled selection - is a smaller lever. Three institutional implications follow:
- Spend the most time on SAA. Most allocators invert this and spend most of their meeting time on manager selection. The math says they should spend most of it on the strategic mix.
- Define the TAA bands ex-ante. Discretion without policy bands turns TAA into mood-driven trading. Wide bands (±10–15%) provide room to tilt; narrow bands (±2–3%) protect against drift.
- Document the regime indicators. Whatever framework drives TAA - valuation, momentum, macro - write it down before you trade it. Otherwise the framework becomes whatever you wish it had been after the fact.
Asset allocation is one of the few areas of investment management where the academic literature has been consistent for decades. The discipline of acting on it is rarer than the knowledge.
References
- Ilmanen, A. (2011). Expected Returns. Wiley.
- Ang, A. (2014). Asset Management. Oxford University Press.
- CFA Institute. Asset Allocation. CFA Program Curriculum.
Frequently asked questions
Does the 90% number really hold up?
It depends on what you measure. The original Brinson study measured variance of returns; subsequent work (Ibbotson and Kaplan 2000) found 100% for variance, 40% for level. The directional claim — allocation dominates security selection over long horizons — is well-established.
What is risk parity and how is it different?
Risk parity sets weights so each asset class contributes equally to portfolio risk, not equally to portfolio dollars. Bonds get larger dollar allocations (because they are less volatile), often via leverage. The approach pioneered by Bridgewater earned attention by handling 2008 well, then suffered in 2022 when bonds and equities both fell.
How often should the SAA be reviewed?
Most institutional investors review SAA annually as a check, with a formal full review every 3–5 years. Mid-cycle changes are usually reserved for material changes in the investor's circumstances (new liabilities, regulation, mandate) rather than market conditions, which are TAA decisions.
What is the difference between TAA and market timing?
TAA operates within defined policy bands around the SAA and uses systematic signals. Market timing is the attempt to enter and exit the market entirely based on directional forecasts. TAA is empirically value-additive when disciplined; market timing has shown to destroy value on average.
Should I allocate by liability or by return target?
Liability-driven investing (LDI) sets the allocation to match the liability cash-flow profile — natural for pension funds and insurers. Return-target allocation optimises against a stated CAGR. The choice depends on the mandate: if you have specific outflows, LDI; if you have a long-horizon growth objective, return-target.
Stay informed
Market commentary, firm news and research from EC Assets - direct to your inbox.