Risk Parity vs Equal-Weight: Choosing a Strategy Allocator

Two ways to combine N strategies into a portfolio:

Equal-weight: allocate the same dollar amount to each.
Risk parity: allocate so that each strategy contributes the same volatility to the portfolio.

These produce materially different portfolios. The choice between them is more consequential than most allocators realize.

The mechanic of each

Equal-weight

weight_i = 1 / N

Simple. Robust. Doesn't require any estimation of strategy properties. The high-volatility strategies dominate portfolio risk; the low-volatility strategies are along for the ride.

Risk parity

weight_i ∝ 1 / σ_i

Each strategy contributes equally to portfolio variance (under independence). Lower-volatility strategies get larger dollar allocations; higher-volatility strategies get smaller. Often combined with vol-targeting to a portfolio-level vol target.

Why risk parity earns its reputation

Risk parity dominates equal-weight when strategies have meaningfully different volatilities. In a portfolio of, say, equity strategies (15% vol) and bond strategies (4% vol), equal-weight gives the equity strategies ~93% of portfolio risk despite ~50% capital allocation. The "diversified" portfolio is, in risk terms, an equity portfolio with a small bond position.

Risk parity rebalances toward equal risk contribution: roughly 4× more capital to the bond strategies than to the equity strategies, producing a portfolio where both contribute equally to variance.

Where risk parity fails

1. Estimation error

Risk parity requires estimates of strategy volatility (and ideally correlation). Estimation error in σ propagates directly into allocation. Strategies that look low-vol because of a quiet sample period get over-allocated; the next vol regime change is brutal.

2. Leverage

Risk parity portfolios systematically use more leverage than equal-weight portfolios — that's how the low-vol legs get to comparable risk contribution. Leverage means funding costs and tail risk. The 2008 unwind hit risk-parity strategies disproportionately.

3. Correlation regime change

Risk parity allocations are usually computed assuming roughly stable correlation structure. When correlation decays — say, when previously uncorrelated strategies start co-moving in stress — the portfolio's true risk exposure spikes far above target.

4. The volatility-as-risk assumption

Risk parity uses volatility as the risk metric. For asymmetric strategies, this is the wrong metric (see drawdown vs volatility). A portfolio that risk-parities short-vol strategies looks fine on volatility but is disasters-in-waiting on drawdown.

When equal-weight is better

Strategies with similar volatility profiles. If σ is comparable across strategies, equal-weight and risk parity produce similar portfolios with less estimation noise on the equal-weight side.
Untrusted volatility estimates. Few observations, regime-suspect data, asymmetric returns.
Defensible-to-allocators portfolios. Equal-weight has a clean explanation; risk parity requires explaining the model.

What we do

Sentivue's portfolio construction uses risk parity within strategy clusters and equal-weight across clusters. Strategies are grouped by archetype (trend-following, mean-reversion, carry, etc.). Within each cluster, risk parity. Between clusters, equal-weight to enforce diversification across strategy types even if their realized volatilities are similar.

The motivation: clusters tend to share regime sensitivities. Within a cluster, vol-based weighting is informative. Across clusters, structural diversification is more important than local risk equalization.

Practical takeaways

Risk parity is the right default when strategy vol differs meaningfully. Equal-weight under-diversifies in those cases.
Risk parity requires you to trust your vol estimates. When you don't, equal-weight is safer.
Mixed approaches (cluster-level risk parity, strategy-level equal-weight, or vice versa) often beat either pure approach.