Selection effect
What is Selection effect?
Selection effect is the portion of benchmark-relative return attributed to portfolio holdings performing differently from their benchmark group.
A selection result should be shown with bucket weight, holdings, benchmark constituents, coverage, and interaction convention. Concentrated exposures and off-benchmark securities can dominate the number. Positive selection can reflect factor tilts or luck rather than security-specific insight, requiring further attribution and repeated evidence.
Review should compare repeated outcomes across periods and control for sector, style, size, currency, and concentration.
Holdings-level drill-down should reconcile security contributions to the bucket result and identify unmapped, proxy, and residual observations.
Core idea
Selection evaluates bottom-up results within sectors, regions, or other groups. It asks whether the securities held in a bucket outperformed that bucket's benchmark return, separating this from the decision to overweight or underweight the bucket. A broad index holding should have little selection effect relative to the same correctly measured index.
Calculation
A common Brinson formula multiplies benchmark bucket weight by portfolio bucket return minus benchmark bucket return. Some variants use portfolio weight or combine an interaction term with selection. The exact convention changes bucket values, so reports must state the model and reconcile allocation, selection, interaction, and residual to active return.
Example
If healthcare is 15% of the benchmark and the portfolio's healthcare holdings return 4 percentage points more than benchmark healthcare, simplified selection adds 0.6 percentage point. The result does not show which security was responsible until holdings-level contribution is examined, and it can change with the classification or return proxy.
Look-through and data
Fund sleeves require historical holdings and weights or a proxy, while direct securities use actual returns. Corporate actions, currency, derivatives, cash, and trades must align. Using today's fund composition for earlier periods introduces look-ahead bias. Incomplete coverage should produce an explicit residual or confidence indicator rather than false precision.
Interpretation
Positive selection can arise from fundamental insight, factor exposure, concentration, luck, or benchmark mismatch. Compare results over time, across decisions, and after transaction costs. A manager can show positive selection but negative total active return because allocation detracted, and a portfolio with no security-selection mandate should not be judged primarily by this effect.
Sources and further reading
- Portfolio Performance Evaluation, CFA Institute