Brinson attribution
What is Brinson attribution?
Brinson attribution decomposes portfolio return relative to a benchmark into effects associated with allocation and security selection across defined groups.
Palance should label the exact Brinson variant, grouping taxonomy, benchmark, return frequency, weight convention, interaction treatment, and residual. Look-through estimates and proxy returns need visible coverage and confidence. Attribution explains a measured return difference; it does not prove skill, causation, or that the same decision will add value again.
A complete report should also separate data residuals and transaction effects instead of forcing them into allocation or selection.
Multi-period linking should state how arithmetic effects are compounded and how the chosen method preserves reconciliation over time.
Purpose and framework
The method explains active return after the fact by comparing portfolio and benchmark weights and returns within sectors, regions, or other buckets. Brinson-Hood-Beebower and Brinson-Fachler use related but different allocation formulas. Results depend on the chosen model, taxonomy, benchmark, frequency, and treatment of an interaction term.
Allocation effect
Allocation asks whether overweights and underweights were placed in groups that performed favorably relative to the benchmark. In the Brinson-Fachler form, a bucket's active weight is multiplied by its benchmark return relative to the total benchmark. A positive effect can arise from overweighting a winner or underweighting a laggard.
Selection and interaction
Selection compares portfolio and benchmark returns within a bucket, commonly scaled by benchmark weight. Interaction captures the combined effect of active weight and within-bucket active return. Some reports combine interaction with selection or allocation, changing displayed totals. The convention must be stated before comparing systems or managers.
Data and look-through
Accurate weights, returns, classifications, corporate actions, cash, derivatives, and intra-period flows are required. Funds need look-through or a documented proxy. Palance can use actual direct-holding returns and sector proxies for fund sleeves, but the result is indicative when underlying holdings, historical weights, or benchmark decomposition are incomplete.
Interpretation and limitations
Attribution reconciles what happened under a grouping model; it does not prove why, distinguish luck from skill, or forecast persistence. A sector effect can hide style, country, currency, or stock factors. Review multiple periods, decisions, transaction cost, and risk. Residuals should remain visible rather than assigned arbitrarily to make the chart reconcile.
Sources and further reading
- Portfolio Performance Evaluation, CFA Institute