On this page
Brinson Performance Attribution: Allocation vs. Selection
Brinson attribution is the standard method for decomposing a portfolio's active return against its benchmark into contributions from asset allocation, security selection, and their interaction. It originated with the 1986 Brinson, Hood, and Beebower paper in the *Financial Analysts Journal*.
Key Takeaways
- Brinson attribution splits active return into three effects: allocation (sector weights), selection (stock picks within sectors), and interaction (the combined effect of both).
- In the worked example, 68 basis points of active return broke roughly half to allocation and half to selection, demonstrating that the source of alpha matters for evaluating manager skill.
- A common mistake is adding single-period attribution effects across months or quarters, multi-period attribution requires a linking algorithm because simple addition does not match compounded active return.
- Attribution is a diagnostic tool, not a scorecard; a manager's output must be compared to their stated process, not to another manager with a different mandate.
Key Takeaways
- Brinson attribution splits active return into three effects: allocation (sector weights), selection (stock picks within sectors), and interaction (the combined effect of both).
- In the worked example, 68 basis points of active return broke roughly half to allocation and half to selection, demonstrating that the source of alpha matters for evaluating manager skill.
- A common mistake is adding single-period attribution effects across months or quarters, multi-period attribution requires a linking algorithm because simple addition does not match compounded active return.
- Attribution is a diagnostic tool, not a scorecard; a manager's output must be compared to their stated process, not to another manager with a different mandate.
What It Is
Brinson attribution answers a specific question: when a portfolio beat (or missed) its benchmark, how much came from weighting sectors or asset classes differently, and how much came from picking different securities inside each sector? The original framework splits active return into three effects: allocation, selection, and interaction. A later variant by Brinson and Fachler treats each segment against the benchmark segment return rather than the total benchmark return.
The method assumes the portfolio and benchmark share a common segmentation, typically by sector, country, or asset class. Returns and weights for each segment must be available for both the portfolio and the benchmark.
The Intuition
Suppose a US equity manager beats the S&P 500 by 2 percent over a year. The allocator wants to know why. Did the manager overweight technology stocks during a tech rally, or did they pick winners inside each sector regardless of weight? These are different skills, and they should be rewarded or hired for separately.
Brinson attribution isolates those effects arithmetically. The allocation effect asks what would have happened if the manager had picked index returns inside each sector but used their own weights. The selection effect asks what would have happened if the manager had used benchmark weights but their own security picks. The interaction effect captures the cross term, the piece that depends on both weights and picks at once.
How It Works
Let wp be portfolio weights, wb be benchmark weights, rp be portfolio segment returns, and rb be benchmark segment returns, each indexed by segment i.
The three classic effects, summed across segments, are:
Allocation effect = sum over i of (wp_i - wb_i) * rb_i
Selection effect = sum over i of wb_i * (rp_i - rb_i)
Interaction effect = sum over i of (wp_i - wb_i) * (rp_i - rb_i)
The three sum to the total active return (portfolio return minus benchmark return). Many consultants use the Brinson-Fachler variant, which subtracts the total benchmark return from each segment benchmark return in the allocation formula:
Allocation (Brinson-Fachler) = sum over i of (wp_i - wb_i) * (rb_i - R_b)
where R_b is the overall benchmark return. Brinson-Fachler makes the allocation effect intuitive: an overweight to a segment that outperformed the total benchmark produces a positive allocation effect, while an overweight to a segment that underperformed the total produces a negative one.
Multi-period attribution requires a linking algorithm (Carino, Menchero, Frongello) to avoid the arithmetic of single-period effects drifting away from total compound active return.
Worked Example
A US large-cap portfolio is compared to the S&P 500 over one quarter. The benchmark is split into three sectors.
| Sector | Benchmark weight | Portfolio weight | Benchmark return | Portfolio return |
|---|---|---|---|---|
| Tech | 30% | 40% | 8.0% | 9.0% |
| Health | 15% | 20% | 2.0% | 1.5% |
| Other | 55% | 40% | 4.0% | 4.2% |
Benchmark return = 0.30 x 8.0 + 0.15 x 2.0 + 0.55 x 4.0 = 2.40 + 0.30 + 2.20 = 4.90% Portfolio return = 0.40 x 9.0 + 0.20 x 1.5 + 0.40 x 4.2 = 3.60 + 0.30 + 1.68 = 5.58% Active return = 5.58% - 4.90% = 0.68%
Allocation effect: (0.40 - 0.30) x 8.0 + (0.20 - 0.15) x 2.0 + (0.40 - 0.55) x 4.0 = 0.80 + 0.10 - 0.60 = 0.30%
Selection effect: 0.30 x (9.0 - 8.0) + 0.15 x (1.5 - 2.0) + 0.55 x (4.2 - 4.0) = 0.30 - 0.075 + 0.11 = 0.335%
Interaction effect: (0.40 - 0.30) x (9.0 - 8.0) + (0.20 - 0.15) x (1.5 - 2.0) + (0.40 - 0.55) x (4.2 - 4.0) = 0.10 - 0.025 - 0.03 = 0.045%
Sum: 0.30 + 0.335 + 0.045 = 0.68%, which matches the active return.
The manager's 68 basis points of alpha came roughly half from allocation (overweighting Tech) and half from selection (picking Tech and Other names that beat their sectors).
Common Mistakes
-
Using the wrong benchmark segmentation. If the portfolio is classified by GICS sector and the benchmark by custom industry groups, the attribution is meaningless. Segments must be defined identically on both sides.
-
Ignoring the interaction term. Some presentations combine interaction into selection to give a cleaner two-effect chart. That is defensible only if it is disclosed. Silently burying interaction hides whether large selection wins were amplified by a weight bet.
-
Single-period math over long windows. Effects do not sum cleanly across periods when returns compound. Multi-period attribution requires a smoothing or linking algorithm. Simple addition of monthly effects will not equal the compound active return.
-
Attributing skill to currency by accident. International portfolios must separate the currency effect from local allocation and selection. Otherwise a manager who picked a great German stock while the euro fell gets blamed for the currency move.
-
Treating attribution as a performance grade. The output is a diagnostic, not a scorecard. Low tracking-error mandates will show small effects by design. A manager's attribution should be compared to the stated process, not to another manager with a different mandate.
Frequently Asked Questions
Q: What is Brinson performance attribution in simple terms? It is an arithmetic framework that splits a portfolio's excess return over its benchmark into the portion that came from weighting sectors differently (allocation effect) and the portion that came from picking different securities within each sector (selection effect), plus a cross term called the interaction effect.
Q: How does Brinson performance attribution affect investment decisions? It tells you whether a manager's returns came from a sector bet or from genuine stock-picking skill, which are different capabilities. Allocators use this to decide whether to hire a manager for tactical asset allocation versus bottom-up security selection, and whether the fee paid is appropriate for the source of value.
Q: What is a real-world example of Brinson performance attribution? A US equity manager beats the S&P 500 by 68 basis points in a quarter. Attribution shows 30 basis points came from overweighting technology (allocation) and 34 basis points from picking outperforming names within technology (selection). The remaining 4 basis points is the interaction term. Both skills contributed, but the tilt was the larger driver.
Q: How can investors use Brinson attribution when evaluating managers? Ask for multi-year attribution reports, not just a single period. Check whether the allocation or selection effect dominates. If allocation explains most of the active return, the manager may be paid an active fee for what is essentially a factor tilt available more cheaply elsewhere.
Q: How is Brinson performance attribution different from Fama-French factor attribution? Brinson attribution explains returns by sector weights and stock picks within a common segmentation. Fama-French attribution uses statistical regression to decompose returns into systematic factor exposures, market, size, value, and others. Brinson is segment-based and intuitive; factor attribution identifies which risk premia explain the pattern.
Sources
- Brinson, G.P., Hood, L.R., and Beebower, G.L. (1986). "Determinants of Portfolio Performance." Financial Analysts Journal, 42(4), 39-44. https://www.tandfonline.com/doi/abs/10.2469/faj.v42.n4.39
- Brinson, Hood, Beebower (1986). Full text PDF. https://pfa.kz/images/download/100331-brinson_Determinants_Portfolio_Performance.pdf
- CFA Institute. "Overview of the Global Investment Performance Standards." https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/overview-of-the-global-investment-performance-standards
- Commonfund. "Determinants of Portfolio Returns: It Depends." https://www.commonfund.org/blog/determinants-of-portfolio-returns-it-depends-
Disclaimer
This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.