NISM-Series-XXI-A: Portfolio Management Services (PMS) Distributors · Performance Measurement and Evaluation of Portfolio Managers

Performance Attribution and Evaluating Portfolio Managers

Updated 11 October 2026 · Fact-checked

Performance attribution splits a portfolio's return versus its benchmark into sources, mainly asset allocation (over or underweighting asset classes or sectors) and stock selection (picking better or worse securities within them). Evaluating a manager then combines these numbers with qualitative factors like process, team, consistency and compliance.

Understand Performance Attribution and Evaluating Portfolio Managers

Knowing that a PMS portfolio returned 15% is not enough. You need to know why. Did the manager do well because of a smart call on which sectors or asset classes to hold, or because of good stock picks? Performance attribution answers this by breaking the difference between portfolio return and benchmark return into its causes.

The two main sources are asset allocation (also called the allocation effect) and stock selection (the selection effect). Asset allocation measures the value added by holding different weights from the benchmark. If the benchmark holds 20% in banks and the manager holds 30%, and banks beat the overall benchmark, the extra weight added value. Stock selection measures the value added by picking securities that did better or worse than the benchmark's securities in the same sector. A third piece, the interaction effect, captures the combined impact of weight differences and selection differences. Some textbooks fold it into selection.

The approach is often called Brinson attribution, after the researchers who developed it. It compares the portfolio with a benchmark segment by segment. The sum of all effects equals the total active return, which is portfolio return minus benchmark return. Attribution only works if the benchmark is appropriate and the segments are defined the same way for both portfolio and benchmark.

Numbers alone do not tell the whole story. A good evaluation also looks at qualitative factors: the investment philosophy and whether the manager sticks to it, stability and experience of the team, quality of the research and risk-control process, consistency across market cycles, size of assets under management, portfolio turnover, compliance record and transparency of reporting. Quantitative factors include returns over several periods, risk-adjusted measures, tracking error and comparison with benchmark and peers.

As a distributor, you use this to explain to clients where returns came from and whether the manager's skill is repeatable. Strong recent returns from one lucky bet are weaker evidence than steady, process-driven outperformance.

Key formulas to remember

Active return: Active return = Portfolio return − Benchmark return; The total value added or lost versus the benchmark. Attribution effects add up to this.
Allocation effect (per segment): (Wp − Wb) × (Rb_segment − Rb_total); Wp and Wb are portfolio and benchmark weights. Rb_segment is benchmark return of the segment. Some versions use (Wp − Wb) × Rb_segment; both are used, so follow the data given.
Selection effect (per segment): Wb × (Rp_segment − Rb_segment); Uses the benchmark weight. Rp_segment is the portfolio's return in that segment.
Interaction effect (per segment): (Wp − Wb) × (Rp_segment − Rb_segment); Combined effect of different weight and different selection.
Total attribution: Allocation + Selection + Interaction = Active return; Use this to check your arithmetic.

How to solve Performance Attribution and Evaluating Portfolio Managers questions

Use this method for any question on attribution or manager evaluation.

1Read what is asked: a source of return (allocation or selection), a calculation, or a judgement on a manager.
2Identify whether the question is about weights (allocation) or about returns within a segment (selection).
3For allocation questions, compare portfolio weight with benchmark weight and check whether that segment beat or lagged the benchmark.
4For selection questions, compare the portfolio's return in the segment with the benchmark's return in the same segment.
5If calculating, work segment by segment, then add. Check that the total matches portfolio return minus benchmark return.
6For evaluation questions, separate qualitative factors (process, team, consistency, compliance) from quantitative ones (returns, risk-adjusted measures, tracking error).
7Choose the option that matches the definition exactly and reject options that rely on a single period of returns alone.

Quickest way: Weights or picks?

When to use it: Use when an MCQ asks which effect explains the outperformance.

If the difference is in how much was invested in a sector or asset class, it is asset allocation.
If the difference is in how well chosen securities did within the same sector, it is stock selection.
If the question says the portfolio matched benchmark weights, allocation effect is zero.
If the question says the portfolio held the same securities as the benchmark in each segment, selection effect is zero.
For manager evaluation, prefer answers that combine long-term, risk-adjusted numbers with process and consistency.

Common mistakes in Performance Attribution and Evaluating Portfolio Managers

Confusing allocation with selection.
Both explain outperformance and the terms sound similar.
Fix: Allocation is about weights. Selection is about which securities. Ask: did the difference come from how much or from which?
Using portfolio weight in the selection effect formula.
It feels natural to use the manager's own weight.
Fix: In the standard Brinson split, selection uses benchmark weight; the leftover goes to interaction.
Judging a manager only on the latest one-year return.
Recent returns are easy to see and attract clients.
Fix: Look at several periods, risk taken, consistency and process. Short-term returns can be luck.
Ignoring the benchmark choice.
Students assume any index will do.
Fix: Attribution is only meaningful against a benchmark that matches the strategy. A wrong benchmark distorts every effect.
Treating qualitative factors as unimportant in an objective exam.
Numbers feel more testable.
Fix: Questions do ask about team stability, philosophy adherence and compliance record. Learn the list.
Not checking that effects sum to active return.
Students rush the arithmetic.
Fix: Always add allocation, selection and interaction and compare with portfolio minus benchmark return.

Worked examples

Example 1

A benchmark has two segments: Equity (weight 60%, return 10%) and Debt (weight 40%, return 5%). A portfolio holds Equity 70% earning 12% and Debt 30% earning 5%. Find the active return and the selection effect using the Brinson split with benchmark weights.

Show the solution

Benchmark return = 0.60 × 10% + 0.40 × 5% = 6% + 2% = 8%.
Portfolio return = 0.70 × 12% + 0.30 × 5% = 8.4% + 1.5% = 9.9%.
Active return = 9.9% − 8% = 1.9%.
Selection effect: Equity = 0.60 × (12% − 10%) = 1.2%; Debt = 0.40 × (5% − 5%) = 0%. Total selection = 1.2%.
Check: allocation Equity = (0.70 − 0.60) × (10% − 8%) = 0.2%; Debt = (0.30 − 0.40) × (5% − 8%) = 0.3%; allocation total = 0.5%. Interaction Equity = 0.10 × 2% = 0.2%; Debt = 0. Sum = 0.5% + 1.2% + 0.2% = 1.9%, which matches.

Answer: Active return is 1.9%, and the selection effect is 1.2%.

Example 2

A PMS manager had the highest one-year return among peers. A distributor wants to judge whether to recommend the manager. Which approach is most appropriate?

A. Recommend because of the top one-year return
B. Review returns over multiple periods against the benchmark, risk-adjusted measures, attribution, and qualitative factors such as process and team stability
C. Judge only by the size of assets under management
D. Judge only by the manager's fee level

Show the solution

Single-period return can come from luck or a concentrated bet, so A is weak.
Assets under management alone says nothing about skill, so C is weak.
Fees matter to net returns but do not show skill, so D is incomplete.
B combines quantitative and qualitative evidence and shows where returns came from.

Answer: B

Exam tips

Know the definitions cold: allocation is weights, selection is picks, interaction is the combined effect.
If a numeric question gives segment weights and returns, compute benchmark and portfolio return first, then check your effects sum to the difference.
In evaluation MCQs, the best option usually mixes long-term, risk-adjusted numbers with process and consistency.
Watch for options that rely on one period of return or on one factor only; these are usually traps.
With negative marking of 10% per question in this exam, eliminate clearly weak options and answer when you can narrow to two.

Practice questions from Performance Measurement and Evaluation of Portfolio Managers

Performance Attribution and Evaluating Portfolio Managers in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Performance Attribution and Evaluating Portfolio Managers: frequently asked questions

What is the difference between asset allocation effect and stock selection effect?

Asset allocation effect comes from holding different weights in asset classes or sectors than the benchmark. Stock selection effect comes from choosing securities that perform better or worse than the benchmark's securities in the same segment.

What is Brinson attribution?

It is a method that splits a portfolio's active return versus its benchmark into allocation, selection and interaction effects, segment by segment. The effects add up to the total active return.

How do you evaluate a PMS manager?

Use both quantitative and qualitative factors. Quantitatively, look at returns over several periods, risk-adjusted measures and tracking error against a suitable benchmark. Qualitatively, assess philosophy, process, team stability, consistency, compliance record and transparency.

Why is the benchmark important in attribution?

Every effect is measured relative to the benchmark. If the benchmark does not match the manager's strategy, the allocation and selection results will be misleading.