How is sector allocation alpha calculated?

Calculate the return contribution of holding each sector at portfolio weight, with stocks held at benchmark proportions within each sector. Subtract this from total portfolio return; the residual is stock selection alpha. The sector allocation contribution is what the manager added (or lost) by overweighting/underweighting sectors relative to benchmark.

Can I do attribution myself, or do I need the AMC to provide it?

AMCs provide attribution in their quarterly portfolio statements. For independent verification, sophisticated investors can run their own attribution using the disclosed positions and benchmark data. Trustner's annual review includes attribution discussion using both AMC-provided and independently-verified analysis.

What if a manager's alpha all comes from timing?

Caution. Timing alpha is the most volatile and least repeatable. Managers who appear to consistently time markets often had a few lucky calls that drove headline numbers. Sustained timing alpha across 5+ years is rare; unless the manager has exceptional macro-research infrastructure, expect timing alpha to mean-revert toward zero.

PMS Advanced — Manager Risk, Performance Attribution & Exit Decisions (1/3)

Topic 1 of 3~5 min read

Performance Attribution — Decomposing Manager Skill

Definition

PMS performance attribution decomposes total returns into stock selection (security-level alpha), sector allocation (over/underweight calls), and timing (entry/exit decisions vs benchmark). Disciplined attribution separates lucky streaks from durable skill, and identifies which alpha sources are repeatable versus circumstantial.

In Simple Words

Three-factor attribution. (a) Stock selection alpha: portfolio return minus benchmark return holding sector weights constant. Measures the manager's skill in picking stocks within each sector. (b) Sector allocation alpha: difference attributable to over/underweighting sectors vs benchmark. Measures the manager's top-down sector calls. (c) Timing alpha: difference attributable to entry/exit decisions (buying low, selling high relative to benchmark). Measures market-timing and execution skill. A skilled manager generates alpha across at least two of the three; a manager generating alpha only from one source has narrow skill that may not persist. Attribution discipline: review at year-end across the past 3-5 years. Look for consistency — managers showing positive stock-selection alpha in 4 of 5 years are more credible than those showing 80% alpha one year and -30% the next. The pattern of rolling 3-year attribution often surfaces concerns that headline returns hide.

Real-Life Scenario

A multi-cap PMS's 5-year attribution: stock selection +2.1% annualised (consistent across years), sector allocation +0.8% (variable, sometimes negative), timing -0.4% (negative — the manager is a poor market timer). Net alpha 2.5% before fees; fees absorb 3.5%; net-of-fee alpha is -1.0%. Attribution identifies that the manager's strength is bottom-up stock picking, but fees absorb the entire alpha. Recommendation: hold position, monitor for fee compression OR alpha widening; do not add to allocation.

Key Points to Remember

Three-factor attribution: stock selection, sector allocation, timing.

Skilled managers generate alpha across at least two factors.

Single-factor alpha may not persist — narrow skill warning.

Review across rolling 3-5 year periods, not single year.

Net-of-fee alpha (after subtracting fees) is the structural test.

Pattern matters: consistency > lucky-year peaks.

Trustner's framework includes annual attribution review per empanelled PMS.

Frequently Asked Questions

Test Your Knowledge

3 questions to check your understanding

Question 1 of 3Score: 0/0

A manager showing alpha only from timing across 5 years is:

Summary Notes

Three-factor attribution: stock selection, sector allocation, timing.

Net-of-fee alpha is the structural test, not gross headline returns.

Consistent multi-factor alpha > single-factor lucky year.

Trustner's framework reviews attribution annually per empanelled PMS.