NISM Professor

Security selection effect

In performance attribution, the part of excess return earned by choosing better securities within a sector, measured at the benchmark's own sector weights, separate from allocation and interaction effects.

In plain language

A portfolio can beat its benchmark two different ways: by holding more of the sectors that do well, or by picking better stocks within whichever sectors it holds. Security selection effect isolates the second source.

The workbook's own words: it "measures how well the portfolio's chosen stocks performed relative to the benchmark within the same sector." Even if a manager's sector weights were identical to the benchmark's, picking stronger stocks within each sector would still generate excess return, and that excess is exactly what this effect measures.

How it works

Formula (Chapter 10, section 10.5), per sector, then summed.

Security selection effect = sum of Wb × (Rp − Rb)

Wb is the benchmark's weight in a sector, Rp is the portfolio's return in that sector, Rb is the benchmark's return in that sector.

Why benchmark weight, not portfolio weight? Using Wb isolates the stock-picking decision from the allocation decision. If portfolio weight were used instead, an overweight sector's stock-picking would get extra credit simply for being large, exactly the double-counting the third component, interaction effect, exists to capture separately.

The three components sum to the total excess return: total excess return equals the asset allocation effect plus the security selection effect plus the interaction effect.

A worked example

The workbook's own three-sector example, a portfolio and its benchmark each with three sectors:

SectorPortfolio weightPortfolio returnBenchmark weightBenchmark return
Technology40%12%35%10%
Healthcare30%8%40%7%
Financials30%5%25%6%

Security selection effect, sector by sector (Wb × (Rp − Rb)):

  • Technology: 35% × (12% − 10%) = +0.70%
  • Healthcare: 40% × (8% − 7%) = +0.40%
  • Financials: 25% × (5% − 6%) = −0.25%
  • Total security selection effect = 0.70% + 0.40% − 0.25% = 0.85%

Out of the portfolio's total 0.9 percentage point outperformance, 8.7% portfolio versus 7.8% benchmark, 0.85 percentage points, the bulk of it, came from picking better stocks within each sector, not from how the sectors were weighted. On a ₹1,00,00,000 portfolio, that is ₹85,000 of the ₹90,000 total excess return.

Why NISM asks about it

Chapter 10 (Performance Measurement and Evaluation of Portfolio Managers), section 10.5 (Performance Attribution Analysis), works through this exact three-sector, three-effect example. Expect a calculation question isolating the security selection effect given sector weights and returns, and a question on which weight, portfolio's or benchmark's, belongs in this formula versus the allocation effect's.

Common exam traps

  • Security selection uses benchmark weight and the return difference — the allocation effect uses the weight difference and the benchmark's return instead. Swapping the two is the most common error.
  • A sector can have a negative security selection effect even if the portfolio outperformed overall. Financials contributed −0.25% here despite the portfolio beating its benchmark by 0.9% in total.
  • The three effects must sum to the total excess return. Use that identity to check or back out a missing figure.
  • This is distinct from interaction effect, which captures what happens when overweighting and better stock-picking occur in the same sector together.

Where this is taught

Free preparation for NISM Series XXI-A

Related terms

← All terms
Something look wrong? Report it