Performance and attribution
In this chapter: Time-weighted returns, attribution by source · Benchmark selection
Performance measurement uses TWR (neutral to cash flows) compared against a benchmark. Attribution decomposes returns into allocation effect (over/under-weighting sectors), selection effect (stock picks within sectors), and interaction. The benchmark choice matters — it must match the mandate.
Brinson attribution: Allocation = (PortfolioWeight − BenchmarkWeight) × BenchmarkReturn; Selection = BenchmarkWeight × (PortfolioReturn − BenchmarkReturn); Interaction = (PortfolioWeight − BenchmarkWeight) × (PortfolioReturn − BenchmarkReturn). Sum = total active return. Benchmark choice: NIFTY 50 for large-cap mandates, NIFTY 500 for multi-cap, NIFTY Midcap 150 for mid-cap, NIFTY Smallcap 250 for small. Custom benchmarks (50-50 NIFTY/Midcap) for hybrids. SEBI now requires PMS to compare against a SEBI-defined benchmark.
A nuance: the difference between "benchmark-agnostic" PMS and "benchmark-aware" PMS. Some managers explicitly run portfolios with no benchmark constraint (concentrated, contrarian). For these, comparing to NIFTY 50 over 3 years can show massive over- or under-performance — neither is meaningful in isolation. The right metric is risk-adjusted: Sharpe Ratio over 3-5 years, including drawdowns. Clients evaluating PMS need to look beyond simple alpha numbers.