Measuring value added
Active return RA = RP − RB = Σ Δwi × Ri
where Δwi = portfolio weight − benchmark weight. Value added can come from asset allocation (over- or underweighting classes) and security selection within each class.
Active risk (tracking error) = standard deviation of active returns.
Sharpe ratio and information ratio
- Sharpe ratio = (RP − rf) ÷ σP.
- Information ratio = RA ÷ σA. Unchanged if the active position is scaled up or down.
Combining the benchmark with the active portfolio:
SR_P² = SR_B² + IR²
Optimal active risk σA* = (IR ÷ SR_B) × σB
SR_B = 0.40, IR = 0.30: SR_P = 0.50; with σB = 16%, optimal active risk = 12%.
Investors should pick the manager with the highest IR, then choose how much active risk to take.
The fundamental law
IR ≈ TC × IC × √BR
E(RA) ≈ IR × σA
- IC (information coefficient): correlation between forecasts and actual returns. Skill.
- BR (breadth): number of independent decisions per year.
- TC (transfer coefficient): how well forecasts become weights (1 if unconstrained).
A small edge applied many times can produce a high IR, which is why quantitative strategies emphasize breadth. Constraints reduce TC and waste skill.
Applications
- Market timing: few independent bets a year, so breadth is low; a high IC is needed.
- Security selection: many bets, so a modest IC can work.
- Practical limits: estimating IC ex ante is hard, bets aren’t fully independent, and costs eat into returns.
Exam focus
Compute active return, IR and the optimal combined Sharpe ratio, apply the fundamental law, and explain the role of each component.