Two approaches
| Fundamental | Quantitative | |
|---|---|---|
| Inputs | Company research, management meetings | Data, factor scores |
| Breadth | Fewer, deeper positions | Many positions |
| Process | Judgment | Systematic rules, backtests |
| Key risks | Behavioural bias, concentration | Model risk, crowding, data mining |
Where active return comes from
Active return = factor contributions + stock selection. Decompose active risk the same way: a portfolio with large factor bets can have high tracking error even if stock picks are neutral.
Active share (how different the holdings are) and tracking error (how different the returns are) together describe the style:
| Low tracking error | High tracking error | |
|---|---|---|
| High active share | Diversified stock picker | Concentrated stock picker |
| Low active share | Closet indexer | Factor bettor |
The fundamental law
IR ≈ TC × IC × √BR. A modest edge applied across many independent decisions can beat a large edge on a few. Constraints that block the manager from acting on views reduce TC.
Sizing positions
Size by conviction and risk contribution, cap by liquidity (e.g., no position larger than a few days’ trading volume), and watch sector and factor concentration.
Essay practice
Case. Manager A runs 40 stocks with active share of 90% and tracking error of 8%. Manager B runs 300 stocks with active share of 45% and tracking error of 7%, mostly from a value tilt.
Task. Characterize each manager’s style and state one implication for fees.
Model answer
A is a concentrated stock picker: holdings differ greatly from the benchmark, and risk comes from stock selection. B is a factor bettor: most active risk comes from the value tilt. The value exposure in B could be bought cheaply through a smart-beta fund, so B’s fee should be judged on returns beyond the factor tilt.
Exam focus
Contrast fundamental and quant approaches, apply the fundamental law and interpret active share versus tracking error.