Active Management

The mechanics begin with the active weight: for each security, the portfolio weight minus that security's weight in the benchmark. Because both weight sets sum to one, active weights necessarily sum to zero, so every overweight is funded by an underweight somewhere else. Active return over a period is then the sum, across all securities in the combined universe, of each active weight multiplied by that security's return. Equivalently, it is portfolio return minus benchmark return.

That identity produces the decomposition candidates are expected to recognise. Active return can be split into an allocation component, arising from over- or underweighting sectors, countries or factors relative to the benchmark, and a selection component, arising from holding different securities within a sector than the benchmark does. A manager who owns only index constituents but weights them differently is still generating active weights, and therefore active return, entirely through the allocation and selection channels.

Deviating from the benchmark creates active risk, measured as the standard deviation of active returns over time and commonly called tracking error. The information ratio scales mean active return by active risk, giving return earned per unit of benchmark-relative risk taken; it is the natural yardstick for comparing managers who run mandates of different aggressiveness. An approximate relation links the information ratio to the manager's skill, expressed as the correlation between forecasts and realised outcomes, and to the number of genuinely independent decisions taken, implying that modest skill applied across many uncorrelated bets can rival high conviction applied narrowly.

One structural constraint frames all of this. Since the aggregate of all investors holds the market, active returns across the market must sum to approximately zero before costs and are negative in aggregate after fees, trading and taxes. Active management is therefore a competition for a fixed pool, not a source of collective excess return.

Real mandates sit along a spectrum. Enhanced indexing accepts a small tracking-error budget and adds modest tilts; core active mandates run wider deviations; concentrated and long-short strategies may bear little resemblance to the index. Active share, the fraction of the portfolio that differs from the benchmark in holdings terms, complements tracking error: a portfolio can have high active share yet low tracking error if its deviations are spread across many correlated names.

Investment policy statements typically constrain active management explicitly, through tracking-error limits, sector deviation bands, position caps and permitted use of derivatives or leverage. Fees matter disproportionately because they are certain while alpha is not, which is why fund boards scrutinise closet indexers charging active fees for near-benchmark exposure, and why fee structures with layered management and performance charges face a higher hurdle.

Expect to compute active weights from a pair of holdings tables, aggregate them into an active return, and attribute that return between allocation and selection. Expect to calculate an information ratio from a series of active returns and to interpret it alongside the Sharpe ratio, recognising that the Sharpe ratio evaluates total risk while the information ratio evaluates benchmark-relative risk.

Equally important is the qualitative argument. You should be able to state the conditions under which active management is rational, principally the existence of mispricing that the manager can identify at a cost lower than the value recovered, and link this to market efficiency: the more efficient the segment, the harder positive net active return becomes. You should also be able to justify a benchmark as appropriate, since an unsuitable or unrepresentative benchmark makes every active statistic derived from it meaningless.