On this page
A portfolio can meet its headline tracking-error budget and still take risk in the wrong places. Risk-adjusted performance monitoring therefore compares the realised outcome with the mandate, the forecast risk with its budget, and the sources of both risk and return. A single attractive return or acceptable total risk figure cannot establish that the manager followed the intended process. For FRM® Part II, the useful habit is to investigate what produced the number before judging it.
What is being monitored?
Risk planning defines the objective and acceptable exposure. Budgeting translates that objective into limits or allocations. Monitoring compares current activity and outcomes with those expectations. The three stages answer different questions: what risk should be taken, how much is assigned to each source, and whether the actual portfolio still behaves as intended. GARP’s Risk Principles for Asset Managers says a portfolio should have a benchmark or other objective and that managers should examine both what happened and why it happened. The latter inquiry is performance attribution.
Tracking error is the variability of a portfolio’s return relative to its benchmark. A forward-looking tracking-error estimate is a model forecast based on current positions, exposures and assumptions. A realised tracking-error measure comes from observed active returns over a stated historical period. Treating them as interchangeable is a common mistake. GARP cautions that forward risk metrics are estimates and that past realised risk is not necessarily a good forecast. If a portfolio has forecast tracking error near its budget, you still need to check whether the model captured the important exposures and whether the positions match the mandate.
A risk budget can be stated at the total portfolio level and decomposed by sector, factor, security or other source. The total number is a summary of interacting exposures; it does not reveal which bets created it. Offsetting changes can leave the total near its original value even when a concentrated active position appears. Think of a value manager whose stock-selection risk falls while a growth-factor tilt rises. The reported total may remain stable, yet the strategy no longer represents the same process. That is why a risk analyst looks at composition as well as aggregate magnitude.
Work through a deliberately simple example
Take an invented benchmark-relative portfolio with a forecast tracking-error budget of 4%. The current forecast is also 4%, so the headline measure is on budget. Suppose the risk report originally assigned most active exposure to company selection within a value mandate. A later report shows growth and momentum positions dominating the exposure, while selection risk has declined. No real manager or market data are being described here. The comparison is designed to show how equal totals can conceal a change in the driver of risk.
First, confirm the method: use the same benchmark, risk model, horizon and assumptions for both forecasts. A change in the definition can mimic style drift. Next, inspect the factor and position breakdown. Ask whether the growth exposure is allowed by the mandate and whether it is deliberate or incidental. Then compare the current forecast with any predefined tolerances for each component. The correct conclusion is conditional: a 4% total does not itself prove a breach, but it cannot clear the portfolio either. A material, unexplained mandate departure warrants investigation and escalation under the firm’s policy.
Now add an invented performance outcome. The portfolio earns 8% while its benchmark earns 6%, an active return of 2 percentage points. Suppose an attribution report assigns 1.5 points to a concentrated growth position and 0.5 point to other choices. The pieces sum to 2 points; no return was created by the arithmetic. Even if the growth position was profitable, attribution shows that most of the active result came from the very exposure that may be outside the intended value process. A good outcome does not retrospectively make an unapproved exposure consistent with the mandate.
The example also shows why “skill versus luck” needs care. Attribution identifies the source of realised return, not the manager’s intention or the future repeatability of that source. A single period cannot prove that an exposure was skillful, accidental or persistent. An analyst would inspect the documented investment rationale, the timing of the position, the risk taken to earn the return and evidence from other periods. CFA Institute’s performance evaluation material likewise warns that appraisal ratios depend on limited, imperfect return data and need to be considered with broader analysis.
Choose the comparison that fits the question
A raw return asks how much the portfolio gained or lost. Active return compares the portfolio with its benchmark. As the CFA Institute comparison explains, a Sharpe ratio scales excess return over a risk-free reference by total return volatility. An information ratio scales active return by tracking error. These are different denominators and therefore different questions. The CFA Institute risk-adjusted performance overview identifies Sharpe and information ratios as common appraisal measures. For a benchmark-relative mandate, active return and tracking error are directly relevant, but even a favourable information ratio can coexist with an unacceptable source of risk.
Consider the invented active return of 2 percentage points and a realised tracking error of 4% measured over a compatible period. The simple information ratio would be 0.5, found by dividing 2 by 4. This calculation is illustrative only, and it should not be mixed with the earlier 4% forward forecast as though one were the other. If the active return and realised tracking error use different horizons or inconsistent return definitions, the ratio is not interpretable. If the benchmark is unsuitable, the ratio can reward an exposure the mandate never intended to take.
A benchmark is not just a convenient market index. It is the reference against which the manager’s assigned job is judged. If a value mandate is assessed against an unrelated broad portfolio, the comparison may blur the very style drift the monitor should detect. The question to ask is whether the benchmark reflects the investable universe and objective established before the outcome was known. CFA Institute’s portfolio-performance reading warns that benchmark misspecification undermines measurement and attribution. That warning matters for FRM monitoring because a precise ratio computed against the wrong reference is still a poor control.
From observation to action
A sensible monitoring process defines thresholds and responsibilities before an unusual result appears. A small forecast change within documented tolerance can remain under routine observation. A larger or unexplained change triggers discussion, model checks and a review of positions. A clear mandate breach requires escalation through the agreed risk governance process. The colour of a dashboard is only shorthand for those underlying criteria. It should not be inferred from whether a return is positive or negative: unusually high performance can be the symptom of an unintended exposure.
When investigating, distinguish model error, genuine market movement and manager choice. A factor estimate may move because market volatility changes while positions stay fixed. A portfolio can also drift because a holding appreciates relative to others, even without a new trade. Alternatively, an intentional position change may violate the risk budget. These causes imply different responses. Reconcile positions and data first, rerun the risk estimate if necessary, then discuss the economic exposure and mandate. Jumping straight from a red-looking metric to a conclusion about skill or misconduct would outrun the evidence.
For exam purposes, answer in this order: state the observed variance, identify the relevant budget or benchmark, decompose the risk and return, then recommend the next investigation or escalation condition. Do not let a good return excuse a risk breach, or a bad return alone prove poor process. The FRM programme page gives programme context, and the Omni how-it-works page describes the learning format. The central exam distinction remains simple: a total measure tells you how much; attribution and decomposition tell you where it came from.
Sources
- Risk Principles for Asset Managers | GARP · garp.org · Retrieved
- Risk-Adjusted Performance Measures | CFA Institute · rpc.cfainstitute.org · Retrieved
- Portfolio Performance Evaluation | CFA Institute · cfainstitute.org · Retrieved
- The Sharpe Ratio and the Information Ratio | CFA Institute · rpc.cfainstitute.org · Retrieved


