Learning Outcome Statement
Describe attributes of an effective attribution process.
On the exam: The question gives portfolio and benchmark returns, attribution effects, process controls, and benchmark facts. It asks you to identify an effective-process attribute or diagnose a flaw in a short constructed response or selected alternative.
Attribution is the audit trail between a portfolio's result and its benchmark's result. It separates the portfolio's own result from the result earned by the benchmark. You care because manager appraisal is only as credible as the comparison, arithmetic, and controls behind it.
| Concept | Reference | Check |
|---|---|---|
| Active return | Portfolio and benchmark | Portfolio minus benchmark |
| Contribution | Portfolio return | Components sum to total return |
| Attribution | Benchmark return | Effects reconcile to active return |
| Comprehensive | All material effects | Nothing material omitted |
| Accurate | Data and calculations | Results reconcile |
| Timely | Reporting timetable | Available when needed |
| Transparent | Methods and assumptions | Reviewers can trace results |
| Repeatable | Rules and controls | Same process across periods |
| Decision-useful | Mandate and appraisal | Supports a decision |
Contribution describes a component of total portfolio return, while attribution explains active return relative to the benchmark. On the arithmetic percentage-point basis used here, attribution effects must add to active return before they support manager appraisal.
Keep the two checks distinct when reading a report. First ask whether the portfolio components reconstruct its total return; then ask whether the attribution effects explain the difference from the benchmark. A report can pass one check and fail the other. That distinction helps you locate an omission or calculation error before judging the manager.
| Comparison point | Contribution | Attribution |
|---|---|---|
| Reference | Portfolio return | Benchmark return |
| Arithmetic check | Components sum to total portfolio return | Effects add to active return |
To measure active performance, start with the portfolio return, $R_P$, then subtract the benchmark return, $R_B$. The difference is the active return, $R_A$.
$$ R_A = R_P - R_B $$
Start with $R_P$ because it is the result you measured. Subtract $R_B$ because it is the comparison return. A higher $R_P$ raises $R_A$, while a higher $R_B$ lowers it. The sign shows outperformance or underperformance, not persistent skill.
These attributes work together rather than as isolated labels. A comprehensive, accurate result can still fail a client if it arrives late or cannot be traced. A repeatable process makes comparisons across reporting periods meaningful, while decision-useful output connects the analysis to the mandate.
Benchmark quality comes before manager appraisal. A sound benchmark is specified in advance, investable, measurable, relevant to the mandate, and consistently applied. A positive effect can reflect a poor comparison rather than manager skill, and a result from one period cannot establish persistence.
Use the sequence below as a gate. If benchmark quality or reconciliation fails, investigate before appraising.
Read the diagram from left to right as a sequence of evidence gates. Benchmark suitability comes first because every later comparison depends on it. Reconciliation then checks that the portfolio records and attribution explain the reported result. Only after those checks pass should you compare periods for persistence. This order prevents a clean-looking appraisal from resting on a weak reference point or unexplained arithmetic.
Loading diagram…
Only a reconciled result from a repeatable process can support a stronger appraisal conclusion.
Worked example
You are handed one reporting period for a growth-equity manager. The portfolio, broad growth benchmark, and arithmetic attribution effects are listed in the report. You need to decide whether the process supports appraisal, not merely praise a positive selection effect.
| Input | Value |
|---|---|
| Portfolio return | 12 percentage points |
| Benchmark return | 10 percentage points |
| Allocation effect | 1 percentage point |
| Selection effect | 1.5 percentage points |
| Interaction effect | -0.5 percentage points |
| Attribution basis | Arithmetic percentage-point scale |
| Benchmark facts | Specified in advance, investable, measurable, relevant, consistently applied |
| Process review | Completeness, reconciliation, timeliness, transparency, repeatability, usefulness |
The bars show the running arithmetic return after each effect is added to the benchmark.
Follow the running value from the benchmark through each effect, rather than treating the bars as separate performance results. Allocation raises the starting return, selection raises it again, and interaction brings it back down. The final bar should match the portfolio return, while the total of the effects should match active return. Use both checks to see whether the attribution is complete and internally consistent.
Loading diagram…
| Stage | Running return |
|---|---|
| Benchmark | 10.00 percentage points |
| After allocation | 11.00 percentage points |
| After selection | 12.50 percentage points |
| After interaction | 12.00 percentage points |
| Portfolio | 12.00 percentage points |
- Calculate active return because the benchmark is the comparison base. $R_A = 12 - 10 = 2.00$ percentage points.
- Add the effects because attribution must explain the entire active return. $1.00 + 1.50 + (-0.50) = 2.00$ percentage points.
- Walk the bridge because the order exposes unexplained movement. Starting from $10.00$, the running returns are $11.00$, $12.50$, and $12.00$ after allocation, selection, and interaction.
- Confirm the final running return because the bridge must land on the portfolio return. It lands at $12.00$ percentage points and matches the portfolio return.
The effects reconcile, but the one-period result supports investigation rather than a conclusion of manager skill.
Changed condition
Keep the same manager, benchmark, reporting period, and arithmetic method. Change only the interaction effect.
| Measure | Before | After |
|---|---|---|
| Interaction effect | -0.50 percentage points | 0.50 percentage points |
| Active return | 2.00 percentage points | 2.00 percentage points |
| Attribution effects total | 2.00 percentage points | 3.00 percentage points |
Before looking at the totals, predict that the effects total will rise while active return stays unchanged. The effects total now exceeds active return, so the effects no longer reconcile. That break points to a process or classification issue, not new evidence of skill. Now add the changed effects and compare them with active return before reading on.
Common error
You see a positive selection effect and conclude that the manager has skill. That feels right because selection is an active decision and the effect is positive. The result comes from one period or rests on an untested benchmark instead of a reconciled, repeatable process, so treat it as evidence to investigate.
Go deeper
Arithmetic and geometric attribution use different ways to combine effects.
| Method | Effect combination | Result compared |
|---|---|---|
| Arithmetic | Add percentage-point effects | Active return |
| Geometric | Compound relative returns | Relative performance |
Use the arithmetic method when effects are reported as percentage-point additions to active return. A geometric method compounds relative performance across periods. Do not add geometric effects as if they were arithmetic effects.
Check yourself
Practice
Which statement best describes an effective attribution process?
Revision snapshot
A sound attribution process is comprehensive, accurate, timely, transparent, repeatable, and decision-useful. Active return equals portfolio return minus benchmark return, $R_A = R_P - R_B$. Attribution explains active return relative to the benchmark. Contribution describes a component of total portfolio return.
| Exam rule | Candidate action |
|---|---|
| Reconciliation | Match effects to active return before appraisal |
| Positive effect | Investigate, then test benchmark quality and persistence |
| Method | Keep arithmetic and geometric attribution separate |
The exam gives returns, effects, process controls, and benchmark facts. Diagnose the process before calling a manager skilled.
Unlock the full lesson and continue reading
Signup to continue reading this lesson and unlimited access questions, flashcards, AI notes, and more