Choose the question before the winner
A broad equity index can help describe the stock market. A bond index can help assess a specified fixed-income exposure. A policy blend can help evaluate a portfolio intended to hold several asset classes. There is no single benchmark that answers every question.
Choosing the easiest comparison after seeing the outcome can make a result look stronger without improving the strategy. A defensible benchmark is stated in advance where possible, alongside the reason it fits the mandate.
Different exposure can explain an apparent gap
Imagine a fictional policy portfolio starts 50% in stocks and 50% in bonds. Stocks gain 10% and bonds are unchanged over one period. Ignoring costs, the policy mix gains 5%. Comparing it only with the 10% stock result would show a five-point lag but would not isolate implementation skill.
If an implemented portfolio with that mandate gains 4.8%, the 5% policy reference asks a more focused question. Comparing with stocks can still describe the opportunity cost of lower equity exposure, provided that is the stated question. Both comparisons can be informative when their purposes are clear.
Hold conventions steady
Match dates, currency, distribution treatment and costs as far as the question requires. A total-return portfolio compared with a price-only index can receive credit simply because income was counted on one side. A gross strategy compared with a net product introduces another mismatch.
A specialized fund may also differ in size, sector, geography or risk from a familiar headline index. Those differences are not automatically flaws, but they need to be identified before a return spread is labeled skill or alpha.
One reference may not explain every contribution
A blended portfolio can benefit from several comparisons: the policy mix, individual sleeves and a broad market reference. These should be labeled rather than combined into a convenient narrative. Attribution can then examine allocation and selection effects more carefully.
An index is also distinct from an actual product that follows it. A realistic alternative can involve fund fees, taxes or trading frictions. State which comparison is theoretical and which approximates an implementable choice. That transparency makes the result easier to assess.
A stock/bond policy mix earns less than an all-stock index during a strong stock period. Has that alone proved poor security selection?
Show the explanation
No. Different allocation can explain much of the gap. Compare with the intended policy mix to ask a more focused implementation question.
Sources and editorial notes
Original explanations and hypothetical examples prepared with AI assistance. Primary references checked October 16, 2026. Numerical examples are hypothetical unless explicitly identified otherwise.
- CFA Institute: Performance Measurement and Attribution
Benchmark selection before evaluation, excess return, attribution and risk analysis.