Advanced3 min read

Performance Attribution

Decomposing a return into the decisions that produced it. Without it, a good year and a lucky year look identical.

MadStockAlerts Research · Updated August 29, 2026

What to take away

  • Attribution separates allocation decisions from selection decisions.
  • A return can be entirely explained by exposure to something that rose.
  • Currency is a separate contributor in any international portfolio.
  • Attribution is descriptive; it does not establish skill.
  • The most useful version compares against what would have happened doing nothing.

MAD Academy Training Video · 0:46

Where the Return Actually Came From

Attribution splits a result into the decisions behind it, which usually shows the return came from somewhere nobody intended.

This lesson is part of a Stock Alerts + Tools plan.

See the library

The basic decomposition

EffectThe question it answers
AllocationDid being overweight or underweight a sector or asset class help?
SelectionWithin each sector, did the specific holdings beat that sector?
InteractionThe combined effect of being overweight a sector where selection was also good
CurrencyHow much of the return came from exchange rates rather than from assets

The distinction between the first two is what makes the exercise worth doing. A portfolio that beat its benchmark because it happened to be heavily weighted in the sector that rose most has a different explanation from one that beat it by holding better companies within every sector.

Why a headline return explains nothing

Consider a portfolio that returned twenty percent in a year when its benchmark returned twelve. That single comparison is consistent with several completely different explanations.

  • A large overweight in one sector that rose, with no selection effect at all.
  • Consistent selection ahead of the benchmark within every sector.
  • One position that rose enormously, with everything else lagging.
  • A currency movement in an international allocation.
  • Higher beta, which produced a higher return in a rising market and will do the reverse.

The fifth is the one that most frequently masquerades as skill. Holding more risk produces more return in a rising market by construction, and attribution against a risk-matched benchmark is what separates the two.

One return, five explanations
One return, five explanations02468More risk produces more return in arising market, by constructionSector overweightSelectionOne positionCurrencyHigher betaContribution to the 8 point gap

Scroll the chart sideways to see all of it.

A portfolio up 20 percent against a benchmark up 12 is consistent with all of these. Attribution is what separates them, and the last one is not skill.

What it cannot establish

Attribution is a description of what happened, not evidence about what will. It identifies which decisions contributed and says nothing about whether those decisions were repeatable.

  • A positive selection effect over one year is well within the range of chance.
  • Attribution over a period of one market regime describes that regime.
  • It cannot distinguish a good decision with a good outcome from a bad one with a good outcome.
  • It is computed from realised returns, so it inherits every sample-size problem that applies to them.

The version worth doing

For an individual portfolio, a full institutional attribution is usually more machinery than the question requires. A simpler decomposition answers most of it.

  1. 1Compute the do-nothing returnWhat a single index fund matching the overall exposure would have returned over the same period.
  2. 2Compute the actual returnNet of every cost, including tax where it applies.
  3. 3Identify the largest contributorsBoth directions. Usually a small number of positions explain most of the difference.
  4. 4Ask whether those were the intended decisionsA result driven by one position that was never a deliberate bet is a description of luck.

The fourth step is the one that produces the finding. A portfolio's outcome frequently traces to a small number of positions that were not the ones the process was designed around.

Attributing an individual's results

Two effects dominate most individual portfolios' outcomes and appear in no standard attribution model, because institutional portfolios do not have them.

  • Cash drag and timing of contributions: money added or withdrawn at particular moments changes the realised return relative to a time-weighted figure.
  • Behaviour gap: the difference between a holding's return and the return its holders actually earned, which studies consistently find is negative.

The second is the more consequential. Research comparing fund returns against dollar-weighted investor returns in the same funds finds investors earning materially less, because money arrives after good performance and leaves after bad.

That gap is a decision effect rather than a selection effect, and it is frequently larger than any difference between the holdings that were chosen. It is also the one thing on this list that is entirely within a holder's control.

Educational content only. MadStockAlerts provides market commentary, research, and educational content. It is not personalized investment advice, and nothing here is a recommendation to buy or sell any security. Trading and investing involve substantial risk, including loss of capital. See the Risk Disclosure and Customer Agreement.