Arithmetic of Active Management
Compare active returns with the market before and after costs
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 99%
The arithmetic starts from a market identity rather than a forecast. Broad market indexers hold their share of the market and earn the market return before costs. Everyone else collectively holds the remaining securities, so the active group must also earn the market return before costs. Once management fees, research expenses, trading, and other costs are deducted, that group necessarily earns less in aggregate. An individual active manager can outperform, but another participant must underperform by a corresponding amount before costs. Sharpe uses this mechanism to challenge smart-beta marketing and attractive historical backtests: a tilt can win for long periods, but the arithmetic does not show who can identify the winning tilt beforehand. The framework turns an outperformance claim into an accounting question about costs, counterparties, and repeatability.
Origin
Sharpe says he wrote this simple market-aggregation argument years earlier in a piece titled Arithmetic of Active Management.
Core principles
- 01All investors collectively hold the market portfolio.
- 02Indexers receive the market return before costs.
- 03Active investors as a group must also receive the market return before costs.
- 04After costs, the active group must trail the lower-cost market portfolio.
- 05Persistent winners require corresponding losers on the other side.
How to run it
- 1
Define the market whole
Choose the market whose securities and investors are being analysed. Treat all holdings in that market as one complete portfolio.
Pro tip Keep the market definition consistent across index and active groups.
Watch out A partial universe can exclude counterparties and break the accounting identity.
- 2
Separate indexers and active investors
Identify broad market indexers and group all non-index positions as active. Recognise that the groups together must hold the whole market.
Pro tip Classify factor tilts and smart-beta strategies as active departures from market weights.
Watch out A rules-based label does not make a non-market tilt passive for this arithmetic.
- 3
Apply the before-cost identity
Assign the market return to the aggregate holdings before costs. If one active subset earns more, another subset must earn less.
Pro tip Ask explicitly who is on the losing side of any persistent edge.
Watch out Outliers do not invalidate the aggregate identity.
- 4
Subtract all costs
Deduct management fees, trading costs, research expenses, and implementation drag from each group. Compare the resulting net returns.
Pro tip Use realised all-in costs rather than headline expense ratios alone.
Watch out Ignoring costs removes the central practical implication of the framework.
- 5
Interrogate the claimed exception
For a strategy claiming repeatable outperformance, ask how it identifies mispricing in advance and why the losing counterparties will persist. Distinguish genuine prospective evidence from a fitted backtest.
Pro tip Look for a mechanism that survives different periods, not just one strong historical window.
Watch out A successful decade selected after the fact does not establish forecasting ability.
In the wild
If the total market earns ten percent before costs, broad market indexers earn that market return before their small costs. Active investors collectively own the rest of the same market, so their aggregate before-cost return must also reconcile to the market; their higher combined costs then reduce the net result.
→ The comparison exposes a structural net-return disadvantage for the active group without predicting any individual fund.
A factor strategy shows that a value tilt would have produced exceptional returns over the last decade. The evaluator classifies the tilt as active, adds implementation costs, asks which investors supplied the offsetting underperformance, and looks for evidence that the winning period could have been identified beforehand.
→ An appealing historical chart becomes a testable prospective claim rather than a marketing conclusion.
Common mistakes
Looking only at winners
A manager's outperformance is incomplete evidence unless the analysis also accounts for the investors who underperformed.
Calling a factor tilt passive
A systematic portfolio that departs from market weights remains an active bet for this market arithmetic.
Trusting an attractive backtest
Sharpe warns that extensive data, fast computers, and marketing can produce strategies that look compelling after the fact.
Is it for you?
Best for
It is best for investors evaluating active funds, factor tilts, backtests, and indexing alternatives.
Not ideal for
It does not identify in advance the rare individual manager who may outperform or model every issuance and market-friction edge.
From the transcript
“They have to earn before costs the same that the other guys do, and after costs they're gonna earn less.”
“But if so, somebody's routinely doing worse.”
“There's very little evidence that people can tell in advance what's coming next”
From the episode
Interview With William Sharpe: Masters in Business (Audio)
William Sharpe