High active share funds are now the worst performers

Active share measures how far a fund's holdings stray from its benchmark index. Since it was introduced in 2009, a high reading has been treated as evidence of a manager worth paying for. New research covering US fund data from 1984 to 2024 finds that the relationship has now reversed.

The case for indexing does not rest on the claim that no active manager can beat the market. It rests on the claim that picking the winners in advance is close to impossible. Active share was the industry's answer.

Martijn Cremers and Antti Petajisto introduced the measure in a 2009 paper in The Review of Financial Studies. It is the proportion of a fund's holdings that differs from its benchmark, running from zero for a tracker to one for a fund holding nothing its index holds. From 1990 to 2003, they found, the funds that deviated most from their benchmarks outperformed.

The measure spread quickly. Morningstar reports it as a fund evaluation metric. European regulators took it up as well. In 2016 ESMA used it, alongside tracking error, to screen more than 2,600 UCITS equity funds, and concluded that between 5 and 15 per cent of them could potentially be closet indexers. High conviction became a selling point in its own right.

High active share now points the other way

Since 2010 the relationship has run in reverse. Hannah Unterberg, a doctoral candidate at the University of California, Irvine's Paul Merage School of Business, sorted US equity funds into quintiles by active share. Between 1990 and 2009, she found, the most active quintile beat the least active by 0.85 percentage points a year gross of fees, measured against a four-factor model. Between 2010 and 2024 the most active quintile trailed the least active by 1.11 percentage points before fees, and by 1.51 points after fees. The gross spread moved by 1.96 percentage points between the two periods.

Her paper, Passive Flows, Active Woes, covers US domestic equity mutual funds and ETFs from 1984 to 2024, using CRSP fund data merged with Thomson Reuters holdings. It is a working paper and has not been peer reviewed, though it won the Two Sigma award for the best paper on investment management at this year's Western Finance Association conference.

The reversal coincides with a change in where investors keep their money. On Investment Company Institute data cited in the paper, index funds and index ETFs have grown from 19 per cent of US equity fund assets to more than half since 2010.

The premium was shakier than it looked

The pre-2010 premium was not statistically significant. The spread of 0.85 percentage points carried a t-statistic of 1.16, short of the conventional threshold. And in 2016 Andrea Frazzini, Jacques Friedman and Lukasz Pomorski of AQR re-examined the original data. They concluded that active share correlates with benchmark returns rather than predicting fund returns, and that within individual benchmarks it was as likely to correlate positively with performance as negatively. Cremers has disputed that reading.

So the fairest description is not that a proven premium vanished. It is that a premium which was already shaky has turned negative. The post-2010 gross figure is itself significant only at the 10 per cent level, though the net figure reaches 5 per cent.

The reversal does survive the obvious objection. When Unterberg repeats the test comparing funds only with others tracking the same benchmark in the same month, and separately for funds tracking large-cap and small or mid-cap indices, the negative relationship holds.

The most active funds take the hardest hit

Unterberg's explanation is mechanical rather than behavioural. When investors move money out of active funds and into index funds, two things happen at once. The index money is invested in benchmark weights, whatever the price. The departing active money forces managers to sell what they already hold, roughly in proportion to how much of it they hold. Stocks that active managers overweight face selling pressure; stocks they underweight get bought. The further a portfolio sits from its index, the harder it is squeezed.

The most persuasive evidence comes from the calendar. US retirement plan contributions arrive in the first few trading days of each month and cannot be responding to that month's performance. In those days, Unterberg finds, when passive flows are one standard deviation above average, funds in the lowest active share quintile earn 2.6 basis points more per day than they do over the rest of the month, and those in the highest earn 2.1 basis points less. Flows into and out of active funds produce no comparable pattern.

Flow pressure is another route by which active funds can underperform without the manager having picked badly.

Structural does not mean temporary

Unterberg concludes that the shortfall 'reflects structural demand headwinds rather than declining manager skill'. That is a finding about causes, and the active industry will read it as exoneration: we are not worse, we are being squashed, and when the flows turn we will shine again. The claim is a more rigorous version of an argument some fund managers have been making for years.

Her own data makes the second half of it harder to sustain. Price pressure from active fund flows largely reverses within one to three years. The price impact of passive flows remains at around half its initial size after three years, because a reallocation towards index funds is a lasting change in who owns what rather than a temporary liquidity shock.

The simpler point is that the cause of a shortfall does not alter its size. Nor is the drag confined to the most active funds. Before fees, the asset-weighted active sector produced a four-factor alpha of minus 0.01 per cent a year up to 2009 and minus 0.66 per cent from 2010. The sector now trails the market before costs as well as after.

Neither end of the range is the answer

For years I told people that if they were set on using an active fund, they should at least pick one with a high active share. I would not say that now. Between 2010 and 2024, funds in the lowest active share quintile lost 0.90 per cent a year after fees on the four-factor measure, compared with 2.41 per cent for the highest. They were less exposed to the drag, largely because their holdings resemble the index, the feature that makes them poor value at an active price.

UK funds still advertise the number. One currently sets a minimum active share target of 90 per cent and presents it as evidence that it is genuinely actively managed. The claim is accurate as far as it goes.

The measure still tells you how different a fund is from its index. It no longer tells you whether that difference is worth paying for.

Resources

Cremers, K. J. M. (2017). Active share and the three pillars of active management: Skill, conviction, and opportunity. Financial Analysts Journal, 73(2), 61–79.

Cremers, K. J. M., & Petajisto, A. (2009). How active is your fund manager? A new measure that predicts performance. The Review of Financial Studies, 22(9), 3329–3365.

European Securities and Markets Authority. (2016). Supervisory work on potential closet index tracking (ESMA/2016/165). ESMA.

Frazzini, A., Friedman, J., & Pomorski, L. (2016). Deactivating active share. Financial Analysts Journal, 72(2), 14–21.

Unterberg, H. (2026). Passive flows, active woes: Passive investing and the decline of active mutual fund alpha [Working paper]. Paul Merage School of Business, University of California, Irvine.

Subscribe to the TEBI Digest

Evidence-based investing journalism, free to your inbox. Add your email address to the 'Subscribe' box on the TEBI homepage and we'll do the rest.