Over 90% of active funds fail to beat their index

arvy's Teaser: "Over 90% of active funds fail to beat their index." That number ends most investment discussions before they start. In May 2026, three finance professors recalculated the report it comes from — and found three calculation rules that systematically distort the picture. Corrected, the figure is 55%. But the real insight is a different one, and it hits Swiss investors harder than anyone else.

The number everyone knows

Talk about investing anywhere and you'll hear it within two minutes. In every YouTube video, every Reddit thread, on every finfluencer slide:

"Over 90% of active funds fail to beat their index."

The source is almost always SPIVA — "S&P Indices Versus Active," published twice a year by S&P Dow Jones Indices since 2002. It is carefully built and consistent across decades. It settled a debate that had been open for a very long time, and it taught the industry a dose of humility it badly needed.

But almost nobody has examined how the number is produced. Until now.

Three rules, three problems

In May 2026, Martijn Cremers (University of Notre Dame), Jon Fulkerson (Dayton) and Timothy Riley (Arkansas) recalculated the 2024 SPIVA report. Cremers is not a fringe figure — he created "Active Share," the metric the industry uses to measure how active a fund actually is.

They found three choices that systematically pull the result in one direction.

One: every fund that disappears counts as a loser

Funds get closed or merged — constantly. SPIVA treats every fund that leaves the sample as an underperformer, regardless of how it actually did. Over 20 years, a large share of all funds disappears. That makes the 90% figure, to a meaningful degree, a statement about fund closures rather than investment results.

Two: a 50 million fund counts the same as a 50 billion fund

SPIVA counts funds, not dollars. In reality most capital sits in a small number of large funds. A niche product and a flagship move the statistic identically — even though a thousand times more investor money rides on the second. What you care about isn't how many funds lose, but how much money loses.

Three: the benchmark isn't buyable

An index costs nothing. No fee, no transaction costs, no tracking error, no tax. You can't buy it either. What you can buy is an ETF, and that costs money. So SPIVA measures against a fiction, not against your actual alternative.

What changes

Source: Cremers, Fulkerson, Riley (SSRN, May 2026), own illustration.

"Essentially nobody makes it" becomes "roughly every second dollar doesn't." That is a different claim. Not the opposite — but different.

But that isn't the real point

Before anyone opens the champagne for active management: even after the correction, the majority of investor money in US equities still loses. 55% is less than 92% — but 55% is more than half. Anyone turning this study into "active wins" hasn't read it.

The more interesting question is a different one: who exactly is SPIVA measuring?

Answer: everything that calls itself active. And that's where the problem starts.

Most "active" funds aren't active

Cremers studied this across 30 countries. The measure is Active Share.

The Swiss result is uncomfortable:

Share of assets in equity funds with Active Share below 60%. Source: Cremers, Ferreira, Matos, Starks (JFE), own illustration.

What that means: a large part of what SPIVA measures as "active management" is an index with a markup. These products have to lose. They hold approximately the index and deduct 1 to 1.5% in fees. That isn't a failed investment strategy — it's arithmetic.

And it is exactly the product most likely to be sold to you across a bank counter.

Separate the two groups and the picture sharpens:

The second condition: patience

High Active Share alone isn't enough. Cremers' work on "patient capital" shows that trading frequently loses — even with high Active Share. A manager holding 60 high-conviction positions and swapping them every four months burns their own edge on transaction costs and timing errors.

Only the combination pays: high deviation from the index plus holding periods beyond two years. That group beat its benchmark by more than 2% per year.

The dividing line isn't active versus passive. It's conviction versus arbitrary.

That's the real finding of the whole debate — and it's uncomfortable for both camps. For the fund industry, because most of its products sit on the wrong side of the line. For the index camp, because "active loses" simply doesn't hold as a blanket statement once you strip out the closet indexers.

The gap no index closes

Now the number almost nobody talks about — and it matters more to your return than the entire SPIVA discussion.

Dalbar has spent decades measuring what investors actually earn — not what the market delivers. J.P. Morgan publishes the analysis regularly. Over the 20 years from 2003 to 2022:

Read that again. Not 6.4% less — 6.4 percentage points, every year, for twenty years. Gone, without a single fund manager making a mistake. The gap comes from buying and selling at the wrong moment. In after the rally, out after the drop. And from money that sits in cash for years after a crash while the recovery happens without you.

For context: the entire SPIVA debate turns on return differences of one or two percentage points. The behaviour gap is several times larger than anything active-versus-passive has ever argued about.

Illustrative calculation based on the J.P. Morgan / Dalbar QAIB returns (2003–2022), extended over 20 years. Not a forecast. On the more conservative Morningstar yardstick (8.2% vs 7.0%) the gap is about CHF 97,000.

Two details from the Morningstar analysis matter most, because they show where the gap opens up. It was largest in funds with high tracking error and volatile flows — that is, where investors chase the product. And smallest in allocation funds, where investors captured nearly 97% of the fund's return. Not because allocation funds are better, but because they're duller. People forget about them.

Translated: the product doesn't determine your return. Your behaviour with the product does.

An ETF is cheap. But an ETF has never stopped anyone from selling in March. It doesn't explain why it fell 30%. It has no face, no reasoning, no story. It's a basket of 1,500 anonymous positions — which is exactly why it's so easy to let go of in a crisis.

In fairness: the exact size of this gap is contested in the literature — the same authors who criticise SPIVA argue the 15% figure is overstated. That the gap exists is not contested.

What actually closes a gap

You don't hold on because someone tells you to "stay invested."

You hold on because you know what you own.

If you can explain why a particular company is in your portfolio — what its moat is, why its customers don't switch, how it came through the last three recessions — you don't sell it over a red week. If you own 1,500 names you've never looked at, you sell the whole basket.

Understanding is the cheapest form of risk management there is. It's the same logic we described in The Tortoise Problem — and the same reason we write more about mastering your own emotions than about return forecasts.

Skin in the game

And then there's the question that comes before any return discussion: who owns the risk?

A manager running your money while keeping their own somewhere else holds an asymmetric position. They earn the fee. You take the loss.

The research on whether managers with their own money invested actually perform better is mixed — there are studies pointing both ways. So we don't claim skin in the game guarantees returns.

What it does guarantee is different: it changes what can be done to you. Someone with their own wealth in the same portfolio won't run a closet index. Won't take risk they're unwilling to carry themselves. Won't hold a position because selling it would be awkward to explain.

It isn't a promise about performance. It's a promise about alignment. And in this industry that's rarer.

When passive makes sense, when active does — the decision framework

Enough theory. What does this mean for your portfolio? Here's our framework — and it is explicitly not "active always."

Note the second-to-last point. The theoretically optimal investor is 100% passive, holds forever, never gets nervous. The real investor sells at the bottom — and that's exactly where the 6.4 percentage points a year go. A portfolio you understand reduces the odds that you let go of it at the worst possible moment.

Why we built arvy

We're three CFA charterholders. We worked at Pictet and UBS Investment Bank and saw how the business runs from the inside: products that get sold rather than explained. Portfolios nobody can justify. Fees for work a machine does better.

At the same time we watched the opposite among friends — people with good incomes not investing at all. Not out of laziness. Because nobody had explained what they'd actually be buying.

So we built what we wanted ourselves:

This isn't a robo-advisor. It isn't an index with a markup. It's closer to a private investment club — only regulated, transparent, and open from one franc.

The shared journey

The reason this works has less to do with investment technique than you'd think.

More than half our clients come through friends and family. Not through advertising. Because someone at a kitchen table told them what they own — and why. You can't do that with an ETF.

Building wealth takes twenty years. Twenty years alone is hard. Twenty years when you know what you own, when the people who picked it carry the same risk, and when every two weeks you hear what happened — that's a different task.

We can be wrong. We have been wrong, and we wrote it up. What we promise isn't that we win every year. It's that you always know what you own and why — and that we carry the same risk you do.

FAQ — The most common questions on active vs. passive

Is it true that over 90% of active funds fail to beat their index?

That number comes from the SPIVA report by S&P Dow Jones Indices. A study by Cremers, Fulkerson and Riley (May 2026) identifies three calculation rules that distort the result: funds that disappear automatically count as losers, every fund counts equally regardless of size, and the benchmark is free and therefore not buyable. Correct for those and the 20-year rate falls to 55% of investor assets. Still more than half — but far less than 90%.

What is a closet indexer?

A fund that sells itself as active but effectively holds the index. It's measured with Active Share: below 60% deviation from the index, a fund counts as a closet indexer. These products charge active fees of 1 to 1.5% for passive portfolios — they have to lose mathematically. In Switzerland, over 50% of assets in Swiss equity funds sit in closet indexers, versus 13% in the US.

What is Active Share and why does it matter?

Active Share measures how far a portfolio deviates from its index, position by position. 100% means no overlap, 0% means exactly the index. Funds with Active Share above 90% delivered a positive benchmark-adjusted return in 60.2% of years (average +3.64%); closet indexers in only 23.5% (average −0.13%). Active Share alone isn't enough though — only combined with holding periods beyond two years did that group beat its benchmark by more than 2% per year.

Why do active bond funds do better than active equity funds?

Because bond indices weight by outstanding debt: the biggest borrower gets the biggest index weight. That's a design flaw an active manager can exploit and an index cannot. After correcting the SPIVA methodology, the 10-year underperformance rate for bond funds falls from 71% to 37% — and in high yield, 86% of investor assets outperform.

What is the behaviour gap and how big is it?

The behaviour gap is the difference between the return that was available and the return investors actually achieve. There are two yardsticks: J.P. Morgan / Dalbar compares the average investor with the index and finds ~10% versus ~3.6% a year over 2003–2022 — 6.4 percentage points. Morningstar compares investors with the same funds, isolating timing alone, and finds 8.2% versus 7.0% over ten years to end-2024 — 1.2 percentage points. So the honest range runs from 1.2 to over 6 percentage points. The exact size is contested; the existence of the gap is not.

Does an ETF protect you from the behaviour gap?

No. An ETF is cheap, but it doesn't stop anyone from selling. If anything the opposite: Morningstar found the gap was largest in products with high tracking error and volatile flows, and smallest in dull allocation funds, where investors captured nearly 97% of the fund's return. A basket of 1,500 anonymous positions gives you no reason at all to hold in a crisis.

So should I invest actively or passively?

The question is framed wrongly. The relevant dividing line isn't active versus passive, it's conviction versus arbitrary. A broad, cheap ETF is clearly superior to a closet indexer. A genuinely active portfolio with high Active Share, long holding periods and rationales you can follow is a legitimate alternative — especially if that understanding keeps you from selling in a crash. What never makes sense: active fees for a passive portfolio.

What does "skin in the game" mean for an asset manager?

That the people responsible hold their own wealth in the same portfolio. Whether that improves returns is contested in the research. What it does change is the alignment: someone invested alongside you won't run a closet index, won't take risk they're unwilling to carry, and won't hold a position just because selling it would be awkward to explain. It isn't a promise about performance — it's a promise about alignment.

The honest conclusion

Pulling it all together:

"The most-quoted number against active investing is overstated. But even corrected, the majority of the money loses. The reason isn't that thinking doesn't work — it's that most 'active' funds don't think."

Our read — simplified:

One last thought: this isn't the article where we tell you active management wins. On average it doesn't. What the data shows is something more useful — namely where the dividing line actually runs, and that the biggest performance gap doesn't happen inside the product at all.

The question was never "active or passive." The question is: do you know what you own — and will you hold it?