Institutional Investors' Trades Add No Value. We Checked 268,000 Fund-Quarters.

June 30, 2026 · FindataFox

Everyone asks whether you can profit by copying institutional 13F filings. We asked a different question: do the institutions profit from their own trading? We rebuilt 268,724 fund-quarters, compared each fund's actual post-trade portfolio against a "frozen" version that made no changes, and measured both forward. The trades helped 49.8% of the time — a coin flip — and that's before the costs of trading. Net of those costs, the funds would have done better sitting still.

TL;DR

Every hedge-fund-tracking site argues about whether you can ride along with the smart money. Our other studies say you can't. But that leaves a more basic question hanging: do the institutions themselves get anything out of all that trading?

So we ran a clean counterfactual. For every institution and every quarter:

• ACTUAL = the fund's portfolio after its trades, carried forward.
• FROZEN = the fund's portfolio from the quarter before, held with no changes at all, carried forward over the exact same window.

If a quarter's trades were smart, ACTUAL beats FROZEN. Across 268,724 fund-quarters:

A coin flip, at every horizon. The value added wobbles around zero by fractions of a basis point and is never statistically distinguishable from nothing. And this is the generous version — read on.

The cost twist that turns "zero" into "negative"

Those numbers are gross of trading costs. But trading isn't free — funds pay spreads, commissions, and market impact to execute every one of those trades. The frozen portfolio paid nothing; it just sat there.

So the honest scoreboard is: the trades added zero value before costs, and funds paid real money to make them. Net of what it actually costs to trade, the do-nothing portfolio wins. The average institution would have ended up slightly richer by making no changes at all.

"But trades pay off over years, not next quarter"

It's the obvious objection, and a fair one — a fund might buy for a multi-year thesis that one quarter can't capture. So we extended the forward window to two quarters and a full year. The answer didn't budge: +0.08 pp/yr at two quarters, −0.06 pp/yr at one year, both indistinguishable from zero. (We corrected the statistics for the fact that longer overlapping windows inflate confidence — the same trap we caught ourselves in on a related test.)

And separately, we showed that holding institutional favorites longer doesn't beat the market either. So the "the value just shows up later" escape hatch is closed from both sides: the trades don't pay off later, and holding the names longer doesn't help.

The funds that bet biggest add the least

Here's the part that should give pause. We split funds by concentration. The diversified funds (>30 holdings — closer to index-huggers) were a clean coin flip, ~50.8% helped. The concentrated funds (≤30 holdings — the high-conviction, "we have an edge" bettors) were worse: their trades helped only 47% of the time, with a slightly negative value-add, at every horizon.

The managers who trade with the most conviction — the ones whose 13Fs people most want to copy — are exactly the ones whose trades, in aggregate, subtract a little value. It lines up with our finding that institutional skill doesn't persist and that cloning concentrated managers' best ideas is a null.

What this means

• The information in a 13F is in the stable core, not the quarterly churn. What a fund keeps holding tells you more than what it just bought or sold — because the buying and selling, in aggregate, is noise.
• It's not that institutions are bad at investing — it's that their trading, on the visible long-equity book, doesn't add value over standing pat. Most of their return is just owning the market (mostly mega-caps), which they'd capture frozen.
• This is the cleanest argument yet for indexing that you can build from 13F data: a quarter-million data points showing that the pros' trades are, collectively, a coin flip they pay to play.

The honest caveats

• 13F shows the long U.S.-equity book only — no shorts, options, leverage, or international. We're measuring the value of their visible trading, not their entire strategy. A market-neutral fund's edge can live entirely in what 13F doesn't show.
• This is aggregate. It says the average trade adds nothing — not that no individual fund ever made a great trade. Someone is always on the right side; the point is you can't tell who in advance, and as a group it nets to zero.
• We don't observe each fund's actual execution costs, so "net of costs they'd do better frozen" is a directional claim (trading isn't free), not a precise per-fund figure. Survivorship cancels in the difference, since both portfolios carry the same delisted names.
• Research and education, not investment advice. Past performance does not predict future results.

Part of our 13F research series: we backtested the popular strategies — none beat the market · skill doesn't persist · holding longer doesn't help · the famous pickers are closet indexers · beating the S&P isn't an edge. NOT investment advice.