The Small-Cap Rally Is 13F Data's Biggest Blind Spot

The best-performing corner of the US market in 2026 has been small caps. The Russell 2000 outran the Nasdaq for fifteen straight sessions in January, IWM put up +11.7% in April, and the "Great Rotation" that strategists had been forecasting since 2023 finally arrived.

Now go read the Q2 13F filings and try to find it. You mostly can't. That is not because institutions missed the trade. It is because Form 13F is structurally bad at showing you small caps, and knowing exactly how it fails is more useful than any single position you'll pull out of it.

Three reasons the data undercounts

First, the reporting threshold. A manager only files if they hold at least $100 million in Section 13(f) securities. That catches every large fund but excludes the entire universe of smaller specialists — and small-cap value investing is disproportionately a small-shop business. The managers with the deepest edge in a $900 million industrial are frequently the ones running $60 million and filing nothing at all.

Second, position sizing. A $40 billion fund that takes a genuinely high-conviction 2% position in a small cap has to either accept an illiquid stake it cannot exit or cap the position at something that rounds to noise in the portfolio table. So the large filers you can see are mechanically pushed toward mega caps, and any screen built on them inherits that bias. Sorting a 13F database by position size is, in effect, sorting by market cap.

Third, the ownership base itself. Small caps are held far more heavily by index funds, retail and non-13F vehicles than mega caps are. The institutional layer that 13Fs illuminate is simply a thinner slice of the shareholder register down there.

What the filings did show

The one clear Q2 signal was subtractive. Hedge funds collectively cut their iShares Core S&P 500 ETF holdings by roughly 63% during the quarter, from about 840 million shares to roughly 311 million. That is an enormous reduction in generic large-cap beta.

Money leaving broad index exposure has to go somewhere, and the visible destinations in Q2 were semis, AI infrastructure and selective cyclicals — the rotation everyone covered. But a meaningful portion of that unwind likely went into positions that either fall below reporting thresholds or sit in vehicles 13Fs don't reach. Absence of evidence, in this dataset, is very often just absence of coverage.

You can see the visible half of that shift in aggregate on the quarterly flow view. What you're seeing there is the part of institutional behaviour that happens to be reportable.

The valuation case, and its catch

The rotation has a straightforward driver. Entering 2026 the Russell 2000 traded near 18x forward earnings against roughly 26x for the S&P 500 — a discount north of 30%. Add a Fed funds rate that settled into the 3.50%–3.75% range after a run of cuts, and small caps get direct relief, because they carry far more floating-rate debt than the cash-rich mega caps do.

That is a clean thesis. It is also, by now, a widely held one, which is the standard problem with any trade that has been outperforming for eight months and generating headlines the entire time.

The harder question is quality. The Russell 2000 contains a large share of companies with negative earnings, and index-level valuation multiples obscure the split between profitable small caps and the unprofitable long tail. A discount to the S&P is not automatically a bargain if a chunk of the index is levered and loss-making — the rate relief argument cuts both ways.

How to use 13Fs here anyway

The filings aren't useless below mega-cap. They just require a different question.

Instead of asking "what are big managers buying," ask "where does a small position from a credible manager represent unusual conviction?" A $30 million stake is background noise in a $40 billion book, but if it represents 8% of the target company's float, someone did real work. Percentage-of-company is the more informative field than percentage-of-portfolio when you go down the cap scale.

The second question worth asking: which small-cap-focused managers file at all? A handful do, and their filings are far more legible than a multi-strategy giant's, because the whole book is in the strategy you care about. The investor directory is the place to identify them, and the overlap tool will tell you where two or more of them landed on the same name independently.

And keep the timing caveat in view. Q2 filings describe June 30 positions. If the small-cap move accelerated in July and August, the November filings are the first place it can appear. The broader positioning backdrop is tracked on the macro consensus page, but no aggregate view fixes a 45-day lag.

The reasonable conclusion is not that institutions are absent from small caps. It's that 13F data cannot answer the question, and pretending otherwise is how a research tool turns into a false negative.

---

Data sourced from public SEC 13F filings. Educational research only — not investment advice.