← Back to Blog

Q2 13F Filings Land August 14 — But July's Chip Rout Already Made Them Stale

Hedge funds dumped chip and memory stocks in July, so the June 30 snapshots arriving on August 14 need careful reading — here's how to extract real signal anyway.


The Q2 2026 13F filing deadline is August 14, and under normal circumstances the next two weeks would be the most informative stretch of the quarter for anyone tracking institutional money. This quarter comes with an asterisk. Between June 30 — the date every position in those filings is frozen — and today, hedge funds executed one of the fastest deleveraging waves in recent memory, concentrated almost entirely in the chip and memory names that dominated their Q1 portfolios.

If you read the incoming filings at face value, you'll be looking at a photograph of a portfolio that, in many cases, no longer exists.

What happened in July

The Nasdaq fell roughly 3.5% in July, semiconductor stocks led the decline, and the Fed's "hawkish hold" — rates unchanged at 3.50%–3.75%, but with several members openly arguing inflation warrants more — sent yields climbing and risk appetite falling. Signs of progress in China's advanced chipmaking compounded existing worries about whether the AI capex boom can keep paying for itself, and the memory-chip complex that was the smart money's favorite trade in Q1 got hit hardest.

The scale of the unwind shows up in the leverage data. JPMorgan analysts estimate that hedge fund deleveraging in tech, semiconductors, and memory stocks advanced faster than they anticipated, and the asset size of leveraged ETFs tied to memory-chip stocks dropped about 34% from its June peak. That's not trimming. That's an exit.

Which raises the obvious problem: the 13Fs landing on August 14 will show funds holding exactly the positions many of them spent July aggressively selling.

How to read a stale snapshot

This doesn't make Q2 filings useless — it changes what questions to ask of them.

Look at what was already being sold before the rout. Any fund that was trimming semis or memory names during Q2 — visible as a quarter-over-quarter reduction in the filing itself — was ahead of the July unwind, not caught in it. That's the difference between a manager who saw the risk building and one who got margin-called out of a crowded trade. Quarter-over-quarter position changes across every fund we track are visible on our flow page, and it's the single best tool for separating those two groups this season.

Crowding is now the most important metric in the filing. The July unwind was violent precisely because so many funds were in the same chip and memory trades with leverage. The overlap tool shows which stocks appear across multiple top portfolios — and this quarter, high overlap in a semiconductor name should be read as a risk flag as much as a conviction signal. When everyone owns the same thing, everyone sells the same thing at the same time.

Watch what didn't get sold. JPMorgan's read is that the forced deleveraging is largely complete, even if full normalization takes months. If that's right, the chip positions that survive into Q3 filings will belong to holders with genuine conviction rather than leverage. In the meantime, the Q2 filings' non-tech positions — the defensive rotations, the energy and healthcare adds, the cash-like moves — are far less likely to have been unwound in July and are correspondingly more trustworthy as current signal.

Check the macro tilt, not just the tickers. A fund's sector-level positioning shifts more slowly than its individual names. Our macro consensus page aggregates the directional bets implied across tracked portfolios, which is a more durable read than any single position in a fast-moving tape like this one.

The contrarian setup

There's a familiar pattern worth flagging. The banks themselves are split — JPMorgan is calling for summer dip-buying in semis, Goldman thinks the selloff is nearing its end, and Morgan Stanley prefers hiding in the cloud giants. Meanwhile, the smartphone market is forecast for its worst year-on-year decline on record, which keeps a real fundamental cloud over the memory names specifically.

Historically, the moments when 13F data looks most "wrong" — when the filings show heavy exposure to something that just crashed — are also when the next quarter's filings get most interesting. The funds that add to chips in Q3, after the deleveraging flushed out the leveraged holders, will be making a very different kind of bet than the ones who rode the trade up in Q1. Browse the full investor list now, note who's most exposed on paper, and watch what they do next quarter. That comparison, not the August 14 snapshot alone, is where the signal lives.

Q2 filings start trickling in now and flood in around the deadline. We'll be ingesting them as they land.


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


Explore the full data behind this analysis on InvestorLens.

View Investor Portfolios →