Most Bought and Most Sold: Why the Same Mega-Caps Top Both 13F Lists

If you scan Q2 2026 13F rankings, something looks broken. Microsoft, Nvidia and Amazon sit near the top of the most-bought lists. Apple, Alphabet, Microsoft, Amazon and Nvidia also appear among the most-sold names. How can a stock be both?

It is not a data error. It is one of the most useful things to understand before you trust any "smart money is buying X" headline.

The Numbers Behind the Paradox

According to aggregated Q2 2026 filings tracked by Arkolith, Microsoft was added by roughly 2,105 funds, Nvidia by about 1,970 and Amazon by about 1,957. Apple still saw about 1,644 funds add, even though around 1,891 funds trimmed it, making it the most-trimmed name in that dataset.

Those are counts of filers, not dollars, and they overlap heavily. With thousands of 13F filers holding the same handful of mega-caps, any given quarter will have thousands of funds adding and thousands trimming. Ranking by the number of funds acting in either direction mostly measures how widely held a stock is.

Popularity Is Not Direction

A stock held by 5,000 institutions will naturally have the largest absolute number of buyers and sellers. The raw count rewards size and ubiquity. The more informative question is the net: did the buyers outweigh the sellers, and by how much capital?

Index funds and ETFs complicate this further. The Vanguard Growth ETF drew buyers among 1,635 funds in the same data, and the iShares Core S&P 500 ETF showed about $132.7 billion in aggregate holdings across roughly 1,200 filers. Flows into these vehicles mechanically push money into the same mega-caps, which looks like conviction in a 13F but is really allocation.

How to Separate Signal From Noise

First, look at who is acting, not how many. A concentrated manager adding to a top-five position is a different event from a quant fund rebalancing a 0.02% weight. On our investors page you can see each manager's actual changes and weights, which is where conviction shows up.

Second, look at overlap among concentrated portfolios. The overlap tool shows which names several high-conviction managers hold together, a far smaller and more meaningful group than the raw buyer count.

Third, look at direction in aggregate. The flow page tracks net buying and selling rather than headline participation, and the macro consensus view shows where macro-oriented funds agree.

Where the Real Newsworthy Moves Were

The more telling items in Q2 were not the mega-caps at all. Honeywell Aerospace drew about 1,139 funds opening first positions, SpaceX about 931, and Intel about 685. New positions, especially across many funds at once, say more about changing sentiment than a trim of a stock everyone already owns.

On the other side of the ledger, the largest outright exits in broader institutional data were in names like Carnival and Coterra Energy, where hundreds of institutions closed positions entirely. A full exit is a stronger statement than a modest trim.

What to Keep in Mind

13F data is a snapshot of quarter-end holdings filed up to 45 days later. A fund that bought in April and sold in June can show up as a net change near zero. Short positions are not reported, so a trim may simply be a hedge being rebalanced elsewhere. And aggregated lists from third-party sites can differ depending on which filers and security types they include, so treat any single ranking as a starting point.

The Takeaway

When a stock tops both the most-bought and most-sold lists, read it as "widely owned and actively managed," not as a contradiction. Weight the moves by who made them, how big the position is relative to the portfolio, and whether the change is a new position or an exit. That is how 13F data becomes research rather than a headline.

---

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