Where Insiders Are Cluster-Buying: June 2026 Form 4 Data

Why clusters beat single buys

Corporate insiders file a Form 4 within two business days of trading their own stock. Most filings are noise: option exercises, tax withholding, scheduled plans. The signal subset is the open-market purchase (transaction code P): an executive or director spending personal cash at market price. And the strongest version of that signal is the cluster, where several insiders independently buy in the same window. Academic work on insider trading has long found purchases more informative than sales, and consensus among insiders more informative still.

The clusters, last 60 days

Source: SEC Form 4 filings, open-market purchases (code P, non-derivative), amendment-corrected, trailing 60 days through 2026-06-12, minimum 3 distinct buyers. Ingested by Arkolith.

Reading the table honestly

The TSM row is a lesson, not a screaming buy signal. Thirty-one distinct buyers sounds extraordinary until you see the dollars: $0.4M total, an average around $13,000 per insider. That pattern reads as broad-based small purchases, plausibly program-like behavior, rather than executives backing up the truck. Always read buyer count with dollar size.

The conviction rows are Ryan Specialty and Charter. Four buyers each, with meaningful money: $4.5M and $3.8M respectively. Mid-seven-figure aggregate purchases by multiple individual insiders is the classic cluster profile.

Small banks keep appearing. Bankwell, Marathon Bancorp, and Primis Financial (3 buyers, just under the dollar table) all show multi-insider buying. Regional-bank insiders buying their own depressed stock has been a recurring pattern in our data this spring.

S&P Global is the blue-chip outlier. Three insiders putting $2.6M into a mega-cap is rarer than it sounds; large-cap executives mostly receive stock, they do not buy it.

Combining insider data with fund flows

Insider buys are a 2-day-lag signal from people who know the company; 13F filings are a 45-day-lag signal from institutions allocating to it. They answer different questions, and the interesting setups are where they agree or sharply disagree. You can see both signals side by side on any stock page, check what funds hold a name in the 13F data, or pull the joined dataset through the API with each transaction's source filing attached.

Frequently asked questions about insider cluster buying

What counts as a cluster?

Our threshold here is three or more distinct insiders making open-market purchases in the same company within 60 days. Stricter definitions also require minimum dollar amounts; the table shows dollars so you can apply your own.

Why only purchases and not sales?

Insiders sell for many reasons (diversification, taxes, scheduled plans) but mostly buy for one. Sales carry some information; purchases carry much more.

How fresh is this data?

Form 4s are due within two business days of the trade, and we ingest filings continuously. This table covers trades through June 12, 2026, and will age accordingly: treat it as a dated snapshot, not a live screen.

Do you include 10b5-1 scheduled trades?

The table counts all open-market purchase transactions (code P), which can include trades made under pre-scheduled 10b5-1 plans. Scheduled purchases carry less signal than discretionary ones, so for any name you care about, read the source Form 4s before drawing conclusions.

This article reports public regulatory filings. It is not investment advice.