SEC Form 4 insider activity
Section 16 reporting by officers, directors, and large holders — purchases, sales, grants, exercises, withholding, and gifts, read by transaction code rather than by headline.
Track insider activity without treating every filing as a signal.
Buteon follows publicly disclosed insider transactions at the companies you track, then reads each one with the evidence attached: what the filing actually reports, the actor and market context around it, what weakens the read, and what happened after comparable cases.
First, the term
This is lawful, publicly disclosed insider activity.
“Insider trading” is used two ways, and only one of them is what this page is about. Corporate insiders — officers, directors, and large shareholders — are required to report their transactions in their own company’s stock to the SEC, usually on Form 4 and generally within two business days. Those filings are public, and reading them is ordinary research.
That is different from illegal insider trading, which means trading on material nonpublic information. Buteon tracks the disclosed, lawful kind: transactions their filers were legally obliged to publish. Filing a Form 4 is routine compliance, not an accusation.
Why the distinction changes the reading
Because these filings are compulsory, most of them are unremarkable — scheduled sales, equity compensation, tax withholding, administrative transfers. Treating every disclosed transaction as a deliberate market statement is the most common way to misread the public record. A tracker that only lists filings leaves that problem with you.
What is covered
The boundaries are part of the product.
A tracker is only useful if you know what it does not see. These are the current boundaries, not a roadmap.
Form 144 proposed-sale notices, where collected
A Form 144 announces that shares may be sold; it is limited context around a Form 4, not a second live feed and not a record of what was done.
Watchlist by company
You choose the companies you want followed. Watchlist size depends on your plan.
Scoped to tracked names
Buteon does not claim universal company coverage. A company that is absent is a coverage limit, not a judgment about that company.
Delayed on Free
Free access sees activity on a delay. Buteon does not present itself as a real-time market feed on any plan.
Schedule 13D and 13G
Large-ownership disclosures are not a live tracking lane today. Where they appear in public research, they are labeled as such.
The problem with raw filings
The same filing can mean opposite things.
A large disposition can be a sale, a gift, shares withheld to cover taxes, or a transfer back to the issuer — all reported on the same form, distinguished by a single letter. An increase in holdings can be an open-market purchase with the insider’s own money, or a grant that cost them nothing. A sale can be a decision made this week, or the mechanical execution of a plan adopted months earlier.
None of that is hidden. It is in the transaction code, the footnotes, and the related rows — which is exactly the material a filing feed reproduces without interpreting. These guides work through the specific ways the public record gets misread:
How a read is formed
A filing is the input. Everything after it is the work.
The filing is the input. Everything after step one is the work a raw feed leaves to you.
- 01
Classify what actually happened
The transaction code decides the event: an open-market purchase, an award, an option exercise, tax withholding, a gift, or a transfer. Direction alone does not settle it — acquired is not always purchased, and disposed is not always sold.
- 02
Resolve the actor and the company
Who filed, in what role, at what company, and how that person has appeared in prior filings. A chief executive committing personal capital and a routine administrative transfer are not the same evidence.
- 03
Weigh timing and market context
When the transaction happened, when it was disclosed, and what surrounded it. A same-day disclosure and a filing that lands near the end of its window carry different weight.
- 04
Surface what weakens the read
Ordinary explanations stay visible instead of being edited out: a pre-existing trading plan, an offsetting sale, a single isolated actor, a size that is immaterial to that person, or a thin evidence base.
- 05
Attach confidence and freshness
A read carries how strong the evidence is and how current it is. Low confidence stays labeled low, and an aging read is not presented as a fresh one.
- 06
Compare with measured historical cases
Where comparable cases exist, Outcome Memory shows what followed them, with the sample size stated. It is historical context — not a forecast, and not a performance claim.
Worked example · published Case FilePreview Only
SOFI • Open-market purchase
Read
Bullish
A fresh, low-confidence bullish read driven by a single, isolated insider.
Confidence
Low
Confidence reflects how much independent evidence supports the read — not how notable the headline is.
What weakened the read
- No independent cluster confirmed the purchase. A lone transaction can be notable without establishing broader agreement.
- Buteon did not convert the CEO title or the purchase amount into high conviction. The isolation of the event remained visible.
Preserved as of May 11, 2026. The read is preserved as it stood at the time rather than rewritten with hindsight, and it is not a prediction or performance claim. Read the full Case File →
More than a filing feed
The same filing, read two ways.
Getting the filing quickly is a solved problem — the SEC publishes it. The unsolved part is deciding what it means, and that is the part Buteon is built for.
| Step | A raw filing stream | A Buteon read |
|---|---|---|
| The filing | Listed as a row: ticker, insider, code, shares, price. | Treated as an input, then classified into what the transaction mechanically was. |
| The actor | A name and a title. | Role and prior filing behavior at that company, so a first purchase and a routine pattern read differently. |
| The context | Left to the reader. | Transaction timing, disclosure timing, and surrounding market context are part of the read. |
| The doubt | Absent — a row cannot argue with itself. | What weakens the read is published next to what supports it. |
| The confidence | Implied by inclusion. | Stated explicitly, including when the evidence only supports a low-confidence or non-actionable read. |
| The history | Each filing arrives without memory. | Comparable past cases are recorded with their sample size, as context rather than as a promise. |
The comparison is with a raw filing stream as a category, not with any named product.
The filing
A raw filing stream Listed as a row: ticker, insider, code, shares, price.
A Buteon read Treated as an input, then classified into what the transaction mechanically was.
The actor
A raw filing stream A name and a title.
A Buteon read Role and prior filing behavior at that company, so a first purchase and a routine pattern read differently.
The context
A raw filing stream Left to the reader.
A Buteon read Transaction timing, disclosure timing, and surrounding market context are part of the read.
The doubt
A raw filing stream Absent — a row cannot argue with itself.
A Buteon read What weakens the read is published next to what supports it.
The confidence
A raw filing stream Implied by inclusion.
A Buteon read Stated explicitly, including when the evidence only supports a low-confidence or non-actionable read.
The history
A raw filing stream Each filing arrives without memory.
A Buteon read Comparable past cases are recorded with their sample size, as context rather than as a promise.
What this is not a claim about
Buteon does not claim to see filings before they are public, to cover every company, or to know what an insider was thinking. It claims something narrower and testable: that a disclosed transaction is worth classifying, contextualizing, and qualifying before it is treated as a signal — and that the doubts belong on the page next to the evidence.
Free, no access required
Open the Form 4 decoderStart with the transaction code.
The letter in the transaction row decides what the filing reports — a purchase, an award, an exercise, withholding, or a gift. The free decoder explains each covered code, what it does not prove, and the misreads it invites, with the official SEC definition and a primary filing alongside.
Evidence · published Case Files
Three filings that would have been misread as rows.
Each one follows a single public filing from transaction code to a preserved read, including where the evidence ran out. Every fact links to the primary document.
Open beta
Track the filings that are worth your attention.
Buteon is in open beta: create a free account and start reading today, or begin with the free decoder — it stays useful either way.
Buteon provides market data analysis and signal interpretation for informational and educational purposes. It is not a broker, custodian, or investment adviser, and this page is not investment advice or a buy/sell recommendation. Insider activity is evidence to weigh, not a trade to copy. Always verify the official filing before relying on any description of it.