Insider Purchase Cluster Monitor - SEC Form 4

Follow disclosed insider purchases

Build an issuer research timeline from SEC Form 4 purchase disclosures. See who reported a purchase, the shares and reported price, and whether multiple officers or directors appear in the inspected purchase window. Designed for analysts and research teams reviewing insider activity alongside company filings.

Open this Actor in Console · First run · Results · Saved comparisons · Full guide

Try this Actor

Start with one historical Occidental Petroleum purchase. The June 23, 2026 transaction reports 4,770 shares at 52.38 per share. This fixed SEC example demonstrates the output; its date stays attached to the observation. No SEC key or paid source subscription is required.

Inspect an SEC Form 4 purchase — paste this input into the Console JSON editor.

{
  "filingUrls": [
    "https://www.sec.gov/Archives/edgar/data/797468/000162828026045313/wk-form4_1782342238.xml"
  ],
  "maxResults": 5
}

Open the dataset to review the purchase and follow sourceUrl to the filing. Then open Run report or SUMMARY to see what was inspected. This one-insider sample does not establish a cluster.

Read the code correctly: SEC transaction code P covers an open-market or private purchase. It does not establish which of the two occurred.

Result preview

One row is one purchase-coded, non-derivative transaction. The overview brings the research essentials together:

Research question	Output fields
Which issuer and when?	ticker, issuerName, transactionDate
What was reported?	shares, pricePerShare, reportedValue
Who reported it?	reportingOwners, ownershipNature
Were other eligible insiders observed?	observedDistinctInsiders, clusterDetected
What changed since the saved observation?	comparisonStatus, changedFields, previous
Where is the evidence?	sourceUrl, accessionNumber, observedAt

Cluster counts cover the inspected filings. The price’s currency is not inferred, and missing or nonnumeric prices remain null.

Usage and pricing

Base result rate on the FREE tier: $5 per 1,000 saved purchase observations. The Pricing tab shows your plan’s result rate and the Actor Start charge. Set maxResults and a platform spending limit before running a larger watchlist. A result cap is an upper bound, not a promise of that many results.

Saved unchanged observations are billable. onlyChanges skips unchanged rows after a baseline; skipped rows, source errors and summary files are not dataset-item charges.

Build a repeatable workflow

1. Choose a small issuer watchlist. Use the Occidental watchlist example below, then replace its CIK with your tickers or SEC CIKs. Review filing coverage before expanding the scope.
2. Save a baseline. Use the saved-comparison example below with comparePrevious: true. Keep the same input scope, snapshotName and named history store between runs.
3. Review the changes. Check comparisonStatus, changedFields, ownership details and the linked Form 4. A new accession/transaction line has its own identity.
4. Deliver the evidence. Export the completed dataset to your research sheet or connect it through your own workflow. Use id as the transaction key and retain the source link. After the baseline, onlyChanges: true can keep unchanged transactions out of later exports.

Run a comparison namespace sequentially. Check SUMMARY before interpreting an empty result or routing changes; source gaps cannot establish that an insider stopped buying.

Full guide

Choose your evidence

Use either companies or filingUrls. A specified company query never silently falls back to the historical sample. Direct documents are inspected regardless of age. Company discovery searches only the SEC’s returned recent submissions index; it does not traverse archived index files.

• companies: up to 10 ticker symbols or SEC CIKs.
• filingUrls: up to 30 raw SEC archive XML URLs. SEC XSL viewer prefixes are normalized to raw XML.
• lookbackDays: bounds company discovery by filing date and company-watchlist purchases by transaction date.
• maxFilingsPerCompany: bounds matching filings before XML inspection. A company can appear as a reporting owner in another issuer’s filing; issuer mismatches are excluded after parsing, so a narrow cap can miss matches.
• clusterWindowDays, minimumInsiders and onlyClusters: select the observed cluster definition and optional output filter described below.

Field reference

Field	Meaning
issuerCik, issuerName, ticker	Issuer identity reported in the ownership XML. Ticker may be missing or historical.
transactionDate, transactionCode	The transaction date and SEC code P.
shares, pricePerShare, reportedValue	Reported shares and price; value is shares multiplied by price, rounded to six decimals. Missing or nonnumeric price remains null. Currency is not inferred.
reportingOwners	Filing owners, CIKs, director/officer flags and officer titles.
ownershipNature, sharesOwnedFollowing	Direct/indirect ownership and the reported post-transaction amount.
footnotes, remarks, rule10b51Flag	Source qualifications and the filing’s plan flag when present. Footnotes apply at filing level; review the original for precise references.
clusterDetected, observedDistinctInsiders	Conservative count within the observed transaction window.
sourceUrl, accessionNumber	Raw SEC evidence and filing identifier.
filingDate, acceptedAt, observedAt	SEC discovery metadata when available, and our observation timestamp. Direct URLs have null filing/acceptance metadata; transaction dates are not publication times.

Grants, sales, withholding, exercises, derivative transactions and holdings-only entries are excluded.

How the cluster count works

For each purchase, the window ends on its transaction date and includes that date plus the preceding clusterWindowDays - 1 days. Only observed purchases for the same issuer and case-insensitively matching security title are considered. Counts use distinct owner CIKs, not numbers of transactions.

A counted owner must be a director or officer, be the sole reporting owner in that filing, and report direct ownership. Joint filings, indirect ownership and institutional-only/other ownership remain visible as purchase rows but do not increase the cluster count. Separate security titles are not assumed to represent the same class.

Form 4/A amendments are excluded and recorded in the run summary to prevent counting the same acquisition again. This release does not resolve amendment chains: an original filing may later be corrected. The output exposes that limitation. It does not estimate total company insider spending or infer that an insider stopped buying.

Saved comparison behavior

Enable comparePrevious and keep the same input scope, snapshotName and named store across scheduled runs. The first saved observation is new. Reobserved transactions are changed or unchanged, with previous values and changed fields. A newly filed transaction has its own accession/line identity.

onlyChanges skips unchanged rows and can correctly produce zero output. onlyClusters can also produce zero output—including the default one-insider example. Neither mode fabricates replacement rows. Saved state advances only after a row is confirmed written; spending limits do not consume unsaved history. Use one writer per comparison namespace.

Coverage, limits and failures

Open Run report for the readable outcome, SUMMARY for source-by-source coverage, or export JSON/CSV from the dataset. Filings are inspected before cluster enrichment so counts use the full inspected set even when fewer rows are saved. The filing cap, recent-index boundary, excluded amendments, failed sources and transaction-date bounds all limit what a cluster can establish.

Requests identify Maydit to the SEC and are serialized at approximately one per second. Requests have bounded time and response size; temporary server errors receive one retry. Access refusals are reported without bypassing them. An upstream failure never becomes a paid data row. An all-failed run fails honestly; an empty valid purchase filter succeeds with an explicit summary.

Three example inputs

The first exact example is Inspect an SEC Form 4 purchase above. The other saved examples are:

Review a bounded Occidental insider purchase window

Inspect up to 12 recent Form 4 filings for Occidental and extract eligible purchases within 365 days.

{
  "companies": [
    "0000797468"
  ],
  "maxFilingsPerCompany": 12,
  "lookbackDays": 365,
  "maxResults": 10
}

Save an insider purchase comparison baseline

Save the documented Form 4 purchase and compare its observed cluster details on later runs. First run returns a new observation.

{
  "filingUrls": [
    "https://www.sec.gov/Archives/edgar/data/797468/000162828026045313/wk-form4_1782342238.xml"
  ],
  "comparePrevious": true,
  "snapshotName": "oxy-purchase-example",
  "maxResults": 5
}

Research use and sources

Use the observations to prioritize filing review, maintain issuer timelines, or combine disclosed purchases with earnings and financing evidence. The Actor does not calculate buy/sell recommendations or predict share-price moves.

SEC ownership transaction codes · SEC Developer Resources · SEC automated access guidance

Help with your run

Open Run report in the run’s Output tab for a readable outcome and SUMMARY for detailed coverage and source issues. Export CSV for flat spreadsheet columns or JSON to preserve nested evidence.

For help, open this Actor’s Issues tab with the run URL, relevant input and expected result. Remove credentials and private information before posting.

Explore more Mayd-It Actors