SEC Form 4 Insider Trading Scraper — Cluster Buys & Signals
Pricing
from $10.00 / 1,000 insider transaction returneds
SEC Form 4 Insider Trading Scraper — Cluster Buys & Signals
Scrape SEC Form 4 insider trades: open-market buys only, 10b5-1 flags, cluster-buy detection and conviction scoring.
Pricing
from $10.00 / 1,000 insider transaction returneds
Rating
0.0
(0)
Developer
Samik Sarkar
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
19 hours ago
Last modified
Categories
Share
SEC Insider Trading Signals turns raw SEC Form 4 filings into a clean, ranked feed of insider buying and selling — with the compensation noise stripped out. It reads every Form 4 disseminated by EDGAR, keeps only the transactions that reflect an actual decision, flags pre-scheduled 10b5-1 trades, and detects cluster buys where several insiders at the same company purchase within days of each other.
Most Form 4 scrapers hand you every row in the filing. Roughly 80% of that volume is option exercises, share grants and tax withholding — mechanical compensation events that tell you nothing. This Actor classifies every transaction so you can filter to the ~20% that carries information.
What SEC Insider Trading Signals does
- Scans all Form 4 and Form 4/A filings across the entire US market, or just the tickers you care about
- Classifies each transaction: open-market buy, open-market sale, grant, option exercise, tax withholding, gift, disposition
- Flags Rule 10b5-1 pre-scheduled trades so you can separate discretionary conviction from a plan set up months earlier
- Detects cluster buys — two or more distinct insiders buying the same issuer inside a rolling window, counting named individuals only so a fund complex co-filing under a dozen affiliated vehicles cannot manufacture a signal
- Collapses joint filings to one row per transaction, so a trade co-filed by 13 entities is one event with a co-filer count, not 13 duplicate rows
- Scores each purchase 0–100 on a transparent conviction heuristic: seniority, dollar size, how much the buyer grew their own stake, whether it was discretionary, and whether colleagues bought alongside
- Resolves ticker symbols, insider roles (CEO, CFO, director, 10% owner) and links straight back to the source filing on EDGAR
Why use SEC Insider Trading Signals
- Quant and systematic research — a clean daily feed of open-market insider purchases, already de-noised, ready for a backtest
- Financial media and newsletters — surface the notable buys of the day without reading a thousand filings
- Investor relations and compliance — monitor Section 16 activity at your own company or your peer group
- Fintech products — power an insider-activity screen, alert or dashboard without building an EDGAR pipeline
- Watchlist alerts — schedule a daily run over your portfolio and get notified when an executive buys
How to use SEC Insider Trading Signals
- Click Start — the defaults sweep the last two days of the whole US market for open-market buys over $25,000.
- Wait for the run to finish. Progress appears in the log as filings are scanned.
- Open the Storage / Dataset tab and switch between the Insider signals and Cluster buys only views.
- Export to JSON, CSV, Excel or XML, or pull it from the Apify API.
- To watch specific companies, set Scan mode to Specific companies and list your tickers.
- Use Schedules to run it every morning, and Integrations to push results to Slack, a webhook, Google Sheets or your database.
Input
Every field has a working default, so you can run it without configuring anything.
{"mode": "market_sweep","lookbackDays": 2,"signalTypes": ["open_market_buy"],"minTransactionValueUsd": 25000,"excludePlannedTrades": false,"detectClusterBuys": true,"clusterWindowDays": 7,"clusterMinInsiders": 2,"maxResults": 1000}
To follow a watchlist instead:
{"mode": "companies","tickers": ["AAPL", "NVDA", "JPM"],"lookbackDays": 30,"signalTypes": ["open_market_buy", "open_market_sale"],"minTransactionValueUsd": 0}
Output
{"convictionScore": 84,"ticker": "ACME","issuerName": "ACME ROBOTICS INC","issuerCik": "0000320193","insiderName": "DOE JANE A","insiderRole": "CEO","officerTitle": "Chief Executive Officer","isDirector": true,"isOfficer": true,"isTenPercentOwner": false,"signalType": "open_market_buy","transactionCode": "P","transactionCodeLabel": "Open market or private purchase","transactionDate": "2026-08-14","shares": 10000,"pricePerShare": 42.5,"transactionValueUsd": 425000.0,"sharesOwnedAfter": 60000,"holdingsIncreasePct": 20.0,"isPlanned10b5_1": false,"insiderIsEntity": false,"coFilerCount": 0,"clusterIndividualCount": 3,"clusterTotalValueUsd": 1250000.0,"accessionNumber": "0000320193-26-000075","filedAt": "2026-08-15","filingIndexUrl": "https://www.sec.gov/Archives/edgar/data/320193/..."}
You can download the dataset in various formats such as JSON, HTML, CSV or Excel.
Data fields
| Field | Description |
|---|---|
convictionScore | 0–100 heuristic for open-market buys. Higher = more informative on its face. |
signalType | open_market_buy, open_market_sale, award_grant, option_exercise, tax_withholding, gift, disposition_to_issuer, other |
transactionCode / transactionCodeLabel | The raw SEC Table I/II code and its plain-English meaning |
ticker, issuerName, issuerCik | The company whose stock was traded |
insiderName, insiderCik | The Section 16 reporting person |
insiderRole, officerTitle | CEO, CFO, President/Chairman, Senior officer, Director, 10% owner |
isDirector, isOfficer, isTenPercentOwner | Raw relationship flags as filed |
shares, pricePerShare, transactionValueUsd | Size of the transaction |
sharesOwnedAfter, holdingsIncreasePct | Position after the trade, and how much the buy grew it |
isPlanned10b5_1 | True when the filing indicates a Rule 10b5-1 trading plan |
isDerivative | True for Table II (options, RSUs, warrants) |
securityType | common, preferred, derivative or other, from the security title |
dataQualityFlags | Markers for figures that cannot be right as filed — see the FAQ |
clusterId, clusterTotalValueUsd | Cluster-buy membership, when detected |
clusterIndividualCount | Named individuals in the cluster — the number that qualifies it |
clusterInsiderCount | Everyone in the cluster, individuals and entities |
insiderIsEntity | True when the filer is a fund or holding vehicle rather than a person |
coFilerCount, coFilerNames | Other entities that co-filed the same transaction |
accessionNumber, filedAt, periodOfReport | Filing identifiers and dates |
filingUrl, filingIndexUrl | Direct links to the source document on SEC EDGAR |
How much does it cost to track SEC insider trading?
This Actor uses pay per event, so you pay for results rather than for compute time:
| Event | Price |
|---|---|
| Actor start | $0.01 |
| Each insider transaction returned | $0.01 |
A default run (two days of the whole market, open-market buys over $25k) typically returns a few hundred transactions, so most daily runs land between $1 and $4. Watching a handful of tickers costs cents. Results are sorted by conviction score before the maxResults cut, so a smaller budget still keeps the most significant activity.
Tips
- Set
excludeEntityFilersto true to drop fund and holding-vehicle filers entirely and keep only named individuals. - Set
excludePlannedTradesto true to remove Rule 10b5-1 activity. This is the single most effective noise filter after the transaction-code classification. - Cluster buys are the highest-signal output. Set
clusterMinInsidersto 3 for a much shorter, much stronger list. - Widen
lookbackDayson Mondays — weekends and market holidays produce no filings. - Lower
minTransactionValueUsdto 0 when watching small-cap companies, where meaningful insider buys are smaller in dollar terms. - Raise
maxFilingsToScanwhen scanning more than a few days of the full market. - Schedule a daily run just after market close; Form 4s are due within two business days of the transaction, so most arrive in the evening.
FAQ
Where does the data come from? Directly from SEC EDGAR — the daily index and the ownership XML inside each filing. Nothing is bought, cached or resold, and the Actor respects the SEC's fair-access rate limits and identifies itself with a contact address as the SEC requests.
Is this legal? Yes. SEC filings are US federal government records published for public use and are not subject to copyright. Insider transactions are required to be disclosed publicly under Section 16(a) of the Securities Exchange Act, usually within two business days.
Is the conviction score investment advice? No. It is a transparent screening heuristic computed only from fields the insider themselves disclosed — seniority, size, stake change, plan status, cluster membership. It is not a prediction, a recommendation, or a signal to trade, and it has not been validated against forward returns. Do your own research and speak to a licensed professional before making investment decisions.
Why do I see fewer results than filings scanned? By design. Most Form 4 rows are grants, option exercises and tax withholding. The defaults return only open-market purchases above $25,000. Joint filings also collapse to a single row per transaction rather than one per co-filer.
A fund filed under a dozen entity names — do I get charged for each?
No. One transaction is one billable row regardless of how many affiliated vehicles co-filed it. The extras appear in coFilerCount and coFilerNames.
I see an insider "buy" with an absurd dollar value.
Form 4 is self-reported and filers do mis-key it. One real example: 1,000 shares of a penny-stock preferred reported at $100,000 per share, producing a $100,000,000 "purchase". Rows like this are kept, because that is genuinely what the filing says and filingIndexUrl takes you straight to it — but they carry a dataQualityFlags entry and are pushed down the conviction ranking so they never lead your results. Preferred and derivative securities are demoted for the same reason: they are usually negotiated placements, not open-market conviction.
Some transactions have no price.
A minority of filings report the price only in a footnote, or report a range. Those rows come through with pricePerShare and transactionValueUsd set to null rather than a guess.
Can I get historical data?
Yes — set startDate and endDate to any past range and raise maxFilingsToScan. EDGAR daily indexes go back many years.
Support
Found a bug, a filing that parses badly, or need a field that isn't here? Open an issue on the Issues tab and it will get looked at. Custom variants of this Actor for a specific workflow can be arranged.