How FilingPulse turns a raw SEC Form 4 into a transparent, explainable signal.
Each filing receives a heuristic score, clamped to the range −100 to +100, where positive leans bullish (insider accumulation) and negative leans bearish (distribution). The score is a simple, additive sum of the factors below — there is no machine-learning black box. Every contribution that applied to a filing is recorded and shown in the filing's signal breakdown, so you can always see why it scored the way it did. Filings with no genuine open-market trade (grants, option exercises, tax withholding) have their role-based noise damped toward zero.
| Factor | Weight | Why |
|---|---|---|
| Open-market purchase (code P) | +30 | Insider committed personal capital — the strongest single input. |
| Cluster buying | +20 | ≥3 distinct insiders bought the same issuer on the open market within 30 days. |
| Large net open-market value | up to ±25 | Log-scaled by the net buy/sell dollar amount; signed by direction. |
| Senior officer (CEO/CFO/COO/Chair/President/Founder) | +15 | Role weight when a real trade is present. |
| Company officer | +8 | Non-senior officer role weight. |
| Board director | +6 | Director role weight. |
| 10%+ owner | +5 | Large-holder role weight. |
| Discretionary open-market sale (code S) | −22 | Sells are weaker signals than buys and can have many motives. |
| Pre-planned (Rule 10b5-1) sale | −8 | Scheduled sales carry far less information than discretionary ones. |
| Late filing | −5 | Filed more than 2 business days after the transaction — a governance flag. |
Institutional managers report long US equity positions quarterly on Form 13F-HR, with up to a 45-day delay. FilingPulse compares each manager's consecutive quarters to classify every position as new, add, trim, exit, or hold, and aggregates per-ticker holders into a “net accumulating / net reducing” read. 13F data shown today is illustrative sample data; the pipeline is built to swap in live SEC filings without UI changes.
A Form 144 is a notice of proposed sale of restricted or control securities, filed before the sale. It is a forward-looking read on insider selling intent that often precedes the Form 4 recording the actual trade. FilingPulse flags whether each notice is discretionary or executed under a pre-arranged Rule 10b5-1 plan — routine plan sales, option exercises, and RSU-vesting sales carry less signal than discretionary ones. A proposed sale is intent, not a completed transaction, and may not be fully executed. When a database is configured, Form 144 notices are ingested live from EDGAR (share count, aggregate market value, broker, and approximate sale date come straight from the structured filing); otherwise illustrative sample data is shown.
An 8-K is filed when something material happens at a company, tagged with one or more standardized “item” numbers — earnings (2.02), a material agreement (1.01), an acquisition (2.01), an executive change (5.02), or a financial restatement (4.02), among others. FilingPulse flags the high-impact items so corporate events sit alongside insider and institutional flow on each ticker. Not every 8-K is market-moving — many are routine. When a database is configured, 8-K events are ingested live from EDGAR (the item numbers come straight from the SEC submissions API); otherwise illustrative sample data is shown.
The convergence score fuses every disclosure stream we ingest for a ticker — insider trades (Form 4), 13F institutional holdings, Schedule 13D/G stakes, Form 144 planned sales, and 8-K events — into one transparent, additive total, floored at 0 and capped at 100. As with the signal score, every contribution is an attributable factor, shown as a chip on the signals board and the per-ticker disclosure radar. 8-K events are counted as context (they don't move the score); Form 144 planned selling counts against bullish conviction. The classification highlights convergence: when independent disclosures point the same way on one name.
| Factor | Weight | Side |
|---|---|---|
| Open-market insider buying | +12 | Insider |
| Cluster buying (≥3 insiders) | +14 | Insider |
| Cluster buying (2 insiders) | +7 | Insider |
| Strong insider signal (≥50) | up to +20 | Insider |
| Net insider buying | +6 | Insider |
| Institutions accumulating | +12 | Institutional |
| New institutional position(s) | +8 | Institutional |
| Broad ownership (≥3 funds) | +6 | Institutional |
| Net institutional share increase | +4 | Institutional |
| Activist 13D stake (5%+) | +14 | Activist |
| Passive 13G stake (5%+) | +6 | Activist |
| Planned insider selling (Form 144) | −6 to −10 | Sell pressure |
| Material 8-K event (recent) | context | Events |
SEC Form 4 reports each transaction with a one-letter code. FilingPulse maps each to a normalized buy/sell direction and flags whether it is an open-market (discretionary) trade.
| Code | Meaning | Direction | Open-market |
|---|---|---|---|
| P | Open-market purchase | buy | Yes |
| S | Open-market sale | sell | Yes |
| A | Grant / award | buy | No |
| D | Disposition to issuer | sell | No |
| F | Tax withholding | sell | No |
| M | Option/derivative exercise | buy | No |
| X | Exercise of in-the-money derivative | buy | No |
| C | Conversion of derivative | neutral | No |
| G | Gift | neutral | No |
| V | Voluntary early report | neutral | No |
| J | Other acquisition/disposition | neutral | No |
FilingPulse is an informational tool that summarizes public SEC filings. The signal score is a heuristic, not a prediction, recommendation, or solicitation to buy or sell any security. Insider buying and selling happen for many reasons that this score cannot capture. Always read the underlying filing and do your own research. FilingPulse does not execute trades.