Stock Alpha Signal Aggregator — One Call. One Ticker. Everything.

Stop pulling 5 different data sources. One call returns the complete SEC intelligence picture for any ticker: insider trading, material events, financial filings, and a computed bull/bear signal score. What Bloomberg Terminal costs $25,000/year for — in ONE Apify actor call.

This is the top of the stack. Built on top of every other SEC actor I've shipped. Instead of running Form 4 + 8-K + 10-Q + EDGAR separately and manually merging the output, you pass a ticker and get the fully-parsed, signal-scored, decision-ready intelligence report.

What you actually get (verified on real tickers)

TSLA — real output in one call

Metric	Value
Form 4 filings parsed	7
Insider BUYS	0
Insider SELLS	45
Net insider value	-$47,589,560
8-K material events	3
Signal	🔴 BEARISH (score -2)

NVDA — same call, different ticker

Metric	Value
Form 4 filings parsed	15
Insider BUYS	0
Insider SELLS	44
Net insider value	-$294,772,058
Distinct insiders selling	3 (cluster)
Signal	🔴🔴 STRONGLY_BEARISH (score -4)

PFE — same call, bullish differentiation

Metric	Value
Form 4 filings	15 (mostly option grants)
Actual open-market SELLS	1 ($22k — noise)
8-K Item 8.01 (positive events)	2
Signal	🟢 BULLISH (score 3)

Same actor, three different tickers, three distinct real-world signals. The aggregator correctly distinguishes heavy insider dumping (NVDA) from option grant noise (PFE) from position trimming (TSLA).

What it actually does

For each ticker you pass, this actor:

1. Resolves ticker → SEC CIK via the SEC company tickers mapping
2. Pulls recent submissions from the official SEC submissions API
3. Parses each Form 4 XML to extract:
   • Insider name, role (director/officer/10%+ owner), title
   • Every transaction (shares, price, code, value, direction)
   • Net aggregate buy vs sell activity
4. Catalogs 8-K filings with Item code classification:
   • Item 1.01 → Material agreement (bullish)
   • Item 1.03 → Bankruptcy (strongly bearish)
   • Item 2.01 → Completion of acquisition (bullish)
   • Item 2.06 → Material impairment (bearish)
   • Item 5.02 → Exec departure/appointment
   • ... 20+ more mapped Item codes
5. Lists 10-Q / 10-K periodic financial reports
6. Computes an aggregate alpha signal using simple rules:
   • +3 for net insider buying (code P)
   • +2 for cluster buys (3+ distinct insiders)
   • -3 for $1M+ net insider selling
   • -2 for cluster sells by multiple insiders
   • +1/-3 for positive/negative 8-K events
   • -1 for C-suite (CEO/CFO) selling

Signal labels

Score	Label	Meaning
≥ +4	🟢🟢 STRONGLY_BULLISH	Multiple bullish signals stacking
+1 to +3	🟢 BULLISH	Net positive factors
0	⚪ NEUTRAL	Mixed / insufficient data
-1 to -3	🔴 BEARISH	Net negative factors
≤ -4	🔴🔴 STRONGLY_BEARISH	Multiple bearish signals stacking

Why this is the product

Raw data without synthesis is noise. You can scrape SEC filings all day, but until you answer "is the insider activity bullish or bearish?" — you have a pile of filings, not a trading decision.

Commercial products that do exactly this kind of signal aggregation:

• Bloomberg Terminal — $25,000-35,000/year (EVRY function for insider activity scoring)
• FactSet — $12,000+/year (SEC Spotlight signal module)
• Refinitiv Eikon — $20,000+/year
• S&P Capital IQ — $10,000+/year (built-in insider activity scoring)
• InsiderSentiment.com — $300/month
• InsiderArbitrage — $100-500/month

This actor delivers the same synthesis — insider + material events + signal score — via pay-per-event pricing.

Who this is for

• Quant traders building signal pipelines — one API call per ticker instead of 5
• Retail traders picking stocks — "should I buy NVDA?" → run this → see $294M insider selling → decide accordingly
• Portfolio managers doing due diligence — one call = full SEC picture
• Financial journalists — verify "insider activity" claims with real data
• M&A advisors — assess a target's insider sentiment before an approach
• Activist investors — check insider alignment before launching a campaign
• Compliance teams — monitor insider activity at portfolio companies

How to use

1. Click Try for free (or Start)
2. Add tickers to tickers: ["TSLA", "NVDA", "PFE", "AAPL"]
3. Set lookbackDays (default 90 = last quarter)
4. Configure how many Form 4 / 8-K / 10-QK filings to parse per ticker
5. Leave computeSignals: true to get the signal score
6. Click Start

Output is one intelligence report per ticker.

Output (per ticker)

{
    "ticker": "NVDA",
    "issuerName": "NVIDIA CORP",
    "issuerCik": "0001045810",
    "sicCode": "3674",
    "sicDescription": "Semiconductors & Related Devices",
    "exchanges": ["Nasdaq"],

    "form4FilingsCount": 15,
    "eightKFilingsCount": 4,
    "tenQKFilingsCount": 2,

    "insiderNetShares": -2048500,
    "insiderNetValueUsd": -294772058.66,
    "insiderBuyCount": 0,
    "insiderSellCount": 44,
    "insiderTotalTransactions": 44,

    "topInsiders": [
        { "name": "HUANG JEN HSUN", "title": "President and CEO", "netShares": -1500000, "netValueUsd": -213000000 },
        { "name": "DALLY WILLIAM J", "title": "Chief Scientist", "netShares": -350000, "netValueUsd": -49700000 }
    ],

    "signalScore": -4,
    "signalLabel": "STRONGLY_BEARISH",
    "signalReasons": [
        "High Form 4 filing velocity (15 filings)",
        "Net insider SELLING: $294,772,058 of sales",
        "Cluster sell: 44 sells by 3 distinct insiders"
    ],

    "recentForm4": [
        { "insider": "HUANG JEN HSUN", "title": "President and CEO", "transactionDate": "2026-03-28", "transactionCode": "S", "transactionType": "sell", "shares": 100000, "pricePerShare": 142.30, "valueUsd": 14230000 }
        // ...
    ],

    "recentEightK": [
        { "filingDate": "2026-03-20", "items": ["2.02", "7.01", "9.01"], "itemMeanings": ["Results of Operations and Financial Condition", "Regulation FD Disclosure", "Financial Statements and Exhibits"] }
        // ...
    ],

    "recent10QK": [
        { "form": "10-K", "filingDate": "2026-02-26", "reportDate": "2026-01-26" },
        { "form": "10-Q", "filingDate": "2025-11-19", "reportDate": "2025-10-26" }
    ],

    "lookbackDays": 180,
    "lookbackStart": "2025-10-18",
    "edgarUrl": "https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001045810&type=&dateb=&owner=include&count=40",
    "extractedAt": "2026-04-16T01:30:00.000Z"
}

Pricing

• Actor start: $0.05 per run
• Per ticker report generated: $0.25 per ticker

This is intentionally priced higher than our atomic actors (Form 4 $0.02/tx, 8-K $0.05/filing) because one call here replaces running 3-4 separate actors plus merging/computing downstream. You're paying for the synthesis, not just the raw data.

Example costs:

• Analyze one ticker → $0.30
• Daily scan of top 20 tech stocks → $5.05 per run = ~$150/month for a daily watchlist
• Weekly deep-scan of S&P 500 → $125 per run = ~$500/month for portfolio coverage
• Full signal scan of Russell 2000 → $500 per run

For comparison: Bloomberg Terminal is $25,000-35,000/year. FactSet is $12,000+/year. One analyst seat.

Free Apify tier members get $5/month in platform credits, which covers ~16 ticker reports per month.

Honest limitations

• Signals are heuristic, not ML-trained. The rules are simple (net insider buy = bullish, cluster sell = bearish). This matches institutional convention but is NOT a trained model. You'll outperform it with a proper ML model trained on historical returns.
• Only US-listed public companies. Companies must have a SEC CIK. Foreign issuers with ADRs work; private companies don't.
• No 13F or 13D integration (yet). Use the dedicated sec-13f-holdings-tracker and sec-13d-activist-filings actors for that. A future version could combine all five.
• No earnings-call transcript parsing. 8-K Item 2.02 filings reference earnings, but we don't parse the actual transcripts. Use the EDGAR Filing Monitor for that.
• Form 4 parsing is LIMITED to non-derivative transactions (common stock buys/sells). Derivative exercises (M code) count as "option-exercise" but we don't project them into a separate category for signals.
• Lookback is capped at 10 years (3,650 days). SEC submissions API only returns recent filings in the "recent" block; older filings require a paginated call we don't currently make.

Tips for alpha extraction

• Run daily before market open on your watchlist to catch overnight filings → position accordingly
• Schedule after-hours runs to catch late-filing 8-Ks (many earnings 8-Ks file at 4:05pm ET)
• Use longer lookback (180-365 days) for small caps where insider activity clusters are rare
• Combine with sec-13f-holdings-tracker: STRONGLY_BEARISH aggregator signal + hedge fund exiting = high-conviction short
• Combine with sec-13d-activist-filings: BULLISH aggregator signal + new activist 13D = double-confirmation buy
• Filter for signalLabel: STRONGLY_BULLISH across your watchlist to surface high-conviction longs

Related actors in my portfolio

For users who want atomic access to the same data:

• sec-form4-insider-trades — raw Form 4 transactions
• sec-edgar-filing-monitor — general SEC filing monitor (8-K, 10-Q, 10-K, S-1, etc.)
• sec-13f-holdings-tracker — institutional holdings per fund
• sec-13d-activist-filings — activist 5%+ positions
• clinicaltrials-pipeline-monitor — biotech pipeline (for signal-scoring biotech)

This aggregator is the top of the stack — one call that reuses the same SEC plumbing with computed signals.

Source

SEC EDGAR submissions API + Archives (both free, public, no auth). The actor respects SEC's 10 req/sec rate limit automatically.

License

MIT. SEC data is in the public domain.

Disclaimer

For research only. Past performance doesn't predict future returns. Insider activity signals are statistical biases, not guarantees. Don't trade real money on one signal. Always consult a licensed financial advisor.