The SEC Signals Trilogy
Three SEC filings. Three freshness cadences. Three branded scores. Read together, they give you a more complete picture of any public company than any single filing can.
The problem with reading SEC filings in isolation
Most investors who follow SEC filings stop at one type. They watch quarterly 13F filings for what hedge funds own. Or they watch Form 4 filings for what insiders are doing. Or they watch Form 8-K filings for material events.
Each one in isolation answers a useful question — but a different one. Together, they answer the question that actually matters: what is happening at this company, right now, as a system?
The three cadences
What makes the three filings complementary is that they have different freshness windows. They cover different timescales. They answer different questions.
| Filing | Cadence | Lag | Question it answers |
|---|---|---|---|
| 13F | Quarterly | 45 days | What did the smartest investors decide to own at the end of last quarter? |
| Form 4 | Per-event (daily) | 2 business days | What are this company's own officers actually doing with their stock right now? |
| Form 8-K | Per-event (intra-day) | 4 business days | What material things has this company itself just announced? |
The three scores
HoldLens computes one branded composite score from each filing type. Each is a deterministic transformation of public SEC data — fully reproducible from the source filings, no machine learning, no opinion, just a formula. The exact formulas live in our methodology page.
ConvictionScore — the 13F score
A signed score from −100 to +100 capturing the strength of a superinvestor's conviction in a single position, computed from 8 quarters of 13F filings. Inputs: position size, quarter-over-quarter change, persistence across 8 quarters, peer overlap. Use it when: you want to know which positions reflect deep, durable conviction vs. mechanical rebalancing.
InsiderScore — the Form 4 score
A signed score from −100 to +100 measuring the signal quality of corporate-insider transactions. Inputs: officer role (CEO > CFO > director > 10%-owner), action type (discretionary buy > sell; 10b5-1 sales discounted), transaction size relative to historical norm, cluster detection (multiple officers trading the same direction within 30 days), recency decay. Use it when: you want to see what the company's own officers think about their stock right now — much fresher than 13F.
EventScore — the 8-K score
A signed score from −100 to +100 measuring the material significance of an SEC Form 8-K filing. Inputs: item-type severity (bankruptcy > CEO departure > material agreement), market-cap weight, recency decay, event-cluster bonus. Use it when: you want to know what the company itself has just told the market about — the freshest signal of the three.
How to read the trilogy together
Reading any single one is useful. Reading all three for the same company at the same moment is where the picture gets sharp. A few patterns to look for:
- High ConvictionScore + cluster buy + positive 8-K event: a superinvestor's quarterly thesis is being confirmed by both insider behavior and management action. Strong triangulated signal.
- High ConvictionScore + cluster sell + negative 8-K event: the superinvestor's stale 13F is being undermined by fresher signals. Worth re-reading the 13F position with skepticism.
- No 13F position + heavy insider buying: insiders see something that the smartest outside money hasn't reacted to yet. Watch the next 13F cycle.
- Bankruptcy 8-K + heavy insider 10b5-1 sales in prior weeks: 10b5-1 plans don't generally indicate signal — but a clean cluster of pre-arranged sales right before a bankruptcy 8-K is a pattern that has shown up in more than one corporate failure.
- Restatement 8-K + recent activist 13D filing: activist pressure forced earnings re-examination. Activist's thesis is being validated by management's own disclosure.
None of these are guaranteed signals — every market regime breaks every pattern eventually. But reading the trilogy together filters far more noise than reading any single filing in isolation.
Where the three live on HoldLens
Each filing type has a dedicated surface, and each per-company page (e.g., for AAPL) cross-links to all three for that ticker.
- /investor/ — per-superinvestor 13F holdings + ConvictionScore trends across 8 quarters
- /insiders/ — daily insider transactions + InsiderScore + cluster-buy detection
- /events/ — material 8-K filings + EventScore + per-item-type drilldowns (cybersecurity, bankruptcy, M&A, etc.)
Honest about what the three CAN'T do
The trilogy is a research tool, not a predictive system. Every signal can be wrong. Every cluster buy that looked smart in retrospect was preceded by ten that looked smart at the time and wrong six months later. Every restatement 8-K that crashed a stock was preceded by ten that didn't.
What the trilogy does do is reduce the surface area where you have to guess. Instead of reading one stale 13F and trying to extrapolate, you have three independently-sourced data streams to triangulate against. That's a better decision-making position than any of the three alone, but it's not a crystal ball.
See our disclaimer for the full statement on data lag, accuracy, and the explicit not-investment-advice framing.