What Insider Buy Signals Actually Tell You (And What They Don't)

Every day, the SEC's EDGAR system publishes thousands of Form 4 filings — the mandatory disclosures that corporate insiders (officers, directors, and 10%+ shareholders) must file within two business days of buying or selling their company's stock.

Most people ignore this data. That's a mistake.

Why insider buys matter more than insider sells

Here's the thing about insider transactions: sells are noisy, buys are signal.

Insiders sell for a million reasons — diversification, taxes, buying a house, their financial advisor told them to rebalance. An insider selling stock tells you almost nothing about their conviction in the company.

But when an insider buys stock on the open market with their own money? That's different. There's exactly one reason to do that: they think the stock is going up. They already have massive exposure to the company through their salary, options, and existing holdings. Voluntarily adding more risk? That's conviction.

The data nerd's edge

The raw Form 4 data from EDGAR is... not pretty. It's XML wrapped in SGML wrapped in a filing index that looks like it was designed in 1996 (because it was). Parsing it reliably at scale is a genuinely interesting data engineering problem.

At kibble.shop, we're building a clean, queryable dataset from these filings. Every transaction type, every derivative instrument, every footnote that explains why an insider got shares. The goal: make this data as easy to work with as checking the weather.

What the research says

Academic studies have consistently found that insider purchases outperform the market by 4-8% annually, depending on the study period and methodology. The signal is strongest for:

• Cluster buys — multiple insiders at the same company buying within a short window
• Large purchases — transactions that are meaningful relative to the insider's compensation
• Small and mid-cap companies — where information asymmetry is highest
• CEO and CFO buys — the people with the broadest view of the business

What they don't tell you

Insider buy signals aren't magic. A few caveats:

• Timing is approximate. Insiders are often early. Sometimes very early. A cluster of insider buys might precede a turnaround by 6-12 months.
• Context matters. An insider buying after a 50% drawdown is a very different signal than an insider buying at all-time highs.
• Not all insiders are created equal. A director who sits on 12 boards and buys a token amount is less informative than a COO making the largest purchase of their career.
• It's one input, not a strategy. The best use of insider data is as a confirming signal alongside fundamental and technical analysis.

What we're building

Our SEC Insider Trading data product on kibble.shop parses every Form 4 filing in near-real-time and enriches it with company metadata, transaction classification, and historical context. You'll be able to:

• Screen for cluster buys across the entire market
• Filter by insider role, transaction size, and company characteristics
• Get alerts when unusual buying patterns emerge
• Access everything via dashboard or API — whichever fits your workflow

The data is free while we're in early access. We're building this in public, and we want people who actually use this stuff to help us make it great.

Want to dig into insider trading data yourself? Sign up for early access and poke around. It's free, and I'd love to hear what you think.