——
Level — · Streak —
— XP
Mastery—weighted
Quizzes 30d—last 30 days
Cert progress—active level
Rank—global
Loading...
The page behind this dialog is live. Create a free account or sign in and you'll land right back on it.
Form 4 filings, cluster buys, and what historically beats the market vs. noise
Every officer and director who buys or sells their own company's stock has to tell the SEC within two business days. Those filings are public. The Insider tab pulls them in. The trick is knowing which trades carry signal and which are noise.
Insiders — officers, directors, and 10%+ shareholders — file when they trade their company's stock. The data is structured, machine-readable, and available within 48 hours of the trade. The signal value depends entirely on the type of trade.
Insiders sell for many reasons that have nothing to do with their view of the business: tax obligations, divorce, scheduled diversification, college tuition, buying a house. Insiders buy for essentially one reason: they think the stock will go up. Academic studies consistently find insider buying carries far stronger predictive power than insider selling — because the noise-to-signal ratio is much lower on the buy side.
A single insider buying $50K of stock is mildly interesting. Three different officers each buying $200K within the same two weeks is much stronger evidence. Multiple insiders, large dollar amounts, and short time windows together are called a , and they have outperformed isolated trades in academic studies by a wide margin.
Many insider sales come from , which let executives schedule sales months in advance. These pre-arranged sales carry almost no signal — the insider didn't choose the timing. The Insider tab labels 10b5-1 sales separately so you can filter them out. Discretionary opportunistic sales are a different matter.
Josef Lakonishok and Inmoo Lee analyzed 20+ years of insider trades and found that companies with heavy insider buying outperformed those with heavy insider selling by roughly 4.8 percentage points per year before transaction costs. The signal was strongest in small-cap stocks, where institutional analyst coverage is thin and insider information has more incremental value. Cluster buys delivered roughly double the alpha of isolated trades. The study became the academic foundation for taking insider buying seriously as a quantitative signal. Source: Lakonishok & Lee, 'Are Insider Trades Informative?', Review of Financial Studies, Vol. 14 No. 1, 2001.
Open the Insider tab for any ticker. The default view shows the last 24 months of Form 4 transactions, color-coded by buy vs. sell, with 10b5-1 sales filtered into a separate column so you can ignore the pre-arranged ones. Look for: (1) cluster buys — multiple insiders buying within a two-to-four-week window, (2) the dollar size of the buys relative to the insider's prior trading history (a CFO buying $5K is meaningless; a CFO buying $5M is a signal), (3) buys at multi-year price lows (insider conviction at a discount), and (4) the absence of any selling at all over a 12-month window in companies where insiders historically sell freely.
The most common beginner mistake is reacting to insider sales as if they were a sell signal. Most insider sales are 10b5-1 pre-arranged transactions, taxes, diversification, or charity. The CEO selling $20M of stock that's already 5% of their net worth doesn't tell you anything about the business. Insider BUYING is the high-signal channel because it has only one explanation. Filter 10b5-1 sales OUT of your analysis or you will be misreading noise as signal.
There is no better tip-off to the probable success of a stock than that people in the company are putting their own money into it. Insiders may sell their shares for any number of reasons, but they buy them for only one: they think the price will rise.
Insiders file Form 4 within two business days of any transaction; the data is public on SEC EDGAR. Insider buying carries far stronger signal than insider selling because the buy decision has only one explanation. Cluster buys (multiple insiders, large dollar amounts, short window) historically outperformed isolated trades by roughly 2x. Lakonishok & Lee 2001 found roughly 4.8 percentage points of annualized alpha from insider-buying portfolios. 10b5-1 plan sales are pre-arranged and carry essentially no signal — filter them out. The Insider tab labels and filters 10b5-1 sales so you can focus on discretionary trades.