Peter Lynch wrote one of the most memorable lines in value-investing literature: "Insiders might sell their shares for any number of reasons, but they buy them for only one: they think the price will rise."
That asymmetry is the entire reason insider buys are a high-information signal and insider sells are usually noise. Yet most retail investors either ignore SEC Form 4 filings entirely or, worse, react to every insider sell as a bear signal.
This post walks through what's actually useful in Form 4 data, which transactions to weight, how to spot the high-conviction signals, and how to integrate them into a value-investing workflow without getting whipsawed by noise.
What is Form 4 and why does it exist
Form 4 is the SEC filing that corporate insiders - officers, directors, and 10%+ shareholders - must file within 2 business days of any transaction in their company's stock. The form discloses:
- Reporting person's name and title
- Transaction type (purchase, sale, gift, option exercise, etc.)
- Number of shares
- Price per share
- Date of transaction
- Total shares held after the transaction
The filings are public on SEC EDGAR within 24-48 hours of the transaction. They're the cleanest disclosure regime in equity markets - the data is structured, the timing is fast, the rules are strict.
The reason it exists: the Securities Exchange Act of 1934 wanted to prevent insiders from secretly trading on inside information. By forcing rapid public disclosure, the SEC turns insider knowledge into a (slightly delayed) public signal.
The asymmetry: buys vs sells
Why insider sells are mostly noise:
- Diversification - an insider's wealth is concentrated in company stock. Selling some to fund a house, pay taxes, or rebalance is rational and unrelated to outlook.
- Tax planning - long-term capital gains realizations, charitable donations of appreciated stock, estate planning.
- 10b5-1 plans - pre-scheduled selling programs filed months in advance. These execute mechanically regardless of current outlook.
- Compensation cycles - vesting events trigger automatic sales to cover withholding tax.
A CEO selling $20M of stock could mean "I think the stock is overvalued" or "I'm buying a yacht." You usually can't tell from outside.
Why insider buys are signal:
- Insiders already own a lot - they don't need MORE concentrated exposure to the company they already work for. Adding to the position is a deliberate choice.
- There are no tax-driven reasons to buy - sells have tax reasons, buys do not.
- There are restrictions - insiders can only buy during open windows (after earnings releases, before the next quarter's results). Every buy is intentional and time-windowed.
- The downside is concentrated - the insider knows the business better than anyone. Their buy is a vote with their own money.
Lynch's asymmetry holds: buys mean they expect the price to rise. Sells could mean anything.
Which insider buys matter most
Not all insider buys are equally informative. The hierarchy:
Tier 1: CEO and CFO open-market purchases
These are the highest-signal buys. The CEO has the most information about the business and the most to lose if the stock drops. A CEO buying $1M+ of stock on the open market is one of the cleanest bullish signals available.
The CFO's signal is even stronger in one specific case: when the CFO buys in the days following an earnings release where guidance was lowered. The CFO knows the numbers cold. If they're buying after a bad print, they think the market overreacted to the downside.
Modern example: a high-conviction CEO buy pattern
When Carlos Brito, then-CEO of Anheuser-Busch InBev (BUD), bought $5M+ in open-market purchases during the 2018 stock drawdown, it preceded the multi-year recovery. The CEO's confidence in the long-term franchise was the signal the market hadn't priced in.
Tier 2: directors (especially independent ones)
Directors have less day-to-day information than executives, but they sit on board meetings where strategic plans, competitive positioning, and major decisions are discussed.
The signal is strongest when:
- An independent director (not an executive, not a major shareholder) buys with their own money
- The director already serves on multiple boards (their reputation is at stake)
- The buy is opportunistic - shortly after a stock drop, not on a calendar schedule
Tier 3: 10%+ shareholders
These are activist investors, family offices, or strategic buyers building a position. Their buys can move stocks because the disclosure of a 10%+ stake (filed on Schedule 13D, parallel to Form 4) signals a thesis.
The signal here is the identity of the buyer, not the buy itself. A buy by Cliff Asness's fund means one thing. A buy by Bill Ackman means another. A buy by Berkshire Hathaway means a third.
Tier 4: routine insider buys
A VP-level executive buying $50K of stock at a quarterly window is noise. They participate in employee stock purchase plans, they got a small bonus, they did a routine top-up. Their information edge over the market is low.
Don't weight these heavily.
What to filter out: the noise patterns
Several Form 4 patterns are commonly mistaken for signal:
Option exercises
When an insider exercises stock options, the form shows a "buy" line item - they're buying at the exercise price. But this is mechanical. The option was granted years ago; the exercise is a tax-and-timing decision, not a fresh vote of confidence.
Filter: ignore Form 4 entries with transaction type codes M (option exercise), F (payment for option exercise), and S (sale to cover taxes).
10b5-1 plan transactions
10b5-1 plans are pre-scheduled trading programs filed at least 90 days in advance, specifically designed to be defensible against insider-trading claims. The trades execute mechanically.
Form 4 doesn't always flag 10b5-1 transactions explicitly, but a regular monthly pattern (e.g., insider sells 10,000 shares on the 15th of every month) is a tell. These are noise.
Gifts and inheritance
Transaction type G (gift). Mechanical, no information about outlook.
Cluster buying that's actually one decision
Sometimes a major insider buy is reported across multiple Form 4 filings as the trade prints fill across days. Three sequential Form 4s of $1M each isn't three signals - it's one $3M decision. Aggregate before weighting.
The cluster signal: when multiple insiders buy together
This is the strongest pattern in the data. When 3+ insiders buy independently within the same 30-day window, the signal is much stronger than any single buy.
The mechanism: insiders don't coordinate buys (that would be illegal). When you see three independent insiders all deciding to buy at the same time, they're all reacting to the same underlying assessment that the stock is mispriced.
Modern example: cluster buying signals
Several major buyout-thesis stocks in 2022-2023 showed cluster buying patterns 60-120 days before the deal was announced. The insiders weren't trading on deal information (that would be illegal) - they were buying because they believed the standalone business was undervalued, and the same belief separately motivated the buyout offer.
The cluster pattern is one of the inputs we use in invest-like's big_insider_buy alert detection - when the total dollar value of insider purchases in the last 3 days crosses $1M USD, the watchlist alert fires.
How to integrate insider data into a value-investing workflow
Three patterns I find consistently useful:
Pattern 1: insider buy + stock down 20%+ from 52-week high
This is the cleanest contrarian setup. The market thinks something's wrong. Insiders think the market is overreacting. When they buy at the bottom, it's high-conviction.
The trade isn't to mimic the insider - it's to use the buy as a trigger to do your own fundamental work. The insider sees something. Your job is to figure out what.
Pattern 2: insider buy + improving fundamentals
When insider buys happen as fundamentals are turning, the insider sees the inflection before the market prices it. Look for buys in companies where:
- Revenue growth re-accelerated last quarter
- Margins are stabilizing or expanding
- Capital allocation just shifted (new buyback authorization)
The insider knows what the next quarter's trend will be. The buy is the confirmation signal.
Pattern 3: cluster buying in a sector under pressure
When insiders across multiple companies in the same sector all buy in the same window, the signal is sector-wide rather than company-specific. The 2020 oil-and-gas insider buying during the COVID demand collapse was the cleanest example - executives across XOM, CVX, COP, EOG all bought materially within a 60-day window. The sector recovered 200%+ over the following 18 months.
How to monitor Form 4 filings without drowning in noise
The volume is high. SEC receives ~50,000 Form 4 filings per year for S&P 500 companies alone. Most are noise. You need filters.
Three approaches:
- Direct SEC EDGAR - free, complete, no filters. Useful for one-off research on specific companies. Tedious for monitoring.
- Aggregation services - several free and paid sites (OpenInsider, InsiderScore) aggregate Form 4 data and provide basic filters. Quality varies.
- Watchlist-integrated alerts - invest-like's watchlist alert system sends a digest email when a meaningful insider buy ($1M+ value, filing within 3 days) hits any ticker on your watchlist. Built specifically to filter out the noise patterns above.
Pick whichever workflow matches your time budget. The data is the same. The filtering is where you win or lose.
What insider data CAN'T tell you
To balance the above: insider buys are a strong signal, but they're not a complete thesis. Specifically:
- Insiders can be wrong about their own company. CEOs of Lehman, Enron, and Bear Stearns all owned massive amounts of stock right up to the collapse. Insider conviction isn't infallibility.
- Buy timing isn't perfect. Insiders buy on quarterly windows, often weeks after the fundamental setup is already in motion. The signal is "this is a reasonable price" not "buy at the open tomorrow."
- Insider buys don't tell you about competitive threats. The insider knows their business. They may not know what the competitor's next product launch looks like.
Use insider data as one input to a fundamental thesis. Not as a standalone strategy.
The integration with framework-based scoring
The most useful integration: combine insider buy signals with framework-based stock screening. The intersection is small but high-signal:
- Stocks that score "Strong Fit" on Buffett's framework AND have a meaningful insider buy in the last 30 days: ~5-10 names at any time.
- Stocks where the Munger framework scores high AND multiple directors are buying: rarer still, ~2-5 names.
These intersections are watchlist gold. They're the situations where both the documented framework AND the people closest to the business agree.
See current intersection picks on the watchlist - the digest email rolls these together: insider buys hitting tickers that also crossed framework thresholds get prioritized in the daily/weekly summary.
What to do tomorrow morning
If you don't currently monitor insider activity on your holdings, three concrete steps:
- Pick your top 10 holdings. Pull their Form 4 history from SEC EDGAR for the last 6 months. Note any insider buys >$500K. That's your baseline.
- Set up automated alerts on those 10 tickers. Use either OpenInsider's RSS feeds or invest-like's watchlist digest - the latter filters out the noise patterns automatically.
- Build a simple decision rule: when an insider buy >$1M hits a ticker on your watchlist, that's a trigger to re-read the most recent 10-Q + earnings transcript (here's how to read a transcript well). The buy isn't the decision - it's the prompt to do the work.
The discipline that beats the market over decades is systematic attention to high-signal data + the willingness to act when the signal confirms a fundamental setup. Insider buys are one of the cleanest sources of that signal. They're free, they're public, and most retail investors aren't paying attention.