Insider Buying vs Selling: Why Only One Predicts Stock Returns

If you follow insider trading data, you've probably noticed something: insiders sell far more often than they buy. In any given month, insider sales outnumber purchases by 5-to-1 or more. Does that mean insiders are bearish? Should you worry when a CEO sells? The answer, backed by decades of research, might surprise you.

The fundamental asymmetry

Insider buying and insider selling are not opposite signals. They carry completely different informational weight:

The reason is simple: insiders sell for many reasons, but they buy for only one.

Why insiders sell (and why it usually doesn't matter)

When a CEO sells shares, it could mean any of the following:

• Diversification — Most CEO compensation is heavily weighted toward company stock. Financial advisors universally recommend reducing concentration risk. A CEO selling 10% of holdings is just basic portfolio management.
• Tax obligations — Stock option exercises trigger tax events. Many executives sell shares specifically to cover the tax bill.
• Liquidity needs — Buying a house, funding education, estate planning, divorce settlements. Life happens.
• Pre-scheduled plans — 10b5-1 plans are set up months in advance and execute automatically. The CEO may have forgotten the plan was even running.
• Expiring options — Stock options have expiration dates. If they're in the money, the executive exercises and often immediately sells. This is rational regardless of their view on the stock.

What the academic research shows

The asymmetry between buying and selling signals is one of the most consistent findings in insider trading research:

• Lakonishok & Lee (2001) — Found that insider purchases predict positive abnormal returns over 12 months, but insider sales have no statistically significant predictive value for future stock performance.
• Jeng, Metrick & Zeckhauser (2003) — Confirmed that insider purchase portfolios earn significant abnormal returns, while insider sale portfolios show no abnormal returns.
• Seyhun (1998) — In one of the most comprehensive studies covering 20+ years of data, found that aggregate insider buying is a reliable market-timing indicator, but aggregate selling is not.
• Ravina & Sapienza (2010) — Even among independent directors (who have less information than executives), purchases predict positive returns while sales do not.

When insider selling actually matters

While individual insider sales are usually noise, there are specific patterns that do carry informational value:

1. Cluster selling — When 3+ executives at the same company sell large amounts within the same week, that's unusual and potentially informational. A single sale is nothing; coordinated selling is a pattern.
2. Unusual size — A CEO who normally sells $200K per quarter suddenly dumping $5M is worth noting. The deviation from normal behavior is the signal, not the sale itself.
3. Selling immediately after buying — If a CEO bought shares 2 months ago and is now selling, something may have changed in their outlook. This is rare but meaningful when it occurs.
4. Selling before bad news — This is actually illegal (insider trading in the traditional sense), but it does happen. If you notice heavy selling before an earnings miss or FDA rejection, it may indicate the executives knew what was coming.

Why we only track buying

CEO Trader deliberately ignores insider selling. Here's why:

• The academic evidence is clear: buying predicts returns, selling doesn't
• Tracking sells would flood the signal with noise (5x more transactions)
• Every sell requires contextual judgment (is it a tax event? diversification? 10b5-1 plan?) that can't be reliably automated
• Buying is a clean, high-conviction signal that can be scored and tracked systematically

By filtering exclusively for C-suite open-market purchases, we isolate the one transaction type that academic research consistently identifies as predictive. Each purchase then goes through a 6-factor scoring system that evaluates the signal in the context of fundamentals, financials, technicals, conviction level, CEO history, and cluster activity.

The bottom line

Don't panic when you see a CEO sell. Do pay attention when a CEO buys. The asymmetry is real, well-documented, and forms the foundation of every signal on our track record page.

The next time someone tells you "insiders are selling!" remember: they're always selling. That's the default state. What's unusual — and informative — is when they buy.