AlphaYouDo Corporate Insiders Really Beat the Market? What the Research Shows
The academic answer is a qualified yes, but only for purchases, only modestly, and mostly in smaller companies. Lakonishok and Lee, Jeng Metrick and Zeckhauser, and what a third of Form 4s being pre-scheduled does to the signal.
"Insiders know something" is one of the most durable beliefs in retail investing. It is also one of the few that has been studied properly for decades, so we do not have to guess.
Direct answerResearch consistently finds that insider *purchases* predict modest positive abnormal returns, while insider *sales* predict essentially nothing. Lakonishok and Lee found around 4.8% excess returns for strong-purchase portfolios; Jeng, Metrick and Zeckhauser found roughly 52 to 68 basis points a month over six months. The effect is real, small, concentrated in smaller companies, and diluted by the fact that around a third of insider transactions are pre-scheduled.
- Two patterns repeat across essentially all of this literature, and both matter more than the headline percentages.
- Every study above finds the same asymmetry. Purchases predict something; sales predict close to nothing.
- Note what these numbers are not. Nobody found that following insiders doubles your money. An excess return in the low single digits, before costs and taxes, is a modest edge that could easily be eaten by trading frictions or by the position sizing of someone chasing it.
What the Research Actually Found
Two patterns repeat across essentially all of this literature, and both matter more than the headline percentages.
| Study | Finding |
|---|---|
| Lakonishok and Lee (2001), *Are Insider Trades Informative?* | Portfolios built on a strong insider **purchase** signal showed excess returns of about **4.82%**. Strong **sell** signal portfolios showed no significant excess returns. |
| Jeng, Metrick and Zeckhauser (2003) | Insider purchase portfolios earned abnormal returns of roughly **52 to 68 basis points per month** over the first six months. Sale portfolios showed no significant negative returns. |
| Later small-cap work | Insider buys in smaller companies associated with roughly **7.4% abnormal returns** over the following twelve months, with no meaningful underperformance following insider sales. |
Pattern One: Only Buying Carries Information
Every study above finds the same asymmetry. Purchases predict something; sales predict close to nothing.
The reason is structural rather than statistical. Insiders sell for dozens of reasons unrelated to their view of the company: tax bills, diversification, buying a house, divorce, scheduled compensation vesting. Our own data shows how lopsided this is - we logged **11,544 insider sells against 3,538 buys**, about 3.3 to 1. Selling is the background hum.
Buying is the unusual act. An executive already dependent on the company for salary, bonus and existing equity choosing to concentrate *further* into it with their own after-tax money has one obvious explanation and few alternatives.
Pattern Two: The Effect Is Small and Concentrated
Note what these numbers are not. Nobody found that following insiders doubles your money. An excess return in the low single digits, before costs and taxes, is a modest edge that could easily be eaten by trading frictions or by the position sizing of someone chasing it.
The effect also concentrates in smaller, less-covered companies, which is intuitive: in a megacap followed by forty analysts, an insider knows relatively little that the market has not already priced. In a thinly covered smallcap, they know a great deal more.
The Modern Complication the Old Studies Predate
Roughly **32% of the insider transactions in our data were pre-scheduled 10b5-1 plan trades** - executed on a timetable set months earlier. A plan trade reflects a decision made in the past, under different information, and carries close to zero signal about what the insider thinks now.
If you build a strategy on all Form 4 activity without filtering plan trades, you are diluting whatever real signal exists with a third that has none by construction. The plan status is on the filing, which makes this a solvable problem rather than an inherent limit.
Cluster buying is the other well-known refinement: several distinct insiders at one company buying in a short window, none on a plan. One person can be idiosyncratic or wrong. Four independently is harder to explain away.
What This Means Practically
Insider buying is best understood as a filter, not a trigger. It is a reason to look harder at a company you can already form a view about, not a reason to buy something you have never heard of.
And the honest framing on all of the above: these are historical statistical relationships measured across large portfolios over long periods. They describe averages, not individual outcomes, and none of them promise anything about the next trade you see. Past performance does not predict future results.
FAQ
Do insider purchases beat the market?
Research consistently finds a modest positive effect, roughly 5% excess returns in Lakonishok and Lee, and about 52 to 68 basis points a month in Jeng, Metrick and Zeckhauser. Real, but small.
What about insider selling?
The same studies find little to no predictive value. Sells outnumber buys more than three to one and happen for personal reasons unrelated to company outlook.
Where is the effect strongest?
Smaller, less analyst-covered companies, where insiders hold more of an informational advantage over the market.
Does every Form 4 carry signal?
No. Around a third are pre-scheduled 10b5-1 plan trades reflecting decisions made months earlier. Filtering those out matters before drawing any conclusion.
Get the filings this guide is about.
Funds, Congress and insiders, read the moment they file. Free, and one unsubscribe click.