The Insider Cluster Buying Signal: Why It’s the Strongest Free Signal

Most retail investors who follow insider activity make the same mistake: they treat every Form 4 purchase as equally meaningful. A director buying $15,000 worth of stock shows up in the feed right next to a CEO buying $2 million. Both are code P. Both are open-market purchases. But they are not the same signal—and neither of them individually is as powerful as what happens when multiple corporate insiders at the same company all buy independently within a two-week window: the insider cluster buying signal.

That pattern is called an insider cluster buying signal, and it is the highest-conviction publicly available indicator in the entire SEC disclosure ecosystem. This post explains precisely what qualifies, why the academic evidence backs the signal’s predictive power, and how to run a free screen to find live cluster buys without writing a line of code.

If you’re new to Form 4s entirely, start with our plain-English guide to reading SEC Form 4 filings. This post assumes you can already find and parse a Form 4—what it teaches is the signal-interpretation layer on top of that mechanical knowledge.

What Is an Insider Cluster Buy?

The definition has two hard requirements and one important qualifier. A cluster buy event is two or more unique corporate insiders independently purchasing shares within a 7–14 day window—each purchase recorded as an open-market transaction (transaction code P on Form 4), not a prescheduled or compensation-related event.

The qualifier is independence. A cluster buy requires that each insider made a separate decision to buy using personal capital. That is entirely different from a company-wide equity grant (code A) or a coordinated block purchase arranged by the board. The signal comes specifically from multiple people, across different roles and levels, each individually deciding that the stock is worth buying at the prevailing price.

The signal is strongest when insiders span different roles—a board director and an operating officer buying in the same window indicates conviction that is not confined to one function or level of the organization. Ten directors buying is notable. Three directors and two C-suite executives buying is more notable still.

Why Cluster Buys Outperform Single Trades

The intuition is straightforward, and it holds up under academic scrutiny. When one insider buys, you are watching a single person make a single bet. That person may have idiosyncratic reasons: they want to demonstrate confidence publicly, they are rebalancing their personal portfolio, or they genuinely believe the stock is cheap. Any of those could be true. A single purchase is one data point.

When three insiders buy independently within two weeks, you have three separate people, with different roles, different information flows, and different personal financial situations, each arriving at the same conclusion: this stock is worth buying with personal cash right now. The probability that all three are acting for idiosyncratic or coincidental reasons drops sharply with each additional independent buyer. Cluster buy signals historically produce roughly double the excess return of single-insider buy signals for exactly this reason.

The asymmetry between buys and sells also matters here. Insider sells are noisy signals because there are many valid reasons to sell—diversification, liquidity needs, tax obligations, estate planning—none of which reflect a view about the company’s future value. Insider sells are significantly less informative than insider buys precisely because motive is ambiguous. Buys have a single motive: the insider believes the stock is undervalued relative to what they know. That is why the cluster-buy signal works, and why the cluster-sell equivalent is a categorically different and less reliable indicator.

The Academic Evidence: What the Research Actually Says

Four decades of peer-reviewed research consistently find that insider purchases predict positive future returns. Three landmark studies are worth knowing in detail.

Seyhun (1986): The Foundation

H. Nejat Seyhun at the University of Michigan analyzed insider transactions from 1975 to 1981 and found 4.3% abnormal returns over the subsequent 300 trading daysfor firms with net insider buying, after adjusting for size effects. For firms with net insider selling, the equivalent figure was—2.2%. The asymmetry established what would become the consensus finding of the field: insider purchases carry information; insider sales largely do not.

Lakonishok and Lee (2001): Large Scale Confirmation

Josef Lakonishok and Inmoo Lee examined every NYSE, AMEX, and Nasdaq company from 1975 to 1995—one of the most comprehensive datasets in the insider trading literature. Their findings, published in the Review of Financial Studies, showed that firms with heavy insider buying beat firms with heavy insider selling by 7.8% over the following 12 months. After controlling for size and book-to-market effects, abnormal returns dropped to 4.8%—still economically significant and statistically robust across the full sample period.

Lakonishok and Lee also confirmed the contrarian pattern: insiders in aggregate buy after price declines and sell after price appreciation. They predict market movements better than simple contrarian strategies, which suggests genuine information rather than value-timing as the primary motive. In small-cap companies specifically, insider buys in open-market purchases delivered 7.4% abnormal returns over the subsequent 12 months—the small-cap premium attributable to lower analyst coverage and greater information asymmetry between insider and market.

Cohen, Malloy, and Pomorski (2012): The Sharpest Lens

The most practically useful study for screening purposes is “Decoding Inside Information,” published in the Journal of Finance (2012). Lauren Cohen, Christopher Malloy, and Lukasz Pomorski introduced the distinction between routine and opportunistic insiders. A routine insider trades in the same calendar month each year, suggesting mechanical, compensation-based patterns. An opportunistic insider trades irregularly.

The performance gap is stark: a value-weighted portfolio of opportunistic insider purchases earns 82 basis points per month in abnormal returns. The equivalent routine-insider portfolio earns returns indistinguishable from zero. Strip the routine trades from the universe and focus only on opportunistic buyers and the signal becomes roughly four times larger than using all insider purchases indiscriminately. The authors also found that opportunistic traders are significantly more likely to face SEC enforcement action—confirming that the classification captures genuine information-driven trading rather than a statistical artifact.

Opportunistic vs. Routine Insiders: The Filter That Doubles the Signal

The Cohen-Malloy-Pomorski framework is straightforward to apply in practice. Before trusting any cluster buy, check each insider’s historical filing pattern on EDGAR.

A routine insiderbuys every January regardless of price. You’ll see the same month of year appear repeatedly in their filing history. That regularity tells you the purchase is likely driven by a compensation program, a dollar-cost averaging plan, or a personal financial calendar—not by a conviction about current valuation.

An opportunistic insider buys irregularly. Their filing history shows no pattern by month or season. When they do buy, it tends to coincide with periods when the stock has pulled back or when they appear to have formed a strong view about near-term prospects. This is the profile you want to see in a cluster.

The practical check takes about two minutes per insider on EDGAR. Search the filer’s name, pull their Form 4 history for the issuer, and look at the dates of prior code-P purchases. If every prior purchase happened in January and February, treat the current purchase as routine. If there is no historical pattern and this is an irregular buy, particularly after a price decline, it is more likely to be opportunistic.

Reading the Form 4: The Fields That Matter for Cluster Screening

You don’t need to parse every section of a Form 4 to run a cluster screen. The full field-by-field guide covers everything—but for cluster identification specifically, four fields carry essentially all the weight.

Case Study: UnitedHealth’s 10-Director Cluster, April 1, 2026

On April 1, 2026, ten UnitedHealth Group board directors filed Form 4s showing open-market purchases on the same day—described by analysts as “the cleanest cluster-buy signal in a U.S. large-cap this year.” The April 1 cluster was not the first insider buying at UNH that month. The CFO and the CEOs of Optum and UnitedHealthcare had all made open-market purchases on March 17, two weeks earlier—none of them under a 10b5-1 plan.

Every element of the cluster met the screening criteria. All transactions carried transaction code P. None had the 10b5-1 checkbox ticked. The purchases used personal capital at prevailing market prices. And the buying occurred while UNH was approximately 46% below its 12-month high—exactly the contrarian pattern the academic literature identifies as the highest-signal environment for insider purchases.

What would a retail investor following cluster screens have seen, and when? The March 17 purchases would have appeared on EDGAR within two business days of the transactions—so by March 19. The April 1 cluster would have been visible by April 3. Both windows gave investors at least a brief period to observe the signal before it became widely reported.

Building a Free Cluster-Buy Screen in Under 10 Minutes

OpenInsider provides a free, real-time screener of SEC Form 4 filings with filter options including transaction code, minimum dollar amount, cluster-buy grouping, and insider role. The platform aggregates EDGAR data and allows filtering to “cluster buys” defined as multiple unique insiders buying at the same company within a user-specified number of days—updating within hours of EDGAR publication. Here is the four-step workflow:

What Cluster Buys Cannot Tell You (And What to Pair Them With)

The cluster-buy signal has real predictive power, and it also has real limitations. Being clear about both is what separates disciplined use from overconfidence.

Cluster buys predict direction, not timing or magnitude.The academic studies measure returns over 6–12 month windows. A cluster does not tell you whether the stock will move up next week, next month, or in six months. It also does not tell you how far it will move. Small-cap and mid-cap insider purchases produce materially larger post-filing returns than mega-cap purchases—which means the UNH example, involving a mega-cap with massive analyst coverage, should be expected to produce more modest returns than an equivalent cluster in a small-cap with limited coverage.

The signal lags by up to two business days.Form 4 must be filed within two business days of the transaction, but that is the maximum—not an average. By the time you see a cluster on EDGAR or OpenInsider, the trades are already in the past. In liquid large-caps, the market may have partially digested the information; in illiquid small-caps, the lag matters less.

Industry context matters. In technology and banking, insider purchase signals are significantly more informative than in utilities, where regulatory and dividend constraints mean insider buying more frequently reflects compensation-based accumulation rather than information-driven conviction. The same cluster pattern at a biotech pre-approval carries different predictive weight than the same pattern at a regulated utility.

A cluster at a company with serious fundamental problems is still a cluster at a company with serious fundamental problems. Insider conviction is not a substitute for basic fundamental analysis. Insiders can be wrong. Insiders at companies heading toward bankruptcy have bought stock and lost everything. Even sophisticated machine-learning models trained on insider Form 4 data achieve only AUC 0.70—statistically significant but far from certain. The signal raises probabilities; it does not guarantee outcomes.

The most robust approach pairs cluster-buy screening with two additional lenses: institutional flow data from SEC Form 13F filings, which shows whether large institutional holders are building or reducing positions, and basic fundamental filters to confirm the company is not in structural decline. Cluster conviction plus institutional accumulation plus intact fundamentals is a stronger combination than any single signal on its own.