U.S. corporate insiders — executives and directors of publicly traded companies — sold $77.6 billion of stock during the first half of 2026, a 20% increase from the prior year, according to data from EPFR Global Market cited by Yahoo Finance on July 17 (Yahoo Finance). The Gurufocus Overall Market Insider Buy/Sell ratio, a widely tracked aggregate of insider transaction activity, now stands at 0.25 as of July 2026, down from 0.27 the prior month and below the platform's historical average of 0.35 (Gurufocus).

Think of the buy/sell ratio as a thermometer for insider sentiment across U.S. stocks. It divides total insider purchases by total insider sales. A reading of 0.25 means that for every dollar of insider selling, insiders bought just 25 cents' worth. The historical average of 0.35 tells us the current tilt toward selling is steeper than what Gurufocus's dataset typically shows, though the metric has been lower in past market cycles.

This first-half figure builds on a trend visible earlier in the year. In February 2026, Reuters reported 2,260 recorded instances of U.S. corporate insiders selling shares against 543 instances of buying, producing a seller-to-buyer ratio of 4.2 — the highest level in 20 months — based on data from the Washington Service (Reuters). That February spike coincided with elevated market volatility, though the data does not establish that volatility caused the selling.

MarketWatch, reporting on the February figures in an April 6 article, pushed back against the bearish narrative that dominated headlines at the time. Its analysis found that insiders were "slightly more bullish in March than in February," a characterization that tempers the read-across from the February spike to a sustained trend (MarketWatch). A separate MarketWatch article from July 2024 had headlined that "Corporate insiders are dumping stock at the fastest rate in more than a decade," suggesting the elevated selling pattern has persisted across multiple reporting windows (MarketWatch).

Not all insider sales carry the same signal, though. Walmart disclosed in an SEC filing that CEO Doug McMillon is scheduled to sell 19,416 shares each month from June 2026 through January 2027 under a pre-arranged trading plan, commonly known as a 10b5-1 plan (SEC Filing). These plans are set up before the trading window opens, insulating executives from accusations of trading on material non-public information — that is, information not yet available to the public. The practical effect is that scheduled, automatic sales can inflate aggregate selling volumes without necessarily reflecting a discretionary bearish view.

That distinction matters when interpreting the Gurufocus and EPFR figures. A 20% year-over-year increase in insider selling dollar volume could reflect a mix of 10b5-1 plan executions driven by equity compensation vesting schedules, valuation-driven discretionary sales, and the market's price action during the period. The data does not separate these motivations.

The broader context here is the tension between the aggregate sell-side skew and the MarketWatch counter-narrative from April. A buy/sell ratio of 0.25 sits below the 0.35 historical average, and a 4.2 seller-to-buyer ratio in February was a 20-month high. Both data points lean bearish. Yet the March improvement MarketWatch flagged, combined with the mechanical nature of 10b5-1 selling visible in filings like Walmart's, cautions against reading the aggregate as a clean directional signal.

The dollar-volume figure carries its own caveat. A 20% increase in selling value does not necessarily translate to a 20% increase in selling urgency, since price appreciation alone can inflate nominal dollar sales without any change in the number of shares sold. Without adjusting for share counts, the $77.6 billion figure reflects both the decision to sell and the market price at which those sales occurred.

For anyone who tracks insider data as one input among many, the current picture offers a directional but noisy signal: aggregate selling is elevated by multiple measures, the buy/sell ratio sits below its historical norm, and at least some portion of the volume is structurally pre-committed rather than discretionary. The data warrants attention; it does not, on its own, constitute a timing signal.