Capital Wealth
TUE CLOSE · OCT 6   S&P 500 7,818.93 ▲0.58%  ·  DJIA 51,521.28 ▲0.49%  ·  NASDAQ 27,599.79 ▲0.45%  ·  10-YR 5.27%  ·  2-YR 4.79%  ·  WTI $89.44 ▲0.01%  ·  GOLD $4,159.20 ▲0.75%  ·  VIX 15.01 ▼3.3%
Heard on the Street · Insider Trading · IN04

Bank Insiders Stopped Buying Their Own Stock — Lowest Rate in 23 Years — a Week Before Earnings

Executives can't tell you when to sell, but their filings talk. This quarter the loudest message came from the banks and the refineries. Read insider tables as a mood ring for a sector, not a clock for your trades.

By Sean Anees Saifi · Capital Wealth · Published Wednesday, October 7, 2026 · Source: The Wall Street Journal, Wednesday, October 7, 2026, whose market figures are the Tuesday, October 6 close (Heard on the Street)
Key Points
23 yrs
financial-exec insider buying at its lowest rate in almost, Q3 (VerityData)
$250M
Broadcom co-founder Samueli's Sept. 23 sale, Monday's insider table
206%
PBF shares this year through Tuesday — while executives sell
0.5×
consumer-staples sell/buy ratio vs its long-run average, the lowest sector
A man in a long dark overcoat, seen from behind, walks past the revolving door of a glass-and-steel office tower across a wide granite plaza, the low sun at the end of the street throwing long shadows.
Sellers weren't unusually numerous in finance this quarter; it was the buyers who went missing, at a rate not seen in almost 23 years.
In one line: Bank executives bought their own stock at the lowest rate in almost 23 years before next week's earnings and refiner bosses sold into records; insider tables are a sentiment check on a sector, not a timing tool, and the books hold the banks at weight.

“Blue Horseshoe loves Anacott Steel” is the movie version of insider trading. The legal version is duller and more useful: executives report their buys and sells to the SEC within two business days, and the Journal prints the biggest ones every Monday. This week's table and Spencer Jakab's Wednesday Heard column read together like a mood ring. Financial executives bought their own stock in the third quarter at the lowest rate in almost 23 years, per VerityData — a week before the big banks report. Refinery bosses are selling into all-time highs. And Broadcom (AVGO) co-founder Henry Samueli sold $250 million of stock on a single September day.

Start with the banks. Goldman Sachs (GS), Morgan Stanley (MS) and JPMorgan Chase (JPM) each earn roughly four times per share what they did a decade ago, and the KBW Nasdaq Bank Index set a record this summer. Ben Silverman of VerityData says the number of financial-industry buyers fell to a near-23-year low in the quarter; sellers weren't unusually many, but the ratio of sellers to buyers was the most stretched of any sector against its long-run average. Return on tangible equity is the best in years at some big banks. What the insiders may be weighing that the market isn't: strained consumers, trouble surfacing in private credit, and AI companies borrowing on a scale never seen before.

Energy is the other lopsided sector. Executives at the smaller refiners Par Pacific and PBF have been selling with their shares up 147% and 206% this year through Tuesday and just under records. Jakab's explanation: $100 crude is the headline number, but the margin that matters is what a refinery earns turning it into diesel — and on that basis a barrel fetched more than $200 last week, until Europe agreed to tap its stockpiles. People with decades in the business know that doesn't last; the crowd may be slower. Monday's table shows the same instinct up and down tech — Samueli's $250 million Broadcom sale, Mark Zuckerberg's roughly $21 million of Meta (META) at prices above $775, CrowdStrike's finance chief at $125 million, Ajit Jain parting with about $20 million of Berkshire (BRK.B). By sector, tech insiders sold roughly 210 million shares against under 11 million bought.

Where the money's going the other way

Consumer staples has the lowest sell-to-buy ratio of any group, at half its long-run average, and consumer discretionary is subdued too; Jakab flags recent purchases by executives at Altria (MO), Tyson Foods and Celsius. Monday's buyer list is thinner and smaller: GameStop's Ryan Cohen bought about $10.6 million, and an Oracle (ORCL) director put in $3.5 million at 139.35 with the stock down 27% this year. Jakab's own caveat is the important one — executives aren't great judges of Wall Street's mood, so the sector signal is weaker than the company-specific one. It isn't a timing device. It's a sentiment check.

Our read

File this under Investments/Risk (IN04) and Behavioral. What insider tables are good for is catching a divergence between the people who run a business and the people who price it. Right now the divergence is in financials and refiners, and the books are on the long side of both: JPM in 26 books, GS in 19, MS in 16, plus the energy sleeve. Nothing changes because of a ratio — the banks report next week, and that's when the earnings do the talking — but it's a fair reason to look at how big the bank and energy lines have grown after a strong year. Broadcom (29 books) and Meta (20) are held at weight; nothing added, nothing trimmed on a founder's diversification sale.

The transferable rule: insiders sell for a hundred reasons and buy for one, so weight the buys and read the sells as a crowd barometer. Then borrow their discipline. If a position has doubled, you don't need anyone's signal to take it back to weight — that's what the review is for, and it's better done before the earnings than after.

What It Means For Your Portfolio

Hold — read insider tables as mood, not as a clock

No portfolio action — JPMorgan (26 books), Goldman (19), Morgan Stanley (16), Broadcom (29) and Meta (20) are held at weight ahead of next week's bank earnings; insider ratios are a sentiment check, and the move they justify is a weight check, not a trade.

General planning principles, not advice for anyone in particular. Treat sector-level insider data as one sentiment gauge among several: it tells you where the people closest to the business are more cautious than the crowd, not when the crowd will catch up. Company-specific clusters of buying carry more information than any sector ratio.

The useful personal version is the trim rule: positions that have run far past their target weight get taken back on a schedule, not on a hunch. Write the rule down before earnings season, when it's easiest to follow.

Book a 15-Minute Review → Back to Edition No. 180 →