How insider filings predict stock moves

Company insiders sell for a hundred reasons — taxes, a divorce, a new house. They buy for only one: they think the stock is going up. Here's how to read that signal before it shows up in the price.

Every time an officer, director, or 10%-plus owner of a public company buys or sells that company's stock, U.S. law gives them two business days to tell the world. They file a Form 4 with the SEC, and it becomes public the moment it lands. No analyst, no paywall, no edge required — just a filing anyone can read.

Almost nobody reads them. That's the opportunity.

The single most useful idea in this entire area fits in one sentence, and a version of it is usually credited to Peter Lynch:

"Insiders might sell their shares for any number of reasons, but they buy them for only one: they think the price will rise."

A CEO might sell because she's funding a divorce, paying an estimated tax bill, diversifying a net worth that's 95% tied up in one ticker, or simply because a pre-set plan triggered. Selling tells you very little. But when an insider takes money out of their own pocket and puts it into open-market shares of the company they run, there's really only one motive — and they know more about that company than you ever will.

What a Form 4 actually tells you

A Form 4 is short and structured. The fields that matter:

• Who filed it, and their role — CEO and CFO buys carry more weight than a junior VP's.
• Transaction code — this is the whole game. Code P means an open-market purchase (real cash, real conviction). Code S is a sale. Codes A, M, and G are awards, option exercises, and gifts — not the same thing at all (more on that below).
• How many shares, at what price, and the total dollar value.
• Shares owned after — did this meaningfully increase their stake, or just nudge it?

The mistake everyone makes: a "buy" isn't always a buy

This is where most insider-tracking tools fall on their face, and it's worth slowing down on because it's the difference between a real signal and noise.

When a company grants its board their annual equity, every director gets shares on the same day — and a naïve scraper records all of them as "buys." It looks like the whole board just piled in. It isn't. Nobody spent a dollar; the company handed out stock as compensation.

Because one Robinhood director, Meyer Malka, kept showing up separately — writing real checks:

That's over $55 million of personal money into HOOD across eight days — open-market, code P, real cash. That's not a grant. That's a director who has seen the inside of the business and is betting his own net worth on it. Three weeks later, HOOD made a quiet ~22% move (we'll come back to that in 5 signals that moved before the news).

The rule: ignore grants. Watch for cash. A buy only counts when someone chose to spend money they didn't have to.

The patterns that actually carry signal

Once you're only counting real purchases, a few patterns separate the meaningful filings from the trivial:

1. The C-suite open-market buy

A CEO or CFO buying shares on the open market is the gold standard. They have the most information and the most to lose from looking foolish. When SoFi's CEO Anthony Noto bought 13,888 shares (~$251K) on June 16, that's a leader putting personal money behind his own guidance. When Strategy's CEO Phong Le bought ~$999K of MSTR on June 22, same read.

2. The cluster

One executive buying is interesting. Several different insiders buying within a short window — independently, with their own money — is a far stronger signal. It's hard for a whole leadership team to be wrong about their own near-term prospects at the same time. (The key word is different people spending real cash — not the same grant hitting everyone's account.)

3. Size relative to the person

A $20,000 buy from someone earning $400K a year is a gesture. A multi-million-dollar buy that visibly increases an insider's stake is conviction. Always read the dollar value against who's writing the check.

How Sort Brick uses this

We pull Form 4 filings continuously and fold "real insider buying" straight into the Signal Score — our 0–100 ranking of which names have the strongest mix of leading indicators behind them. Insider buying is one of the heaviest factors, precisely because it's cash conviction from people who know, not chatter.

Crucially, an asset can't score on hype alone. There has to be a real catalyst underneath — an insider cluster, a contract, an 8-K, an FDA approval — before a name even qualifies. Insider buying is often the first one to show up.

What insider buying does not tell you

We'd be lying if we sold this as a crystal ball, so here's the honest part:

• Timing is loose. Insiders are early, sometimes by months. They know the business, not the tape. A buy tells you they think it's cheap — not that it bounces next week.
• They're not always right. Executives are talking their own book and can be too optimistic about their own company. Conviction isn't omniscience.
• One buy is a data point, not a thesis. The signal lives in the pattern — repeat buys, clusters, size — cross-checked against price and everything else.

That last point is the whole reason Sort Brick exists. A single Form 4 is a clue. Stack it against chatter, price momentum, and the other leading indicators, and you get something far more useful than any one filing alone: a read on whether a move is real, fake, or quietly building before the crowd notices.

Information & analysis only — not financial, investment, or trading advice. Form 4 data is sourced from public SEC filings. Dollar figures are drawn from Sort Brick's filing database as of June 24, 2026 and reflect reported open-market transactions; we exclude grants and option exercises from "buying" signals. Markets are risky; do your own research. © 2026 Sort Brick.