Not every insider buy is a buy.
Every week someone posts it: "Insider buying on SEDI!" Screenshot, rockets, the room gets excited.
Half the time, nobody bought anything.
SEDI records every change in an insider's holdings — including plenty that involve no decision and no money at all. The signal is in there. You just have to know which lines to read.
Here's how I read an insider filing.
Start with the code, not the headline
Every SEDI transaction carries a nature-of-transaction code. That number tells you what actually happened. The ones that matter most:
→ 10 — bought or sold on the public market
→ 11 — bought or sold privately
→ 16 — acquired under a prospectus exemption, usually a private placement
→ 46 — shares received as compensation for services
→ 50 / 51 — options granted / options exercised
→ 53 / 54 — warrants granted / warrants exercised
→ 90 — change in the nature of ownership, like shares moving into a holding company
A code 50 shows up as an "acquisition." It's compensation. Nobody spent a dollar.
A code 90 isn't a trade at all.
Open-market buys are the signal
A code 10 buy is the purest line in the file. Full market price. The insider's own money. No discount, no free warrant, no tax angle — just a decision that the stock is worth more than they're paying.
A private placement (16) is still real money, and still a good sign. But it came at a placement price, often with a warrant attached. Less pure.
An option exercise (51) near expiry is usually mechanical: use them or lose them. The real tell is what happens next. Held, or sold within days?
Size it against the person, not the company
SEDI shows the balance held after every transaction. Use it.
→ 150,000 shares bought, 180,000 held after: they just grew their position sixfold.
→ 150,000 shares bought, 9 million held after: a rounding error.
Same trade. Completely different message.
Who's buying matters too. A CEO knows the asset from the inside. A large fund holder is an insider on paper, but it's often trading for its own portfolio reasons.
One insider is an opinion. Several is a signal.
One director buying could mean conviction, or just optics. Three or four insiders buying on the open market within a few weeks is a much stronger read.
Routine trades tell you nothing
A well-known study by Cohen, Malloy and Pomorski split insider trades into "routine" — insiders who trade on a predictable calendar — and everything else. The routine trades, over half the total, carried essentially no information. The predictive power was in the trades that broke pattern.
It's US data, but the logic travels. SEDI lets you pull an insider's past filings. The question isn't just "did they trade?" It's "is this normal for them?"
Sales are noisier than buys — but not always noise
Insiders sell for a hundred reasons: taxes, a house, diversification, covering the cost of options they just exercised. They buy for basically one.
So most sales deserve less weight. The ones worth a second look: a big share of their personal holding, several insiders at once, or someone who never sells suddenly selling.
Mind the timing
Insiders generally have five calendar days to file. The transaction date and the filing date aren't the same — check both.
Most companies also block insiders from trading around material news. That's why insider buying often shows up in bursts right after news comes out — and why a quiet SEDI during a blackout tells you nothing.
Insiders sell for a hundred reasons. An open-market buy has one.
Before you react to "insider buying," check four things:
→ The code: was it actually an open-market buy?
→ The size: against what they already held
→ The crowd: is anyone else buying?
→ The pattern: is this normal for them?
SEDI is free, public, and updated within minutes of a filing. Two minutes there beats any screenshot.
Previously: "Ownership trumps results" (reading financings) and "An ounce isn't an ounce" (reading resource estimates).
I am an ambitious economics and finance student sharing whichever findings I deem relevent. stay tuned
