What insider buying can (and can't) tell you
27 July 2026 · 5 min read
Corporate insiders – officers, directors, large shareholders – have to disclose their trades in their own company's stock, in the US within two business days. That fast disclosure window is what makes insider data interesting: unlike a 13F filing, which can lag by up to 45 days, you're looking at something close to real time.
The distinction that matters: 10b5-1 plans
Not every disclosed insider trade tells you something. A large share of insider selling – and a meaningful share of buying – happens through Rule 10b5-1 plans: pre-scheduled trading arrangements set up months in advance, often for entirely mundane reasons (diversification, an option exercise deadline, a pre-planned liquidity event). A 10b5-1 sale by an executive says essentially nothing about what they think of the stock today, because they didn't choose today – a plan they wrote long ago did.
That's why any serious reading of insider activity has to explicitly exclude 10b5-1 transactions and focus on open-market purchases specifically – trades where an insider looked at today's price, today's information, and decided to spend their own money. That's the signal with actual information content, if there is one at all.
Even then: it's a data point, not a verdict
One insider buying a modest amount isn't a strong signal – insiders buy for all sorts of idiosyncratic personal reasons that have nothing to do with the company's prospects. What's more informative is a cluster: multiple distinct insiders making genuine open-market purchases within a relatively short window. Clustering reduces the noise from any one person's individual reasons, though it never eliminates it.
What this is not
It's not a trading signal on its own, it's not a prediction, and a cluster of insider buying is not a recommendation to buy the same stock. It's one input, disclosed honestly, among several that go into a broader read on a company or a market – the same posture as every other piece of research on this platform: show the real data, be explicit about its limits, and let you draw your own conclusions.
Educational content only – not investment advice or a recommendation to trade. Nothing here estimates or implies future returns.