The case for paying attention to insider purchases is narrow and worth stating precisely. Directors sell for many reasons — tax, a house, diversification, divorce — so selling carries little information. They buy for essentially one: they expect the shares to be worth more. That asymmetry, not any claim of prescience, is the whole argument.
What it does not mean is that a purchase predicts the share price. Insiders are frequently early, sometimes by years, and are as capable as anyone of being wrong about their own company. They are also subject to a well-documented bias toward optimism about the thing they run.
The features that carry information are mostly not the headline. Whether it was a genuine open-market purchase rather than an award. Whether several people bought independently. How large it is relative to that person's existing holding and their pay, rather than in absolute pounds. Whether they bought into a falling price or a rising one. And whether the buyer is someone with a track record worth anything.
Treated as one input among several, that's useful. Treated as a recommendation, it isn't — which is why everything here is information rather than advice, and why the performance figures published alongside it include the purchases that didn't work.
Every explainer · Browse companies
Information only, not investment advice.