Active share is computed by summing the absolute differences between portfolio and index weights and halving the total. A fund holding 40 stocks from a 500-stock index, in weights unrelated to the index, will typically score above 80%.
It answers a different question from tracking-error. A fund can hold very different names yet track closely if those names have similar risk characteristics, giving high active share and low tracking error. The reverse also happens with concentrated sector bets.
Low active share combined with active fees is the structural problem described under closet-indexing: the fee is charged on the whole portfolio while only a small slice can possibly outperform. High active share is not evidence of skill, only evidence that skill has room to show up.
Related: closet-indexing, tracking-error, alpha, style-drift, concentration-risk, benchmark