Your data on changing stock prices for five pharmaceutical companies from 1997 to 2010 (Nature 480, 16–17; 2011) erroneously indicate a 39% drop in aggregated share value, when in fact it would have risen by 82% (see Correction, Nature 480, 425; 2011). The error was due mainly to the selection of unadjusted, rather than adjusted, stock-market closing prices. (Correspondence)