By asahifile 
Two numbers have been circulating this week to describe the same event, and they differ by a factor of five. Reuters reported on October 26 that China’s STAR50 index of AI and semiconductor leaders had fallen 40% in less than four months, vaporising about $480 billion in market value. The Guardian, citing Bloomberg, reported a day later that $2.4 trillion had been wiped from Chinese stocks since May.
Both figures are accurate. They are not measuring the same thing, and the gap between them is the most useful fact in the story.
$480 billion, $2.4 trillion, and the $1.92 trillion in between
Subtract one from the other. If the whole Chinese market lost $2.4 trillion since May while the AI and semiconductor complex inside the STAR50 lost $480 billion, then $1.92 trillion — four-fifths of the damage — happened outside the AI trade that the headlines blame. AI names are, by Goldman Sachs’ count, 42% of the market capitalisation and 57% of the turnover of mainland A-shares plus Hong Kong H-shares. On these numbers they absorbed one-fifth of the losses.
That ratio argues for reading the selloff as a broad de-rating with an AI face rather than an AI bubble popping in isolation. The Guardian’s reporting runs the same way: weak household demand, thinner-than-expected margins, higher US rates and the threat of Western bans on Chinese data-centre components and humanoid robots. The global comparison is unflattering to both narratives — the Nasdaq is about 6% off its record, so the STAR50 has fallen roughly 6.7 times as far.
Reuters (via Telegraph India), The Guardian, Goldman Sachs figures as reported by both. Index levels and multiples are as reported, not calculated by this site; the derived totals are.