With AI stocks falling recently, is it time to move some capital back to the Hong Kong stock market?

Now the AI bubble has clearly burst, high-flying stocks are all falling outside. Shouldn't foreign capital free up some positions and return to Hong Kong stocks?

Yesterday, as soon as Seres released its earnings forecast, many people laughed that Aito loses over 20,000 yuan per car sold. But to be honest, Xiaomi's Q2 probably won't be much better. Memory price hikes have been too fierce. Whether it's phones or cars, Xiaomi is all about packing in large memory, so profits on both sides are bound to be under pressure. Just thinking about the financial report is worrisome. Moreover, Xiaomi lumps AI investment and automotive R&D together in its accounting, making it impossible to separate. Who knows how much the cars are actually losing and how much money AI is burning? It's simply impossible to calculate in detail.

Compiled the Q2 delivery data for both companies for your reference:

2026 Q2 Delivery Data

Xiaomi Auto (CPCA Retail Sales Basis)

• April: 36,702 units
• May: 32,759 units
• June: 34,738 units
• Q2 Total: 104,199 units
• Profit/Loss per Vehicle: No official separate data, can't calculate

Aito Series

• April: 23,152 units
• May: 34,320 units
• June: 30,199 units
• Q2 Total: 87,671 units
• Q2 Net Loss Attributable to Parent: 1.9B ~ 2.15B yuan (earnings forecast figure, includes asset impairment provisions, not all from car sales losses)
• Calculating with 1.9B: 1.9B ÷ 87,671 ≈ 21.7k yuan/unit
• Calculating with 2.15B: 2.15B ÷ 87,671 ≈ 24.5k yuan/unit
• On average: roughly a book loss of about 23k yuan per car

To be honest, it's not about which is better than the other; all the new EV players are struggling on the profitability front. The divergence at AI's high levels is growing stronger. Funds shifting to lower-priced sectors is inevitable. Hong Kong auto stocks have been adjusting for so long; a capital return should be coming.

$XIAOMI-W(01810.HK) $SERES(09927.HK)