The Shanghai Composite rebounds and turnover rises: does this show capital is starting to shift?
Exchange data support a rebound in the indices from their daily lows and higher stock turnover on both exchanges, but they are insufficient to confirm that institutions increased positions, leverage declined, or ETFs absorbed selling. The more defensible judgment is that prices recovered; a shift in capital flows and its persistence still require independent evidence. The decline in the overseas semiconductor index also provides only context and cannot directly explain A-share fluctuations.
The rebound is established; a shift remains unproven
On October 13, 2026, the Shanghai Composite Index reached a low of 3,741.53 points and closed at 3,800.86, up 0.10%; the ChiNext Index reached a low of 2,860.80 and closed at 2,958.34, down 0.18%. Both closed above their daily lows and below their daily highs. These data support a limited but clear judgment: prices rebounded during the day, with different degrees of recovery by the close. Interpreting this as evidence that capital is starting to shift goes beyond what index quotations can tell us.
Intraday highs and lows do not tell us who the buyers were
Both the Shanghai Composite and the ChiNext Index opened below their previous closes and reached intraday highs above those closes, but one ended higher and the other lower. This supports an inference that prices reversed direction substantially during the day, based on the relationship between the opening levels, highs, lows, and closing levels. The quotation table does not give the times of the highs and lows, so it cannot be used to reconstruct the full intraday sequence, much less determine whether the rebound came from institutional allocation, short-term trading purchases, or other activity. Closing below the daily high is also insufficient on its own to prove that investors were unwilling to chase rising prices.
Higher turnover means more trading
On October 13, 2026, stock turnover was RMB 780.423 billion in Shanghai and RMB 897.631 billion in Shenzhen, compared with RMB 715.758 billion and RMB 823.634 billion, respectively, on October 12, 2026. Under each exchange's own statistical definitions for stocks, turnover was higher than on the previous trading day. These stock statistics include both A-shares and B-shares and cannot simply be labeled as A-share-only turnover. Turnover records the value of trades that have already occurred; it cannot identify the buyers and sellers or be read directly as a net inflow of new capital. A price rebound alongside higher turnover is still insufficient to confirm sustained buying support from capital allocated to investment positions.
The explanation for divergent capital flows needs a separate test
The WeChat account argues that deleveraging and buying through ETFs on dips can occur at the same time, using this to explain divergent directions in market capital flows. This offers a testable hypothesis: different accounts and trading instruments may follow different strategies. But the exchanges' stock turnover tables provide neither outstanding margin financing balances nor ETF subscription and redemption information, so they cannot confirm that this explanation holds. Testing it requires financing data with clear statistical definitions, fund subscription and redemption data, and corresponding time windows. Even if divergent capital flows are confirmed, their contribution to that day's index rebound would still require further assessment.
The overseas decline provides context; causation is still missing
The Nasdaq page lists the PHLX Semiconductor Index at 12,236.79 points on 10/12/2026, down 2.67% for the day. This record can provide context for observing the overseas semiconductor market, but it covers a specific index and cannot be extended into a claim of a broad correction in global technology stocks. The WeChat account links the overseas correction to the market reassessing the sustainability of AI capital expenditure; that is the account's interpretation. The index decline itself does not reveal why capital expenditure expectations changed, nor does it prove that this drove the subsequent fluctuations in A-shares.
Price recovery and investment value require separate answers
Taken together, these quotation and turnover records lead me to conclude that we can currently confirm a price recovery that day and increased trading value, but cannot yet judge that the market has changed direction. The limits of this judgment are specific: index quotations contain no information on company orders, profits, capital expenditure, or valuations. They therefore cannot tell us whether AI hardware companies' fundamentals are stable, nor can a price correction alone establish that the risk-return relationship has improved. Only if subsequent evidence includes sustained price performance, corresponding capital-flow data, and company operating disclosures will there be a basis for linking market behavior to operating changes. That evidence could also overturn the initial expectation that the rebound may persist.
What to watch next
- Whether subsequent index quotations using the same statistical definitions show the rebound continuing or prices falling again; a one-day recovery is insufficient to establish persistence.
- Whether official disclosures of outstanding margin financing balances and ETF subscriptions and redemptions support the explanation that deleveraging and buying on dips occurred simultaneously, and whether their timing aligns with market movements.
- Whether cloud providers' capital expenditure and disclosures of orders and profits from relevant companies change, providing independent evidence for a link between the overseas semiconductor correction and A-share performance.
Sources and verification
Topic originated from 潘大的投研笔记, published on 2026-10-13 17:18 (UTC+8). This article was independently organized using the primary sources below. Read observations separately from interpretation.
← All research notes · Research sources and review method · Editorial standards