This morning, all major A-share indices were still in the green. The ChiNext Index fell more than 3% during the session, the STAR 50 Index once dropped over 3.5% and lost the 1,400-point mark, while the Shenzhen Component Index and the Shanghai Composite Index also declined, pushing market sentiment to a freezing point.

Unexpectedly, the situation changed dramatically in the afternoon. The Shanghai Composite Index was the first to turn positive, and then the Shenzhen Component Index, ChiNext Index, and STAR 50 Index all recovered their losses. Although they pulled back again near the close, the full-day declines narrowed significantly, with the Shanghai Composite Index and STAR 50 Index closing in the green, while other indices only edged slightly lower.

From a panic-driven selloff in the morning to a V-shaped reversal in the afternoon, the dramatic shift was completed within half a day.

Where did the money come from?

When the market turned positive in the afternoon, statistics had already been released showing that the turnover of several broad-based ETFs today far exceeded their full-day levels yesterday. Among them, E Fund ChiNext ETF exceeded 9.2 billion yuan, China AMC STAR 50 ETF exceeded 8.1 billion yuan, Southern A500 ETF exceeded 4.1 billion yuan, Huatai-PineBridge CSI 300 ETF reached 4.6 billion yuan, and Southern CSI 1000 ETF exceeded 4 billion yuan.

This was not achieved in a single day. Wind data shows that stock ETFs have recorded net inflows for the sixth consecutive trading day. Last Friday, single-day net inflows reached 22.782 billion yuan, yesterday's net inflows were 4.338 billion yuan, and cumulative net inflows since September 28 have reached as high as 81.735 billion yuan.

Broad-based ETFs were the absolute main force, with net inflows of 3.847 billion yuan yesterday. E Fund ChiNext ETF, Huatai-PineBridge CSI 300 ETF, and Southern CSI 1000 ETF ranked in the top three by net inflow.

Judging only from publicly available data, broad-based ETFs were indeed the most direct capital driver behind today's afternoon reversal.

Now let's look at another source of funds — quantitative funds. In today's afternoon rally, the footprints of quantitative capital were very obvious. Looking carefully at the trigger points, they were actually quite concentrated: aside from some stocks that accelerated to limit-up at the open, many targets suddenly surged at roughly the same time in the morning.

This is most likely because quantitative models captured a combination signal of "earnings catalyst + low open," and programmatic buying followed immediately. More quantitative models then captured the signal of capital entering the market, and chase-buying orders were triggered in batches.

Once such a signal is simultaneously captured by multiple models, buying pressure is released in a concentrated manner within a short period, forming a pulse-like rally. Therefore, today's volume-driven rise may not necessarily mean institutions were systematically replenishing positions, but more likely that trading-oriented capital was using the information window to play a rebound.

Previously, the PCB sector, along with technology growth stocks, experienced a round of sharp declines. After sufficient chip turnover, selling pressure itself was weakening. At this time, positive earnings landed and stocks opened low but moved higher, making it very easy to trigger the buy conditions of quantitative models.

Deep declines, sufficient chip turnover, and earnings support — when these three conditions come together, it is not hard to understand why quantitative capital entered the market.

In addition, there may have been structural position adjustment by main funds. In today's morning session, main funds recorded a net outflow of more than 6.5 billion yuan from the electronics sector. But by the afternoon, the electronics sector as a whole turned to a net inflow of 2.916 billion yuan, with printed circuit boards (PCB) recording a net inflow of 2.188 billion yuan, integrated circuit manufacturing 1.788 billion yuan, and semiconductors 1.197 billion yuan.

This shows that main funds were not indiscriminately selling technology stocks, but were intensely rebalancing within the sector — shifting from directions with high valuations and a short-term lack of earnings support to sub-sectors with clear third-quarter earnings expectations.

As for the fourth category — foreign capital — there is no data support yet, so it can only be treated as inference for reference only. Bank of America strategists recently issued a report pointing out that global active long-only funds have adjusted their average allocation weight to Chinese stocks from "underweight" to "benchmark neutral" since June, ending a four-year underweight status.

Why buy at this time?

Many people may wonder: with market sentiment so low and macroeconomic pressure showing no obvious signs of fading, why did funds choose to enter and drive a rally at this time?

If viewed from a contrarian investment perspective, several characteristics of the current market become apparent:

At the valuation level, the overall A-share market currently shows a typical pattern of high PE and low PB. The Wind All A price-to-earnings ratio is 20.33 times, at the 76.6th percentile historically, but the price-to-book ratio is only 1.71 times, at the 39.8th percentile. The CSI 300 PE percentile has fallen to 25.9%, and the ChiNext fully dynamic PE ratio is only at the 18th historical percentile.

A Shenwan research report shows that the ChiNext Index PE is 34.8 times, at the 19th historical percentile; the CSI Dividend PE is only 8.7 times and PB only 0.9 times, at the 30th historical percentile.

The margin of safety for equity assets is gradually improving, and the downside space for the index may be limited. This partly explains why funds dared to absorb selling during the decline.

At the earnings level, the wave of positive third-quarter earnings pre-announcements provides a tool for separating the genuine from the false. As of October 12, more than 80 companies had disclosed third-quarter earnings forecasts, with a positive pre-announcement rate of about 77.6%. Twenty-one companies had upper limits of net profit growth doubling, Lianxun Instruments expected growth of more than 20 times, and Yuanjie Technology expected growth of more than 9 times.

Today's limit-up PCB stocks such as Ultrasonic Electronics and Xunjiexing were direct beneficiaries of earnings exceeding expectations. Huazheng New Material is expected to see net profit attributable to shareholders grow by as much as 2,658% year-on-year, Nanya New Material is expected to grow 605%, and Shengyi Technology is expected to grow 175%.

In the PCB manufacturing segment, Jingwang Electronics expects third-quarter single-quarter net profit to grow 205% to 265% year-on-year, with single-quarter profit already exceeding the total for the first half of the year.

The main source of earnings growth is very consistent: accelerated mass production and shipment of high-performance PCBs in high-speed communications and AI data infrastructure, combined with reasonable cost-price transmission in general-purpose products. Many companies have full order books in AI servers, data centers, and high-speed optical modules, with capacity utilization and yield rates at relatively high levels.

CITIC Securities expects that in the third quarter of 2026, the electronics industry's overall performance will maintain high prosperity under the AI trend, with PCB/CCL being one of the best-performing sub-directions. When earnings are well delivered, there is an opportunity to attract capital inflows. Moreover, this earnings-driven logic is more solid than pure thematic speculation.

Finally, there is policy and macroeconomics. On October 10, CSRC Chairman Wu Qing held a symposium to listen to market opinions. Among the recommendations from participating experts, "market stabilization mechanisms" were listed first. Compared with the July symposium, the wording upgraded from "strengthen" to "continuously strengthen," changing from an action to a mechanism.

At the same time, the 2026 Financial Street Forum Annual Meeting will be held from October 19 to 22, with Pan Gongsheng, Wu Qing, and others attending and delivering keynote speeches, focusing on "building first-class investment institutions to serve high-quality development of the capital market."

On the macroeconomic side, U.S. Treasury yields are already at historical highs and some turning-point signals have appeared. Rate hikes have also moved from expectation to fact. Although this cannot be taken as conclusive evidence of a reversal, the absence of marginal deterioration is itself enough to ease tense market sentiment and provide the possibility for a technical rebound.

How should we view the market outlook?

Today's deep V-shaped reversal did indeed boost confidence, but investors still need to remain clear-headed, especially given the late-session spike and pullback today. If macroeconomic pressure is not resolved and earnings and valuations are not recalibrated, the sustainability of the rebound remains doubtful.

Although broad-based ETFs and quantitative funds are relatively active, in essence, one is supporting the market rather than driving a rally, and the other can enter at low levels today and cash out tomorrow. To truly support sustainability, more types of capital and larger-scale capital need to enter the market.

In the short term, the AI track, which previously fell deeply, underwent sufficient chip turnover, and has earnings support, does indeed have room for a repair rally, and its elasticity will likely remain. As long as the earnings disclosure period has not ended, similar "earnings + low open" combination signals will repeatedly appear, and programmatic buying will also be repeatedly triggered.

But funds that took positions at previous highs have not yet broken even. Each rebound to resistance levels will encounter selling pressure from investors trying to get out. Without more capital entering, relying only on passive funds and quantitative funds will make it difficult to support a reversal行情.

To see more certain signals, it is necessary to focus on third-quarter earnings data. On the one hand, closely watch overseas technology giants. This month it is TSMC, Broadcom, Microsoft, Google, Amazon, Apple, and so on; next month it is Nvidia; the month after that it is Micron. Their statements on next-quarter capital expenditure and order visibility are especially the biggest judgment signals.

If the guidance from U.S. technology giants continues to exceed expectations, risk appetite for global AI hardware can remain stable, and institutions' willingness to re-enter the market will also increase. Conversely, if overseas guidance falls short of expectations, the upside of A-share rebound may be significantly constrained.

On the other hand, there is A-share companies' own financial report data. For funds that have already entered early to bet on earnings, good data may lead them to cash out, while bad data would make them even more determined to leave. In other words, the market will continue to fluctuate repeatedly with high volatility.

Finally, more attention should be paid to A-share turnover. The current daily average is around 1.5 trillion yuan, and today it expanded to 1.68 trillion yuan, but it is still far behind the 2 trillion or 3 trillion yuan scale seen at times in the first half of the year. If turnover cannot make a big leap upward during the rebound, it indicates that incremental capital's willingness to enter is not strong, and the rebound height will be limited.

(End of article)

Previous review

Gelonghui statement: The views in this article are all from the original author and do not represent the views and positions of Gelonghui. Special reminder: Investment decisions must be based on independent thinking. The content of this article is for reference only and is not intended as any practical operational advice. Trading risks are borne by the reader.