In a previous analysis, we examined that high-growth industries on the A-share market often exhibit an "M-top" pattern at their price peaks, and the current tech rally may follow a similar trajectory. Since late June, AI-related tech sectors such as electronics and communication have experienced significant corrections, with recent signs of stabilization. This has sparked growing investor discussions about the timing of a second wave for the tech sector. This article explores the historical patterns of the second wave in high-growth industries' M-tops and the key factors to watch for the next opportunity in AI tech stocks.
Stock prices in high-growth industries typically peak 1 to 1.5 years ahead of fundamental performance, often forming an M-shaped top. In our earlier report, "Lessons from Six Major High-Growth Industry Rallies Over Two Decades: When Will the Current AI Rally Peak?" we analyzed six cases over the past 20 years, finding that the stock price peak preceded the ROE peak by an average of about 13 months. This occurs because the market front-runs orders, prices, and future profits during the upswing; once expectations are linearly extrapolated to high levels, stock prices become desensitized to positive news, even as earnings continue to grow. Consequently, by the time fundamental data officially confirms a peak, stock prices have already completed a major adjustment.
The M-top pattern in high-growth industries is consistent, with the secondary peak offering strategic positioning opportunities. Historical data shows that after reaching an absolute price peak, these six industries underwent notable corrections, accompanied by a contraction in valuations and crowded trades. However, after a pullback, the industries still delivered strong earnings, market sentiment turned optimistic again, and stock prices rebounded sharply, forming the secondary peak of the M-top. In terms of upside, the average gain from the M-top trough to the secondary peak across these six industries was 54.3%, with a median of 60.5%. The ratio of the secondary peak to the absolute peak averaged 86.9%, with a median of 87.1%, underscoring the strategic value of the secondary peak. The two peaks of the M-top correspond to two distinct market valuations of the same industry cycle: the absolute peak reflects overvaluation due to concentrated pricing of high-growth expectations, while the secondary peak validates whether the industry trend can be sustained after the correction.
What causes the transition from a boom-driven cycle to a contraction in expectations at the M-top's absolute peak? Before high-growth industries can offer a re-entry opportunity and move toward the secondary peak, specific conditions need to be met. The absolute peak forms primarily due to adverse industry developments that compress long-term expectations, coupled with extreme trading conditions. During the main rally phase of a high-growth industry, the market linearly extrapolates strong order growth and profit releases into the future, driving up valuations and crowding. At the absolute peak, the price-to-book ratio of these six industries reached extreme levels within a five-year period, and the share of turnover and turnover rate quantiles were in the 95-100% range, indicating high heat. As industry disruptions cause market consensus on long-term prospects to diverge, stock prices become desensitized to positive news and enter a distinct correction, with an average decline of 41.5% and adjustment periods ranging from 1-2 months to 7-8 months. For example, in 2015, some operators' 4G capital expenditure marginally declined; in 2022, a sharp rise in lithium carbonate and polysilicon prices squeezed midstream and downstream profits, pressuring profitability in the photovoltaic module sector and fueling concerns about cost increases in the photovoltaic and lithium battery supply chains dampening demand. When crowding is high and long-term earnings expectations diverge, stock prices begin to desensitize to positive news and undergo significant adjustments.
After the absolute peak of the M-top, the average decline across these six industries was 41.5%, with a median decline of 46.6%. The average adjustment period was 4.6 months, with a median of 4.1 months. If the absolute peak coincided with a bull-bear market transition, the correction was deeper and longer. Among the six cases analyzed, four instances—communication equipment in 2015, photovoltaic equipment in 2021-2022, batteries, and energy metals—saw their M-tops align with a bull-bear shift. Specifically, in mid-2015, the A-share market entered a bear phase due to deleveraging, coinciding with the peak in communication equipment. In 2022, the Russia-Ukraine conflict and pandemic disruptions triggered a bull-bear transition in the A-share market, with the battery, energy metals, and photovoltaic equipment sectors entering their M-top zones. The average post-peak decline for these four industries was 50.6%, with an adjustment period of 6 months, significantly higher than the 23.2% decline and 1.8-month adjustment seen in the consumer electronics sector in 2010 and the semiconductor sector in 2021. Clearly, during a bull-bear transition, the price correction and valuation compression in high-growth industries are amplified by declining risk appetite.
How do we move toward the secondary peak? Valuations and trading metrics need to return to reasonable levels, and industry concerns must be alleviated or new catalysts emerge. The secondary peak in the M-top still offers substantial gains relative to the trough, and this rally is not merely a rebound from oversold conditions but a re-pricing of earnings trends, industry catalysts, and improved risk constraints after valuations and crowding have normalized. Valuations and crowding in high-growth industries need to return to reasonable levels. After the absolute peak of the M-top, valuations decline alongside stock prices, with a deeper percentage drop, reflecting a significant correction in overly optimistic expectations. Across the six industries, the average decline in PE was 52%, and in PB was 50%, both exceeding the average stock price decline of 41%. After this sharp pullback, the five-year quantile of PE and PB fell to 40-60% and 50-70%, respectively, indicating moderate-to-high levels. Trading heat also cooled, with the share of turnover and turnover rate declining by an average of nearly 40%.
The decisive factor for the transition from the M-top trough to the secondary peak is the elimination of industry concerns or the emergence of new demand or technological catalysts, leading to an upward revision of earnings expectations. Even after the absolute peak, a significant valuation compression alone is not sufficient to stop the decline and trigger an upward reversal. The market must shift from fearing unsustainability in high growth to believing that earnings growth can be sustained or accelerated. For instance, in early 2011, strong growth in smartphone shipments catalyzed consumer electronics, halting the decline and forming the secondary peak of the M-top in March 2011. In 2022, strong overseas demand for photovoltaic products and price increases facilitated cost pass-through for polysilicon, while market expectations for polysilicon capacity release improved, driving a rally in the photovoltaic sector. During the same period, the acceleration of domestic production resumption led to new energy vehicle sales growth rates exceeding 100% from May to July 2022, driving the battery and energy metals sectors to a secondary peak in July 2022. If the M-top coincides with a bull-bear transition, the secondary peak may only appear after the macro and micro shocks of the bear market subside. For example, in 2015, deleveraging triggered a negative feedback loop in A-share micro-funding, leading to a bull-bear transition. The communication equipment sector only stabilized after the proportion of margin buying in total A-share turnover fell to a relatively healthy level of around 8% by the end of 2015. In 2022, after the Russia-Ukraine conflict, the Federal Reserve tightened monetary policy, and domestic fundamentals were impacted by the pandemic, pushing the A-share market into a bear phase. Subsequently, in April 2022, pandemic disruptions gradually eased, and domestic policy support helped stabilize the stock market, with high-growth industries like photovoltaic and lithium batteries showing greater fundamental flexibility and moving toward the secondary peak of the M-top.
AI computing hardware may already be in the M-top zone, with conditions for a second wave gradually accumulating. Since June, market concerns about the sustainability of CSP capital expenditure have intensified, leading to a significant correction in global AI tech hardware stocks, including A-share communication equipment and semiconductor sectors. If historical patterns of high-growth industry rallies are any guide, AI computing hardware may have entered the top zone, with the left peak of the M-top potentially already formed. The key question now is whether conditions for the right peak are in place. First, the correction in AI computing hardware has been significant, but trading indicators have not yet pulled back substantially. Since late June, the communication equipment sector has seen a maximum drawdown of 45%, and the semiconductor sector 42%, indicating a notable correction in overly optimistic expectations. However, valuation and trading heat indicators remain at historically extreme levels. As of August 7, 2026, the PE, PB, share of turnover, and turnover rate for communication equipment and semiconductors remain in the 90th percentile or above within a five-year period. Second, corporate earnings reports from CSPs have slightly eased concerns about the sustainability of capital expenditure. The latest quarterly reports show that capital expenditure by North American CSPs continues to grow strongly, with the four major CSPs' capital expenditure in the second quarter of 2026 up 87% year-on-year, and full-year 2026 capital expenditure guidance raised to over $730 billion. At the same time, CSP cloud business revenue has grown significantly, with Google, Amazon, and Microsoft's cloud business revenue growth in the second quarter of 2026 reaching 38.5%, up from 33.6% in the first quarter. Data on remaining performance obligations disclosed by cloud providers also indicates that customer demand for computing power still exceeds investment in AI infrastructure, leading to a slight stabilization in the AI computing supply chain both domestically and internationally. However, under high capital expenditure, CSP cash flow continues to deteriorate, and market concerns about the sustainability of cloud provider capital expenditure have not been fully resolved. Overall, the conditions for a second wave in AI computing hardware are still accumulating, and the market may currently be in a bottoming phase. Going forward, it will be important to monitor whether the high growth in AI revenue can be sustained and whether new narrative catalysts, such as the Agent explosion seen in early 2026, can emerge to drive a new round of expectations for the entire tech sector.
Following the recent disruption, the A-share bull market is expected to continue, with balanced allocation being the optimal strategy for the mid-to-late stages of the bull market. Since July, the A-share market has experienced a correction, primarily due to increased external disruptions, which is a normal adjustment in the mid-to-late stages of a bull market. Historically, a bull-to-bear transition in the A-share market has occurred when the overall market sentiment is overheated, liquidity conditions are persistently tightening, and the macro environment is clearly weakening—none of which are currently present. Looking ahead, the July Politburo meeting has clearly signaled an increase in counter-cyclical adjustment, which is expected to support improvements in macro and micro fundamentals. Additionally, external demand remains resilient, with exports in dollar terms growing by 23.9% year-on-year in July, with strong growth in semiconductors, high-end manufacturing, and other sectors. Inflation in the US continues to decline from its peak, and with political pressure from the midterm elections, the Federal Reserve may find it difficult to raise interest rates in September. Supported by these positive macro and micro factors, the A-share bull market is likely to deepen. In terms of structure, focus on the diffusion within the growth sector and opportunities in dividend-paying and domestic demand-related areas. First, within the tech sector, look for opportunities in lower-priced areas. As we analyzed earlier, with positive factors like the digestion of crowding and industry-level catalysts accumulating, the AI computing supply chain may still have upward potential. The tech sector is expected to see a diffusion toward lower-priced areas, with opportunities in AI application segments and Hong Kong-listed tech stocks that have fundamental expectations and positive industry catalysts. Second, with the entry of long-term capital and market volatility, the attractiveness of dividend-paying assets is increasing. Policy guidance for long-term capital entry, combined with the current market environment potentially still experiencing periodic volatility, makes dividend-paying assets with relative value appealing. Third, sectors with strong fundamentals, such as healthcare and securities, as well as consumer staples and real estate, which are supported by policy-driven expansion of domestic demand, are also worth watching.
Risk Warning: A significant increase in Federal Reserve tightening expectations or a slowdown in domestic economic recovery could pose risks.