Written by: Little Cake

On September 19, 2014, Alibaba listed on the New York Stock Exchange, closing on its first day at $93.89. On that day, Alibaba's market value was $231 billion, exceeding the combined total of Oracle and Intel.

On June 25, 2026, Alibaba closed at $95.07.

There is a full twelve years between the two numbers.

At the same time, Meituan closed at HK$65.45, slipping below the 2018 IPO issuance price of HK$69.

Pinduoduo hovered around $79, returning to June 2020 levels.

Tencent’s price-to-earnings ratio shrank to 12 times, almost halving from the ten-year historical average of 25.7 times.

As for those younger Chinese internet companies, Bilibili fell from a peak of $156 to $18, a 89% drop; Kuaishou fell from its IPO peak of HK$417 to HK$44, evaporating nearly 90% of its market value; iQIYI, Zhihu, Douyu, and Huya each experienced pullbacks of between 85% and 98%.

A whole generation of Chinese internet assets has undergone a collective valuation reset. What framework does the market use to price these companies? Or has the framework itself already died?

Establishing and Removing Anchors

Looking back, the valuation logic of Chinese internet has undergone an exceptionally clear process of "establishing and removing anchors."

From 2014 to 2017, the core narrative in global capital markets regarding Chinese internet was “U.S. benchmarks at a discount.”

Alibaba is China's Amazon, Tencent is China's Facebook plus China's Visa, and Baidu is China's Google.

This methodology is concise and powerful: first, find the valuation multiples of corresponding U.S. companies, then multiply by a growth premium for the Chinese market and a governance discount, yielding a reasonable price. Within this framework, Chinese internet companies generally enjoy a price-to-earnings ratio of 20 to 40 times.

Foreign capital flowed in, and Chinese concept stocks became a must-have asset; this was the first anchor point.

In 2018, the U.S.-China trade war erupted. For the first time, global capital was forced to contemplate a question it had previously deliberately avoided: if U.S.-China relations shifted from cooperation to competition, would the legal structures of companies operating in China and listed in the U.S. still be reliable? The VIE structure has never received clear recognition under Chinese law, but no one minded during the bull market. The trade war exposed this sore spot to the light of day for the first time. The valuation anchor began to loosen but had not yet been removed.

In October 2020, Ant Group's IPO was halted, and the pricing of “Chinese regulatory risk” in international capital markets changed from a vague discount factor to an explicit core variable. The antitrust storm of 2021 pushed this logic to extremes. Alibaba was fined 18.2 billion yuan, Didi was scrutinized the day after its IPO, and the education and training industry was wiped out overnight. Chinese concept stocks shifted from “growth premium” to “regulatory discount.”

By 2022, panic over Chinese concept stocks delisting reached its peak.

The SEC listed Alibaba, Baidu, JD.com, and hundreds of other Chinese concept stocks on the “pre-delisting list.” Although the U.S. and China ultimately reached a compromise on the auditing documentation issue, the damage was already done. Global index funds began systematically reducing their weights in Chinese concept stocks, and some institutional investors directly liquidated their positions due to compliance requirements. This structural exit of funding turned the compression of valuations from sentiment-driven to capital-driven.

At the beginning of 2025, the emergence of DeepSeek briefly sparked a wave of hope. Deutsche Bank referred to it as China’s “Sputnik moment," predicting that the valuation discount of Chinese assets would disappear.

In the first two months of 2025, Alibaba and Tencent's stock prices rebounded more than 60% at one point. However, this wave of revaluation driven by AI narratives fizzled out in less than half a year. Entering 2026, the Pentagon listed Alibaba and Tencent among "Chinese military-related enterprises," and Anthropic publicly accused Chinese companies of launching large-scale distillation attacks on its Claude model. Nasdaq introduced new listing rules for Chinese concept stocks, tightening liquidity thresholds. Every attempt to rebuild valuation anchors was swiftly undermined by new geopolitical shocks.

At this point, the “U.S. benchmark at a discount” valuation methodology had completely failed. The market no longer priced these companies based on their business models, growth rates, or profitability.

But it’s not that simple.

Old Tencent Stocks on Both Sides of the Pacific

Shifting the focus from Chinese concept stocks on the New York Stock Exchange to the U.S. tech giants trading in the same building reveals that: what the market has abandoned extends far beyond Chinese internet.

Microsoft in 2026 is the worst-performing stock among the "Magnificent Seven," with a decline of over 20% during the year, plummeting from a high of nearly $490 at the end of 2025 to around $360. Its price-to-earnings ratio shrank from a five-year median of 34 times to 22 times, the lowest level in three years.

The company’s fundamentals remain intact: Azure cloud revenue grew 39% year-on-year, AI business annualized revenue exceeded $37 billion, and quarterly net profit of $31.8 billion hit a historical high.

The market does not care about these numbers but is more concerned with another number: $190 billion, Microsoft’s capital expenditure budget for the entire year of 2026, almost entirely directed toward AI infrastructure. Single-quarter capital expenditure exceeded the total amount for the entire year five years ago. Free cash flow dropped from $20.3 billion to $15.8 billion, widening the gap between profit and cash.

Microsoft's plight is not an isolated case.

In 2026, all seven tech giants lagged behind the S&P 500. The combined capital expenditures of the four large-scale cloud providers (Amazon, Microsoft, Alphabet, Meta) nearly reached $700 billion this year. The GPU clusters and data centers bought with this money will not generate income until the future 3 to 5-year depreciation cycle, pushing investment forward while returns lag behind, and crushing free cash flow in between.

A deeper issue is: these companies are using massive capital to chase a technological paradigm that might disrupt their own business models.

Microsoft's core revenue comes from Office subscriptions and Windows licenses, which is a SaaS model charged by number of users, nearing growth limits. The business logic of the AI era is billed based on consumption, where you pay for the number of tokens used.

CEO Nadella has publicly acknowledged that every user-based charge business of Microsoft will transition to a “user + usage” mixed model. GitHub Copilot switched to completely usage-based pricing in June 2026, but the market's concern lies precisely in: the old model had very high margins, whether the new model can maintain the same level is unknown.

From a distance, this picture forms a structural mirror image with Alibaba and Tencent's predicament.

Alibaba's core e-commerce business is a highly profitable advertising machine, as stable as Microsoft's Office, yet the valuation multiple given by the market is steadily declining. Tencent’s WeChat ecosystem remains the strongest fortress in Chinese internet, yet game revenue growth is slowing, and the advertising business faces encroachment from short video platforms, echoing Microsoft's situation being squeezed by Alphabet in search advertising.

Both old giants are desperately investing in AI in an attempt to save themselves; Alibaba is pouring $55 billion into building AI infrastructure, while Microsoft invests $190 billion, but markets on both sides are casting doubt on "whether this money can be recouped."

Chinese internet practitioners tend to attribute their company's decline to regulatory suppression and geopolitical issues, while U.S. tech practitioners blame Microsoft's decline on "spending too aggressively." Peeling back the surface narrative reveals the same underlying issue: AI-native companies are redefining the entire value chain of the tech industry, while the previous generation of platform giants, regardless of nationality, is transforming from "companies defining the future" to "companies needing to prove they won't be eliminated by the future."

In the Chinese internet, these stocks have gained a precise nickname — old Tencent stocks.

Nikkei: A Lesson on the Death of Valuation Systems

This phenomenon of "the valuation coordinate system being replaced" is not unprecedented in the history of global capital markets. The closest parallel is Japan post-1989.

On December 29, 1989, the Nikkei 225 Index closed at 38,915 points, setting a historical record.

That year, eight of the ten largest companies globally by market value were Japanese. After its IPO in 1987, NTT’s stock price soared to 3 million yen per share within two months, surpassing the combined market value of the eight largest American companies at the time. Land prices in Tokyo were 350 times those in Manhattan. Sony acquired Columbia Pictures, and Mitsubishi purchased Rockefeller Center.

Investors in that era, like the Chinese internet workers in 2020, sincerely believed that their system would lead the global economy's future.

The trigger for the bubble's burst was the Bank of Japan raising interest rates. But the degree of the downturn was merely the shallowest characteristic of this crisis; the duration and nature of the decline were truly suffocating.

The Nikkei fell half in the first half of 1990 and had halved to 14,000 points by 1992. Had it ended there, it would have been just an ordinary bubble burst and valuation adjustment. However, the Nikkei did not stop there. It continued to decline for ten years, dropping to 7,600 points by 2003, an 80% retraction from its peak.

The core reason for this decade-long sustained decline was not the collapse of the competitiveness of Japanese companies.

Toyota remained the best car manufacturer in the world, and Sony continued to create groundbreaking consumer electronics. The issue lay at a deeper level: global capital no longer believed in “Japanese premium.”

Before 1989, the valuation framework for Japanese companies was "the most efficient manufacturing civilization in the world + perpetually growing domestic demand market + unique corporate governance advantages."

After the bubble burst, these three assumptions were successively denied. Manufacturing advantages were chased by South Korea and China; the domestic market fell into deflation and an aging population; and corporate governance was proven to be a breeding ground for shielding inefficiency. The old valuation framework died, but a new framework had yet to be established.

In 1989, 32 out of the top 50 companies in the world by market value were Japanese. By 2018, only Toyota remained.

How long did this period of vacuum last? About 25 years. The Nikkei only began a genuine trend recovery in 2012, returning to 38,915 points by February 2024. And the catalyst for this revaluation was not a comprehensive revival of the Japanese economy.

A specific individual redefined “why one should buy Japanese assets” with a new language.

In the summer of 2019, Buffett began to buy stocks of Japan's five major trading companies. The logic of this investment is completely different from the way the market has viewed Japan over the past thirty years. Buffett does not discuss GDP growth rates, population trends, or technological innovation. His reasoning is extremely simple: these five companies are undervalued, have high dividends, stable cash flow, and are advancing genuine corporate governance reforms. He uses yen-denominated bonds to hedge currency risks and his credibility to back Japanese assets. By 2025, Berkshire's stake in the five trading companies was nearly 10%.

Buffett provided a whole new valuation language for Japanese assets; the old language was “Japan will dominate the global economy”; the new language is “low valuation + high dividend + corporate governance reform.”

Where Is the "New Language" for Chinese Internet?

Aligning Japan's timeline with the Chinese internet's situation reveals several structural similarities that cannot be ignored.

The old valuation framework has died. The failure of the "U.S. benchmark at a discount" model resembles the collapse of the narrative that “Japan will dominate the globe.” In both cases, the fundamentals of the companies have not completely deteriorated; it is the macro assumptions supporting the valuation premium that have been negated. The macro assumption for Chinese internet is “the deep integration of the Chinese market with global capital markets will continue,” while Japan's macro assumption was “the Japanese model represents the most efficient form of capitalism,” both of which have been falsified.

A new valuation framework has yet to be established. Currently, the market's pricing of Chinese internet assets is essentially a discount on the ruins of the old framework. Just like in Japan in 1995, the market knows the old prices are wrong but does not know what the new price should be.

From Japan's experience, this vacuum period may last much longer than most people expect. Japan took about 25 years from the bubble burst to the new valuation framework becoming accepted by the market. The valuation system of Chinese internet began to systematically collapse in 2020; now it has only been six years. If Japan's timeframe holds any reference value, the current position may just be the beginning of the revaluation process.

However, key differences exist between China and Japan. Japan's asset revaluation was accompanied by long-term deflation and population decline, resulting in a substantial deterioration in corporate profitability after the bubble burst. The leading companies in Chinese internet are still profitable, with Tencent's annual net profit exceeding 220 billion yuan, and Alibaba's core e-commerce business maintaining robust cash flow. This means that if a new valuation language can be constructed, the speed of revaluation may surpass that of Japan.

What might become the "new valuation language" for Chinese internet?

AI is the most noticeable option, but also the most contradictory.

For the past twenty years, the underlying business models of global internet companies have largely converged: competing for users' attention, aggregating traffic onto platforms, and then monetizing through advertising, e-commerce commissions, or in-game purchases.

AI is shaking the foundation of this business.

When AI agents can compare prices, place orders, and plan itineraries on behalf of users, users no longer need to open pages on Taobao one by one. When AI can directly recommend content or even generate content based on preferences, the time users spend "browsing" on a single platform will shrink. Attention shifts from human eyes to AI agent interfaces, altering the points of entry for traffic and undermining the strategic position of platforms as intermediaries. This poses a threat to nearly every core internet sector, including e-commerce, search, social media, content, and gaming.

If any Chinese internet company can be the first to complete the transition from "attention platform" to "AI infrastructure and service provider," it may gain a whole new valuation language.

The brutal reality of this path is that actively overturning means intentionally dismantling the most profitable old businesses.

Taobao’s advertising revenue is built on merchant bidding rankings; if AI agents bypass the rankings and directly help users select products, that revenue will dwindle. Every step of the transformation erodes existing profits, while the profitability of the new model has yet to be validated.

If you pursue AI, you have to endure the immense capital expenditure pressing down on free cash flow. Microsoft's PE ratio dropped from 34 times to 22 times; that is the conclusion of this story; if you do not pursue AI, the market will deem you as being abandoned by the times.

Microsoft wagered $190 billion on a rewrite of its revenue structure; winning is the infrastructure of a new era, losing is the largest capital misallocation in history.

Shareholder returns are the second candidate option. Both Tencent and Alibaba are conducting large-scale buybacks; Tencent's dividend yield has risen to 1.25%. This is highly similar to the logic Buffett used to price the Japanese trading companies: since the market is unwilling to pay for growth, use actual cash buybacks and dividends to build a valuation floor. However, the current buyback efforts are limited relative to the market value decline and are still insufficient to become an independent pricing anchor.

The current plight of Chinese internet assets is highly similar to that of Japanese assets around 1995: the old framework has died, the new framework has yet to be born, and the market is waiting in a vacuum for someone or something to redefine "why one should buy."

The current position may very well just be the mid-section of this long revaluation.

This article only represents the analytical perspective of潮向研究 and does not constitute any investment advice. The individual stock analysis mentioned is based on public information, and investors should make independent judgments and bear their own risks.