主題
Recently, China has been continuously hyping up AI-related topics, including the so-called company DeepSeek, which has also received a lot of attention and hype. At the same time, the stock prices of related sectors in China have been steadily rising. As a result, some people have concluded that the Chinese economy is improving. Does this argument hold water? Of course not, because the fundamentals of the Chinese economy are currently completely disconnected from the Chinese stock market.
The Chinese stock market cannot reflect the fundamentals of the Chinese economy. Therefore, equating the short-term rise of a certain theme in China with the economy transitioning from decline to recovery is a completely incorrect way of thinking. It’s also a very typical "citizen" mindset, with focusing only on the surface without looking at the substance. However, what I find more interesting is this: Why has China chosen this particular moment to hype up AI? What’s the purpose behind it, and how does it work? I think these are the questions truly worth our attention.
Let’s approach today’s topic from this perspective. First, let’s examine the impact of the AI bubble on the Chinese stock market through recent financial data. Some time ago, the stock market in China’s Hong Kong free trade zone experienced an unusual surge, with mainland stocks (primarily tech companies) driving the momentum. The Hang Seng Tech Index rose from less than 5,000 points a month ago to 5,900 points. During the same period, China’s mainland A-shares also saw a corresponding increase. According to a report from the South China Morning Post, this upward trend indicates that global funds are undergoing a rebalancing process.
China’s market is regaining favor, with capital shifting from the United States and India toward China. In this wave of AI investment frenzy in China, the most critical factor is undoubtedly the development of DeepSeek. It has recently garnered attention because it allegedly developed an AI product comparable to those in the U.S. at an extremely low cost. Currently, all Chinese tech companies’ stocks are capitalizing on the hype surrounding this topic to reap its dividends. However, at the same time, DeepSeek has sparked some controversy, primarily related to accusations of copying or imitating similar American products. Western media claim that DeepSeek’s development costs are less than one-tenth of its competitors’, with the fundamental reason being its reliance on hijacking OpenAI’s model data to develop its own technology. Trump’s advisory team reportedly shares a similar view, indicating that this isn’t just the perspective of American media but also an official stance. Of course, I have no intention here of diving into a deep technical discussion about whether DeepSeek is genuinely involved in copying or imitation.
In fact, this is how I see the issue: whether it’s copying and imitation or an above-board technology transfer, it all boils down to one core issue: the global diffusion of technology. This trend of technological diffusion is, in reality, an inevitable feature of modern human society. The world we live in today is essentially a diffused version of the modern Western world. We’re not living in a multicultural mosaic of diverse ethnicities but rather in a singular, unified culture. Here, I’d like to reference a famous historian and Stanford University classics professor, Ian Morris. In his book Why the West Rules. For Now, he presents a theory based on his research team’s detailed examination of data from the development of various civilizations throughout history. They found that no hunter-gatherer society or agrarian village society could leapfrog the imperial stage and achieve modernization unless influenced by another society that had already passed through that phase. Similarly, no society could jump directly from a nomadic steppe phase into an industrial society unless conquered by an agricultural empire or an imperial state.
By 1900, Europeans and their overseas colonizers had effectively unified nearly the entire world into a single economic entity. The process of globalization has been unfolding for centuries. Once a society crosses the threshold of fossil fuel usage, the entire world inevitably moves toward modernization. The passage above addresses the inevitable technological and economic diffusion in modern human society. So, what does this have to do with the AI bubble we’re discussing today? In fact, in the early days of this channel, I often used a phrase: "Go with the flow." The "flow" here refers to an objective law that operates independently of human will.
People in finance love to capitalize on this kind of "flow" to make their moves. Specifically, technological diffusion is inherently a global megatrend. As I just quoted from historian Ian Morris, after human history entered the 20th century, the technology and economy of the West were destined to spread globally, it’s like an irreversible fungal infection. Technological diffusion naturally spreads from the center to the sides. Only the central regions of the global capitalist system have the potential to develop truly original technology. In this system, the United States is a typical "center" country, while China is a typical "sides" country.
So-called AI technology is a quintessential invention of modern Western civilization, and its diffusion outward to surrounding regions is an inevitable trend. Whether you’re the core financial clique on Wall Street or the comprador financial faction in China, they’re all leveraging this megatrend of AI technological diffusion for speculation, specifically, what’s known as stock market theme speculation. In the process of AI technology spreading from the center to the sides, these financial cliques exploit it for hype, claiming that a side region like China can "corner overtaking" and surpass the central region. This kind of hype serves a purpose: it triggers international capital flows. In the short term, this speculation can drive capital from the center to the sides, which, to some extent, boosts the asset prices in those side regions. As long as capital flows shift and asset prices fluctuate, the financial clique’s goals are achieved. Their aim is nothing more than to "harvest the citizen" or offload the Chinese assets they’re holding.
This, then, is the big picture of China’s AI bubble. That’s why I keep emphasizing that the key issue lies in the capital interests behind it and how power operates, these are the parts worth understanding. So far, I’ve laid out the background and the broader trends for you. Next, let’s look at some specifics: How exactly does this AI bubble in China operate? What’s its relationship with the Chinese economy itself, and with the people orchestrating this bubble? The major premise for discussing these questions is that China’s markets, its real economy and its financial sector, are completely disconnected, as I’ve mentioned before. Under this disconnect, China’s real economy remains in terrible shape, which leads to a critical issue: an asset shortage. There simply aren’t enough good assets to speculate on.
When an asset shortage occurs, a predictable phenomenon emerges: deliberate and even malicious speculation. China’s current AI bubble falls squarely into this category. So, let’s move on to discuss China’s asset shortage and the endgame of this AI bubble. As for the so-called AI theme in China, its stimulating effect on the stock market is undeniably massive.
Within this wave of stimulation, a key target is the company we just mentioned, DeepSeek. It has become an iconic tool for speculation. A certain financial institution in mainland China created a table of AI concept stocks, and this chart very intuitively demonstrates the chain reaction DeepSeek has triggered in the stock market. As shown in the chart, on January 20, 2025, after DeepSeek released its next-generation reasoning model, it took just about 10 days for the entire Chinese stock market to kick off a frenzy of AI concept speculation, followed by a surge in prices. Many companies have been using storytelling tactics, trying to align their businesses as closely as possible with DeepSeek’s industry. Through the methods of narrative economics, they’ve driven up their stock prices.
According to statistics, 15 major sub-industries and over 150 companies have engaged in conceptual speculation around DeepSeek, directly benefiting from this bubble. The chart clearly illustrates this, showing industries like computing power, servers, smart driving, information security, large-scale models, and even advertising e-commerce, healthcare, and media entertainment, all finding ways to tie themselves to DeepSeek’s upstream or downstream ecosystem, thereby boosting their asset prices. In the two weeks following early February, their returns generally ranged from 24% to 42%. These figures alone highlight just how massive this AI bubble in China has become, relying on a single concept to push up the stock prices of one or two hundred companies, a number that will likely continue to grow.
What’s happening in the Chinese stock market right now reminds me of the various thematic speculations we’ve seen in China’s stock market over the past few years. These speculations always start with a lot of fanfare in the short term but struggle to fundamentally reverse the trajectory of the Chinese stock market. For example, in recent years, we’ve seen hype around Chinese white wine (with brands like Moutai), the military-industrial sector (especially aerospace), semiconductors, and so-called new energy industries, including China’s new energy vehicles. Last year, I even did a special episode about China’s electric vehicle bubble. And now, it’s the turn of AI concept speculation. This time, compared to the past, I don’t think there’s any fundamental difference, it’s like a gust of wind blowing through, unable to stir up much of a wave.
Why no waves? Because the Chinese stock market, and even the Chinese economy as a whole, is like a stagnant pool of water. Yet, we still need to ask: Why is it like this? Why is it that despite China switching between all sorts of themes over the years, the stock market and the economy just can’t be ignited? This is because no matter how much speculation takes place, the most fundamental problems of the Chinese economy cannot be resolved. I’ll break down these fundamental issues into three aspects.
First, confidence in the Chinese economy has completely collapsed. This is directly reflected in the extremely sluggish consumer confidence index. At the same time, China’s broad money supply has been continuously declining year-over-year, and foreign investment has plummeted almost off a cliff.
Second, the Chinese economy is experiencing deflation. A key indicator of deflation is the continuous decline in China’s GDP deflator. As early as 2024, Bloomberg News reported on this issue in English, predicting that this deflationary situation might persist into 2025. The GDP deflator is essentially the ratio between nominal GDP and real GDP, reflecting the true direction of price levels. In real-world terms, this deflation manifests as sluggish consumer demand and a massive buildup of unsold inventory for businesses, forcing them to sell at discounted prices. In China right now, aside from daily necessities with inelastic pricing, many goods are being sold at a discount. The real reason behind this is deflation, or to put it more bluntly, an economic depression.
Third, China’s housing market collapse and stock market slump. Housing market has historically been China’s most valuable industry, but times have changed dramatically. Looking at the comprehensive index of Chinese housing market development, it’s clear that it has hit its lowest point in the past decade. While housing market collapses, the investment returns in China’s stock market are also highly unsatisfactory. Especially after the trade decoupling between China and the U.S., Chinese company stocks have almost been regarded as toxic assets, heavily dumped by foreign investors. The simultaneous decline of both the stock and housing markets signifies that China’s financial investment landscape is lackluster.
From these three aspects, it’s undeniable that the current crux of China’s economic problems cannot be addressed through AI. So, let’s look at it: the collapse of housing market, the breakdown of consumer confidence, poor stock market investment returns, widespread unemployment, worsening deflation, these represent the crashing of China’s housing and stock markets. In all these areas, so-called AI is fundamentally incapable of helping. Under such baseline conditions, China faces an asset shortage. Looking across this vast country, there simply aren’t any good assets to be found, with only poor-quality ones remain in sight.
Since high-quality assets can’t be found, the only option is to rely on propaganda and deception to forcibly inflate "premium asset." In this round, the premium asset being hyped up is DeepSeek. As mentioned in an earlier section, China’s comprador financial clique is driven by profit motives and an instinct for gain, engaging in financial speculation around AI topics. But it takes two hands to clap, the financial clique needs someone to collaborate with, to echo their efforts, in order to successfully pull off the speculation.The ones teaming up with China’s financial clique are Chinese companies, including both state-owned enterprises and nominally private firms. A key reason for this is profit, which remains a constant pressure on businesses. With China currently experiencing deflation and sluggish market demand, if companies continue to invest and produce as they normally would, there’s a high likelihood their products won’t sell, or if they do, it’ll be at a loss. So how can companies make a profit? At times like this, they have to get creative. Instead of channeling funds into production, they redirect them into non-productive areas, such as the financial sector. This is reflected in the data as a significant increase in companies’ non-operational investments. A large portion of these investments is likely tied to AI-related financial assets, which have a higher chance of rising in price. This allows companies to generate profits amid the harsh winter of the real economy.In other words, the pressure to make profits drives companies to team up with the financial clique to engage in speculation. The most direct consequence of this speculative feast is a societal misallocation of capital. A huge amount of corporate funds is no longer being used for genuine productive investment but is instead funneled into non-operational speculation. If this continues, it will lead to the hollowing out of the Chinese economy.
The degree of bubble formation will intensify, and the Chinese economy will die faster and more thoroughly. In the end, no matter how aggressively financial institutions and companies hype things up, China’s bubble won’t last long, it will inevitably burst. So when will it burst? I can point to a specific trigger: as soon as the United States takes countermeasures, that’s when the countdown to the bubble’s collapse begins. In fact, if we look at recent reports from Bloomberg News in the U.S., they’ve already highlighted this issue. The decoupling between China and the U.S. is bound to accelerate, and the Trump administration’s restrictions on China in the AI sector, such as technology limits and trade embargoes, will only increase in the future.
China’s AI technology might sound impressive in its propaganda, but in reality, it still relies on American chips for hardware. If the U.S. imposes an embargo in this area, China’s AI efforts will completely collapse. On the software side, China also depends on American AI prototypes for simulation and training. If the U.S. uses technical measures to enforce a blockade here as well, China’s AI development will face a cliff-like stagnation. As discussed earlier, technology spreads from the center to the sides. But if the center cuts off its connection with the sides, the related tech industries in the side regions will turn into a stagnant pool. At that point, both capital and technology will flow back from the periphery to the center.
This is the mechanism by which China’s AI bubble will burst. When will this happen? My answer is: it could happen at any moment. This year, 2025, is an especially sensitive year. If we take a quick look back at the history of stock markets, we’ll notice a pattern of “chaos every decade.” Roughly every 10 years, with a possible one-year margin of error, there’s a highly sensitive trend. At these moments, based on the historical patterns of China’s and Hong Kong’s stock markets, a stock market crash becomes almost inevitable every decade or so.
Because the United States could take action at any moment, imposing sanctions and decoupling from China in the areas of semiconductors and AI technology, the U.S. national security agencies will undoubtedly crack down hard. They’ll strictly investigate and punish smuggling, reselling, and illegal technology transfers related to AI, among other activities. The U.S. is moving quickly. Overall, this current AI bubble will likely come fast and go just as fast, much like China’s previous bubbles, leaving no substantial change behind. The Chinese economy will continue sliding along its predetermined track.
As mentioned before, there’s an unchanging rule in the Chinese stock market: "fast bull, slow bear." The bull market doesn’t last more than a quarter, while the bear market drags on for a decade. That’s roughly the rhythm. A stock market with this kind of "fast bull, slow bear" pattern is clearly not a normal market, it’s a scam market. It’s not much different from the fraud rings in Myanmar. But then again, Chinese stock investors, and even Chinese people as a whole, are incredibly easy to deceive. That’s because the collective memory of Chinese people is extraordinarily short. They quickly forget even the most horrific events that happened just three months ago, events that might have caused significant harm to their personal interests. They wipe it all clean from their minds. People like this deserve to be scammed.
Reference:
