I view market index adjustments as opportunities. When prices were rising, many lamented missing the boat; now that a correction is here, doesn't it offer a chance to board?

Since 2024, I have consistently emphasized that adjustments are beneficial—they test your conviction in a bull market. Those who believed in the bull market from the start of 2024 have likely profited handsomely. I am no different; my steadfast belief in the bull run has consistently yielded returns.

During downturns, the most stable focus areas are the CSI 300 and CSI A500 indices. This bull market is projected to be lengthy, potentially surpassing the 2007 peak of 6000 points, especially since the index is already above 4000 yet many companies remain undervalued.

Let's discuss artificial intelligence today, a sector that has recently surged, prompting many questions from followers.

We first mentioned the ChiNext AI ETF (159363) and the STAR AI ETF (589520) back on July 16th, in a post titled, 'Wow, a new high!':

We have reiterated their potential since then, even introducing a new thematic concept called the 'AI Twin Stars - The New Bull Market Leaders':

The relevant articles are all here for verification: 'New High!', 'Tech Takes Off, Gaining 10% in a Week!', 'The Bull Market Isn't Over!'. I wonder if any followers profited—feel free to comment below.

Many are puzzled by AI's significant gains, so let me clarify the underlying logic.

The investment theses behind these two ETFs are fundamentally distinct.

First, consider the ChiNext AI ETF (159363; OTC: 023408), which operates on a "picks and shovels" logic.

This ETF currently boasts a scale of 5.3 billion yuan, a historical high, leading its peers in both size and liquidity. Over 70% of its portfolio is concentrated in computing hardware, particularly optical modules. Essentially, regardless of which AI company ultimately wins, the suppliers of these essential components—the "shovel sellers"—are poised to profit.

Why? Because whether it's NVIDIA's GPUs or Google's TPUs, building hyperscale computing centers necessitates optical modules to overcome communication bottlenecks. Optical modules are the data center's expressways; without them, even the most powerful chips are hamstrung.

With global cloud providers consistently raising capital expenditures, and driven by dual demand from NVIDIA and Google, the need for high-speed optical modules is continuously being revised upwards. The appeal of this segment lies in its high certainty. No matter how AI evolves, the demand for computing power will only intensify, directly fueling demand for optical modules.

Furthermore, with recent strong performance in AI applications, the ChiNext AI ETF also offers exposure to this sub-sector. It has recently been officially included in the Stock Connect program, which could potentially attract northbound capital—a fresh source of liquidity that may enhance its market activity.

Now, let's examine the STAR AI ETF (589520; OTC: 024561), which follows a "gold digging" logic.

This ETF focuses heavily on the domestic AI industrial chain, with nearly half its weight in semiconductors and over 30% in AI applications. ByteDance's supply chain alone accounts for over 30% of its weighting. It essentially bets on the entire progression of China's AI sector from "functional" to "effective" and ultimately to "leading."

Why is now a critical window? Three key reasons. First, substantial policy support is materializing, moving beyond rhetoric with clear mandates from the Ministry of Industry and Information Technology for AI breakthroughs. Second, earnings are beginning to materialize. Q3 reports show most constituent companies are profitable, with the majority posting year-on-year growth in net profit attributable to shareholders. Third, external pressure is accelerating technological self-sufficiency. Chip embargoes, rather than being detrimental, have hastened the domestic substitution process.

This pathway offers greater imaginative potential. AI applications are just beginning to reshape business models across industries—a narrative far more compelling than simply selling hardware.

However, it also carries higher risk due to rapid technological iteration; predicting the ultimate winners remains uncertain.

The rise of these two ETFs, in a way, mirrors the ascent of Chinese manufacturing. The stock market reflects the economy, indicating that this AI industrial upgrade is not merely conceptual but a tangible efficiency revolution backed by real investment.

Today, AI is no longer a buzzword but an unfolding reality.

Recent internal revelations from Tesla have starkly exposed the real productivity gap between Chinese and US manufacturing. Upon seeing this, top US investor Louis directly stated: The trade game is essentially over; the US was outmaneuvered.

Why? The story begins with Tesla. Tesla serves as an ideal case study for comparing the US-China gap: it's the same company, producing the same Model 3 and Model Y, with one factory in California and another in Shanghai.

The 2024 data is telling. The Shanghai plant, with 20,000 workers, produced 1 million vehicles. The California plant, with 22,000 workers, produced 464,000 vehicles. Do the math: Shanghai's output per worker is 50 vehicles, compared to just 21 in California. The productivity of a Shanghai worker is 2.4 times that of a Californian worker.

The wage comparison is even more striking. A Shanghai worker's annual salary is $14,000, while a California worker's is $88,000, plus an additional $20,000 in healthcare costs. The healthcare cost alone for a California worker exceeds the total annual salary of a Shanghai worker!

This isn't a narrative about sweatshop exploitation. It's crucial to understand that the Shanghai factory is one of Tesla's most automated globally, surpassing the California facility in robot density, supply chain efficiency, and electricity costs. The truth is, for the same tasks, Chinese workers, equipped with more advanced technology, perform faster and better.

Given such a clear productivity advantage, why have numerous reports over the past decade claimed Chinese manufacturing labor productivity is only a fraction of America's? The issue lies in flawed measurement methodologies, where economists have made two fundamental errors.

The first error is misattribution.

The US Bureau of Statistics categorizes design-focused companies like Apple and NVIDIA—which don't operate production lines—as manufacturing. An Apple employee might generate millions in value, while a Foxconn assembly worker generates tens of thousands. Comparing an Apple designer to a Foxconn assembler inevitably leads to the conclusion of lower Chinese productivity. By the US's own accounting, 30% to 40% of its manufacturing output is actually created by overseas contractors, yet this value is claimed as part of US GDP. China doesn't play this game; it only counts output from domestic physical factories. Comparing these two different accounting standards is fundamentally misleading.

The second error involves price disparities. A Chinese steelworker produces 3.2 times more annual steel tonnage than an American counterpart, but the dollar-value output is only about 20% higher. Why? Because US steel prices are 75% above international market levels, heavily protected by tariffs. Is this evidence of American workers being more valuable, or simply a result of inflated prices due to protectionism?

Let's examine five industries where output can be compared using direct physical units: shipbuilding, steel, electric vehicles, photovoltaics (PV), and cement. In these sectors, there's no room for statistical manipulation—a ton is a ton, a ship is a ship. The results are clear: measured by physical output, the average efficiency of a Chinese worker is 2.4 times that of an American worker.

China now accounts for two-thirds of global shipbuilding output, two-thirds of global EV production, two-thirds of lithium-ion battery production, and a staggering 80% of global PV module production. In recent years, China's installation of industrial robots constitutes over half the global total, with a robot density 50% higher than the US's. There are over 30,000 smart factories, many operating 24/7 as unmanned "lights-out" facilities.

Another revelation concerns the counterproductive effects of tariffs. The US steel industry, protected by high tariffs leading to prices 75% above international levels, has seen its hourly steel output decline by 32% since 2017. Excessive protection creates complacency, stifling the incentive to improve efficiency.

The PV industry tells a similar story. Chinese PV module prices have plummeted 60% since 2020, driven by intense internal competition that forces continuous cost reduction and efficiency gains. The US, relying on tariffs and subsidies to maintain high prices, has an industry with only a few companies languishing in a state of comfortable stagnation.

The price difference in cement is equally dramatic: $148 per ton in the US versus $55 in China. This isn't due to cheaper Chinese labor, but rather higher factory efficiency, massive capacity, and fierce competition. US cement plants, comfortably operating within high price brackets, lack the motivation for technological upgrades. The IMF highlighted this in a 2019 report, concluding that tariff increases can reduce labor productivity by 0.9% within five years.

Trade protectionism, while appearing to save jobs, effectively preserves inefficiency.

So, what has China accomplished in these seven years? When the US imposed a chip embargo on China in 2018, the intent was to stifle its progress. China's response was decisive:

Alright, then we'll become self-reliant! Over the past seven years, China mobilized all available resources and savings, pouring capital into the industrial sector on an immense scale. Despite significant corrections in real estate and stock markets, and local governments tightening their belts, the singular focus was on elevating manufacturing capabilities.

Looking back, this strategy has proven successful! Breakthroughs have been achieved in high-speed rail, nuclear power, industrial turbines, new energy vehicles, drones, and more. One sees Chinese-built trains in Indonesia, Chinese-bid nuclear plants in Saudi Arabia, and Chinese cars filling the streets of Thailand.

Most crucially, consider energy infrastructure. China's current electricity generation equals the combined total of the entire European Union plus the United States and Canada! Data center electricity costs are 3 cents per kWh in China, compared to 7-8 cents in the US.

What does cheap power signify? It enables the deployment of greater computing power, facilitates more experiments, and supports expanded production capacity. DeepSeek's breakthrough is no accident. Denied access to high-end chips, China leveraged cheap electricity to amass computing power and adopted an open-source model to attract global developers. The result is a large language model performing on par with GPT-4, but at a fraction of the cost.

Now, the US faces a dilemma. Option one: Spend trillions of dollars rebuilding rare earth, steel, and chemical industries, while enduring stock market crashes and economic recession. Option two: Seek reconciliation with China and continue benefiting from global specialization. Trump appears to have chosen the latter. Recent high-level meetings have seen US officials discussing topics like rare earth supplies and TikTok deals—essentially asking China for favors. The CEO of Raytheon stated that without rare earths, missile production would halt within three weeks. Ford and General Motors have said they would have to shut factories in two weeks without permanent magnets.

The overall situation has reversed. In 2018, the US threw the punch, and China absorbed it. By 2025, when Trump threatens action, China's stance is: "Bring it on! Let's hurt each other!"—and it's the US that appears hesitant. The reason is simple: the US industrial chain is now dependent on China, whereas China has largely achieved supply chain autonomy.

The Rand Corporation recently published a report titled "Stabilizing Great-Power Rivalry." This think tank, highly trusted by the Pentagon, concluded that the US simply cannot decouple from China and must find a way to coexist.

It's important to recognize that the US model—retaining design, branding, and marketing domestically while outsourcing manufacturing to the most efficient producers—is a strength, not a weakness.

Forcing reindustrialization, compelling American workers into less productive roles, would only reduce US national income. This logic is rather sobering.

During the era of globalization, developed nations enjoyed affordable goods, while developing countries gained employment and technological advancement. Now, some seek to reverse this齿轮 (gear). Who will bear the cost?

China accomplished its industrial upgrade over seven years, at significant sacrifice. Now, if the US attempts the same, its starting point is entirely different. China's debt-to-GDP ratio was only around 30% at the time, providing ample fiscal space for massive investment. The US's current debt-to-GDP ratio exceeds 130%—where will the money come from? Furthermore, implementing industrial policy in the US would inevitably involve navigating complex webs of influence and lobbying.

The Tesla data is highly persuasive: for the same car, under the same standards, the Shanghai plant operates at double the efficiency of the California plant. This isn't a question of who is smarter or works harder; it's the result of a comprehensive ecosystem encompassing industrial infrastructure, policy environment, and systemic efficiency.

Acknowledging this reality is not shameful. Confronting the gap is the first step toward bridging it. Persisting in self-deception will ultimately lead to self-inflicted losses!

MACD golden cross signals have formed, and these stocks are performing well!

Massive information, precise interpretation, all within the Sina Finance App.