• Left-side trading is difficult to execute, and right-side trading is also not good. Xiaomi’s buyback, Lei Bao’s (Lei Jun’s) own buyback, caused the stock price to rise sharply, stabilizing above the 40th level. Alibaba’s financial report saw a significant drop in net profit, and the market predicted that the intensity of takeout subsidies would decrease, leading to a large surge in Meituan’s stock price (6%). The capital logic of AI is not closed, it’s not previous business; first grab users, run around, then slowly make money. AI’s continuous costs cannot be reduced – hardware plus electricity.

In today’s market environment, investors are facing an extreme sense of tearing: Left-side trading (counter-intuitive bottom fishing) often dies before dawn, and right-side trading (following the trend to chase gains) is also prone to standing on mountain tops. The recent cases of Xiaomi, Alibaba, Meituan, and the AI sector perfectly illustrate this game of capital logic.

Xiaomi: Genuine Silver and Gold “Right-Side Signal” – It All Starts with the Boss Leading the Way

• Phenomenon: The stock price breaking through the HK$40 barrier wasn’t instantaneous.
• Deep Logic: Xiaomi’s previous volatility frustrated many left-side ambush funds, losing patience. The real turning point came from a powerful injection of confidence.
  • News Facts: Just days before, Lei Jun personally invested over HK$1 billion to increase his own holdings at an average price of approximately HK$38.58; simultaneously, Xiaomi Group repurchased millions of shares.
  • Market Reaction: This “boss leading the charge + company buyback” double benefit directly sent out the clearest “right-side signal” to the market. The stock price surged, firmly establishing itself above HK$40, and its market capitalization returned to a trillion.
• Trading Insights: For ordinary investors, holding strong faith (believing EVs will succeed, smartphones becoming premium will succeed) is necessary to withstand the left-side ambush of Xiaomi. However, once Lei Jun steps in to do the right side, the cost has increased, although it’s safe, but the profit space is compressed.

Alibaba vs Meituan: “Ceasefire Agreement” in Financial Reports

• Phenomenon: Alibaba’s net profit declines, while Meituan’s stock price surges. This appears contradictory but is logically sound.
• Underlying Logic: This is a game about “ending the intense competition and releasing profits.”
  • News Facts: Recent Alibaba financial reports show a significant drop in net profit (partly influenced by investment losses), but the market keenly observed changes in core business data – the intensity of subsidies for local life (delivery/to-go) is decreasing.
  • Capital Dynamics: As long as Alibaba doesn’t continue to recklessly burn cash subsidizing its fight against “food delivery wars,” Meituan, as the industry leader, stands to benefit most. The market anticipates a shift from “price war” back to “rational profitability.”
• Trading Insights:
  • Short Alibaba on the Left Side of Logic: Panic sell when you see profit declines, potentially missing the logic behind its core e-commerce stabilization.
  • Long Meituan on the Right Side of Logic: Enter only after financial reports confirm subsidy reduction and profit explosion (the “right side”), when stock prices have often already priced in this expectation. Meituan’s massive surge is essentially making money by “predicting others’ predictions.”

AI Sector: A “Black Hole” of Unclosed Capital

• Phenomenon: Regardless of whether it’s the US stock market or the A-share market, AI stocks continue to plummet or experience significant volatility even after demonstrating application implementation.
• Deep Logic: The capital market is transitioning from the “storytelling” phase into a brutal “accounting” phase.
  • Hard Impact (Missing Capital Cycle): Traditional internet models are: burn cash -> acquire users -> diminishing marginal costs -> earn passively.
  • AI’s Achilles Heel: AI’s marginal cost cannot be zeroed out, and it’s even difficult to reduce.
    • Hardware Depreciation: Expensive GPUs depreciate rapidly.
    • Energy Costs: Training and inference not only burn money but also consume electricity. Redwood Capital recently released a report highlighting the “$60 Billion Problem for AI” – meaning that the entire industry needs to earn back $60 billion annually just to cover infrastructure construction costs, while current application revenue is nowhere near even a fraction of that.
  • Current Situation: Only those selling shovels (Nvidia, optical modules) are making money; companies developing applications (acquiring users) aren’t only not earning money, but each additional user requires paying more electricity and computing costs.
• Trading Insights: The left side of AI at this time is like an endless pit because the cost cannot be reduced; the right side of AI (performance fulfillment) remains distant because business models haven’t been fully realized.

Summary

• Xiaomi tells us that certainty often requires gold and silver (repurchase/increase holdings) to confirm, but at the cost of losing low-priced chips.
• Alibaba/Meituan tell us that understanding the “tacit agreement” between giants is more important than simply looking at financial report numbers.
• AI tells us that in this sector, as long as the formula “revenue < (hardware depreciation + electricity costs)” remains unbroken, all rebounds are merely emotional speculation rather than value recovery.