I disagree with this view—historically, every bubble cleanup has never been the result of rational calculation. Markets are not omniscient, rational actors; fear, uncertainty, doubt, and FOMO are the true forces driving market behavior. The perceived explainability of financial markets is merely hindsight bias—selecting and editing from the full set of facts to express individual will or collective consensus. My view is that the AI bubble’s burst marks the end of technological optimism and Pax Americana; humanity will enter an era of massive collapse, and an eternal bear market will begin. Clear signals are already here 📶

Q3 earnings report confirms that Yi Zhongtian is a cyclical industry, essentially an electronics manufacturer. Only DeepSeek among Chinese AI companies is worth long-term investment; all others are bubble assets. What awaits them will be a systemic industry cleanup on the scale of photovoltaics, liquor, and real estate.

Google's free cash flow burn exceeded market expectations, leading to a sharp drop after hours. A new round of valuation compression has begun in the AI computing power chain sector. Note: Memory and semiconductor manufacturers are cyclical industries, not technology companies—it's common for them to post strong earnings this year but face overcapacity, bankruptcy, or restructuring within two years.

Selling now is a masterstroke; Bitcoin under $50,000 is just around the corner.

No matter how strong the quarterly earnings are for companies in the AI computing power chain sectors—such as storage, optical modules, and semiconductors—no matter how robust their order books, as long as the debt yields and CDS spreads of HyperScale and NeoCloud continue to surge (second derivative), you must not attempt to catch a falling knife!

The sole reason for today’s waterfall: Meta’s Q3 earnings report. Meta’s Q3 financial statements slightly exceeded market expectations, but now merely surpassing revenue forecasts isn’t enough—the market demands clear pathways to AI monetization and improved capital efficiency. Meanwhile, Meta’s personal assistant, Muse, remains in its early stages, burning cash to acquire users, with its primary use case still centered on online shopping. Under these conditions, Meta confirmed and raised its full-year 2026 Capex guidance floor to $135–145 billion, while financing its AI initiatives through the issuance of non-voting Class C shares. This further reinforces market concerns about “massive investments, long payback periods, and pressure on free cash flow.” $META

In the U.S. stock market, options are currently pricing in only about 0.8% volatility for the day after the election. Hartnett called for a "Democratic sweep leading to a 10%+ drop in U.S. stocks" back in early September, when only 23% of fund managers predicted a Democratic sweep. Now, prediction markets have priced in a two-thirds probability of a Democratic sweep. Current odds: -- Democratic sweep: Already largely priced in; the outcome becoming reality reduces uncertainty instead. -- Split government (Democratic and Republican): 27% -- Republican sweep: A 9% tail risk—the real unpriced wildcard. Between the election on November 3 and the results being confirmed by November 5, implied volatility from midterm election odds will rise, and crypto market volatility will rise accordingly. Suppose—just suppose—the Republicans manage to retain the Senate: the market might see a rebound driven by repricing, but its sustainability would likely be limited, given it’s nearing year-end and the Fed still has one more rate hike scheduled for December.

True warriors dare to short the U.S. AI computing sector by 3x, even as JPMorgan upgrades its rating. 👻 $SOXS https://t.co/SKX9d2nVia

Will next Monday be Black Monday?

I just saw on Yahoo Finance that tech companies closely tied to Trump are aggressively hiring lobbying professionals to prepare for an impending Democratic sweep of both chambers in the midterm elections. At this moment, the U.S. AI compute chain resembles China’s domestic property developers of the past, and NVIDIA’s Jensen Huang is the equivalent of Evergrande’s Xu Jiayin. Come November 5, let AI compute and crypto brace for the storm—HODL!