Post

What should I do next if I still hold US stock spot positions? Verify this news First, a precaution: this drop shows characteristics of a long-term turning point. Whether it is confirmed depends on whether the subsequent rebound can break through the previous high. Conclusion first: The market is weakening; the worst action is not to cut losses immediately during the downtrend but to wait for a rebound. But what kind of rebound it is—reversal or a bull trap—can only be verified one way. First, why is this drop different from previous ones? On July 1, the Philadelphia Semiconductor Index (PHLX) dropped -6.27% in one day, and on July 2, another -5.44%, totaling over 11% in two days. The trigger was Meta announcing renting out idle AI computing power, directly shaking the core narrative of "permanent scarcity of computing power." By mid-July, the PHLX had retraced nearly 19% from its June high, just one step away from a technical bear market. Marvell dropped nearly 40% from its high, and Kioxia was halved. Such a decline in magnitude and breadth is not a normal correction. The capital logic has indeed changed. Previously, a high-level retracement of three or four points, even seven or eight points, quickly recovered, which was normal oscillation in crowded trades. But a concentrated sell-off of -11% in two days and a halving of the leader requires a completely different recovery time and energy. So after the crash, under what conditions can we be bullish again? There is only one standard: The rebound must effectively break through the previous high. Before breaking the previous high, all rebounds are considered bull traps by default. Does the PHLX still have a chance? The answer is not absolutely yes or no. If it later stabilizes and breaks through the previous high, keep playing; If it doesn't, that is the high point now. Do not enter before confirming a breakout. That position is a trap. Now about the exit timing. Use gold's trend as a reference—note, this is a pattern reference, not a prediction. Gold hit a historical high of 5626 USD on COMEX at the end of January, then dropped to 4090, a retracement of over 20% from the high, and rebounded above 4300 after the US-Iran ceasefire in June. This rebound high is called the escape top. But one key difference must be clarified: Gold's round was driven by interest rate logic—oil prices → inflation → rate hike expectations → real interest rate rise. PHLX's round is driven by chip supply and valuation logic—AI CAPEX narrative shaken + capital withdrawing from crowded trades. Two completely different driving mechanisms. Similar candlesticks do not mean the subsequent path will be the same. So gold's trend is only a pattern reference, not a prediction. If the PHLX later rebounds and repairs, where is the key observation window? One time point worth attention: Anthropic secretly submitted an S-1 in June this year, underwriters are scheduling investor meetings, with the earliest IPO in October 2026. OpenAI also secretly filed an IPO application, targeting a trillion-dollar valuation, but Wall Street generally expects a delay to the first half of 2027. If it follows gold's path, Anthropic's IPO around October is an observation window. If a rebound high appears then, it should be seen as an opportunity to exit in batches. But note two points: First, whether the IPO can be the escape top is itself a probabilistic event, depending on the market environment, valuation acceptance, and secondary market willingness to buy at that time. The correlation between the two events is not that high. Second, if the rebound cannot break the previous high, there will likely be a second bottom test. How much? Do not preset numbers. Q3/Q4 big factory CAPEX guidance, AI downstream demand validation, and fundamentals of stocks in the PHLX index—these are the variables determining the depth of the second bottom. And one difference from gold: AI fundamentals have not collapsed. Anthropic's ARR soared from 9 billion at the end of 2025 to 47 billion in May 2026, with Q2 expected to be the first profitable quarter. As long as downstream AI demand exists, the PHLX has a fundamental anchor. Despite the drop, do not linearly extrapolate it as a crash. Now, there are only two scenarios: Scenario one: PHLX rebounds and breaks through the previous high. Keep playing, keep dancing, the bubble will last longer. The buy point is at the moment of breakout confirmation—note, it may be pushed down by a double top. All other points are bull traps. Scenario two: The rebound fails to break the previous high, following gold's path. The rebound high around Anthropic's IPO is the window for minimizing losses by exiting in batches. One extra note: The Nasdaq and PHLX are diverging. PHLX dropped nearly 19%, while the Dow Jones hit a new high of 52,900 points in the same period. Capital is flowing from the chip sector to finance and retail. Big Tech has visible cash flow support, while the chip hardware chain is facing valuation cuts. QQQ/Nasdaq may continue to survive with subsequent AI giant IPOs, but PHLX and chip equipment tell a different story. Break previous high = continue holding. Fail to rebound past previous high = exit in batches. Two forks left to the market to choose, just prepare response plans. $SNDK $MU $SKHYNIX