Author: XinGPT
The current semiconductor market trend somewhat resembles the script from Nvidia in 2024. Let's review what happened then:
From July 11 to August 5, 2024, Nvidia fell from around $134 to $90.94, a drop of 33%.
On the news front, triggers for the decline included rumors of Blackwell delays starting in July, reports of a US Department of Justice antitrust investigation, and finally, the Bank of Japan's interest rate hike on August 5th, which triggered carry trade unwinding and a panic sell-off in global risk assets, hitting the low. In other words, "a mix of fundamental rumors/small stories + macro-driven asset sell-off."
From August 8th to September 6th, over the course of a month, the stock rebounded from being oversold, coupled with some policy easing on the macro front; but it didn't surpass the previous high. Nvidia's "good but not perfect" earnings report triggered another decline.
Subsequently, on September 11th, Nvidia announced that Blackwell had entered full production and shipments would begin in Q4. Nvidia's stock rose 8% that day. Coupled with the Fed's first 50bp rate cut in 4 years, Nvidia's stock price broke to a new all-time high again on November 8th.
What can we learn from history?
1. What are the characteristics of a阶段性 (stage-wise) top?
Technical: Double top or failure to break through the previous high resistance level; simultaneously, candlestick patterns show bearish engulfing patterns;
Crowded positioning + high leverage: In June 2024, 70% of fund managers believed being long the Mag7 was the most crowded trade, while small-cap short interest approached historical records. The entire market was crowded on one side of the boat;
In the five trading days leading up to August 27, 2024, during another top interval for Nvidia's stock price, retail investors net bought $1.4 billion worth of NVDA. In the same period, the second-place, Tesla, was only at the $100-200 million level, a nearly tenfold difference. NVDL (the 2x Long NVDA ETF) had net inflows of $3.5 billion for the year and repeatedly became the most actively traded instrument intraday.
Fundamentals: On August 28th, Nvidia's earnings report showed revenue and guidance exceeding expectations, but because gross margin fell by 3%, the stock dropped nearly 7% after hours.
So, it's not hard to see that the characteristics of a stage-wise top are the emergence of topping patterns on the technical side, coupled with increased fragility in positioning (high concentration, high leverage). Once there's any hint of trouble in fundamentals or news, the stock price will immediately plummet.
The practical approach is: once technical characteristics appear for identification, and if positioning shows excessive leverage and concentration, one should immediately reduce holdings to hedge risks. Once the downtrend is confirmed, sell.
2. Bottom characteristics
Technical: On August 5th, a long-bodied candlestick after a gap down open, with sufficient volume — 553 million shares was the highest single-day trading volume in that period, exceeding all surrounding trading days. On August 6th, it closed above the previous day's body, confirming an engulfing pattern. On the 7th, a pullback did not break the previous low.
These three days formed the technical signal for a bottom.
We further confirm with positioning: The VIX指数 (VIX index) hit a record that day, reaching 65 intraday (the third highest in history);外围市场 (peripheral markets) like the Nikkei were down -12.4% (largest since 1987), KOSPI down -9% triggering a circuit breaker. All market-level恐慌指标 (panic indicators) were off the charts, also meaning selling pressure reached an extreme;
Fundamentally, the main bearish trigger chain was the BoJ's rate hike on July 31st, yen carry trade unwinding, and the US July non-farm payrolls at 114k, far below expectations. This was a macro-driven sell-off for the broader market, unrelated to Nvidia's fundamentals.
The rebound on September 6th followed similar logic, just not as extreme as on August 5th.
We can also identify bottom characteristics: first, technical patterns begin to show bottoming formations, especially confirmed by volume; second, the logic of the decline does not touch the core business fundamentals of the enterprise. It's more often a one-off macro event or short-term shock for the broader market. Ultimately, quality companies will see their prices修复 (recover from) the erroneous sell-off.
