BlockBeats News, August 19th. Last night, the US stock market experienced a sharp decline, with AI trading at the center of the selloff. The Nasdaq fell by 1.3%, the S&P 500 dropped by 0.7%, the Nasdaq 100 declined by 1.7%, and the Philadelphia Semiconductor Index plummeted by 5.6%. Storage stocks were hit the hardest, with SanDisk falling by about 9%, Micron falling by about 7%, and Western Digital dropping by about 7%. AI chip leader NVIDIA fell by about 2.3%, Broadcom by about 3.2%, and companies along the semiconductor supply chain such as Marvell, Intel, and AMD also experienced widespread declines, with even deeper plunges in the optical communication and AI network equipment sectors. For this current AI bull market, there is currently pressure on both revenue expectations and position structures.

The first pressure comes from the commercialization data of AI labs. OpenAI disclosed to investors that its second-quarter revenue increased from $57 billion in the first quarter to $67 billion, a sequential growth of 18%. However, losses further widened, and the operating profit margin continued to decline. This growth rate is lower than what some shareholders expected, causing the market to reassess OpenAI's path to profitability before its IPO.

Anthropic has also sparked controversy. Its annualized revenue run rate as of the end of July was reported to have reached $650 billion, still showing extremely high growth, but lower than the expectations of $700 billion to $800 billion or more from some third parties and optimistic market views.

This directly hits the most sensitive nerve of the AI market. Over the past year, US stock market bulls have been willing to continue buying GPUs, data centers, storage, power, and cloud capital expenditures, on the condition that top application companies like OpenAI and Anthropic can continuously prove strong end demand. However, when revenue growth is lower than optimistic expectations and losses continue to expand, investors will begin to re-examine the return cycle of the AI capital expenditure chain.

Positioning factors have amplified this downturn. Goldman Sachs Prime Brokerage data has shown that the short interest ratio of typical S&P 500 component stocks has risen to a high not seen since 2011. FactSet and Robinhood statistics also indicate that over the past three months, short positions have increased in most US large-cap stocks. Meanwhile, a large amount of long funds are still concentrated in the AI infrastructure chain, and some high valuation AI companies have become targets of hedge funds' crowded short positions. With both long and short positions crowded, the market has become more sensitive to any negative news.

In this structure, corrections are magnified. The AI chain had a significant increase before, with long positions already having substantial unrealized gains; meanwhile, short positions are waiting for cracks in revenue, profit margin, or capital expenditure logic to appear. Once the growth story of OpenAI and Anthropic falls below the market's highest expectations, funds will first take profits, and shorts will push down valuations, ultimately leading to last night's collective weakness in tech stocks.