According to Woofun AI, Wall Street News points out that the AI boom is facing a severe financial test, hedge funds have drastically cut their exposure, and the Nasdaq 100 Index (NDX) is in turmoil.
In the past two weeks, global technology stocks have received a net sale of about $20 billion, and the intensity of the sell-off has reached an extreme level of -3.3 standard deviations. This intensity reminds the market of a correction in July this year. According to Goldman Sachs data, the TMT sector has accumulated sales of about $20 billion, which is close to $23 billion during the July sell-off period.
Although the S&P 500 index fell only about 2.4% in the current round, down from the 3.4% drop in July, Goldman Sachs risk appetite recorded one of the biggest single declines on record, falling to its lowest level since April 2025. Last Thursday (October 22), the three major strategies of fundamental long and short, systemic and multi-strategy all fell by more than 1% in a single day, the first time since the pandemic. Net fundamental exposure in the TMT sector fell to a low point during the year. Although total exposure is still at the 63rd percentile, this means that although net positions have been compressed, total leverage is still a potential pain point at present. The Nasdaq 100 Index (NDX) is still stuck in the fluctuation range since May. The 29,000-point futures mark and the 200-day EMA of about 28,500 points below it form the most critical technical support band at present.
Notably, fears are highly concentrated in the technology sector. Goldman Sachs analysis shows that the volatility spread between NDX and SPX has rapidly widened to the 98th percentile of the five-year range, and the single-day expansion last Thursday hit the second largest record in five years. Volatility pricing exclusive to the technology sector is already at an unusually expensive level compared to overall market fluctuations, which has spawned potential relative value trading opportunities. Linkage among AI-related stocks has increased significantly, but overall exponential volatility remains relatively restrained.
According to data compiled by Woofun AI, Rich Privorotsky (Rich Privorotsky), head of fixed income and stock trading at Goldman Sachs Europe, pointed out that although AI demand is still strong, electricity supply is becoming a hard constraint, with a gap of about 40 to 50 gigawatts next year (as stated in the “The Markets” podcast on October 22). The financing environment has been tightened at the same time — According to J.P. Morgan Chase statistics, the scale of AI infrastructure debt financing during the year has exceeded 500 billion US dollars, and the financing costs of hyperscale cloud service providers have increased by about 120 basis points compared to the previous period. The financing links between Nvidia (NVDA.US) and its GPU customers are being viewed as a potential credit risk transmission chain. GPU operators, such as CoreWeave (CRWV.US), rely on customer financing and credit support to maintain the expansion of AI infrastructure. Once GPU operators face tight financing, the pressure will be transmitted upward along the supply chain and eventually fed back to upstream suppliers such as Nvidia.
The Philadelphia Semiconductor Index (SOX) is currently testing 200-day EMA support, adding an important technical reference to this risk path. This week, the financial reports of Google (GOOGL.US), Microsoft (MSFT.US), and Amazon (AMZN.US) were intensively disclosed. The focus of the market's attention has gone beyond the capital expenditure guidelines themselves. The real question of concern is whether backlog orders can be effectively converted into revenue and how the next round of investment expansion will obtain financing. AI-related credit default swap (CDS) spreads have clearly widened recently, but J.P. Morgan believes that this is largely not a reflection of deteriorating fundamentals, but rather the result of banking institutions actively buying protection to hedge against their own expanding data center exposure. Interest spreads between CDS and cash bonds are widening, or new relative value opportunities are being created.
Furthermore, while the AI capital expenditure boom is driving domestic demand in the US, large amounts of capital are flowing overseas, creating asymmetric spillover effects. Technology exports account for a share of GDP, with South Korea approaching 20%, which means that if AI capital spending shrinks sharply, Asian exporting economies will bear the brunt. TS Lombard economist Dario Perkins proposed a counterintuitive logic: the end of the AI investment boom is not necessarily a complete economic disaster. He believes that the current rapid expansion of AI capital expenditure is boosting inflation and keeping interest rates at a higher level, putting continuous pressure on interest-sensitive sectors such as real estate. If AI capital spending falls sharply, the Federal Reserve will gain room to cut interest rates more aggressively, thereby reactivating demand in non-tech sectors and buffering the overall economic downturn.
J.P. Morgan maintains a constructive judgment on the fundamentals of the technology sector, but at the same time suggests risks such as insufficient sustainability in semiconductor pricing and emerging cloud computing (neocloud) financing pressure, and the software sector is relatively resilient. The bank believes that there may be a potential buying window in early November, but the conditions have not yet been fully fulfilled: positions must be fully cleared, hyperscale cloud vendors must confirm the direction of capital expenditure, semiconductor profits need to withstand the test, and credit spreads need to return to narrowing. In the absence of any of these conditions, the sustainability of the rebound is questionable. For investors, the current market is neither a clear sign of clearance at a glance, nor an unconditional time to break the bottom — the answer is probably hidden in the earnings reports of several tech giants this week.