NVIDIA, Anthropic IPO, Big Tech capex, fed rates: Why AI trade is facing a brutal sell-off | Explained – Markets


Stock Logo

Written by: Alka Jain

Updated Nov 21, 2026 12:03 IST

Google CTA

AI bubble

NVIDIA, Big Tech capex, fed rates: Why AI trade is facing a brutal sell-off | Explained

The AI trade is facing one of its biggest tests yet. There has been a sharp correction in US equities wiping out trillions of dollars in market value, which has put some of the world’s largest tech companies like NVIDIA, Oracle, AMD and others under pressure.

The Kobeissi Letter, in its latest market analysis published on November 20, has estimated that US stocks have lost approximately USD 6 trillion in market capitalisation in just 45 days. This raises a critical question for investors that if the AI bubble is finally bursting, or the market is going through a correction?

To be more specific, the global capital markets commentary firm which has been bullish on artificial intelligence since 2022 — the launch of ChatGPT — stated that US equities had shed an average of USD 225 billion per trading day over 31 consecutive sessions.

However, it also argued that the recent sell-off does not necessarily signal the end of the AI boom. It rather attributed the correction to concerns over tech companies’capital expenditure, future profitability, interest rates, regulatory uncertainty, geopolitical developments and the NVIDIA controversy.

The analysis comes as NVIDIA witnessed a steep decline in its market valuation after the reports of Jensen Huang’s extra marital affairs surfaced. In a separate post on X (formerly Twitter), The Kobeissi Letter said nearly 40 per cent of the S&P 500’s market-capitalisation loss since its October 6 record high could be attributed to NVIDIA.

AI infrastructure spending crosses USD 1.6 trillion

The main trigger behind the recent technology stock sell-off is the scale of spending on AI infrastructure. According to The Kobeissi Letter, the combined capital expenditure by Amazon, Alphabet, Meta and Microsoft increased by USD 36.9 billion in the third quarter of 2026 to a record USD 201.9 billion.

The figure, as per the firm, had more than doubled over the preceding 12 months and rose 356 per cent from the first quarter of 2024. Cumulatively, the four technology companies have spent USD 1.6 trillion on capital expenditure over the past three years, according to the analysis.

The Kobeissi Letter acknowledged that these concerns were real but argued that infrastructure spending was necessary for the development of AI. It said uncertainty around returns on investment could persist until the first half of 2027, when investors may have greater clarity on the monetisation of AI infrastructure.

At the same time, the firm also highlighted the earnings performance of S&P 500 companies. It cited year-on-year earnings per share growth of 31.9 per cent in the third quarter of 2026, describing it as the strongest growth rate in at least four years.

NVIDIA controversy adds to tech stock pressure

NVIDIA has emerged as a major contributor to the technology sector’s recent volatility. The publication identified the controversy surrounding Huang, the company’s leadership transition and scrutiny involving Singapore-based Megaspeed as factors that have added to uncertainty around the chipmaker.

In its analysis, The Kobeissi Letter said NVIDIA shares had declined around 40 per cent over a matter of weeks. It also argued that the market was pricing in a highly pessimistic outlook for the company, including uncertainty over its leadership and future earnings.

The controversy has developed alongside reports of US investigations involving Megaspeed and questions surrounding alleged diversion of advanced AI chips to China.

Separately, the SEC is examining potential accounting issues at NVIDIA, according to Reuters. These developments have added regulatory and governance risks to an already volatile investment environment.

However, The Kobeissi Letter cautioned against treating allegations involving Huang as established facts. It said it had maintained the position that Huang should be presumed innocent unless proven otherwise and that the company’s board should leave open the possibility of his return until an independent investigation establishes the facts.

It also pointed to developments involving Advanced Micro Devices (AMD), which has faced allegations of chip diversion to China. It noted that the claims had not been substantiated.

The market repricing has also affected valuations. According to The Kobeissi Letter, the technology sector’s forward price-to-earnings multiple had fallen to its lowest level since March 2020. It put NVIDIA’s forward P/E ratio at around nine times, arguing that the market was discounting the company’s future earnings potential.

The publication also referred to NVIDIA’s projected revenue of approximately USD 750 billion for fiscal 2028, arguing that investors were not fully factoring in the company’s earnings outlook.

Market volatility, interest rates & liquidity concerns

Apart from AI spending and NVIDIA-specific developments, The Kobeissi Letter identified monetary policy, political uncertainty and liquidity conditions as additional pressures on US equities. “The US Federal debt situation is even worse,” it said.

It pointed to the US Federal Reserve’s interest-rate stance and the cost of financing large-scale AI infrastructure. Higher borrowing costs can make data centre expansion more expensive, particularly for companies relying on debt financing.

The publication also cited an expected increase in US investment-grade corporate bond issuance. It estimated November issuance at USD 175 billion, up USD 37 billion, or 27 per cent, from a year earlier. This would make it the highest November issuance on record.

Another factor highlighted was the change in the US political landscape following the midterm elections. The Kobeissi Letter argued that the outcome had weakened what it described as the ‘Trump Put’ — the perception among investors that President Donald Trump would intervene to support markets during periods of heightened volatility.

It also flagged concerns about potential AI regulation and taxes, although it believes market expectations of restrictive measures may be excessive.

The publication further pointed to the rise in market volatility and a sharp deterioration in investor sentiment. It said the CBOE Volatility Index had risen more than 100 per cent from its October low.

Liquidity was another concern. The Kobeissi Letter linked the recent volatility to large institutional and retail demand for Anthropic’s initial public offering, which it said had absorbed capital that could otherwise have been deployed in other parts of the market. It is important to note that Anthropic’s USD 81.34 billion IPO in 2026 is nearly 1,937 times larger than Nvidia’s original USD 42 million public offering in 1999.

It estimated institutional demand for the IPO at USD 400 billion and retail orders at USD 100 billion, compared with an initial expected offering of USD 64.4 billion at a share price of USD 78.

Is the AI bubble bursting?

Despite the scale of the sell-off, The Kobeissi Letter does not believe the current correction marks the bursting of the AI bubble.

Its argument is that a substantial amount of negative sentiment has already been reflected in stock valuations, while earnings growth and the long-term investment case for AI infrastructure remain intact.

The publication said it was not ruling out further market declines and cautioned that volatility could remain elevated as liquidity conditions remain constrained.

It also said the market had not necessarily reached its bottom. However, it believes the risk-reward equation is improving for technology companies with strong balance sheets, rising earnings and durable competitive advantages.

The firm expects investor interest to return to high-quality technology stocks once market uncertainty subsides, drawing a comparison with the recovery following the April 2025 market sell-off.

For investors, the distinction between a correction and a structural breakdown in the AI investment cycle will depend on whether the enormous capital being deployed into infrastructure translates into sustainable earnings and cash flows.

The coming quarters, particularly the first half of 2027, could therefore be important in determining whether the current repricing represents a temporary reset in expectations or a more fundamental reassessment of the economics of artificial intelligence.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *