The price of an umbrella becomes expensive during a hurricane, regardless of the wind's direction. For the stock market, this "hurricane" is the Nasdaq 100 Index, and the winds now appear to be shifting. The one-month implied volatility for the Nasdaq 100 has reached 28, while the corresponding figure for the S&P 500 remains below 16. The spread between these two measures has widened to near-record levels. This gap has been expanding throughout the year as stock market returns have become increasingly concentrated among a handful of large-cap technology winners. However, the driver of this widening spread has changed from several months ago. Previously, the distortion in Nasdaq option pricing stemmed primarily from extreme demand for call options. The current expansion, however, is fueled by a rise in demand for put options. As premiums for far out-of-the-money call options have gradually receded, the prices for put options have continued to climb.
According to Nasdaq-compiled data, the spread between the implied volatility of 25 Delta put options for the Nasdaq 100 and the S&P 500—contracts with roughly a 25% probability of finishing in-the-money—has widened from just 3 percentage points in mid-March to 13.6 percentage points currently. This spread reached 13.3 percentage points in 2020. Prior to that, the only time it was higher than current levels was in September 2008. Kevin Davitt, Head of Nasdaq Index Options Content, noted in an interview, "Back then, almost no one was paying attention to puts; everyone was focused on the upside. The market sentiment has now shifted."
"This reflects a growing concern about downside risks for the technology stocks that have been soaring," Davitt added. The rising demand for put options aligns with a slowdown in the upward momentum of AI-themed stocks, which had previously delivered consistent, substantial returns for speculative bulls. On Thursday, the VanEck Semiconductor ETF (SMH) fell 4.5%, dropping below $592 for the first time since late May.
Market Sentiment and the Summer Effect
Nevertheless, while the market has been range-bound for over a month—which could reignite bearish interest—it may not yet be sufficient to signal a full-blown alarm. Demand for call options was exceptionally strong in the first half of the year. Even as investor enthusiasm for further gains has cooled somewhat, overall demand levels remain elevated. According to Nations Indexes' CallDex index, the price of one-standard-deviation out-of-the-money call options for the Nasdaq 100—contracts with about a 16% chance of expiring in-the-money—remains at the 58th historical percentile. Although this is below the 99th percentile peak seen in May, it is still above the long-term average.
Another, more moderate factor potentially widening the volatility spread is the typical summer lull, which tends to suppress S&P 500 volatility. Scott Nations, President of Nations Indexes, stated, "Traders expect the S&P 500 to quiet down over the summer, which is a normal seasonal pattern. However, they don't expect the same for the Nasdaq 100, as they anticipate continued volatility in technology stocks."
Tech Sector Under Pressure
U.S. technology stocks have faced sustained pressure recently. The latest data from Bank of America shows institutional clients have been net sellers of U.S. stocks for four consecutive weeks, with inflows into the tech sector hitting a record low. The immediate catalyst for the recent tech sell-off appears to be a strategic pivot by tech giant Meta Platforms Inc (NASDAQ: META). Reports indicate Meta is planning to enter the competitive cloud computing infrastructure market, aiming to monetize its massive AI infrastructure investments by selling access to its surplus AI computing capacity and models.
This strategic shift follows Meta's enormous capital expenditure on AI infrastructure. The company has raised its 2026 AI-related capital expenditure forecast to a range of $125 billion to $145 billion, nearly double its 2025 projection of $72 billion. In April, Meta also placed an additional $21 billion long-term compute procurement order with AI cloud service provider CoreWeave for the period 2027-2032. Combined with a previous $14.2 billion agreement, their total collaboration now exceeds $35.2 billion.
Market analysis suggests that Meta's transformation from a super-client of compute procurement to a potential competitor with its own low-cost GPU capacity directly pressures the premium pricing power of compute leasing companies. More significantly, it undermines the previously widely held market narrative of an absolute scarcity in compute supply, sparking fears of "peak AI capital expenditure" and a potential "hardware oversupply."
Counterpoints and Industry Realities
However, some analysts caution against overinterpreting Meta's moves as a sign of a global compute surplus. In reality, Meta continues to ramp up its own AI infrastructure investments, having raised its full-year capital expenditure guidance multiple times this year. Its procurement plans for next-generation high-end chips show no signs of contraction. The computing power Meta reportedly plans to release externally primarily consists of older-generation equipment being phased out after upgrades and idle capacity during internal business troughs, not the core, high-end compute power essential for cutting-edge large language model training.
Furthermore, some view the recent tech sector decline as a valuation correction driven by profit-taking after significant prior gains, rather than a fundamental breakdown in the AI investment thesis. The on-the-ground industry reality is more complex than the prevailing panic suggests. Previous reports indicated that Alphabet Inc (NASDAQ: GOOGL) has begun restricting Meta's use of its Gemini AI models because Meta's compute demands exceeded Google's current capacity. This supply constraint directly disrupted the progress of several internal AI projects at Meta, forcing delays in related R&D work. This move by Google starkly highlights that compute supply remains a critical bottleneck for AI industry development. Despite Google's ongoing investments in AI infrastructure, it still cannot guarantee sufficient capacity to meet the market's surging demand.
Comments