By the end of July, six of the US stock market's "Magnificent Seven" had released their latest quarterly earnings, but the AI-led rally did not calm down. Instead, it entered a phase of intense repricing.
A recent report from Bank of America's equity derivatives team revealed a surprising finding: For the first time since OpenAI's ChatGPT launch, almost all of the "Magnificent Seven" stocks experienced actual post-earnings volatility exceeding the pre-earnings implied volatility priced by the options market. This indicates that the options market has, for the first time, systematically underestimated the earnings impact of the major tech stocks.
For example, Microsoft shares surged roughly 16% after its earnings report, marking one of its biggest single-day gains in recent years. In contrast, Tesla plunged about 15% due to disappointing results and outlook. Both cases highlight the increasingly volatile repricing of AI-era leaders.
Bank of America's discovery means that the market is repricing AI leaders by a far greater margin than expected, whether the news is positive or negative. This has pushed the level of divergence among US stocks to its highest point in nearly 35 years.
Benjamin Bowler, head of Bank of America's derivatives business, warned that rising macro uncertainty strongly supports current volatility levels. On a strategic level, Bank of America is positioning for both the medium-to-long-term upside of the tech sector and short-term rotation into value stocks as a hedge. It specifically recommends the SPX December 2026 put spread as a core hedging position.
Tech Giant Volatility Loses Anchor: First Collective Overshoot Since ChatGPT Era
Over the past two years, the ChatGPT-driven generative AI boom has made the "Magnificent Seven" the core driver of the S&P 500's rally. As the market formed a consensus on the path of AI commercialization, the options market became increasingly adept at predicting the post-earnings movement range. In most cases, actual gains or losses did not deviate significantly from implied volatility.
However, Bank of America believes this earnings season has broken that pattern. In a report titled "Market Derailment is a Feature, Not a Bug," the derivatives team noted that, so far, all of the "Magnificent Seven" stocks, except for Nvidia, which has yet to report, have seen actual volatility far exceed the levels implied by the options market. The bank described this as a "first in the ChatGPT era."
This phenomenon reflects a decline in the market's ability to price tech giants effectively. In an AI-narrative-driven market, the options market has consistently underpriced tail risk. The issue of single-stock fragility remains prominent heading into 2026. Bowler estimates that the frequency of "fragility events" for S&P 500 tech stocks could match the historical peak seen in 2025. On July 30, the daily dispersion of returns among S&P tech stocks approached a historical extreme, coinciding with the unwinding of positions from the Situational Awareness fund and the concentrated release of earnings from large-cap tech stocks.
US Stock Volatility Dispersion Hits New High, Bubble Pattern Approaches Internet-Era Extreme
Bank of America's calculations further show that due to the combined impact of tech giant earnings, the unwinding of the Situational Awareness hedge fund, and ongoing uncertainty about the policy path of the new Federal Reserve Chair, Kevin Warsh, the dispersion of individual US stocks has risen to its highest level in nearly 35 years.
Stock dispersion refers to the widening divergence in performance between different individual stocks. Unlike a market where all stocks rise or fall together, a high-dispersion market means that stock selection becomes significantly more important, and index performance becomes a less reliable indicator of actual investor returns. For derivatives traders, this typically means the value of single-stock options rises, while index options may not benefit as much. This also makes market-neutral strategies, pair trading, and volatility trading more active.
On a broader macro level, the realized volatility dispersion of the S&P 500 is steadily converging towards the historical highs seen during the dot-com bubble burst. Bank of America believes this aligns with its earlier assessment that, as the AI bubble accumulates, weighted dispersion could surpass the extreme levels of the internet era. The reason is that the volatility of the ultra-large-cap stocks, which hold a highly concentrated weight in the market, is exceptionally high.
A market structure characterized by low correlation and high rotation further fuels the divergence of individual stocks, both within sectors and between them. Bowler noted that return and volatility dispersion are core indicators for measuring the progression of bubble-like price behavior, a framework he pioneered two years ago.
Bank of America: AI Rally May Not Be Over, But Short-Term Rotation Risk is Rising
Despite the recent surge in market volatility, Bank of America does not believe the AI rally is finished. The report states that historical experience shows that during the formation of a tech bubble, even if the long-term upward trend remains intact, periodic corrections and capital rotations into value stocks are common.
Based on this assessment, Bank of America advises investors to consider using derivatives for risk management rather than simply reducing their equity exposure. The report outlines two representative strategies: First, using call spreads in the healthcare sector. Bank of America believes the healthcare sector currently ranks high on its "bubble risk indicator," but using a spread strategy can provide high return potential while controlling downside risk.
Second, for investors who expect a slow correction in the S&P 500, they could consider put spreads or put spread collar strategies to navigate the current environment of frequent rotation and high demand for protective options. Bank of America emphasizes that in the current AI-dominated market, the index may still appear strong on the surface, but the internal structure is undergoing drastic changes. Future market performance will likely depend more on companies' ability to deliver earnings, rather than the AI concept itself.
Comments