AI's Market Grip Slips as VIXEQ-VIX Spread Sheds 14 Points

The spread between expected single-stock volatility and expected index volatility reached a record 34.14 points on July 9, 2026, according to Penn Mutual Asset Management's reading of Bloomberg data, and Cboe wrote in its note for the week of Aug. 31 that the same VIXEQ-VIX spread had narrowed by more than 14 points in recent weeks. That compression is the measurable evidence behind CNBC's Sept. 10 claim that AI is losing its stranglehold on the U.S. stock market, and it landed in the same stretch in which the 10-year Treasury yield climbed to 5%.

What the VIXEQ-VIX spread measures, and what it did

VIX is the Cboe Volatility Index, the options market's estimate of how much the S&P 500 as a whole will move over the next 30 days. VIXEQ is a separate Cboe index that, as CNBC described it in May, aggregates VIX-like measurements for each individual S&P 500 company and weights them by market capitalization. Cboe itself has also referred to VIXEQ as the Cboe S&P 500 Dispersion Index in its monthly index commentary; both descriptions point at the same idea, which is how violently the individual members move compared with the basket that contains them.

Penn Mutual Asset Management, citing Bloomberg data as of Aug. 14, reported that the gap widened to an all-time high of 34.14 points on July 9, when VIX closed at 15.84 and VIXEQ closed at 49.98 - a VIXEQ reading the firm placed in the 98th percentile of observations since the index's 2014 inception. Its explanation is mechanical rather than narrative: when correlation between stocks is low, individual moves cancel out inside the index, so index volatility can stay calm while single-stock volatility runs hot.

The reversal shows up in exchange data. Cboe's note for the week of Aug. 31 said the VIXEQ-VIX spread had fallen by over 14 points in recent weeks while remaining off its all-time highs, and that the QQQ-SPX volatility spread had more than halved since June, with one-month implied volatility near a one-year low across several U.S., European and emerging-market indices. CNBC reported on Sept. 10 that the trend had gone further, with VIX at its highest since April relative to VIXEQ as traders sold broad index-wide equity exposure.

The verified figures

Why rates, not AI earnings, are setting the tone

CNBC's Sept. 10 piece argued that the bond market may now be usurping AI optimism as the stock market's primary driver. The rate backdrop it points to is dated and specific. CNBC reported on Sept. 14 that the 10-year Treasury note yield rose more than 2 basis points to 5%, its highest since October 2023, with the 2-year at 4.666% and the 30-year at 5.374%, all ahead of the Federal Reserve's decision this week. Cboe's Aug. 31 note recorded the same shift from the rates side: the 2-year yield jumped as investors priced in more hawkish policy, with the overnight index swap-implied probability of a hike at the next FOMC meeting rising from 40% to 60% following Chair Kevin Warsh's Jackson Hole remarks.

Sources put the odds at different levels because they measure different instruments on different days. Yahoo Finance, citing CNBC, reported CME FedWatch pricing for a quarter-point hike at the Sept. 16 meeting at nearly 56% shortly after Jackson Hole and above 60% by the week of Sept. 8, while prediction markets Kalshi and Polymarket sat at 48% and 49% in late August. These are dated, market-implied estimates, not Fed commitments, and they can move again before the announcement.

The practical reading is this: when the gap between single-stock and index volatility shrinks, the market expects stocks to move together more, which is what happens when one macro variable - here, long-term interest rates - is doing the driving. That is a statement about expected correlation, not direction. A narrower spread does not say AI stocks will fall, or that the S&P 500 will.

The Sept. 14 slowdown call is a separate event

On Monday, Sept. 14, CNBC reported that AI-linked stocks slid after Anthropic CEO Dario Amodei called on AI companies to slow the pace at which they advance model capabilities, with OpenAI's Sam Altman saying he agrees that AI companies need to "pace the frontier" and xAI's Elon Musk posting that "Dario is right." South Korean memory-chip makers SK Hynix and Samsung Electronics closed down more than 6% and more than 4% respectively, according to CNBC.

This is a fundamentals shock to the AI trade arriving on top of a rates shock, and the two should not be merged: the volatility-spread shift described by Cboe and CNBC developed across July and August, before Amodei's essay existed. Zoe Gillespie, a senior director at RBC Brewin Dolphin, told CNBC's "Squawk Box Europe" that the equity rally has been based on AI growth and productivity gains, so a derailment could affect equity performance going forward. That is an attributed opinion about a possible path; none of the sources here quantifies any change to AI capital expenditure plans, chip orders or company guidance as a result of the essay.

Two things stay unknown. CNBC did not publish current VIXEQ and VIX levels in the Sept. 10 article, so the exact distance from the July 9 record cannot be verified from these sources, and the identity and motive of the traders selling index exposure is not established, because options flow descriptions do not identify investor types.

The dated catalyst: the Sept. 16 Fed decision

The next verified event is the FOMC decision on Sept. 16, the meeting Yahoo Finance, citing CNBC, identified as the one being priced by CME FedWatch, and CNBC reported Sept. 14 that Treasury yields were rising ahead of it. If the dispersion thesis holds, what to watch afterward is not only whether the Fed hikes but whether index volatility keeps rising relative to single-stock volatility - the VIXEQ-VIX relationship that flagged the shift in the first place.