A Bear Market Trapped Inside a Bull Market?

A horrendous day for the Nasdaq. Double-digit declines in some of the market’s most important companies. And the VIX closes under 20? Is this even possible?

What gives?

This last week was profoundly unsettling; a crescendo of unease that had been building for weeks. Investors increasingly questioned the economics of the artificial-intelligence boom that helped propel technology shares higher. The financial press fretted about mounting pressure on Microsoft, Meta, Amazon and Alphabet as investors scrutinized the scale, financing and prospective returns of their AI investment. By Friday, the Nasdaq was suffering its worst day in months.¹

Intel fell a whopping 28.6%. The semiconductor index lost 7.2%. Alphabet and other AI-related leaders were hit hard as well. These were sudden, painful and very visible repricings of businesses that had been central to the market’s advance. Yet something very different was happening beneath the surface and honestly much of the rest of the market, particularly the forgotten names and sectors, was doing fine.

A bear market in the leaders?

Technology and Communication Services have suffered much more severe drawdowns than most other sectors. That is incontrovertible. But these losses are not occurring uniformly across the market. Energy, Financials, Utilities and other areas have generally held up much better. And despite the carnage in technology, the S&P 500 itself finished Friday only about 3% below the record high it set on October 6.

Perhaps the most striking illustration is the chart below. It compares the Nasdaq-100 with the S&P 500 excluding the 30 index members on our own AI impact list, the companies in our AI Beneficiaries and AI At-Risk baskets. From the October 6 close through Friday, the Nasdaq-100 fell 9.09%. The S&P 500 without those names rose 2.15%.

Chart 1

S&P 500 Excluding AI vs. Nasdaq-100

Price return since the October 6, 2026 close, daily closes through October 30, 2026

S&P 500 excluding AI is the S&P 500 at its October 6 weights, taken from SPY’s holdings, less the 30 index members of Belvest’s AI Beneficiaries and AI At-Risk baskets, about 36% of the index. The remaining names are held at constant share counts from the October 6 close. Source: LucenWorks dataset; Belvest Family Wealth calculations.

This is a very different picture from a traditional broad-market selloff. What feels like a bear market to many investors is actually something we’ve been anticipating for a few weeks: a deliberate and dramatic change in market leadership. And that helps explain one of the strangest features of Friday’s trading: The VIX closed at just 19.42.

The VIX puzzle

For investors heavily exposed to technology, Friday certainly did not feel like a VIX-at-19 kind of day. QQQ, which tracks the Nasdaq-100, fell 3.47%. But SPY, which tracks the S&P 500, fell roughly half of that, or 1.77%. Granted, the VIX did climb from 17.20 to 19.42—an increase of 12.91%—but still finished below 20 and remarkably close to its long-term average.

Table 1

October 30 Market Snapshot

Closing levels on October 29 and October 30, 2026

	Oct 29 close	Oct 30 close	Change
SPY (ETF proxy for the S&P 500)	769.36	755.72	−1.77%
QQQ (ETF proxy for the Nasdaq-100)	715.34	690.49	−3.47%
VIX	17.20	19.42	+2.22 pts (+12.91%)

ETF changes are daily total returns. Source: LucenWorks dataset.

Most investors associate a sharp equity selloff with an explosion in volatility. And declines usually do produce a larger volatility response than gains of similar magnitude. So why didn’t this one?

The first answer is straightforward but important: the VIX is not a Nasdaq volatility index; rather, the VIX is derived from S&P 500 index options. It estimates expected S&P 500 volatility over approximately the next 30 days, expressed at an annualized rate. It does not translate a particular Nasdaq decline into a corresponding VIX level.²

On Friday, QQQ lost almost twice as much as SPY, which tells us a lot. If technology is collapsing while the broader market is relatively resilient, a technology-heavy portfolio can experience a very bad day without S&P 500 options pricing the sort of market-wide panic associated with 2008, 2020 or other systemic episodes. In other words, the VIX was not necessarily missing the selloff, but it was measuring a different market.

Was 19.42 actually unusual?

Yes, a VIX below 20 alongside a 3.47% QQQ decline seems unusual—so we went back and tested it. We began by looking at the 100 worst daily declines in the full available histories of SPY and QQQ and comparing each decline with the VIX close on the same day.

Friday itself did not make either list. In fact, SPY’s Friday loss was 438th worst among 8,496 trading days since February 1993. QQQ’s decline ranked 194th worst among 6,954 trading days since March 1999. Large declines happen; investing isn’t for the faint of heart.

Indeed, the worst-100 charts provide useful perspective. VIX readings on the market’s truly awful days have generally been much higher than Friday’s 19.42.

Chart 2

SPY: VIX Closing Level on the 100 Worst Daily Declines

VIX close against SPY daily total return, 100 worst days from February 1, 1993 to October 30, 2026

October 30 ranks 438th of 8,496 trading days. It is marked for comparison and excluded from the fitted line and the worst-100 statistics. The fitted line is ordinary least squares over the 100 worst days only; October 30 lies outside that range (shaded), so its fitted value is an extrapolation. Returns include dividends on their ex-dates; the VIX is the Cboe spot close. Source: LucenWorks dataset; Belvest Family Wealth calculations.

Chart 3

QQQ: VIX Closing Level on the 100 Worst Daily Declines

VIX close against QQQ daily total return, 100 worst days from March 11, 1999 to October 30, 2026

October 30 ranks 194th of 6,954 trading days. It is marked for comparison and excluded from the fitted line and the worst-100 statistics. The fitted line is ordinary least squares over the 100 worst days only; October 30 lies outside that range (shaded), so its fitted value is an extrapolation. Returns include dividends on their ex-dates; the VIX is the Cboe spot close. Source: LucenWorks dataset; Belvest Family Wealth calculations.

Across SPY’s 100 worst days, the median VIX close was 39.80. Across QQQ’s 100 worst days, it was 28.52. In each sample, only one day had a VIX close below Friday’s 19.42. Interesting—but also a somewhat unfair comparison. In any case, Friday was nowhere near the 100 worst days in either ETF.

A better test is to compare apples with apples. We therefore looked at every other trading day when each ETF fell within half a percentage point of its October 30 decline. Those results are more revealing.

For SPY, there were 525 comparable declines. The median VIX close was 23.51. On 123 of those 525 days, however, the VIX closed below Friday’s 19.42. In other words, Friday’s VIX was low for a 1.77% S&P 500 decline, but hardly extraordinary. It was around the 23rd percentile of comparable observations.

QQQ tells a more interesting story. There were 161 historical days when QQQ declined between 2.97% and 3.97%—roughly comparable to Friday’s 3.47% loss. The median VIX close on those days was 25.90. And only six of the 161 had a VIX close below 19.42. That puts Friday at roughly the 4th percentile.

Table 2

VIX on Comparable Declines: SPY vs. QQQ

Every other trading day within 0.50 percentage point of each ETF’s October 30 return

	SPY	QQQ
October 30 return	−1.77%	−3.47%
Matching band	−2.27% to −1.27%	−3.97% to −2.97%
Comparable days	525	161
Sample dates	Apr 2, 1993 to Jul 29, 2026	Mar 23, 1999 to Jun 23, 2026
Median VIX close	23.51	25.90
Mean VIX close	24.88	29.10
25th percentile	19.78	22.71
75th percentile	27.91	32.02
Minimum	12.13	14.56
Maximum	69.15	79.13
Days with VIX below 19.42	123	6
Days with VIX below 20	135	11
October 30 empirical percentile	23.43	3.73

Comparable days are every trading day within 0.50 percentage point of the ETF’s October 30 return, October 30 itself excluded. Source: LucenWorks dataset; Belvest Family Wealth calculations.

Only 6 of 161 comparable QQQ declines had a lower VIX close. For SPY, the figure was 123 of 525. That is the important distinction. Friday’s VIX was genuinely unusual when viewed through the lens of the Nasdaq-100, but it was considerably less unusual when viewed through the lens of the S&P 500. And this can point us towards the cause of the divergence.

The market is asking a different AI question

The selloff tells us something about how the AI debate is changing. A few years ago the dominant question was whether artificial intelligence would be transformative. That’s no longer being debated. But now investors want to know what companies will earn on the enormous amount of capital required to build it and who the winners will be.

One can believe that AI represents an extraordinary technological revolution and simultaneously question whether every dollar of AI-related capital expenditure will earn an attractive return. Two things can be true. So many times markets start to distinguish between a great technology and a great investment at a particular valuation.

The scale of the current buildout is extraordinary. AI-related corporate borrowing has surged. Capital spending on computing infrastructure has accelerated dramatically. And companies are increasingly being asked to explain when those expenditures will translate into earnings and free cash flow. All of this can be completely consistent with the AI thesis. But, it also can mean that the standards for success are changing, too.

For several years, simply having exposure to the AI investment cycle was often enough to justify super high multiples of revenue. But increasingly, investors appear to be demanding evidence that the capital being deployed will generate free cash flow. And profitability matters, too.

A bear market—or a rotation?

If the S&P 500 were collapsing alongside technology, credit conditions were deteriorating, Treasury markets were signaling economic stress and volatility was exploding across asset classes, the interpretation would be straightforward. But, that is not what we are seeing. Instead, the market’s prior leadership has suffered a substantial correction while other parts of the equity market have remained comparatively resilient. Indeed, this looks more like a bear market in prior leadership inside a still-resilient broader equity market.

There are two very different ways this can resolve. 1) The weakness in technology could spread. What begins as a concentrated repricing could eventually pull down earnings expectations, sentiment and other sectors. If correlations rise and the selling broadens, the market would begin to look much more like a conventional correction.

Or 2) the opposite could occur. In this case, technology valuations could reset while capital migrates toward financials, industrials, energy, value stocks and other areas that participated less fully in the previous advance. In that case, the current correction would represent a broadening of market leadership rather than the end of the bull market.

We do not know which path will prevail but the past few weeks look like more of a rotation than a contagion.

What this means for portfolios

The first takeaway is that the VIX is not a proxy for your portfolio’s risk. A family heavily exposed to the Nasdaq-100, semiconductors or the largest AI beneficiaries could have suffered a sharp loss on Friday while the broad-market “fear gauge” remained below 20. This reflects the concentration risk we’ve been watching for the past few weeks.

The second lesson is that index-level performance can also obscure what is happening underneath the surface. The S&P 500 can remain close to a record while a very large percentage of individual stocks—or an entire former leadership group—are already experiencing bear-market drawdowns. Neither the index nor the VIX tells the whole story.

That is why the useful portfolio questions are more concrete. How much exposure do you really have to the companies and market segments driving actual portfolio risk? How much overlap exists between managers, funds and individual securities? Are recent winners now responsible for a disproportionate share of the portfolio’s potential drawdown? Could the family comfortably meet expected liquidity needs if the correction deepened? Those questions matter more than whether the VIX “should” have closed at 19, 25 or 30. These are all specific questions that LucenWorks was built to help us answer.

Ultimately, the third lesson is perhaps the most important: what feels like a bear market to many investors may actually be a change in market leadership. That does not make the losses unimportant. It doesn’t make them less painful. Nor does it mean the market is safe. But it is very different from a generalized liquidation.

The unusual relationship between QQQ and the VIX supports that point. Historically, a Nasdaq-100 decline of Friday’s magnitude has usually occurred alongside a substantially higher VIX. This time it did not, because the broader S&P 500 was experiencing something much less severe. And, indeed, the options market, sector dispersions, and the performance of the broader market are telling a broadly consistent story. The stress is real, but, at least for now, it’s also concentrated.

We are not prepared to forecast the market’s next leg from the evidence presented here. But, we can start with a more modest conclusion: the current selloff looks less like a rejection of equities generally than an overdue phlegmatic assessment of the heady expectations embedded in the market’s previous leaders.

And that speaks to one of the fundamental laws of investing: diversification matters. The winners of one phase of a bull market do not have to be the winners of the next. And sometimes what looks, from one part of a portfolio, like the beginning of a bear market is actually the market beginning to move on to the next thing.