S&P 500 futures and Nasdaq 100 futures are sharply lower heading into Friday’s session after high-profile disappointments from Amazon and Intel added to the recent run of underwhelming results from tech giants.
As we flagged on Wednesday, Google was the most heavily bought Magnificent 7 hyperscaler by Robinhood traders heading into earnings, making its record one-day decline an acute ouchie for the retail crowd.
But retail traders weren’t hitting stop losses, but rather averaging down into the Gemini maker. Net daily purchases on Thursday were the highest on record, and it’s not particularly close:
As a subsidiary of Robinhood, Sherwood Media is restricted from writing about any company in which Robinhood is or was a selling group member of the IPO during the regulatory "quiet period" for that company.
Spooky szn
In previous lives — for over a decade now — I’ve had the opportunity to both assemble and offer contributions to annual editions of The Charts That Scare Wall Street.
(I am pleased by my musing from 2021; you could do far worse than flagging the potential that “a disorderly rise in real yields could challenge equity valuations at the index level, and the growth-oriented, expensive segments of the market in particular,” heading into the bear market that started early in the next year.)
The most unique element of the trading environment in 2026 is what worries me the most:
Based on data going back to 2014, the average one-month implied volatility of single stocks within the S&P 500 has never run as hot as it has this year. At the same time, implied correlations have never been this low!
As a friend of the program has been fond of posting, “the bubble isn’t AI, it’s the dispersion trade.”
I agree wholeheartedly, while cautioning that this is not exactly the most actionable insight. Bubbles involve extrapolating conditions and growth trajectories that cannot last in perpetuity.
Stocks trade in concert when there is an event or suspected tipping point in the macroeconomic outlook that drives a common reason to question the earnings outlook, valuations, or both.
It is not brave to suggest that something important will happen at some point in the future. And it is only slightly less not brave to point out that the AI boom’s increasing role in economic activity, profit growth, and capital markets issuance means that shifts within this trend alone can spur a correlations-to-one environment.
Of course, this hasn’t happened yet; if anything, the pillars on which market dispersion is built are becoming more blindingly obvious.
Apple’s completely uncorrelated with the Nasdaq 100 right now, for instance, simply because its current operational performance is the least tethered to the AI boom of all the trillion-dollar tech megacaps. But even within Mag 7 hyperscalers, there are big gaps between quarterly performance despite these companies all tying themselves to the same mast.
Moreover, lackluster results from some companies linked to the real estate industries aren’t being met with heavy selling — perhaps because all the bad news was in the price, or maybe just because those stocks’ behavior has been the opposite of whatever AI bellwethers are doing.
The upshot is simply that absent the average stock becoming less volatile in the face of a common shock — which would be unlikely, based on history — any common factor that influences market-wide earnings or valuations would have a mechanically massive impact on index level swings.
To fit the holiday spirit, this would be an Undertaker waking up gif moment, but for index level vol.
Sales fails
“Sales don’t lie” cuts both ways.
Last reporting period, looking at top-line trends for Corporate America was a good way to cut through the noise, since massive paper gains on investments made by megacap tech companies were exaggerating the strength of bottom-line results.
S&P 500 aggregate revenue surprises were above 3%, a multi-year high, reaffirming underlying operational strength. So far, aggregate sales surprises have slipped to just 1.6% this reporting period, which would be the lowest since the Q1 2025 reporting season.
Some of the Mag 7 hyperscalers that have reported in recent days have contributed to this drop-off. Sales for Google and Amazon came in modestly below the consensus estimate for the first time since they reported their Q4 2024 and Q2 2024 results, respectively.
There’s room for nuance, as well as the rebuttal that these misses are rounding errors in the grand scheme of things.
But top-line softness coupled with huge capital outlays accentuate the two critical spots of bother investors have had with the AI trade: a lack of confidence in the ROI on this capex, and how much external capital will be needed to fund future investment plans.
Such concerns become more acute in the event that these initiatives need to make up for signs of fraying in the core business where a company’s held a dominant market position for decades (see: Google, search).
Au-racles
I wanted to react to a typically excellent blog from FT Alphaville’s Toby Nangle earlier this week on the decoupling of the gold price from its chief macro anchor.
Historically, the case for valuing gold lies somewhere between “ou, shiny rock” and its correlation with an asset that inflation can’t debase. The thinking goes: Gold has no yield; TIPS offer an inflation-protected yield, so gold should do well when TIPS yields are lower/negative and vice versa.
As Nangle observes, Russia’s invasion of Ukraine caused an initial big break in this relationship. If you are a nation who thinks they might be running afoul of the international order and subject to US sanctions, you might want to diminish your reliance on US dollar assets.
However, the connection between gold and TIPS reasserted itself this year through July, per Nangle, but has since broken down yet again.
What gives?
Well, the answer, in our view, lies in the increasing role of retail speculation as a driver of gold prices over tactical investment horizons.
As we discussed in early August as gold was making a renewed push higher:
The intense rally in momentum stocks, which really kicked into gear at the end of March, came after the sharp rallies in precious metals had petered out, but also at the time these assets — and risk appetite more broadly — began to show some signs of resilience.
Recently, we have precious metals regaining their mojo after momo stocks collapsed, then carved out a floor.
I don’t think that’s a coincidence.
At the time, Macro Hive strategist Viresh Kanabar spotlighted a turn in speculative demand in both China and the US as a driver of strength in precious metals.
It’s not as compelling an explanation, but given we’ve seen precious metals enjoy several spurts of momentum over the past year that have all the hallmarks of retail excitement, I think it’s safe to say the valuation equation for gold now looks a little something like “real yields + geopolitics + vibes,” with vibes having some increasing – though still variable — explanatory power.
Seen on Socials
How the turn tables — earlier this year, this chart shared by @malinvested on X was initially being used to justify the cratering of software stocks on fears of AI-induced disruption!
