Why the AI Selloff Mattered More Than the Election Bounce
A Goldman Sachs strategist called the AI trade “sold to death” on November 4, 2026, while finding a silver lining in Democrats’ sweep of the midterm elections, according to a MarketWatch article distributed by Dow Jones.
The election is the political backdrop. Democrats won control of both the Senate and the House that day, according to Reuters, which also reported that global stocks and the dollar recovered on November 4 from lows hit overnight. Dow E-mini futures, electronically traded contracts that track the Dow Jones Industrial Average and trade ahead of the regular stock market session, were up 436 points, or 0.84%, as investors assessed the victories (Reuters).
The rebound came after a rough stretch for technology shares. Tech stocks sold off heavily ahead of the vote amid concerns about overspending on artificial intelligence, according to Bloomberg. Read alongside the strategist’s remark, that sequence matters.
In my view, the sequence matters more than the timing of the vote. The selloff was about spending, not the election. The election simply landed on top of it.
A June 25, 2026 Goldman Sachs report titled “An AI Job Apocalypse?” describes continued investment in AI, research and development compute capacity, and talent, while overall headcount contracts (Goldman Sachs). In other words, the firm sees capital and talent being redirected rather than simply cut.
An August 6, 2026 report says AI capital expenditure has contributed to higher server shipments and data center capacity, and notes that Alphabet raised its 2026 capex range by $15 billion (Goldman Sachs). Capex, meaning spending on long-lived assets such as servers and buildings, is the line item that matters most for the bear case on the trade.
For the bear case, capex is the number to watch. Spending on servers and buildings is where a slowdown would show up first, and it is the figure most likely to move the trade.
Older Goldman work points the same direction. Its 2026 Global M&A Outlook, published December 1, 2025, argued that the amount of compute required by AI has ignited a surge in demand across digital infrastructure (Goldman Sachs).
The demand thesis predates the current selloff, which matters if the recent drop is read as a change in the underlying story rather than a shift in sentiment.
Demand has a supply-side mirror. Bloomberg reported on November 4 that Chinese AI developers ordered about 6 million of Nvidia’s RTX Pro 5500 chips, yet expect to receive only some tens of thousands (Bloomberg).
The gap between an order book and a delivery schedule is wide enough to matter for anyone modeling chip revenue. Orders signal intent, and allocation is a separate constraint. Anyone who reads the 6 million figure as revenue should first ask how much of it ships.
The useful distinction for savers and investors is between what moved and what was merely priced. A 0.84% gain in index futures on election night reflects a repricing of expectations over a few hours, not a verdict on AI earnings. The strategist’s silver lining is an interpretation of the political outcome. The Goldman material points to a trade where capex and headcount are moving in opposite directions, and where the largest near-term uncertainty is whether order volumes turn into shipments.
Whether a Democratic-controlled Congress changes the policy backdrop for AI and chips is a question the market will have to price over time. What is known is narrower: the trade was under heavy pressure before the vote, the vote went Democrats’ way, and futures recovered part of the overnight loss. What happens next will depend on spending data and delivery schedules, which arrive more slowly than a single session’s headlines.
