CITIC warns 2028 may be deadline for US tech bull run

The Nasdaq Composite faces more than 10% downside risk from current levels, with 2028 emerging as a potential endgame for the AI bull market, CITIC Securities said.

"Short term, we flag pullback risk — the Nasdaq may fall more than 10% from its highs," said Qian Wei, head of overseas and macro assets at CITIC Securities. "Medium term, 2028 could be a major deadline for US tech."

The warning comes as margin debt has risen 25% above the index's advance, a divergence that preceded the 2000 dot-com crash, the 2008 financial crisis and the 2022 bear market. China's AI models have narrowed the performance gap with US frontier systems to about 5% from nearly 300% at their widest, threatening the capital expenditure thesis underpinning semiconductor stocks. The tech sector's forward valuation has failed twice to break above 30 times earnings since mid-2024, stalling near 25 times in the second quarter.

The analysis identifies four potential paths to a tech bull market peak: a collapse in AI capex as cheaper Chinese models prove competitive, a rotation into traditional economy stocks, political backlash against Big Tech concentration ahead of the 2028 US election, or a Federal Reserve tightening cycle. For the bull market to survive past 2028, AI applications must deliver broad-based productivity gains that justify the spending, the report said.

Leverage and China AI Threaten the Capex Thesis

Margin debt growth outpacing index gains by 25 percentage points has historically signaled trouble. The last three instances of this divergence — 2000, 2008 and 2022 — all preceded significant market declines, the report said.

The rapid narrowing of the AI capability gap between US and Chinese models poses a direct challenge to the capital spending narrative. Kimi K3, developed by Chinese startup Moonshot AI, has closed the performance gap with US frontier models to about 5%, according to industry benchmarks. Chinese models also operate at significantly lower costs, including equipment and energy consumption.

This dynamic raises a fundamental question: if US frontier models cannot sustain a technological lead, why should investors continue to fund massive capital expenditures? Meta's announcement in early July that it would sell computing capacity triggered debate about whether AI infrastructure is being overbuilt. DeepSeek, another Chinese AI model, caused a sharp Nasdaq selloff in early 2025.

The concern has already manifested in stock prices. Since June 22, Micron Technology has fallen about 30%, Advanced Micro Devices has dropped about 10% and Broadcom has slipped about 5%, according to UBS. All three continue to show improving fundamentals under the bank's HOLT framework, suggesting the selloff may have created buying opportunities — but only if the capex thesis holds.

2028 Election Adds Political Risk to Valuation Concerns

The technology sector's forward price-to-earnings ratio has tested the 30 times level twice since mid-2024 and failed both times. After compressing during the first quarter of 2026, valuations rebounded to about 25 times in the second quarter before retreating again. During this period, capital expenditure and earnings growth continued to be revised upward, but the market refused to pay higher multiples.

The K-shaped nature of the AI rally — where gains concentrate among a handful of mega-cap companies while the broader economy lags — creates a political vulnerability. If this divergence persists for another one to two years, anti-Big Tech sentiment could become a campaign issue in the 2028 US presidential election. Historical parallels include the 2000 antitrust case against Microsoft, which contributed to the dot-com crash, and the 2008 campaign where Barack Obama called for stronger financial regulation.

"By 2028, if campaign platforms include stronger AI regulation, antitrust action against Big Tech or higher taxes on technology companies, markets trading near all-time highs would struggle to dismiss these as noise," the report said.

A rotation from technology into traditional economy stocks could also end the tech bull market without a crash. This occurred in 2000, when the Nasdaq fell more than 50% over 12 months while the Dow traded sideways for nearly a year. A similar rotation happened between October 2025 and early 2026, when the Nasdaq corrected while the Dow hit new highs. That rotation reversed when the economic recovery failed to materialize. For a sustained rotation to occur now, traditional sectors would need to show genuine fundamental improvement — a difficult outcome under high interest rates that continue to suppress housing, industrial activity and consumer spending.

This article is for informational purposes only and does not constitute investment advice.