Why Did US Stocks Plunge? Nvidia, SanDisk and Micron Slump as AI Trade Suffers Worst Selloff Since 2020

TradingKey - U.S. stocks suffered one of their most brutal selloffs in recent years, with the AI supply chain hit hardest. On Thursday local time, the Philadelphia Semiconductor Index plunged about 12.3%, marking its largest single-day drop since March 2020; the Nasdaq 100 Index fell about 6.4%, the S&P 500 slumped 4.4%, and the Dow Jones Industrial Average dropped more than 1,800 points. As Nvidia (NVDA), memory chip, and data center-related stocks tumbled across the board, market confidence in the outlook for AI investments was visibly shaken.

Just one trading day after celebrating the U.S. midterm election results, the market abruptly reversed course. After Democrats secured control of both chambers of Congress, investors began reassessing potential changes in data center approvals, electricity costs, environmental restrictions, and AI infrastructure regulation. Policy risk was merely the fuse; what truly magnified the selloff was the excessive concentration in the AI trade, elevated earnings expectations, and the forced liquidation of leveraged positions.

Nvidia and Micron Plunge as Semiconductor Stocks Become Center of Storm

On Thursday, the most pronounced sell-off was concentrated in core segments of the AI supply chain. The Philadelphia Semiconductor Index tumbled about 12.3% in a single day, not only far outpacing the declines in the S&P 500 Index and the Nasdaq 100 Index, but also marking its worst single-day performance since the pandemic-induced market panic in March 2020.

Among them, Nvidia became the focal point of market attention. As a core beneficiary of the AI infrastructure investment cycle, Nvidia had previously been driven by expanding capital expenditures from major cloud service providers. However, as the market began to question the sustainability and future returns of AI buildouts, the world's leading AI chip supplier could not remain unscathed. On Thursday, Nvidia's shares fell about 9.3%, shedding roughly $447 billion in market capitalization in a single day, in one of the largest single-day market cap losses on record.

Memory chipmakers suffered even steeper declines, with SanDisk (SNDK) shares plunging about 22%, while Micron (MU) and Western Digital (WDC) also fell roughly 17% each. This performance reflects investors reassessing the boost that AI infrastructure expansion provides to the memory industry. Previously, sustained growth in AI server demand drove up market momentum for products such as HBM, DRAM, and NAND, prompting repeated upward revisions to earnings forecasts for related companies. However, as the market begins to worry that the pace of data center construction may slow, whether future demand for memory chipmakers can sustain high growth has become a new question.

More importantly, market concerns are no longer limited to AI chip demand itself. As data center projects continue to scale up, land, power, facilities, and supporting infrastructure could gradually become construction bottlenecks. If data center facilities and power supplies fail to expand as expected, even if chip demand remains strong, related companies could still face shifts in order delivery pacing and lower-than-expected capacity utilization.

Election Results Are the Trigger, Crowded Trades Are the Underlying Cause

With Democrats gaining an advantage in the midterm elections, the market immediately began assessing potential policy changes facing U.S. AI infrastructure development. Regions with concentrated data center and power projects, such as Texas and Ohio, came under particular scrutiny. Investors grew concerned that if local governments raise project approval thresholds or tighten oversight of land use, grid load, and environmental impacts going forward, data center construction could slow, dampening upstream chip procurement and related equipment orders.

However, the election results were more of a catalyst triggering the market pullback than the sole reason for the sharp drop. Stocks across the AI value chain had already run up substantial gains, and market expectations for future revenue growth, margin improvement, and order fulfillment were running high. When policy uncertainty began shaking the underlying growth thesis, investors quickly pared back risk exposure, and the selloff broadened from semiconductors to cloud computing, storage, power, and data center construction.

Meanwhile, leveraged trading and systematic strategies further exacerbated the decline. As stock prices continued to fall, certain margin accounts breached risk limits and were forced to liquidate positions to meet margin requirements. Volatility-control funds and other quantitative strategies may have also trimmed holdings due to the sharp spike in market volatility, creating passive selling pressure. Institutional de-risking compounded by retail selling subjected the market to intensified downward pressure.

The VIX, known as Wall Street's 'fear gauge,' rose to as high as 28.36 before closing at 28.18, up nearly 70% on the day. The sharp spike in short-term volatility indicated that investors were aggressively purchasing downside protection, rather than immediately buying the dip as they had during numerous past pullbacks.

As selling pressure continued to spread, the correction ultimately turned into a broad market flush of risk. Sectors across the S&P 500 Index closed broadly lower, with banks, industrials, small caps, and typically defensive consumer staples and utilities also taking a hit. This signaled that the market selloff was no longer limited to a reassessment of AI industry prospects, but had instead transformed into a concerted drive by investors to trim overall risk exposure, further deepening the market's decline.

Is the AI Bull Market Over? Market Needs to Seek Growth Certainty Again

Both the Nasdaq 100 Index and the Nasdaq Composite Index have fallen below their 200-day moving averages, a level widely regarded as a key dividing line for medium- to long-term trends. Meanwhile, the Philadelphia Semiconductor Index has broken below the upward structure formed since its 2025 lows, signaling that the market will require stronger fundamental support to rebuild confidence in the AI supply chain.

From an industry perspective, AI computing demand has not immediately vanished simply because of the stock market pullback. Large model training, AI inference, data center expansions, and enterprise-grade applications still require substantial computing and storage resources.

However, capital markets are altering how they value this industry. In the past, investors were willing to pay a high premium for long-term growth; now, they are focusing more on whether capital expenditures can translate into revenue, profit, and free cash flow.

For Nvidia, Micron, SanDisk, and data center infrastructure providers, future stock performance may no longer depend solely on the scale of AI demand, but also on whether customer capital expenditures can be sustained, order deliveries proceed smoothly, and profit margins can be maintained.

Meanwhile, the market also needs to assess how much of this downturn stems from shifts in fundamental expectations and how much is merely a short-term shock caused by leveraged position adjustments and forced liquidations. If it is primarily driven by deleveraging, the stock prices of certain quality companies could see a rebound as positions are gradually cleared out; however, if data center construction, AI capital expenditures, and earnings forecasts are indeed beginning to be revised downward, this correction could evolve into a more prolonged valuation reset.

AI infrastructure will likely remain a key investment theme over the next few years, but after this round of sharp sell-offs, market expectations regarding growth quality, earnings delivery, and cash flow have clearly become more demanding.