The NASDAQ Trendline That Will Decide Whether the Semiconductor Selloff Spreads

The semiconductor selloff is no longer just a reaction to crowded positioning or isolated earnings disappointments. Gareth Soloway’s central read from this morning’s My Trading Game Plan was that the assumptions supporting the artificial intelligence trade are beginning to face a more serious test. Rising competition, peak-cycle margins, and forced deleveraging are changing how the market values the companies that led the previous advance. The level that now decides whether this remains a concentrated industry unwind or develops into a broader technology correction is the multi-year trendline being tested by the NASDAQ 100.

That distinction is important because the damage beneath the major indices is already substantial. Several AI-adjacent companies have lost roughly half their value within weeks, even while the S&P 500 has remained comparatively stable. The headline is that semiconductor stocks are falling. The deeper signal is that capital is beginning to separate companies with durable earnings support from those whose valuations depended on unusually high margins, uninterrupted demand, and continued investor enthusiasm.

Semiconductor Margins Are Facing Their First Real Test

For much of the AI-led advance, semiconductor companies were treated as though exceptional growth and unusually strong margins could continue without meaningful competition. Gareth’s argument was that this expectation ignored one of the most reliable forces in business: high returns attract new supply.

When a product produces margins that are difficult to find elsewhere, competitors have a powerful incentive to enter the market. That process may take time, especially in technically complex industries, but it eventually affects pricing power. The current pressure from Chinese semiconductor manufacturers suggests that this competitive phase is moving from a theoretical future risk into a factor the market must begin pricing today.

Chinese companies are making progress in areas previously dominated by Western manufacturers, including lithography equipment and memory chips. The source article highlights Shanghai Yulin Cheng as a developing competitor in lithography systems and CXMT as a growing memory-chip producer challenging companies such as Micron and Sandisk. Even when those products do not yet match the highest-end technology available from established leaders, the direction matters because expanding supply can weaken the assumption that current profit margins will remain intact indefinitely.

This is where semiconductor valuation becomes counterintuitive. A low price-to-earnings ratio can look attractive, but cyclical companies often appear cheapest when their earnings are near a peak. If margins are temporarily elevated, the denominator in the valuation ratio is also elevated, making the stock look less expensive immediately before profitability begins to weaken.

The reverse can also be true. Cyclical semiconductor companies may appear expensive, or may not have meaningful earnings at all, near the bottom of the cycle. That is often when excess supply has already been absorbed, expectations have been reset, and future margin expansion becomes possible. The lesson is not that investors should automatically buy high-multiple chip stocks. It is that a low multiple alone does not prove a semiconductor stock is undervalued.

The market is beginning to reassess these companies through that cyclical framework. What looked like a permanent AI earnings boom is being tested by the same forces that have shaped earlier semiconductor cycles: new competition, expanding capacity, lower prices, and pressure on margins.

The NASDAQ Trendline Is the Market’s Decision Point

The semiconductor decline becomes a broader market problem only if it begins damaging the structure of the major technology indices. That is why Gareth focused on the long-term trendline now being tested by the NASDAQ 100.

The line connects a major high from the 2021 bull market with a later pivot high in October 2025. Price came down and tested that structure as the NASDAQ’s decline from its recent peak approached 9%. The market can trade through a trendline intraday without confirming a structural break. The closing price is what determines whether buyers successfully defended the level or whether the broader trend has begun to change.

A close on or above the trendline would keep the structure intact and favor a multi-day relief bounce. That would suggest the semiconductor damage remains severe but has not yet spread far enough to break the broader technology trend. A decisive daily close below the line would carry a different message. It would confirm that selling pressure has moved beyond individual AI names and into the index structure itself.

The next major downside area would then sit near the former highs established during 2025 and early 2026. Based on Gareth’s framework, that could produce another 1,500 to 2,000 points of downside and take the total NASDAQ decline toward approximately 15% from its peak. The exact outcome is not predetermined, but the trendline gives traders a clear way to separate an oversold correction from a more serious technical breakdown.

This is the most important level in the article because it organizes the rest of the market analysis. Semiconductor stocks have already experienced crash-like declines. The NASDAQ has not yet confirmed that the broader technology trend is broken. The closing action around this line will determine whether the current weakness remains concentrated or becomes the next leg of a larger market correction.

Leverage Is Accelerating the Decline

The speed of the selloff cannot be explained by weaker fundamentals alone. Gareth also pointed to the unwinding of leverage accumulated during the euphoric stage of the AI rally.

When prices rise consistently, leverage can appear to improve returns without adding immediate risk. Traders may control several dollars of stock for every dollar of their own capital, while institutions can increase exposure because volatility appears low and market leadership appears dependable. The risk becomes visible only after the trend reverses.

Once heavily owned stocks begin falling, margin calls and risk limits force investors to reduce positions. That selling is not always based on a fresh opinion about a company’s long-term value. It can be mechanical. Falling prices require additional capital, and investors unable or unwilling to provide it must liquidate. Those liquidations create further declines, which generate more margin pressure and another round of selling.

Corning illustrates how quickly that process can unfold. The company plays an important role in AI infrastructure because its fiber-optic products are used in data-center construction. Despite reporting solid earnings, slightly weaker forward guidance triggered a sharp premarket decline. According to the source material, the stock had fallen roughly 56% from its recent high in less than a month.

Sandisk experienced a similarly violent drawdown, losing approximately half its value over roughly the same period. Moves of that size in companies with large market capitalizations justify describing the price action as a crash within those individual stocks, even if the broader index has not experienced the same percentage decline.

The opportunity for disciplined swing traders appears when forced selling reaches major technical support. Gareth identified $120 as the key support area for Corning, with a hold opening the door to a retracement toward $148. The point is not that every stock down 50% should be bought. The edge comes from waiting until extreme downside momentum reaches a level where prior price structure supports a probability-weighted bounce.

Several levels in the original source material for Sandisk and Micron appear internally inconsistent and should be verified before publication. Those figures are not necessary to the central thesis. The more useful takeaway is that severe declines are approaching longer-term support zones, where traders can evaluate whether forced liquidation is becoming exhausted.

Rotation Is Happening Inside Technology

The weakness in semiconductors does not mean capital is leaving technology uniformly. Beneath the index, money appears to be rotating away from hardware and into software companies that were already heavily discounted.

Microsoft, Meta, and Oracle were among the names showing improving relative strength as semiconductor companies continued to fall. Oracle provides the clearest example of how far expectations had already been reset. After declining from approximately $250 to a recent low near $115, the stock began forming a technical setup that Gareth believes favors a snapback toward $140.

That rotation matters because it shows the market is not abandoning growth assets indiscriminately. Investors are distinguishing between groups. Semiconductor companies that benefited most from the AI infrastructure narrative are being repriced as competition and margin concerns increase, while software companies that had already experienced deep declines are beginning to attract capital.

This internal rotation also helps explain why the major indices have not fallen as sharply as some individual AI names. Strength in software and other large-cap companies can partially offset weakness in hardware. The NASDAQ trendline therefore remains the better market-wide signal. As long as the index holds that structure, the market may be rotating rather than fully breaking down.

Apple sits outside that rebound framework. Gareth’s chart read suggests the stock becomes more attractive as a potential short setup if it extends above $340. That reinforces the importance of treating each chart according to its own structure rather than assuming every large technology stock should respond the same way to sector rotation.

Bitcoin Is Showing the Most Important Relative Strength

Bitcoin may be providing the clearest evidence that this is not yet a broad liquidation across every risk asset. Despite the sharp declines in technology and semiconductor stocks, Bitcoin has maintained its structure comparatively well.

Since June, Bitcoin has outperformed the NASDAQ, S&P 500, gold, and silver based on the source article’s comparison. It remains below its recent highs, but it has not followed AI-related stocks into a similar collapse. As long as Bitcoin continues holding its key pivot support areas, Gareth’s technical framework leaves room for another move toward the $71,000 to $72,000 zone.

That relative strength is more useful than making a broad claim that Bitcoin has permanently decoupled from technology. One period of outperformance does not prove that the asset will behave independently during every equity decline. It does show that sellers are currently treating Bitcoin differently from the most crowded parts of the AI trade.

For macro traders, that creates an important confirmation signal. If the NASDAQ loses its long-term trendline while Bitcoin continues holding its structure, the divergence becomes more meaningful. It would suggest capital is not simply reducing all risk exposure. Instead, investors may be rotating toward assets whose positioning, supply dynamics, or institutional use cases differ from those of highly valued semiconductor companies.

The Bottom Line

The semiconductor selloff is exposing the weaknesses that were hidden during the strongest phase of the AI rally. Exceptional margins attracted competition, high earnings made cyclical valuations look deceptively cheap, and widespread leverage increased the speed of the decline once the trend reversed.

The market-level decision has not yet been made. That decision sits at the long-term NASDAQ 100 trendline. A daily close on or above it keeps the broader structure intact and favors a relief bounce, even if individual semiconductor stocks remain damaged. A decisive close below it would confirm that the weakness has spread beyond a crowded sector trade and into the larger technology trend.

Software rotation and Bitcoin’s relative strength show that capital is still distinguishing between assets rather than selling everything equally. The market is not reacting to one earnings report or one headline. It is repricing where future margins, positioning, and technical structure still justify risk.

That is the framework traders should carry forward. The semiconductor crash explains where the pressure began. The NASDAQ trendline will show whether it spreads.