Chip stocks have finally started to fall, say Goldman Sachs traders: cracks are appearing, and the market has underestimated the leverage behind the “frenzied chip speculation.”

Chip stocks have finally started to fall, say Goldman Sachs traders: cracks are appearing, and the market has underestimated the leverage behind the “frenzied chip speculation.”

```

The recent rally in chip stocks has weakened, and internal warnings have been issued within Goldman Sachs.

On Wednesday, Goldman Sachs trader Thilo Deller pointed out that the market's most exuberant sectors "are beginning to show cracks," with semiconductor stocks recently giving back some gains. Meanwhile, Shawn Tuteja, Goldman Sachs Managing Director, released a briefing, directly stating that the market may be seriously underestimating the systemic risk accumulating in leveraged ETFs within the semiconductor sector.

In his briefing, Tuteja said he currently holds a locally negative view on risk assets and specifically highlighted a dynamic he believes the market has not fully priced in. As chip stocks have surged, the scale of leveraged ETFs has expanded rapidly, resulting in a significant accumulation of "short gamma" exposure. If the trend reverses, mechanical deleveraging could trigger a chain reaction of selling.

Trigger for chip stock correction: Disturbance caused by Korea's policy statement

The direct catalyst for this round of semiconductor stock correction came from a policy statement in South Korea. The Korean policy chief publicly stated, Korea should design a national dividend system to return part of AI excess profits to the public, in order to prevent a K-shaped economy from intensifying social inequality.

This statement immediately triggered concerns in the market about the direction of AI industry profit distribution policy, causing chip stocks to fall. However, the official later clarified that his original intent was to use excess tax revenue rather than corporate profits to support redistribution policy. Despite this, sentiments in the semiconductor sector had already been affected, and some of the recent gains were given back.

Leveraged ETF scale surges; short gamma exposure doubles

Goldman Sachs data reveals a deeper structural risk. Since the end of March this year, the Philadelphia Semiconductor Index SOXX has gained over 70%, and the asset management scale of leveraged products tracking individual semiconductor stocks and semiconductor ETFs has surged.

According to Goldman Sachs estimates, these leveraged products hold a total long exposure in semiconductor stocks close to $100 billion. The operating mechanism of these products means they are naturally in a "short gamma" position—when underlying assets rise, they need to buy more; when they fall, they are forced to sell in order to maintain the set leverage ratio.

Tuteja estimates that, in the semiconductor sector alone, the dollar gamma exposure that these leveraged products need to reset daily is about $2 billion. In other words, if the semiconductor sector rises 1% in a day, the leveraged ETF system needs to net buy about $2 billion; if it falls 1%, it needs to net sell about $2 billion. Notably, the scale of this short gamma has nearly doubled in the past 6 to 9 months.

Options pricing fails to reflect deleveraging risk

Tuteja pointed out that compared to S&P 500 index options, semiconductor options pricing shows a significant anomaly: over the past year, the holding cost of semiconductor options has been significantly lower than S&P 500 options, especially in recent months as leveraged ETF scale expanded rapidly and spot prices consistently broke upward, making this divergence increasingly apparent.

More concerning, the left tail pricing of semiconductor options seems to completely ignore the gap risk of one-day sell-offs under deleveraging scenarios. Goldman Sachs regression analysis shows that, using SMH (semiconductor ETF) as a reference, 1-month 10-delta out-of-the-money put options are priced historically low versus at-the-money options. Similarly, regressing 1-month at-the-money options against three times the price of 10-delta out-of-the-money puts also shows tail risk is underpriced—this underpricing is especially prominent given the current scale of leverage in the system.

Tuteja concludes: If semiconductor spot prices fall, leveraged products will be forced into mechanical deleveraging, at a speed dependent on daily volatility—the greater the volatility, the larger the single-day deleveraging scale. And the current options market clearly underprices this gap risk.

Based on the above analysis, Tuteja recommends investors consider tail hedging in semiconductors and AI sectors to guard against potential risks caused by leverage accumulation within the system, especially as downside implied correlation has not fully priced in gap risk.

Goldman Sachs specifically favors holding SMH downside options, as well as downside protection for its AI leader basket GSTMTAIP. Additionally, Goldman Sachs has observed some clients choose to build their own baskets of semiconductor stocks. Because the implied volatility of such custom baskets is generally lower than SMH, holding costs in calm markets are lower, and if deleveraging occurs, their correlation with SMH is expected to remain high, thus achieving the hedging effect.

Risk disclosure and disclaimerThe market carries risks; investment requires caution. This article does not constitute personal investment advice, nor does it take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular situation. Invest accordingly at your own risk. ```