The AI rally accelerated again last week, but several quantitative signals have already issued warnings: the Nasdaq Index showed a combination of "spot price rising and volatility also rising," with the VIX continuing to decline while the VXN rebounded significantly; the skew of U.S. tech stock options dropped rapidly, and the process of quantitative funds covering short positions is nearing its end. In other words, while the AI trade in U.S. stocks has not been declared over, the fuel for further gains from short covering is diminishing.
A core judgment in a Nomura research report released on May 11 is quite direct: "At least for U.S. stocks, the AI rally may be taking a breather." However, this framework does not place the reversal signal on the Nasdaq Index itself, but rather on the Korean stock market: if FOMO (fear of missing out) re-emerges in Korean stocks, it could indicate that the AI trade still has room for another leg of chasing gains.
The signals from the Korean market are clearer than those from the U.S. Non-resident investors who increased their exposure to Korean stocks during the AI rally from September to October last year recently sold to lock in profits as the market surged. Meanwhile, the KOSPI 200 also exhibited "spot price rising and volatility rising," but the call option skew did not follow suit. This does not resemble a volatility expansion driven by call option demand. Whether the foreign selling is bearish or simply making room for subsequent re-accumulation of long positions cannot be easily concluded at this point.
The clues from the Japanese market are more structural: AI and semiconductors still dominate foreign-demand stocks, but the value preference has not disappeared; instead, it remains more within domestic-demand stocks. The upward anchor for the Nikkei 225's floor has also been raised. Calculations based on component stock target prices have increased from the 60,000–61,000 range in late February to above 66,000 points. However, if the Nikkei falls below 61,500 points, pressure from CTA (Commodity Trading Advisor) position reductions and option dealers' gamma turning negative may occur simultaneously.
The AI rally in U.S. stocks is not over, but short-term squeeze fuel is nearly exhausted.
The option skew for U.S. tech stocks has fallen to near historical lows, returning to levels seen around October 2025. Here, skew refers to the difference between the implied volatility of 1-month 25-delta put options and 1-month 25-delta call options. A significant decline in skew means the premium for put protection relative to call options has been compressed, indicating a more crowded market pricing for tech stock upside.
The supply-demand dynamics are no longer as favorable as before. Quantitative funds' equity exposure has recovered to near neutral levels, and the process of forced buying and short covering is largely complete. If AI-related stocks continue to rise, the market can no longer primarily rely on the force of "shorts being squeezed out to buy."
It is important to note that positions involving investor categories such as quantitative funds, CTAs, and macro funds are model estimates, not actual measured holdings. This makes them more suitable as thermometers for marginal changes rather than precise position tables.
Foreign selling in Korean stocks may not be bad news.
The difficulty with the Korean stock market is that the same set of data can be interpreted in two directions.
On one side, non-resident investors are selling. From September to October last year, these funds increased their exposure to Korean stocks during the AI rally; recently, after the Korean market surged, they began to realize profits. Looking solely at fund flows, this is a cooler signal.
On the other side, the call option skew for the KOSPI 200 has declined significantly, indicating that the rise in volatility is not driven by demand for chasing gains via call options. In other words, the Korean market has not yet entered a typical "fear of missing out, scrambling for call options" state. If the AI trade is to continue, the Korean stock market actually retains the potential for FOMO to re-emerge.
Macro assumptions still cannot bypass the Strait of Hormuz. As long as the strait remains obstructed and the U.S. and Iran still disagree on ceasefire conditions, the AI-dominated market environment may last longer than expected.
The real risk lies on another front: a resurgence of inflation concerns forcing global central banks to adopt a more hawkish stance. The event premium for the U.S. CPI release on the 12th this week remains low; the market has not yet paid a high insurance premium for this risk.
The stronger the AI trade, the more Japanese value stocks need to exclude foreign-demand exposure.
U.S. investors remain bullish on the AI trade, with AI-related ETFs continuing to see inflows in the U.S., while Europe and other regions are experiencing outflows. This regional sentiment divergence has already transmitted to stock selection in the Japanese market.
During the reversal from November to December 2025, Japanese AI and semiconductor-related stocks with higher U.S. ownership ratios showed more resilience; stocks with higher European ownership ratios continued to underperform. A high U.S. ownership ratio, in a sense, acted as a buffer against reversal risk.
However, if tech stocks experience a more thorough reversal, the value style may still re-emerge as the main theme. Over the past year, the Japanese market has roughly followed a rhythm: when tech stocks drive index gains, value factor performance is weaker; when tech stock momentum pauses, the value factor recovers.
A more actionable approach is not to broadly buy value, but to tilt towards value within domestic-demand stocks: go long high B/P (Book-to-Price) stocks among TOPIX 500 domestic-demand stocks and short low B/P stocks. This pair still shows slight excess returns currently, similar to the period from September to October last year. Most AI and semiconductor stocks fall under foreign-demand assets, and the supply chain risks brought by the Middle East situation also primarily weigh on foreign-demand stocks. After excluding foreign-demand exposure, the value preference within Japanese stocks still exists.
If the Strait of Hormuz normalizes, lagging stocks may still have room for catch-up.
If the situation in the Strait of Hormuz normalizes, stocks that previously lagged due to Middle East risks may be re-priced. Taking the TOPIX 500 as a sample and dividing it by stock performance from February 27 to April 7, the performance gap between the bottom 20% "losers" and the top 20% "winners" has begun to narrow, but it is not yet halfway to full recovery.
Historical supply chain shocks suggest two paths.
After the 2025 tariff shock, the market quickly shifted from initial selling to betting that Trump would back down, the so-called "TACO trade." After the index itself fully recovered that time, it took about two months for the performance gap between winners and losers from the shock to completely converge. After the 2020 pandemic shock, the performance gap between winners and losers was largely erased about six months after the index recovery.
The 2022 Russia-Ukraine conflict was different. Crude oil prices stabilized gradually after the initial spike, but as global central banks raised interest rates and economic growth momentum continued to weaken, the performance gap between winners and losers only narrowed to about half before stalling. The losers in that shock were mostly foreign-demand stocks, with a median domestic sales ratio of 37%; the winners were mostly domestic-demand stocks, with a median domestic sales ratio of 84%. If the current Middle East situation ultimately drags on global growth, the catch-up of lagging stocks might also stall halfway.
The Nikkei's upper anchor has been raised; 61,500 points is a key support level.
The Nikkei 225 briefly broke through 63,000 points this Monday. The index's rise is not entirely driven by sentiment; the concentrated contribution from heavyweight stocks is evident: the top 5% of stocks by weight alone contributed about 5,000 points of gains, many of which are AI-related stocks relatively less affected by the Middle East situation.
The floor calculation based on the 12-month forward target prices of Nikkei 225 components has risen from the 60,000–61,000 range in late February to above 66,000 points. This level acts more like the current market's upper anchor and may also serve as a short-term resistance reference.
Fund sentiment has not shifted significantly. European investors are not rushing to restart the "flee from dollar assets" trade; macro hedge funds are taking a wait-and-see approach towards both U.S. and Japanese stocks; CTAs have only slightly increased their long exposure to Japanese stocks.
The level that truly needs watching is 61,500 points. Once the Nikkei 225 falls below this level, CTAs may begin to reduce their overall net long positions. Similarly, below 61,500 points, option market makers' gamma exposure is more likely to turn from positive to negative. If the index continues to rise and the call option skew strengthens, market makers might also turn to negative gamma during the stock price ascent, which would amplify market volatility more quickly.
Comments