S&P and other weights plummet 5%: Damo (MS.US) warns of repeated AI bubble

According to Woofun AI, Morgan Stanley (MS.US) revealed in a global strategy report released on October 9, 2026, that the S&P 500 index showed extreme structural differentiation in September: the market capitalization weighted fell only slightly by 0.5%, while the equal-weighted scale fell sharply by 5%.

Resilience under this appearance does not stem from the overall strength of the market, but rather the result of the concentration of capital into high-quality assets, which Damo characterizes as a “shift to quality.” Although the AI and semiconductor sectors are still supported by capital expenditure expansion, high valuations and rising interest rates are eroding the sustainability of the market. Institutions are recommending a gradual reduction in thematic concentration and a tactical shift to low volatility and value stocks.

Deconstructed from the microstructure of the market, the head effect is the core variable that conceals widespread weakness. The top six contributing stocks — Meta (META.US), Apple (AAPL.US), AMD, Nvidia (NVDA.US), Intel (INTC.US), and Micron (MU.US) — collectively contributed about 190 basis points of positive earnings, while the remaining constituent stocks dragged down about 240 basis points in total.

This differentiation has led to a steady rise in US stock quality factor returns, and low volatility factor valuations are already at the 98th percentile in history. Notably, the quality factor of the Japanese market began to strengthen in late September, while the value factor weakened during the same period, reflecting investors' concerns about financial fragility that overwhelmed their preferences for valuation in an environment of rising interest rates.

According to data compiled by Woofun AI, the historical rules from 2011 to 2025 show that foreign investors usually sell the Japanese stock market net in August and September, then switch to net purchases in October and beyond. In the first half of 2026, foreign capital purchases of Japanese AI-related stocks provided key support. Daimo expects that with the end of the net sales cycle in August and September, the return of foreign capital from October will once again form support. This seasonal pattern may be repeated.

In terms of risk depth benchmarking, the historical repetition of the AI valuation bubble and the Internet bubble is becoming more and more obvious. Although Morgan Stanley recommended holding AI and semiconductor stocks in its August 14 report “The Second Half of the AI Market,” it emphasized the need to avoid further focus on the subject. As early as May 14, the report “Are we going back to 1999?” China, Damascus pointed out that there are many similarities between the AI-driven market and the internet bubble. The July 6 report further analyzed the degree of overlap between the two causes. Currently, the median forward price-earnings ratio of core AI and semiconductor stocks is about 32 times. Although it is down from 38 times three months ago, it is still far above the average of about 19 times over the past decade. The comparison between the SOX index and the trend during the internet bubble shows that the two are highly similar in terms of rapid rise, fall, and rebound patterns.

The more critical variable is supply-side pressure. IPOs, new stock offerings, and secondary offerings may diversify investment capital. Just as the number of shares available for sale surged after the lockdown period expired in March 2000, the market is currently facing similar potential supply expansion.

Furthermore, when the internet bubble peaked in March 2000, it was in the middle of the Fed's interest rate hike cycle, and the Federal Reserve recently turned to interest rate hikes. The tightening monetary environment may weaken market sustainability. The controversy over excessive investment in data centers and the credit status of AI companies has also increased uncertainty about the return on investment in the medium term.

In the face of the accumulation of concentration risks, the tactical configuration needs to shift to a defensive one. Morgan Stanley suggests selectively maintaining AI and semiconductor positions, but the concentration of topics must be gradually reduced to increase resilience to market pullbacks. Low-volatility stocks are preferred because they are suitable as defensive allocations and investor positions are relatively light;

At the same time, financially sound value stocks are another important allocation line. Although Japan's value factors have been squeezed by rising AI and quality stocks since late September, and the financial strength of value stocks is generally weaker than growth stocks, when interest rates rise, if AI stocks consolidate and preferences spread, the relative performance of value stocks may improve. The year-end value factor has a seasonal upward trend, and the mid-term allocation value is still there. Damo reminds investors to prepare for a possible transition from December to early 2027. Reducing concentration and increasing the allocation of low volatility and financially sound value stocks is the core direction of the current strategy. Changes in foreign capital flows and profit forecasts will be a key observation point in the future.