Source: Wall Street See

On August 29th, $NVIDIA (NVDA.US)$ Announced impressive second quarter financial report, achieving revenue of $30 billion, a year-on-year growth of 122%. However, Nvidia expects third quarter operating income to grow 80% to $32.5 billion, lower than the highest estimate of $37.9 billion. After the financial report was released, Nvidia's stock price plunged rapidly, falling below 7% at one point.

Morgan Stanley stock analyst Shawn Kim and his team released a report stating that this market performance indicates that the semiconductor industry has entered the late stage of the cycle and is about to peak.

Historically, there have been some signs when the industry enters the late stage of the cycle: despite the market sentiment transitioning from optimistic to euphoric, investors still realize the limitations of overvaluation and future growth potential, leading to a stock price correction. At the same time, the industry's revenue growth begins to slow down.

The semiconductor industry has entered the late stage of the cycle.

Both datacenter revenue and overall revenue of Nvidia in the second quarter exceeded expectations. However, Nvidia's stock price still plunged. Morgan Stanley stated that the semiconductor industry has shown obvious late-cycle characteristics - the market believes that the demand is still strong, and Nvidia's guidance for the third quarter is also positive. However, the market's reaction to the bullish news has been weak.

Morgan Stanley states that typically, the semiconductor industry has either high growth rates or negative growth rates, with very little moderate growth. The semiconductor industry reaches supply and demand equilibrium about every two years, but this equilibrium period does not last too long.

Currently, the semiconductor industry is facing a late-cycle, including the industry conditions and valuations of semiconductors. The market has different views on the future development of the semiconductor industry, balancing fears of the semiconductor cycle reaching its peak and fears of missing out on the AI opportunity.

In 2025, industry revenue growth may begin to slow down.

Morgan Stanley states that the market often anticipates the impact of new technologies before seeing their financial returns. During the Internet bubble period, although in many cases there was not a significant gap between the market's long-term forecasts and the eventual reality, stock valuations could not be maintained at high levels, and the sales multiples quickly returned to normal. The reason for the bursting of the Internet bubble was that the market had absorbed all the Internet infrastructure spending, but companies were unable to achieve high expected returns.

Expanding to the current generative AI. If the market realizes that the potential AI revenue is farther away than expected, then the market's premium on AI valuations may shrink. This is not a pessimistic view, but a more rational perspective, considering that many companies' AI projects are still in the pilot or concept verification stage. But eventually, artificial intelligence will create huge long-term investment opportunities, just like the previous paradigm shift driven by computers.

Morgan Stanley notes that although the current technology cycle may be different from before, there are also signs that the peak of the semiconductor industry cycle is nearing, starting from the fourth quarter of 2022. Although the current market demand for GPUs is obvious, the situation may not be as simple for the next stage of AI trades - the key point is whether AI demand will be unleashed ahead of schedule before 2025-2026.

Morgan Stanley states that historically, the market has been driven by the rate of change. Based on year-on-year sales/pricing growth, the peak of the semiconductor industry cycle is approaching, with global revenue pointing to the peak in the third quarter of 2024. Morgan Stanley believes that tech companies will continue to rise in 2025, and supply may not catch up with demand until the end of 2025, but stock prices will fluctuate with the rate of change, meaning that semiconductor companies will achieve profitability, but stock prices will not reach previous levels.

Over time, more value in the semiconductor industry will shift to the application layer, while the value of infrastructure will decrease. Although AI computing chips are currently in short supply, this trend will not last forever --- eventually, the supply of AI chips will catch up with demand, exhibiting cyclicality and it will be difficult to sustain the current growth rate.

The general expectation is that after 2025, the revenue growth of the semiconductor industry will begin to slow down.

The market sentiment is shifting from optimistic to euphoric.

One thing that remains constant throughout various cycles is human behavior. In the later stages of the industry cycle, investors tend to shift from optimism to euphoria. During this period, the main driving force behind stock increases comes to an end, last year it was interest coverage ratio, this year it is the profit momentum brought by artificial intelligence, which is accelerating and is expected to peak in the next few quarters.

According to Morgan Stanley, the current market psychology will soon shift from focusing on the drivers of the new cycle of growth to fearing the arrival of the cycle top. Looking back at the past 10 cycles, the SOX index has outperformed the S&P 500 index by an average of 13 percentage points in the "pessimistic period" following 3 months; during the approximately 3 quarters of the "doubt period," this outstanding performance expanded to 32 percentage points; in the typically 2 quarters of the "optimistic period," this outstanding performance further accelerated to 76 percentage points, but once entering the "euphoric period," the SOX index's average performance lagged behind the S&P 500 index by 13 percentage points.

Editor / jayden