The 'semiconductor peak theory' resembles a zombie; it refuses to die down. Each time signals emerge indicating that it is not an AI bubble, the market breathes a sigh of relief. However, it is not long before concerns about a bubble resurface, causing fluctuations in the market and stock prices. This year, the KOSPI has been significantly affected by the semiconductor peak theory.
For investors, this is a critical point of concern. Indicators to assess the validity of the semiconductor peak theory or the AI bubble will be released by the end of this month. These include the earnings reports from hyper-scalers (large-scale data center operators). Alphabet (Google's parent company) will announce its second-quarter results on July 23, followed by Meta and Microsoft on July 30, and Amazon on July 31. There are three key points to watch.
- Are they making money with their AI business models?
- Will they continue to increase AI investments (capital expenditures, CapEx)?
- Do they have sufficient cash flow to support these investments?
These three points are closely linked to the stock prices of major semiconductor companies, which are the backbone of the Korean stock market. For the semiconductor supercycle to last, hyper-scalers must profit from AI data centers and continue purchasing semiconductors. Among these points, the second is particularly crucial. This year, semiconductor prices have surged, leading to increased cost burdens for semiconductor consumers. The third point is also important; if expenditures exceed cash inflows, long-term investment plans may be jeopardized.
Now, let’s examine Alphabet's second-quarter results based on these points. The results were remarkable. At first glance, the AI bubble theory seems unfounded. Looking at the details:
First, regarding the question of whether they are making money: Alphabet's second-quarter revenue reached $119.8 billion, approximately 177 trillion won, a 24% increase from the previous year. This exceeded market expectations, marking an 'earnings surprise.' Notably, the revenue from the cloud division, which showcases the growth potential of hyper-scalers, surged 82% year-over-year. Profitability metrics also soared, with net income for the second quarter hitting $112.1 billion (about 165 trillion won), a staggering 298% increase compared to the same quarter last year. Earnings per share (EPS) reached $9.11, tripling market expectations.
Next, concerning capital expenditure plans: Alphabet announced a second-quarter CapEx of $44.9 billion, aligning with pre-announcement market expectations. The outlook for future investments is positive. Alphabet's Chief Financial Officer stated that the annual CapEx would be increased from the previous range of $180 billion to $190 billion to a new range of $195 billion to $205 billion. This indicates a commitment to increasing investments in response to rising AI demand, which translates to purchasing more semiconductors.
Finally, regarding cash flow: If there is a flaw in Alphabet's results, it lies here. The company reported a negative free cash flow of $5.86 billion for the second quarter due to significant investments. This indicates that the cash generated was insufficient to cover investment costs. The shortfall will have to be covered through corporate bonds and loans, raising concerns about the indiscriminate funding practices of hyper-scalers and potential issues arising from this.
How did the market react to Alphabet's results? In terms of stock movement, it was disappointing. Alphabet's Class A shares fell 1.46% during regular trading, with losses exceeding 3% in after-hours trading. Concerns about cash flow outweighed the positive earnings surprise.
In contrast, semiconductor suppliers reacted differently. On the KOSPI, SK Hynix rose 4.86%, and Samsung Electronics increased by 3.65%. The news of Alphabet's increased annual capital expenditures was seen as a positive signal, interpreted as a commitment to purchasing more semiconductors.
The start of the second-quarter earnings season, initiated by Alphabet, has produced mixed reactions. The upcoming results from Microsoft, Meta, and Amazon are highly anticipated. If any of these companies show signs of faltering in cash generation or capital expenditure plans, the market could experience significant volatility again. Conversely, if they meet or exceed expectations, the market may react positively.
For investors in the Korean stock market, another point to consider is 'chipflation' (semiconductor inflation). According to global market research firms, prices for key DRAM products have risen throughout last year and the first half of this year, with projections indicating a further increase of about 20% in the third quarter. While rising semiconductor prices may seem beneficial for Samsung Electronics and SK Hynix, excessive price hikes could dampen market demand. Apple has already expressed concerns about having to raise product prices due to soaring semiconductor costs. Hyper-scalers may face similar challenges.
Returning to the original question: Is the 'semiconductor peak theory' or the 'AI bubble theory' correct or incorrect? Analyses are mixed, and forecasts vary widely. It is not a question that can be easily concluded at this moment. Thus, the market remains volatile, leaving investors confused. It is essential to closely monitor the AI and semiconductor markets and the situations of the key players. Perhaps a longer-term perspective is needed.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.