Summary of Key Points
Recently, the stock markets in South Korea, Japan, and the United States have been experiencing volatility, and there are concerns that the AI bubble may be about to burst. The main question on everyone's mind is whether the storage stocks, which have seen the most dramatic gains, are actually cyclical or growth stocks. If they were cyclical stocks, the bubble would have already burst; if they were growth stocks, the price increases wouldn't last in the short term. This article compares the characteristics of commodities (a typical example of cyclical stocks) and finds that storage stocks are still essentially cyclical, but AI demand has extended their cycle, making them appear more like growth stocks on the surface. However, four arguments put forward by Wall Street to suggest that storage stocks have become growth stocks are not convincing; it's simply the effect of AI exacerbating the natural cyclical nature of these assets.
Why Are Storage Stocks the Key to the AI Bubble?
In the AI boom, storage chips are one of the most critical components (AI servers require large amounts of high-end memory), which is why storage stock prices have risen the most. The nature of storage stocks directly determines whether the AI sector is a bubble:
- Cyclical Stocks: Prices fluctuate sharply, and when they rise too much, they are likely to fall (indicating a bubble is about to burst).
- Growth Stocks: Performance improves over time, leading to sustained price increases (making it harder for a bubble to form).
Therefore, people are closely watching the nature of storage stocks to determine how long the current AI-driven market trend will continue.
What Are Cyclical Stocks?
Cyclical stocks are characterized by a cycle of supply and demand imbalance. A typical example is commodities such as oil and pork. These stocks have three key features:
1. Slow Capacity Expansion: It takes 10 to 2 years to increase production capacity, and by the time it's ready, market demand may have changed.
2. High Volatility in Supply and Demand:
- Urgent demand (even if prices are high, buyers still purchase),
- The bullwhip effect (small demands from downstream markets can be amplified upstream),
- Economic cycles (economic growth leads to surges in demand, while recessions cause a plunge),
- Natural disasters and human-made events (such as pandemics),
- Speculation by investors, all of which contribute to erratic supply and demand patterns.
3. Homogenization of Products: Products are relatively similar; there is no distinction between high-end and low-end versions, so when prices rise, more production is driven, leading to oversupply and subsequent price drops.
These three factors together create a cycle: 价格上涨 → increased production → oversupply → price drop → reduced production → renewed price increase.
Storage Stocks Are the “Commodities of the Chip Industry”
Storage chips fit perfectly into this cyclical pattern:
1. Even Slower Capacity Expansion: It takes 18 to 36 months to build a new storage chip factory and reach full capacity, which is even longer than the time required for养猪.
2. More Extreme Supply and Demand Fluctuations:
- Chips are considered a “21st-century commodity” with high demand; the supply chain is long (taking more than half a year from end-users to chip manufacturers), exacerbating price volatility.
- The global economy has a significant impact on chip prices (for example, the 2008 financial crisis caused a sharp drop in the chip industry).
- Natural disasters and human-made events (such as conflicts in the Middle East) can disrupt supply.
- High levels of financial speculation (chip derivatives are worth over $110 billion and continue to grow in value).
3. Product Homogenization: Storage chip specifications are standardized, with no significant differentiation between different products.
Historical Evidence: The storage industry has gone through nine cycles over the past half century, each following the same pattern: surging demand → increased production → oversupply → price drop → reduced production → renewed demand. For instance, from 2016 to 2019, cloud computing demand surged, driving storage prices up by two times, but subsequent capacity expansion and the bursting of the mining machine bubble caused prices to fall by 79%.
Wall Street Claims Storage Stocks Are Growth Stocks: Four Unconvincing Arguments
Recently, investment banks have raised their target prices for Micron, arguing that storage stocks have shifted from cyclical to growth stocks. However, these arguments are not solid:
1. AI-Driven Demand: AI demand is expected to last 7 to 10 years, but this only extends the cycle, not eliminates it.
2. Long-Term Supply Agreements (LTAs): Micron has signed multiple long-term supply agreements, securing a portion of its sales volume. However, how long will such one-sided agreements remain effective? Could customers switch to other suppliers?
3. Supply Side Constraints: The market is dominated by a few oligarchs (three companies control 90% of the market). Although they may claim to control production, new national semiconductor initiatives in countries like South Korea and China could change this dynamic quickly.
4. Product Customization: Although there is some customization for high-end storage products (such as HBM), the products are still largely homogenized. Oligarchs can still cause oversupply if they increase production.
Conclusion: Short-Term Growth, Long-Term Cyclical Nature—but No Need to Worry About a Bubble Bursting Now
Storage stocks are fundamentally cyclical, but AI demand has extended their cycle to 7 to 10 years. We are currently in an upward phase of this cycle, so stock prices are likely to continue rising in the next one or two years. As Keynes said, “In the long run, we are all dead”—so let's focus on capturing the growth opportunities available now.
(The entire article is written in plain language, making it easy for non-financial professionals to understand the cyclical nature of storage stocks and their relationship with the AI bubble.)