Post

Why are storage prices still rising, but storage stocks are falling first? Because the stock market looks further ahead. Everyone basically knows how high the profits are today; The biggest controversy now is whether there will be an oversupply in two or three years. This scene had already been played out once in the previous new energy vehicle cycle. In 2021, as demand for new energy vehicles exploded, global lithium supply could not keep up, causing the price of battery-grade lithium carbonate to rise from about 60,000 yuan/ton to nearly 600,000 yuan/ton, with the highest increase in two years being nearly tenfold. In the supply chain, whoever is most scarce will have their profits concentrated first. Lithium mining companies are making huge profits, while battery and car manufacturers have to bear ever-increasing raw material costs. In 2022, CATL's gross margin dropped from nearly 28% to around 15%, for a simple reason: lithium prices rose too quickly, and battery price hikes couldn't keep up. High profits quickly attracted a large influx of capital. Mine expansion, rising capital expenditures, and increasing long-term procurement agreements. The market has also begun to worry whether lithium will remain so scarce when this new capacity emerges in two or three years. Therefore, lithium mining stocks often start falling before lithium prices truly peak. When lithium carbonate prices plummeted in 2023, many thought automakers could finally turn all cost reductions into profits. As a result, the automotive industry immediately entered a price war. Tesla is cutting prices, BYD is following suit, and more and more brands are vying for market share. Batteries have indeed become cheaper, but the money saved hasn't all stayed with car manufacturers; a large portion ends up at lower prices. Falling raw material costs and improving downstream profits still create a competitive landscape in the industry. Today's AI industry is somewhat like the new energy vehicles of the past. Storage factories correspond to lithium mines in the current year, cloud manufacturers correspond to vehicle manufacturers. Over the past year, prices for HBM, DRAM, and enterprise SSDs have continued to rise, and memory manufacturers like SK Hynix, Micron, and Samsung have seen significant profit improvements. Meanwhile, AWS, Azure, Google Cloud, and Oracle are all increasing purchases of GPUs, HBMs, and servers, and infrastructure costs are also rising. At this stage, the most scarce segment has taken the most profits first. But the market has already started to look backward. In the next two to three years, if HBM, DRAM, and advanced packaging continue to expand production, how much longer can today's excess profits last? This is also why storage companies' performance is still strong, but their stock prices have already started to adjust. However, a storage peak does not necessarily mean cloud providers will become the biggest winners in the next round. Because on the large model side, the price war has actually begun. OpenAI, Google, Anthropic, as well as Alibaba, DeepSeek, and Moon Darkside, are all continuously lowering model prices. Tokens are getting cheaper, inference costs keep dropping, and some models are even open for free. If computing power supply becomes increasingly abundant in the future, cloud providers may continue to lower prices to attract customers. By then, the cost improvements brought by storage price cuts may not all translate into the profit statements of AWS, Azure, or Google Cloud. Cheaper tokens, lower GPU rental prices, and larger free quotas could all pass on these dividends to customers. So this round of storage stock adjustments can be understood using the new energy vehicle cycle: When upstream is scarce, profits first concentrate upstream; After high profits stimulate expansion, stock prices will worry about supply release in advance; Once raw material prices really drop, how much profit downstream can keep depends on whether the industry starts a price war. AI has now shown signs of this.