Summary in Plain Language
Zhang Jianzhong, the former head of Nvidia's China operations, gave up a annual salary of millions and his high-ranking executive position at the age of 54 to return to China and start his own business. He was determined to create China's own GPU company. After six years of hard work, he successfully launched MooreThread on the STAR Market (China's equivalent to the NASDAQ). On the first day of trading, the company's stock price soared by more than four times, and its market value reached 28 billion yuan, making it the favorite candidate to become China's "Nvidia." However, less than ten months later, the stock price plummeted by 20% on its first major lock-up period expiration in September, and the market value dropped below 20 billion yuan. The reason wasn't a sudden fatal flaw in the company's products; it was because the initial enthusiasm for the "domestic substitution" and the idea of becoming the next computing power leader had faded. Early investors wanted to cash out, and the entire capital market realized that they couldn't continue to support unrealistic promises. Now, all domestic GPU companies are being forced to prove they can actually make money from selling their products, rather than just relying on continuous financing.
---
Detailed Explanation
1. Why a 5% release caused a 20% drop in stock price?
Many were puzzled: Only 5% of the total shares were available for sale, so how could it cause such a significant drop? The problem wasn't the amount of shares but the fact that very few shares were actually tradable before the release. When MooreThread went public, over 93% of the shares were locked up and not available for trading, leaving less than 7% of the total shares (worth over 28 billion yuan) in the market. This was like saying your house was worth 2.8 million yuan, but only a small part of it (less than 7%) was available for sale. If someone bought that small part for 2 million yuan, the entire house's market value was considered 2.8 million yuan, which was highly inflated. With the addition of 25 million more tradable shares, the market capacity increased by 85%. These shares were initially purchased at just over 1.14 yuan each, so even if they were sold at the daily limit down, investors could still make a 2.6-fold profit. Since their profit came from the initial investment, they didn't care about the company's future performance. The top five sellers on that day were all institutional investors, selling for over 1.4 billion yuan, while the top five buyers were retail investors, who only contributed 150 million yuan—far less than the sellers' total. Additionally, the company's semi-annual report revealed troubling figures: a sudden loss of over 40 million yuan in the second quarter, a 18% drop in gross profit margin, 3.5 billion yuan in inventory, and a net cash outflow of over 2 billion yuan. These issues made investors skeptical about the company's fundamentals.
2. It's not just MooreThread; the "story bonus period" for the entire domestic GPU industry is over
This drop wasn't just a setback for MooreThread; it marked a shift in the overall valuation of the domestic GPU industry. Other listed companies also faced trust challenges after their shares were released. For example, Bilrui Technology's stock price dropped 17% on the day of the release and soon after, the company sought to raise another 7 billion yuan through a rights issue, indicating that the industry's rapid growth required more funding than its IPO funds could cover. Tianzhi Chip's stock price only rose slightly but then also needed a rights issue for 7 billion yuan. Even though its profits came from financial investments and stock sales, the market still gave it a higher valuation because its numbers looked positive. In the past, the A-share market was overly optimistic about domestic substitution, giving high valuations regardless of the companies' actual performance. Now, with the tide of share releases, investors are demanding tangible results.
3. MooreThread's attempt to mimic Nvidia's path has become a burden
MooreThread tried to follow Nvidia's path of developing both enterprise-grade and consumer-grade GPUs, hoping to build a user base and software ecosystem like Nvidia did. However, this approach is unfeasible. Its consumer-grade GPUs have sold poorly, with only a few thousand units sold, accounting for just 2.3% of revenue, not even covering the costs of development. Nvidia controls 94% of the consumer-grade GPU market, and game developers typically optimize their games for Nvidia's GPUs. Ordinary consumers don't want to buy MooreThread's products for a small number of users. It's like trying to compete with WeChat, which already dominates the social market. The capital market's previous optimism was based on a promising story, but now investors want to see real profits.
4. The real challenge is ahead
The real test comes in December, when nearly 40% of the shares will be released. These shares were held by early investors at much lower costs, and they are likely to sell quickly. MooreThread's current price-to-sales ratio (market value divided by annual revenue) is 125:1, which is extremely high. This valuation included a large premium for the "China Nvidia" story. As the market becomes more realistic, the focus shifts to actual profitability. The company will have to prove it can generate revenue from selling its products, not just rely on financing. If it can't do this, it will face significant challenges.