Summary of Key Points
OpenAI significantly reduced the prices of its mid-to-low-end AI models over the weekend (Luna by 80%, Terra by 20%), causing a “seesaw” effect in global tech stocks: AI hardware sectors such as chips and storage experienced sharp declines (Korean stock market fell by more than 5%, A-share Sci-Tech 50 index fell by 5%), while Hong Kong internet companies saw relatively strong gains (Alibaba rose 7%, Tencent rose 3%). There is concern in the market that the profitability of large AI models has weakened, which could lead to reduced demand for upstream hardware; however, some believe the price cuts will stimulate the widespread adoption of AI, potentially benefiting software and cloud service providers in the long run. Overall, this represents a revaluation of the AI industry chain from “computing power expansion” to “profitability.” In the short term, hardware companies are under pressure, while application-oriented firms show more resilience.
Detailed Analysis
1. Why Did Hardware Stocks Suffer First? – Price Cuts Trigger Concerns About Upstream Demand
When OpenAI cut prices, investors immediately wondered: “If AI companies earn less money, will they still buy as many chips?”
- Logical Chain: Lower model prices → Reduced profit margins for large AI companies → Possible decrease in purchases of upstream components like NVIDIA GPUs and SK Hynix memory → Declining earnings expectations for hardware manufacturers → Stock price drops.
- Additional Reason for the More Severe Drop in the Korean Stock Market: AI hardware had been highly hyped earlier on, with many investors using leverage to buy stocks. The news of price cuts triggered panic selling, leading to a cascading decline (e.g., those who borrowed money to invest saw their positions forced to be sold at reduced prices by brokers).
- A-share Tech Stocks Followed the Trend: This was mainly due to emotional contagion and the need to adjust valuations after significant gains.
In simple terms, it’s like a milk tea shop suddenly reducing prices; people worry that the shop won’t make enough money and might buy fewer machines, causing the stock price of the machine manufacturer to fall. If many investors bought the company’s stock on credit, their positions were forced to be sold at lower prices, exacerbating the decline.
2. Why Are Hong Kong Internet Companies a “Safe Haven”? – The Application Layer Benefits from Price Cuts
Contrary to hardware companies, Hong Kong internet firms (such as Alibaba and Tencent) saw gains for three reasons:
- Cost Reduction: Lower AI model prices lower the costs for these companies using AI in services like intelligent customer support and content generation, potentially increasing their profit margins.
- Valuation Safety Cushion: These companies had already seen significant declines, leaving their valuations at historical lows. Investors moving away from struggling hardware sectors flocked to them.
- Clear Application Scenarios: Internet companies have real business models (e.g., Alibaba’s e-commerce, Tencent’s social platforms), allowing them to convert AI technology into actual revenue more effectively than pure hardware companies.
For example, if the cost of milk tea ingredients decreases, milk tea shops can make more profit, so investors are more likely to buy their stocks rather than those that sell the ingredients.
3. Why Did OpenAI Cut Prices? – It’s Not About Charity; It’s a Strategic Move
OpenAI’s price cuts were not out of generosity but due to practical pressures:
- Competitive Pressure: Many open-source AI models (e.g., Llama 2) are now available for free or at low costs. If OpenAI doesn’t reduce prices, it may lose users.
- Efficiency Improvement: OpenAI’s own technology has advanced, reducing the cost of running its models, allowing for price cuts.
- Market Expansion: Lower prices make AI more accessible to businesses and individuals, potentially leading to higher long-term revenue through increased usage.
This is similar to how smartphone manufacturers cut prices: either due to competitive pressure, reduced costs, or a desire to attract more customers—ultimately, it’s all about survival and profit generation.
4. What’s the Market Most Worried About? – The Increasing Doubts About AI’s Profitability
The price cuts have raised questions about when AI will truly become profitable:
- Lengthened Profitability Cycle: Lower prices mean that large AI companies won’t earn high profits in the short term, and upstream hardware manufacturers may see reduced orders, casting doubt on the entire industry chain’s financial prospects.
- Valuation Bubble Bursting: The past two years have seen excessive speculation in AI hardware, leading to inflated stock prices. Now that profitability seems more uncertain, valuations need to be adjusted downward.
- Excessively Liquid Market: Many funds were invested in AI hardware, and any market turmoil causes significant price fluctuations.
In other words, if AI was previously seen as a money-making machine, the current situation has raised concerns about its actual profitability, especially for those who borrowed money to invest in it.
5. What’s Next? – Several Factors Will Determine the Future Trend
The market is currently in a wait-and-see phase, with three key indicators:
- NVIDIA’s Financial Report: NVIDIA’s report in late August will provide insights into the actual demand for AI hardware.
- Application Profitability: Whether internet companies are actually generating revenue from AI services (e.g., whether Alibaba’s new AI products can boost sales).
- US Dollar Liquidity: Hong Kong stocks are sensitive to US interest rates; lower rates may attract more investment, while higher rates could put pressure on them.
In summary, the “first half” of the AI industry chain (computing power expansion) may be coming to an end, and the “second half” (application-based profitability) is just beginning. The company that can effectively convert AI technology into real revenue will emerge as the winner.
Conclusion
OpenAI’s price cuts mark a turning point for the AI industry, shifting from hype to practical competitiveness. While hardware companies face short-term challenges, application-oriented firms are showing resilience. For individual investors, it’s better to focus on companies with real use cases and stable cash flows rather than chasing high-profile concepts.