Most people are wrong because they confuse a falling stock price with a failing company. That's lazy analysis. The real signal in the Hong Kong AI sector over the last seven days isn't about Zhipu AI or MiniMax being bad businesses. It's about the market finally doing what markets do: punishing a narrative that got priced for perfection before the financials were written.
Zhipu AI and MiniMax took a beating. Both dropped over 11 percent in a single session on the Hong Kong exchange. The headlines read like obituaries. But I didn't see an obituary. I saw a P&L statement finally being taken seriously.
This is the story of a paradigm shift in how markets value artificial intelligence companies. And it's a story that should be deeply familiar to anyone who watched the crypto market do the exact same thing in 2022.
Context: The Valuation Reality Check
Zhipu AI is a Beijing-based artificial intelligence company backed by Tsinghua University. They build the GLM series of large language models. MiniMax builds consumer-facing AI products like Talkie and Hailuo. Both are considered part of China's "AI Big Four" alongside Moonshot AI and Baichuan Intelligent.
Their stock decline on the Hong Kong exchange is significant. But it's not because of a technical failure or a product bug. It's because the market is finally applying a standard that crypto investors have known for years: Hype is a liability; liquidity is the only truth.
These companies are in the second tier of Chinese AI development. The first tier includes Baidu, Alibaba, ByteDance, and Moonshot AI. Zhipu and MiniMax are trying to carve out their own niches. Zhipu focuses on government and enterprise deals. MiniMax focuses on consumer subscription products. The problem isn't their technology. It's their profit generation. Or lack thereof.
The Hong Kong market has never been kind to unprofitable tech companies. Compare that to the US market, which has shown remarkable tolerance for money-burning AI ventures. Hong Kong investors want to see revenue. They want to see margin. They want to see a clear path to profitability. Zhipu and MiniMax aren't delivering on that front.
I've been in this exact situation before. In 2022, when Terra collapsed, I watched a system that promised algorithmic stability unravel in 48 hours. The technology was elegant. The narrative was compelling. But the fundamentals didn't survive contact with the market. I shorted it with Perpetual DEXs and made a 400% return. The lesson wasn't about being a genius. It was about reading the mismatch between story and substance.
These AI companies have been living in a world where their valuations are based on potential, not performance. That's the same dynamic. When you strip away the narrative, what's left is a business that needs to prove its worth in the real world.
Core: The Order Flow Mechanics
The decline of Zhipu and MiniMax is a classic liquidity event. It's not about the technology. It's about who is holding the tokens and what they're doing with them.
Consider the market structure. Hong Kong is a market where institutional investors dominate. Retail participation is significant, but the institutional players move the volume. These institutions have risk departments and mandate letters. They don't have a tolerance for sustained losses. When a stock falls below a certain level, they get margin calls. They get redemption requests. They need to free up capital. This forces a sell-off. And the sell-off attracts more selling.
The data shows a distinct pattern. The stocks didn't just fall. They fell hard, fast, and without meaningful recovery. That's not a normal market correction. That's a liquidation event. The market is flushing out the speculative excess that had built up in the AI narrative.
I built my career on understanding this exact dynamic. In 2020, I was running a backend development role while writing Python scripts to execute triangular arbitrage on Uniswap. I would watch gas costs, monitor pool imbalances, and move capital in and out of positions within seconds. The principle was the same: understand the mechanics, identify the trigger point, and position accordingly.
The trigger point here is the realization that these AI companies have not yet achieved what their valuations suggested. They are spending heavily on research and development, but the revenue generation is not matching the hype. The market is finally asking the question that I've been asking my entire career: Where is the profit?
The SPAC route for Zhipu and MiniMax is likely. SPAC listings are faster and have a less stringent regulatory burden than traditional IPOs. But they also carry a heavier price: a valuation that is often inflated by the seller's optimism. Historical data shows that SPAC-listed companies tend to see significant declines in their stock price within 6 to 12 months of listing. The market's tolerance for empty promises is limited. The average SPAC has dropped over 50% from its initial offering price. That's not a statistical anomaly. That's a trend.
The Hong Kong AI stock collapse is not an isolated event. It's a reflection of a broader market sentiment shift. Global AI investment enthusiasm has been cooling since late 2024. The US market is seeing the same signs of volatility in AI stocks. The rate environment is a key factor here. High interest rates hurt growth stocks. AI companies are growth stocks. And the market's patience is running out.
A comparison to SenseTime is instructive. SenseTime was the first major AI company to list on Hong Kong exchange in 2021. Since then, its market cap has fallen by more than 70% from its peak. The company has struggled to deliver the growth that the market expected. This is a cautionary tale for Zhipu and MiniMax. The market's memory is long and the appetite for AI stock losses is short.
Contrarian: The Silver Lining in the Carnage
This is the part that most analysts will miss. The sell-off is actually a positive signal for the AI industry's long-term health. This is the market correcting a mispricing. It's not a death knell. It's a reset.
The correction serves a function. It clears out the excess speculation that builds up in any emerging industry. AI companies are being forced to focus on fundamentals. That's a healthy development. It means they have to build real businesses, not just narratives. It means they have to show actual revenue, not just growth potential.
I learned this lesson painfully in 2021 when I launched a generative art project. We raised 500,000 euros in ETH. The floor price was driven by hype, not by utility. When the market turned, the floor price dropped 90% in a week. I had to offer a structured refund plan to the community. The project ultimately failed. But I learned the difference between a market that's pricing in an asset's intrinsic value and a market that's pricing in speculation.
The same dynamic is playing out in the AI sector. The stocks are falling because the speculative premium is being removed. But the underlying technology is still valuable. Zhipu's GLM models are used by enterprise clients. MiniMax's products have a user base. The fundamentals are there. The price is just finally reflecting reality.
This is the position where the smart money gets in. When the market is panicking, the smart money is analyzing. The market is panicking. It's time to analyze.
The key indicator to watch is the cash runway. How much money do these companies have left? At the current burn rate, how many months can they sustain their operations? If they can survive the next 18 months, they can weather the storm. If they can't, they'll be forced to raise capital at a discounted valuation, which will further dilute existing shareholders.
I'm looking at the data. I'm not seeing a collapse in the AI industry. I'm seeing a collapse in the speculative premium. That's a different thing. The technology is still evolving. The market is just being more realistic about the pace of that evolution.
Takeaway: The Trade is the Signal, Not the Stock
The Hong Kong AI sell-off is a wake-up call. It's not just about Zhipu and MiniMax. It's about the entire AI valuation framework.
The market is signaling that the era of narrative-driven pricing is over. The only thing that matters now is the fundamental performance. The company is not the story. The business is the story. The technology is the story. The revenue is the story.
For traders, this means looking beyond the hype. It means analyzing the actual operations. It means understanding the cash flow. It means looking at the balance sheet. It means thinking about the business model, not just the tech.
The core insight is this: AI is no longer a concept stock. It's a real business. And real businesses are valued on real numbers.
We do not predict the storm; we build the ship. This is the time to check the ship's hull. Zhipu and MiniMax are in rough waters. But they're not sinking. The market is just recalibrating its expectations.
Trust the code, verify the chain, own the outcome. The code is the AI model. The chain is the market. The outcome is the stock price. The outcome will be what it will be. The question is whether you're positioned for the reality, not the fantasy.
I didn't predict the fall. But I did predict the underlying pattern. The market is always trying to correct itself. It's always trying to find the true price. The AI selloff is just the market doing its job.
The question is: are you ready to trade the truth? Or are you still holding the fantasy? That's the decision. That's the trade. That's the market.