Breaking Points: South Korea Chip Crash Signals AI Bubble Stress

**Krystal Ball** (0:01)
When markets start shaking in the places that built the AI boom, the warning signs can be hard to ignore. I want to talk about a troubling sell-off we're seeing in tech, chips and South Korea.
This could be the first real crack in the AI bubble.

**Saagar Enjeti** (0:19)
Let's start with that hedge fund story. We're talking about a $20 billion shop founded by a former OpenAI employee that posted a stunning 439% net return through June.
But then July happened.

**Krystal Ball** (0:34)
Right. July nearly erased those gains after a violent downturn in AI-linked stocks, made worse by leverage. Oracle and AMD are down 20% for the month, alongside smaller positions like Nebius, Sound AI, Bloom Energy, and Sandisk. What's remarkable is the fund manager remains upbeat, calling this an attractive buying opportunity. That optimism feels disconnected from what we're seeing globally. South Korea is the bigger story here. The KOSPI has lost roughly a third of its value and hit a three-month low.
Samsung and SK Hynix have both fallen more than 10% as investors dump chip stocks.
South Korea has become one of the clearest pressure points in global semiconductor trade. We're potentially looking at a wipeout for investors who used leverage to chase AI returns, especially retail traders loaded up on margin debt.

**Saagar Enjeti** (1:34)
Building on that point, China is another major force. Chinese chipmaker CXMT reportedly surged 466%.
Beijing is pushing hard for domestic self-reliance, trying to decouple from Nvidia and Taiwanese supply chains. There's also a reported breakthrough from a Shanghai firm in DUV lithography tools.

**Krystal Ball** (1:57)
If that progress holds, China could be moving faster than expected toward chip self-sufficiency, which would reshape the competitive balance.
Meanwhile, American companies are feeling the strain. Meta posted a 10% loss after weaker revenue forecasts and announced even higher capital spending for AI infrastructure.

**Saagar Enjeti** (2:19)
That's the core tension. Companies are pouring hundreds of billions into AI infrastructure, but the revenue gains haven't arrived at anything like the same scale.
These are canaries in the coal mine.

**Krystal Ball** (2:31)
Exactly. The market is being driven by a massive buildout of data centers, debt and investor cash, but that structure looks fragile.
If China develops more efficient open-source AI models that need less compute, the entire business model behind OpenAI and Anthropic could come under pressure.

**Saagar Enjeti** (2:51)
That leaves us with one central question. Even if AI transforms the world, can the financial system supporting it survive the boom?