Over one weekend, Donald Trump rewrote the rules of US immigration. With two executive orders, he simultaneously choked the supply of foreign talent and opened the gates for foreign capital. The first move: a $100,000 fee on every new H-1B visa, the work permit Silicon Valley has relied on for decades.
The second: the launch of the “Trump Gold Card” — a residency program priced at $1–5 million, tailored for wealthy individuals and corporations.
On the surface, it looks like two unrelated policies. In reality, they tell the same story: America no longer wants to import brains; it wants to import money. And that shift could ripple across Wall Street, hitting the one company that has carried the and Nasdaq for the past two years: .
H-1B: The artery of US STEM talent, now taxed more
Formally, the H-1B is a visa for “specialty occupations” — technology, science, medicine, finance. In practice, over 70% of approvals go to Indian nationals and another 12% to Chinese, mostly in IT and engineering. These are the coders, researchers, and PhDs who built the backbone of Google’s search, ’s Azure, ’s LLaMA, and OpenAI’s GPT models.
By slapping a $100,000 toll on every new application, Trump has effectively turned the visa into a tax on innovation. Existing H-1Bs and renewals are exempt, but the pipeline of new talent — the oxygen of the AI boom — is being cut off.
This comes at a fragile moment. The Bureau of Labor Statistics recently revised US payrolls downward by –911,000 jobs for the past year. That means the domestic labor market is weaker than headlines suggested. In other words, there isn’t a deep reserve of local engineers waiting in the wings to replace missing H-1B hires.
The Gold Card: Millions for entry, Millions for hiring
At the same time, Trump rolled out the Trump Gold Card:
- $1 million for individual residency,
- $2 million for corporate placements,
- $5 million for the Platinum tier — offering up to 270 days in the US per year with no tax on foreign income.
But the catch is brutal. Once inside the US, Gold Card investors who want to open businesses face the same new reality: hiring foreign staff now costs $100,000 per head. That leaves no real choice. Either pay the tax, or hire Americans.
It’s not a formal requirement — it’s economic inevitability. America has shifted from importing brains to importing cash, forcing that cash to turn into local jobs.
The core risk: Slower AI R&D
For Big Tech — Google, Microsoft, Meta, — the financial hit is manageable. These companies can afford to pay the $100k fee if they absolutely need to. But money isn’t the only issue. The true cost is time. With fewer engineers flowing in, projects move more slowly, model releases take longer, and data-center buildouts stretch out.
For AI startups, the effect is existential. They don’t have the budget to pay $100k per foreign hire. Without access to international specialists, their ability to scale evaporates. Many of these startups are built entirely around leveraging NVIDIA GPUs in the cloud. If they can’t grow their teams, they can’t grow their computing usage either.
The headline risk isn’t just about talent. It’s about the pace of innovation in AI R&D. And that is where the danger begins for investors.
NVIDIA: Collateral damage with market consequences
NVIDIA doesn’t sponsor armies of H-1B engineers itself. But its customers do. Microsoft, Google, Meta, Amazon, and hundreds of AI startups are the ones buying its GPUs by the thousands. Their demand for compute power is what turned NVIDIA into the world’s most valuable chipmaker and the single largest contributor to S&P 500 and Nasdaq gains in 2023–2025.
If their hiring slows, so does their AI R&D. If R&D slows, so does the deployment of new models and data centers. And if that slows, GPU orders get delayed.
NVIDIA has become the collateral damage of Trump’s visa war, and the implications go far beyond semiconductors. Over the past two years the company has been the single largest contributor to the S&P 500’s performance, effectively carrying the index through the AI hype cycle. While the broader market has advanced at a steady pace, NVIDIA’s stock price has surged by more than 430% in just a few years, turning it into Wall Street’s most important growth engine.
Right now, NVIDIA is trading in the zone between the 1.272 (≈170.9) and 1.5 (≈186.0) Fibonacci extensions. Price keeps stalling here, and the inability to break higher could be an early signal of a trend reversal. In that case, a potential corrective move would target the 1.0 extension around 152.9. If this support fails to hold, the correction could deepen further into the $125-135 range, opening the door to a longer and more painful pullback.
Indices on the edge
In the current environment, I would give my preference to the . Unlike the , overloaded with Big Tech and essentially functioning as a leveraged bet on AI growth and NVIDIA’s continued dominance, or the S&P 500, where technology still makes up more than a quarter of the index and NVIDIA alone has driven a disproportionate share of the gains, the Dow is structured differently.
While it does include Microsoft, , Amazon, , and NVIDIA, the majority of its weight lies in industrials, banks, healthcare, energy, and consumer giants such as , , , , , and . That makes the Dow not only more insulated from an AI slowdown but also positioned to benefit from capital inflows under Trump’s Gold Card program.
Conclusion: A domino waiting to fall
Trump’s weekend orders sent a clear message: America will no longer subsidize the inflow of foreign brains. It will welcome foreign money instead. That shift forces investors to confront an uncomfortable question: what happens when AI innovation slows down?