Why AI Infrastructure Outlasts the Headlines

Why AI Infrastructure Outlasts the Headlines

 In December 1947, physicists at Bell Labs invented the transistor. It changed everything. For decades prior, vacuum tubes powered every computing machine ever built. The hot, fragile, power-hungry glass tubes failed constantly and limited what was possible. The transistor replaced it all, and not gradually. Computing itself suddenly changed. This wasn’t an improvement on the old way; it was a new paradigm.

On Wednesday’s Nvidia earnings call, CEO Jensen Huang said, “inflection point.”

He meant it the same way.

The Noise

Headlines gave investors every reason to feel wary. Trade tensions are back as President Trump slapped 50% tariffs on Canadian goods, threatened vehicles and steel, and for good measure, suggested renaming Lake Ontario to “Lake America.” Canada retaliated with tariffs on 700 American products.

Core personal consumption expenditures (PCE) remained stuck at 3.3%. Federal Reserve Chair Kevin Warsh’s first Jackson Hole speech struck a hawkish tone, warning that inflation remains too high. Rate-hike expectations jumped, short-term yields rose, and stocks reacted with pressure. Every ingredient for a risk-off week was present.

The Signal

But Nvidia remains the biggest story. The company reported revenue of $96.2 billion for the quarter, up 106% year-over-year. Data center revenue reached $89 billion, up 117%. The stock gained 8.74% on Thursday, adding $441 billion in value in a single day.

But the real surprise was guidance – revenue growth of approximately 70% projected for fiscal 2028, nearly twice what analysts expected. “AI has reached its inflection point,” Huang said. “Tokens are productive and profitable. Compute is revenue. And demand is accelerating.”

More confirmation came from memory suppliers that make AI possible:

  • SK Hynix’s CEO said global memory is heading for its worst-ever supply shortage in 2027, with demand exceeding supply well beyond 2030.
  • Micron’s CEO said supply won’t catch up with AI-driven demand until at least 2028.
  • Samsung’s EVP warned constraints will become even more severe in 2027 before persisting through 2028 (Samsung has already locked 60-70% of its capacity into multi-year data center deals).

These three companies together control over 90% of global DRAM supply. They are all saying the same thing at the same time. This is not a “memory trade,” but a multi-year structural reality.

What Flows Said

This is where data does the talking.

The Big Money Index (BMI) closed Thursday at 69.2%. It’s the highest level since before the Situational Awareness blowup. Exchange-traded funds also saw inflows as passive money flooded in simultaneously with active institutional buying. The Nasdaq composite gained 1.57% on Thursday. The CBOE Volatility Index (VIX) fell to 14.51.

The market absorbed every headline this week and kept on chugging.

ETF flows told the most complete story. Gold funds occupied the top six inflow slots for the second consecutive week. Bitcoin and ethereum ETFs attracted multiple days of fresh inflows. International equity ETFs across Europe, emerging markets, and global developed markets all saw buying.

The ARK Innovation and ARK Genomics funds attracted inflows. The lone outflow was a municipal bond ETF. Gold, crypto, international equities, and innovation ETFs all saw inflows in the same week. That is not a risk-off signal.

Flows by Focus | MoneyFlows.com

The sector flows confirmed the rotation.

Inflow Distribution | MoneyFlows.com
Outflow Distribution | MoneyFlows.com

Why AI Infrastructure Outlasts the Headlines

Health care led for the 10th consecutive week with 52 inflows – this is a streak without recent precedent. Materials followed at 43, driven almost entirely by gold miners. Financials saw 45 inflows. Technology recovered to log 43 inflows as confidence returned. Discretionary and industrials saw modest outflows as the late-summer rotation continued.

Health Care Inflows & Outflows XLV | MoneyFlows.com
Materials Inflows & Outflows XLB | MoneyFlows.com
Financials Inflows & Outflows XLF | MoneyFlows.com
Technology Inflows & Outflows XLK | MoneyFlows.com

The top-ranked stocks in our data reinforced the same message from a different angle. Enterprise software, cybersecurity, and networking companies dominate the highest scores. These are businesses generating recurring revenue from AI infrastructure without the direct semiconductor exposure that made the Situational Awareness trade so volatile. These are the names the flows have been pointing at for weeks, and they’re leading the recovery.

Within technology, that recovery shows something interesting. The highest-ranked names in our data are concentrated in enterprise software and cybersecurity. These are businesses generating recurring revenue from the AI buildout without direct semiconductor exposure. But the AI infrastructure stocks that bore the brunt of the forced liquidation are also seeing their model scores recover. Their fundamental scores never deteriorated through the entire selloff. What did compress were technical scores in the form of downward momentum as prices fell under forced selling pressure.

With the Citadel overhang clearing and NVDA confirming the demand picture, those technical scores are now poised to go back toward where the fundamentals always pointed. When the world’s largest AI chip company reports 106% revenue growth and guides 70% higher, it validates the demand for the optical networks that carry the data, the memory that stores it, and the software that monetizes it.

The scores are following the phenomenal earnings. That’s exactly how recoveries begin.

Clearing Out

Citadel has now reduced its exposure from the Situational Awareness portfolio by more than 80%. Griffin said they executed 100-plus block trades representing over $4 billion in market value, including the largest single-session block trades of the year in 10 separate names.

The overhang that suppressed AI names for six weeks should now be functionally gone. Interestingly, the thesis Situational Awareness leveraged into was correct: AI is here and accelerating. The biggest companies’ executives are confirming it. “This time last year, one lab alone was driving the buildout,” Huang noted. “Today, we have a golden age of new AI labs and startups, multiple frontier labs scaling in parallel.”

Situational Awareness’s problem was the leverage, not the thesis.

The Inflection

The transistor didn’t announce itself. It was invented in a lab in New Jersey while the rest of the world was focused on something else. The vacuum tube industry had no idea what was coming. The architecture changed and the old paradigm became irrelevant.

The memory CEOs confirmed the demand. The earnings confirm the revenue. The flows confirm the conviction. The noise this week was loud. The signal was louder.

As the ancient Chinese text Tao Te Ching says, “Knowing others is wisdom. Knowing yourself is enlightenment.”

The market that knows where it is going doesn’t need the headlines to tell it.

MoneyFlows finds the winners early…bringing them to you.

If you’re a serious investor or money manager, get started NOW.

Our PRO subscription allows you to spot daily flows and access our weekly Outlier 20 report.

This is the report that has found EVERY SINGLE ONE of our Outlier stocks.

Enjoy a click-through experience unlike anything out there.

Professional money managers and RIAs looking for additional portfolio solutions including ETF flows & ranks and your own Portfolio Tracking tools, please reach out about our Advisor Solution and Emerging Advisor Program.

Have a great week!