Nvidia Is Now Investing in Power Plants, Not Just Chips.

Nvidia Is Now Investing in Power Plants, Not Just Chips

The Information reports Nvidia is in talks to put as much as $3 billion into SB Energy, a SoftBank subsidiary developing a large planned data centre in Ohio for OpenAI.

Read what that actually means. The most valuable chip company in the world is considering writing a cheque not for silicon, not for research, but for the electricity supply that its customers need before a single one of its chips can be switched on.

It follows a week in which Nvidia was reported to be helping mobilise more than $500 billion of third-party capital for AI infrastructure. The pattern in both is the same: the constraint on this buildout has moved from processors to power.

The arithmetic behind that shift is unforgiving. A single large AI data centre can draw as much electricity as a small city, and the grid it plugs into was designed decades before anyone imagined the load. Building generation takes years. Building a data centre takes months. That mismatch is now the binding constraint on the entire sector.

Investor money is already moving on it, though not in a straight line. Technology funds saw $4.62 billion of outflows last week after six consecutive weeks of inflows, while growth funds took in $8.78 billion and US equity funds $2.58 billion. That is rotation within the theme, not an exit from it.

Here’s why it lands on your desk: for two years the way to own this story was to own the chip makers. If the bottleneck has genuinely moved upstream to fuel, generation and grid, then the next stretch of it gets decided somewhere most portfolios are not currently looking.

Source: The Information / CNBC / TechStock

A Rockefeller Moment, With Coal Instead of Oil

In the 1800s, John D. Rockefeller started refining oil into the world’s most valuable fuel. Now, another innovator is creating its own “Rockefeller Moment” with one of the world’s most abundant energy resources: coal.

This is more important than ever right now, because a perfect storm of operational breakthroughs and policy shifts has the potential to directly impact this company’s valuation.

What’s creating this “Rockefeller Moment” for coal?

Using their patented FASForm technology, Frontieras North America can transform coal into high-value commodities like hydrogen, diesel, jet fuel, and fertilizer, without burning it.

They’re targeting a $2.1 Trillion total addressable market* where demand for these commodities is virtually unlimited.

Reaching just 2% of the global coal market could mean a trillion-dollar valuation for Frontieras.

That’s why the institutional investors are already moving. Frontieras has secured a $150 million investment commitment from GEM and raised over $45 million from private investors.

But here’s why 2026 is shaping up to be such a historic year for this company:

Frontieras is creating what could be a pivotal moment for the future of energy on the world stage.

✱

Silver Hit a Six-Week High and Berkshire Started Buying Again

Spot silver rose 5% on Friday to $64.57 an ounce, its best weekly performance since February. Gold moved higher alongside it.

The drivers were mechanical rather than emotional. Softer hiring data and lower oil prices pushed rate expectations down, and precious metals move inversely to real yields. Futures now price only about a 30% chance of a quarter-point hike in September, down from roughly 55% two weeks ago.

Separately, Berkshire Hathaway became a net buyer of equities for the first time in 15 quarters, purchasing nearly $20 billion more than it sold. After nearly four years of accumulating cash, the most patient large allocator in the market started putting it to work.

The part that gets less attention is what happens inside a metals bull market rather than to the metal itself. Producers carry operating leverage: when the price of what they dig up rises faster than the cost of digging it, margins expand disproportionately. And in the later stages of these cycles, the larger producers have historically bought the smaller ones rather than explore from scratch.

Scale matters here too. Silver at $64.57 is a long way from where it sat two years ago, and the companies that pull it out of the ground did not double their costs over that stretch. That spread between a rising sale price and a roughly fixed cost base is where the leverage lives, and it is invisible if you only watch the commodity quote.

The distinction worth holding: the metal price and the companies that produce it are related but not the same investment, and they behave very differently at different points in a cycle. Which point this is remains an open question, but it is the right question to be asking.

Source: CNBC / Bloomberg

Two Hundred Sixteen Percent a Year, Compounded

I’m not saying this to brag – I want you to understand how much money you can make if you get in early on a massive bull market in gold: Warren Buffett’s best stretch across sixty years never broke 36% a year… my top-performing portfolio returned 216% a year, compounded. Put $10,000 in at the start of 2024 and today it’s worth about $131,000. I’m not smarter than Buffett — I just understand one thing he doesn’t track: what happens to the best small miners when a bull market hits its buyout phase.

SpaceX Clawed Back to Its IPO Price. The Sector Followed.

The shares briefly recovered to $135 after quarterly results reaffirmed analyst confidence — a full round trip from the $225.64 peak in June.

Two months ago that listing was the largest in history. Then the stock fell more than 50% from its high, dropped below the offer price, and spent weeks with roughly a third of its tradeable float sold short. Recovering to breakeven counts as a recovery only against that backdrop.

What the round trip obscures is that the underlying activity never wobbled. Revenue grew 92% year on year in the quarter, the order backlog stood at $47.5 billion, and the AI segment grew 250%. The share price was arguing about valuation, not about whether rockets were launching.

Underneath the headline names, the commercial space economy has kept compounding regardless of what any single listing did. Launch cadence, satellite manufacturing and the supply chain that feeds both have grown through the entire episode, and almost none of that sits in companies most people have heard of.

The pattern is familiar from every infrastructure cycle: the famous name captures the attention and the volatility, while the parts suppliers underneath capture a steadier share of the actual spending. Neither is automatically the better investment. They are simply different bets on the same growth.

Source: CNBC / TradingKey

Six Days After the Filing, Congress Started Buying

Six days after the largest IPO filing in space history, Washington insiders did something unusual.

At least six members of Congress immediately bought shares of a tiny, unknown defense supplier.

Five of them sit on committees overseeing defense, satellite communications, and AI.

They know something Wall Street hasn’t figured out yet:

Space is no longer a science experiment — it’s a $686 Billion critical infrastructure race. Commercial space spending just hit $544 Billion, outspending all world governments 4-to-1.

And U.S. commercial satellite manufacturing is exploding — up 67% year-over-year.

There’s just one massive problem: launching these satellites is still too expensive.

Even Elon Musk admitted it:

“There are only three casting companies in the world that make these parts.”

But a tiny, publicly traded company in Florida just cracked the code to cheap space launches.

Their CEO recently stated:

“We are twice as fast as our competition, and ten times cheaper.”

They’ve created what insiders are calling a “launch cheat code” that SpaceX, Blue Origin, and Rocket Lab desperately need to stay profitable.

The Pentagon is already doubling its space budget to $126 Billion… and this $4 company holds the monopoly.

Robert Kiyosaki’s “Financial 007” believes this stock won’t stay under $5 once the mainstream financial press catches on.

P.S. When politicians on defense committees buy a stock before a major licensing milestone, it’s rarely an accident. See the research before the crowd moves in.

Retail Sales Fell the Most in Over a Year. Sentiment Went With Them.

Census Bureau data released Friday showed the sharpest monthly drop in retail sales in more than a year. Consumer sentiment fell to 51.0 in August from 55.2 in July.

The sentiment reversal is the part that stings. The index had climbed from a historic low of 44.8 in May to 49.5 in June and 55.2 in July. Two months of repair, undone in one reading.

The market has been signalling this for a while, quietly. The retail sector ETF is up roughly 4% in 2026 against a gain of nearly 14% for the S&P 500. Consumer-facing companies have been left behind by a rally driven almost entirely by technology and infrastructure.

This week resolves it. Home Depot reports Tuesday, TJX, Target and Lowe’s on Wednesday, and Walmart on Thursday. Between them they cover home improvement, discount, mid-market and everyday grocery — effectively the full range of American household spending.

If you are drawing income from savings, these four matter more to your year than Nvidia does. Consumer spending is roughly two thirds of the economy, and the gap between an index at record highs and a sentiment reading at 51.0 has to close in one direction or the other.

Source: Census Bureau / University of Michigan / CNBC