FOR PEOPLE WHO WANT TO SEE WHAT BREAKS BEFORE IT BREAKS
OpenAI closed $122 billion at an $852 billion valuation. SpaceX lined up 21 banks for Project Apex. M&A just had its best quarter ever. And a $575 million EV parts factory in Michigan is mostly empty.
THE SETUP
Two economies are running at once.
In one, OpenAI closed $122 billion on Tuesday. SpaceX has 21 banks lined up for June. M&A just posted its best quarter on record.
In the other, GM sold 11,000 electric trucks last year against a plan for 600,000. Magna International runs a mostly empty factory in Michigan. $575 million in robots that haven't run in months. The industry put $200 billion into an EV future that didn't arrive.
The question isn't which economy to trust. It's which assumptions the current one must unwind.
PMD LENS
The EV supply chain wasn't irrational. It ran on assumptions that held until they didn't. The AI capital cycle is making assumptions too. The new factories will either vindicate those bets or become the next cautionary tours.
WHAT MOST WILL MISS
GM paid $400 million to suppliers for canceled EV contracts. That covers only part. The rest sits on supplier balance sheets as stranded assets.
OpenAI opened to retail for the first time. $3 billion through bank channels. Those investors capture the pre-IPO valuation gap instead of buying in after it's locked.
22 megadeals in one quarter while small deals fell 17%. The gap is the sharpest in years.
SpaceX's 21-bank syndicate reflects channel complexity, not just deal size.
Magna needs 18 to 24 months to find new customers for St. Clair. That timeline doesn't bend to a gas price shock or policy shift.
PREMIER FEATURE
The 20-Minute Trading Window Most Retail Traders Miss
According to a veteran trader who’s spent years studying repeatable market behavior, retail traders often do far more work than necessary — and still miss the same daily opportunity.
He says a specific pattern tends to form within a short, consistent window each trading day. When spotted early, it allows trades to be planned calmly — before emotion, headlines, and intraday noise take over.
He’s now breaking it down step-by-step in a free online web class, explaining why this setup keeps appearing and why even beginners are able to follow it once they know what to look for.
IN FOCUS
America's EV Rust Belt
The Factory
In 2022, Magna International broke ground in St. Clair, Michigan. The plant would supply battery enclosures for GM's electric pickup program. Magna spent $575 million. The town added $44 million in tax breaks.
The plant opened. Workers came on. Parts shipped.
Then demand didn't come. GM sold just over 11,000 Silverado EVs against a plan for 600,000. Factory ZERO sits idle. St. Clair runs a fraction of its built-for workforce.
The Scale
The U.S. auto industry put $200 billion into EV and battery plants. GM and Ford built on federal credits and curves that assumed buyers would follow policy. The Trump administration cut the $7,500 buyer credit. It rolled back fuel rules too. GM shipped 8,000 trucks a month against contracts written for 1,000,000 a year.
GM paid $400 million in canceled contracts. That covers only part. The rest is stranded on supplier balance sheets. Magna needs 18 to 24 months to find new customers.
What It Tells the AI Cycle
The EV bet wasn't irrational. Analysts had projections. Policy backed the economics. The problem wasn't the analysis. It was the assumption that policy-driven demand holds like market demand.
AI capital is making similar bets. Capital is building data centers ahead of proven deployment. Investment runs years ahead of revenue. Magna's CEO still believes EVs will take hold. He just doesn't know when. That's the answer AI infrastructure will face too.
Investor Signal
Supply chains built on policy-backed adoption curves are more fragile than those built on market demand. The EV factories aren't wrong about direction. They're wrong about timing. The AI infrastructure cycle is making similar timing bets. Does revenue arrive before the debt does?
FROM OUR PARTNERS
Could the AI Boom End Like the Dot-Com Bubble?
But one market statistic — praised by Warren Buffett as the best measure of valuations — is now flashing a historic warning.
It’s currently higher than it was at the peak of the Dot-Com Bubble.
If the signal proves accurate, the coming AI unwind could shake the entire market.
SIGNALS IN MOTION
The signals below are not forecasts. They are mechanisms already in motion. Each one reveals the same pattern: duration is being financed before economics are fully proven.
Signal 1: OpenAI Closed $122 Billion at an $852 Billion Valuation
The round closed Tuesday at $122 billion, up from $110 billion in February. SoftBank co-led with Andreessen Horowitz and D.E. Shaw. Amazon committed $50 billion. Nvidia and SoftBank $30 billion each.
The structural change is the $3 billion from individuals through bank channels. First time OpenAI has opened to retail before an IPO. Revenue hits $2 billion a month. Business customers are over 40% of that. Still burning cash.
Investor Signal
$852 billion on a cash-burning company is a bet on the adoption curve. The retail channel is the first step toward AI wealth distribution. Watch whether retail access expands or stays a one-time close.
Signal 2: SpaceX Has 21 Banks for Project Apex
The SpaceX IPO has a code name. Project Apex has 21 banks on the syndicate. Morgan Stanley, Goldman Sachs, JPMorgan, Bank of America, and Citi lead. Sixteen more in smaller roles. June target. $1.75 trillion valuation.
The 21-bank structure reflects what the deal requires. Retail, institutional, high-net-worth, and international distribution all run at once. Arm Holdings used close to 30 banks for its 2023 listing. The retail fight is now inside a 21-party structure.
Investor Signal
A 21-bank syndicate distributes across every investor type at once. The retail question isn't Robinhood versus E*Trade. It's where the retail slice sits inside a 21-party structure.
Signal 3: M&A Just Had Its Best Quarter Ever
22 deals at $10 billion or more globally in Q1 2026. Quarterly record. Previous was 21 in Q4 2015. Total value jumped 29% year over year. This week: Unilever and McCormick at $65 billion. Sysco buying Restaurant Depot for $29 billion.
The record came against the worst quarterly stocks in four years and $100 oil. Companies see a window. The DOJ's top antitrust official left in February. PE firms hold AI-threatened portfolios but are waiting on price. Small deals fell 17%.
Investor Signal
Megadeals run when antitrust is calm. The window is open. PE firms will move on software portfolios when prices firm. Watch DOJ posture in Q2. The first major deal the new leadership challenges or approves sets the tone for the rest of the year. That decision tells you whether the window stays open or starts to close.
FROM OUR PARTNERS
7 Stocks That Could Become the Market’s Next Giants
Apple, Google, Tesla…
Sure, they’re household names now, but these companies and the other members of the original Magnificent 7 didn’t start out obvious.
They earned their place over time.
Our analysts believe the next generation of market leaders is forming now…
And we’ve identified the 7 companies that fit the “Magnificent” pattern.
You can see the full list for free today.Don’t wait until everyone’s talking about them.
THE PLAYBOOK
Before pricing infrastructure, confirm demand is market-driven, not policy-driven. Magna's path back is 18 to 24 months if customers appear.
Watch OpenAI's retail access. The $3 billion from individuals is the template for pre-IPO AI wealth distribution.
Watch the SpaceX June timeline against Nasdaq's May 1 fast-entry rule. Filing before May 1 shifts the passive buying wave.
Watch small deal activity in Q2. When PE firms move AI-threatened portfolios, small deal count recovers.
THE PMD REPOSITION
OpenAI raised $122 billion. SpaceX has 21 banks. M&A just broke its own record. And a factory built on $200 billion in promises is hosting tours for its robots.
The capital is moving. The EV rust belt answers what happens when capital moves faster than the adoption curve it's betting on.
That question doesn't have a different answer just because the technology changed.
