The AI Black Paper creative lands on a specific date and a specific time: July 29 at 6:30 PM.
July 29 at 6:30 PM after the bell — specifically chosen, not vague.
July 29 at 6:30 PM — the after-hours earnings window for Meta Platforms.
Jim Rickards has been building toward this date since March. Across 13 press releases and a sustained paid-social campaign, he has argued that the AI debt thesis comes to a head when Meta reports Q2 2026 results. The logic is straightforward: if the biggest spender in the AI arms race stumbles, the whole house of cards shakes.
The question is whether Meta deserves that spotlight — or whether Rickards is picking the wrong company to carry the weight of his entire thesis.
Why Meta and not someone else
Meta is the most leveraged name among the AI hyperscalers. That is a balance-sheet fact.
As of March 31, 2026, Meta carried $58.75 billion in long-term debt on its books. Total debt, including lease obligations, reached $86.77 billion — a 75.2% increase year-over-year from $49.52 billion. The company also disclosed a $107 billion step-up in contractual commitments during Q1 alone.
Nvidia does not have that debt profile. Microsoft does not have that debt profile. Among the Magnificent Seven, Meta is the one that is spending beyond its cash flow and filling the gap with borrowed money.
Rickards’ thesis is about AI debt. If you are looking for a company that best represents the risk of an AI-driven debt buildout, Meta is the natural candidate.
The spending numbers are staggering
Meta’s 2026 capital expenditure guidance sits at $125 billion to $145 billion. That upper bound is roughly double the $72.22 billion the company spent in 2025.
The majority of that spending goes to AI infrastructure: data centers, compute clusters, custom silicon, and the third-party cloud capacity needed to train and deploy Llama, Meta’s open-source AI model. The company is building out its Andromeda ads engine, extending its MTIA custom chip program, and making multiyear commitments to NVIDIA and AMD hardware.
Then there is Reality Labs.
Meta’s AR/VR division lost $4.03 billion in Q1 2026 alone, on top of roughly $19 billion in losses in 2025, bringing cumulative Reality Labs losses since 2020 to nearly $80 billion. The division generated about $2.2 billion in revenue in 2025 — meaning it burned roughly $9 for every dollar it brought in.
Taken together, Meta is on track to spend roughly $145 billion on CapEx and lose another $16 billion on Reality Labs in 2026. That is $161 billion in outflows against a business that generated $83.3 billion in operating income in 2025.
The math works — for now. But it leaves no room for error.
What the market expects on July 29
Meta guided Q2 2026 revenue in the range of $58 billion to $61 billion, assuming a 2% foreign-exchange tailwind. Analyst consensus sits at $60.18 billion with EPS consensus at $7.18, and those numbers matter because the market is already nervous — Meta stock is down roughly 15% year-to-date as of late June 2026, sitting about 29% off its all-time high of $796. The drop came after Meta raised its 2026 CapEx guidance from $115-135 billion to $125-145 billion in April — a move that sent the stock down ~10% in a single day despite the company beating Q1 earnings estimates by a wide margin (EPS of $10.44 versus $6.66 consensus).
The pattern is clear: the market is rewarding Meta’s operational execution but punishing the spending trajectory. Each incremental dollar of CapEx guidance costs Meta market cap.
On July 29, the question is whether that pattern holds or accelerates.
What a miss would mean
Rickards argues that a Meta earnings miss could function as a “dotcom moment” — a single event that confirms the AI debt thesis and triggers a repricing across the sector.
That comparison is worth examining.
The dotcom crash was not one event. As I detailed in the full guide, the Nasdaq peaked on March 10, 2000, and then spent two and a half years grinding lower. It was a credit cycle: the Fed raised rates, venture capital dried up, and companies that depended on cheap capital ran out of runway one by one.
But here is where the comparison has teeth.
The dotcom unwind accelerated when Cisco — the biggest infrastructure spender of the era — warned in December 2000 that it could not see demand. That was the moment the market realized the “infinite growth” narrative was built on vendor financing and circular revenue. I covered that pattern in the Cisco vendor financing breakdown.
The structural similarity is real: a dominant company spending enormous sums on infrastructure, backed by debt, with a narrative that “this time is different.”
If Meta misses revenue guidance on July 29 — or worse, if it guides Q3 below consensus — the market will reprice the entire AI capex cycle. Stocks that depend on Meta’s data-center orders (NVIDIA, AMD, Broadcom) would get hit. The AI debt structures that Rickards has been documenting — the $200 billion in AI-related debt raised in 2025, the off-balance-sheet SPVs, the bundling of data-center leases — would face a credibility test.
That is the “dotcom moment” scenario.
What a beat would mean
Meta has a history of beating estimates. Q1 2026 was a 57% EPS beat. The company grew revenue 33% year-over-year while posting 41% operating margins. The advertising business is firing on all cylinders: ad impressions grew 19% and average price per ad grew 12% in Q1.
A beat on July 29 would not resolve the AI debt thesis — because the thesis is structural, not quarterly. Rickards is tracking $200 billion in AI debt raised in a single year, much of it sitting in off-balance-sheet structures that resemble what Enron and Lehman did before their collapses. A single strong earnings report from Meta does not move that number.
This is what makes the thesis different from a trade. The debt either gets serviced or it does not, and that question plays out across quarters, not in a single print. If Meta beats in July, the next data point is October. If Meta beats in October, the next question is whether free cash flow covers the CapEx trajectory. The structural math accumulates either way.
Where this leaves you
Meta is the right company for the catalyst because it is the most exposed company in the AI debt thesis.
No other Magnificent Seven member combines $145 billion in CapEx, $87 billion in total debt, $80 billion in cumulative Reality Labs losses, and a stock that is already down 29% from its high. The numbers are large enough that a miss would reverberate and the debt structures are complex enough that a repricing would spread. On July 29 after the bell, Meta reports and the market watches revenue, EPS, Q3 guidance, and any commentary on CapEx trajectory. If Zuckerberg signals the spending spree is slowing, the stock may rally. If he signals more spending ahead, the stock may sell off. If the numbers miss across the board, the “dotcom moment” frame gets its first real test.
Rickards has been saying since March that this date matters. On July 10, with the stock down 15% year-to-date and the market already jittery, that prediction looks like it is on track.
Watch the close on July 29. The thesis lives or waits in the numbers.
For the history behind the comparison, see What Was the Dotcom Moment? The Real History.