“Nebius is the company that NVIDIA, Meta, and Microsoft collectively chose to build the power grid for the AI revolution — and they signed $46 billion in contracts to prove it.”

Three years ago, Nebius Group barely existed as an AI business. It was the international rump of Yandex — Russia’s Google — caught in a geopolitical crisis after Ukraine. Its founder, Arkady Volozh, made a decision most executives never face: walk away from everything he had built, divest $5.4 billion in Russian assets, and bet the entire company on AI infrastructure.

That bet has returned +688% since the October 2024 relisting. And the fundamental story — if CapEx execution holds — is still in its early innings.

Join our Discord

The Business

Nebius is an AI cloud provider. It builds massive data centers filled with NVIDIA’s most powerful GPUs and rents that compute to companies training, running, and deploying AI. Think of it as the electrical grid for artificial intelligence — the physical infrastructure that makes all other AI products possible.

What sets Nebius apart: it is entirely GPU-specialized, owns and operates its own data centers, and holds NVIDIA Exemplar Cloud status — the highest tier of NVIDIA partnership — giving it priority access to next-gen chips before the general market.

Nebius is the AWS for AI-native companies — except it is entirely specialized for GPU workloads, runs at higher margins, and is growing approximately 10× faster than AWS ever did at the same revenue level.

The numbers: a financial inflection unlike anything I have seen

The EBITDA margin story is the most remarkable number here. Twelve months ago: negative. Q4 2025: 7%. Q1 2026: 45% on the Nebius AI segment — a 38-point expansion in a single year. One of the fastest documented margin inflections in tech sector history. And it is happening while revenue is simultaneously tripling.

The $46B backlog: the thesis in three contracts

The three relationships that validate everything

• Meta Platforms — $27B, 5-year capacity deal through ~2031
  The largest confirmed neocloud contract. Meta, using Nebius as an overflow AI training compute, cannot build itself fast enough. Recurring revenue with utilization optionality enabling asset-backed financing.

• Microsoft — $17–19.4B, multi-year through ~2030+
  Together with Meta: $44B+ in committed backlog — more than 10× FY2026 revenue guidance. Dedicated capacity built specifically for Microsoft AI workloads.

• NVIDIA — $2B equity investment + Exemplar Cloud status
  Priority allocation of Blackwell Ultra and Vera Rubin GPUs before the general market. When GPU supply is tight — and it is — this partnership is worth more than the $2B check.

The European edge: why Nebius wins deals NVIDIA’s other partners cannot

This is the section most analyses miss — and it may be the most durable competitive advantage Nebius has.

The EU AI Act and GDPR aren’t just compliance hurdles — they are Nebius’s moat. Every regulation that restricts European enterprise data from leaving EU jurisdiction sends another customer to Nebius who cannot legally go to AWS or Azure.— SWP Equity Research analysis

Nebius vs. CoreWeave: the question every investor is asking

Every investor comparing AI infrastructure stocks eventually lands here. Both are GPU neoclouds. Both have massive hyperscaler contracts. Both are burning capital to build data centers. The differences are in the details — and they matter enormously for investment selection.

Who is actually buying — and who is selling

Leopold Aschenbrenner — the former OpenAI researcher whose “Situational Awareness” paper predicted current AI scaling — disclosed a 5.6% stake worth ~$2.8B in May 2026. When the person who wrote the definitive thesis on AI progress puts $2.8B into a single AI infrastructure stock, that is a signal worth understanding.— Form 13F, May 2026

The immediate catalyst: Nasdaq-100 inclusion

Valuation: paying for execution, not the possibility of it

The risks — serious and non-negotiable

My verdict: high conviction, aggressive sizing required

Nebius is the most compelling AI infrastructure pure-play in public markets. Hyperscaler validation, NVIDIA’s deepest partnership, 684% revenue growth, 45% EBITDA margins, a European sovereign moat no US competitor can replicate, and a founder with an extraordinary track record. The thesis is exceptional.

At $229, you are paying for perfect execution. The base case is fairly valued. The bull case — if $7–9B ARR is delivered by December — offers meaningful upside to $310–420.

Read the complete Master Equity Research Report v3.0 — full WACC, DCF, Contracted Backlog NPV, institutional data, and earnings quality forensics at stockswithpawan.com/nbis

⚠️ Disclaimer: For informational and educational purposes only. Not financial advice. NBIS is pre-FCF with negative free cash flow, significant debt, and ongoing dilution. Always DYOR and consult a qualified financial advisor.

Last week, during SpaceX () Q2 2026 earnings, one number that caught investors attention was the Capex increase for the AI segment.

Again this quarter, SpaceX has spent a massive amount of capital expanding its AI segment, just as expected. But it can still be hard to grasp just how aggressively the company has chosen to go all-in on AI overnight.

Before SpaceX merged with xAI, the company’s previous bet had been connectivity during the Covid lockdown and it turned out to be a very successful one. Just as for AI, SpaceX was pouring a lot of capital into Starlink, betting that connectivity could become a massive and highly profitable business, allowing them to fund the Starship program (closing to loop to deploy even more satellites at scale). And that bet worked as the company generated $791 million in net income for full year of 2024.

This quarter, the company reported $1.66B of operating income for its connectivity segment, up 79% from $923M year-over-year.

Before connectivity, SpaceX’s big bet was achieving full reusability. While the Falcon 9 is currently only partially reusable, with only its first stage being recovered and reused, Starship is designed to be the first fully reusable orbital launch vehicle in the world.

During Starship’s 13th flight test last month, the company successfully demonstrated the next generation of Starlink satellite (the V3), with its successful deployment and at the same occasion delivered in a single launch more payload mass than Rocket Lab has launched in its entire history.

The flight test also exceeded engineers’ expectations, as the Ship performed for the first time a very soft touchdown in the Indian Ocean. This gave the recovery team the opportunity to inspect very closely the second stage of the vehicle for the first time to iterate and perfect its design for future missions, as SpaceX plans to catch the Ship with the launch tower (just as it already does with the Super Heavy booster!)

By achieving reusability, SpaceX has access to space in a very cost-effective way that no other company currently has. By achieving full reusability with their Starship rocket, each launch would cost only fuel, maintenance and other operating costs, similar to a commercial plane, while traditional expendable rockets require a new vehicle for every single launch, which is a huge cost disadvantage for these other companies.

SpaceX’s Starmind Could Become a Major AI Profit Driver

SpaceX has recently unveiled Starmind, an AI satellite with the goal of capturing solar energy in space to power low-cost, high-performance AI compute for Earth, while leveraging the existing Starlink constellation.

To manufacture these AI satellites at massive scale, SpaceX is currently building the Gigasat Factory in Bastrop, TX. and expects it to enable rapid production and deployment of thousands of AI satellites starting as soon as late 2027.

A $55 billion manufacturing complex and thousands of AI satellites being launched into orbit as soon as late next year sound like a lot more research & development and capital expenditure coming for the AI segment!

It’s true that SpaceX will be burning a ton of cash to make Starmind possible, but I think we should not overlook the revenue growth potential once these AI satellites are deployed, because it is very comparable to Starlink.

As proven with Starlink and other satellite connectivity constellations like , which has an OEBITDA margin of about 57% and which Rocket Lab agreed to acquire at an $8B valuation, once the infrastructure is deployed, depending on the demand and how cost-effective the infrastructure is to operate on the ground, this can become a passive income generator for the company.

Why Put AI in Space?

Per SpaceX, Starmind’s primary use case is to move heavy AI computation and machine learning workloads off Earth and into low-Earth orbit, bypassing terrestrial limitations such as land availability, water cooling and local power grids.

The bottlenecks of AI datacenters on Earth at the moment are the continuous operating costs required to run the infrastructure, such as water cooling and local power grids. But in space, there are no such limitations, which could significantly reduce the operating costs of running AI infrastructure.

With the growing demand of AI in the years to come, Starmind could become, similar to Starlink, an enormous revenue driver for SpaceX’s AI segment, with manufacturing and deployment being the main operating costs, just as they are for Starlink, the company’s sole profitable segment at the moment.

The demand for SpaceXAI products is already proving strong. This quarter, AI revenue was up 213% from Q1 2026 and 247% year-over-year to $2.6 billion, primarily driven by increased AI solutions and infrastructure revenue from new Cloud Service Agreements with Google, Anthropic and Reflection AI, as well as growth in Grok and X subscription revenue.

The Kardashev Scale

What Elon Musk is ultimately trying to achieve with SpaceX is to bring humanity closer to a Kardashev Type II civilization, being able to harness a fraction of the energy produced by our Sun.

For reference, we are still a Type 0 civilization today, as we are nowhere close to harnessing all the energy available to us on Earth.

Why Would We Want to Harness a Fraction of the Sun’s Energy?

So why does this matter so much for Musk and everyone at SpaceX to reach what seems an impossible goal to achieve? Musk says that energy is what allows civilization to scale. The more energy we have access to, the more we can manufacture, compute, automate, transport and build. The more energy we have, the more technology we can build, and the better our technology becomes, the more efficiently we can use that energy.

For example, having much more available energy could allow us to produce more steel, aluminum, batteries, semiconductors and other energy-intensive materials.

Semiconductor factories in space? Not sci-fi! Varda Space is actively manufacturing drugs in space.

For Musk, unlimited energy is ultimately what could allow civilization to keep scaling without being limited by the amount of energy and resources we can produce on Earth.

Money won’t matter in the future?

Terafab

In addition to the Gigasat factory where Starmind or the AI satellites will be built, the powerful semiconductors that go into them for future generations will later be manufactured in a 100 million square-foot factory called Terafab.

Terafab will be a vertically integrated semiconductor fabrication plant in Grimes County, Texas, in a joint initiative involving SpaceX and Tesla. Per the Office of the Texas Governor, the first phase of the Terafab project represents a capital investment of more than $16.8 billion and will create 3,000 new jobs.

It looks sci-fi but SpaceX actually broke ground on the facility just last week, so things are moving fast!

Until Terafab’s dedicated semiconductor manufacturing lines become fully operational, the first version of Starmind that is expected to be deployed late next year will rely exclusively on NVIDIA Rubin GPUs and Vera CPUs..

What It Takes to Unlock Gigawatt-Scale AI Compute

SpaceX noted three fundamental technological limitations that must be overcome to unlock gigawatt-scale AI compute in space:

Why the Nvidia Partnership Matter

During SpaceX’s earnings call, Musk said that for Starmind the company was committed to sourcing its GPUs from a single supplier… and of course it is Nvidia ()

Both SpaceX and Nvidia have dominant positions in their respective markets (arguably approaching quasi-monopolies). SpaceX is able to launch and deploy massive satellite constellations at scale (as proven with Starlink) and super cost-effectively thanks to their Falcon 9 and soon Starship rocket, while Nvidia dominates AI computing and more specifically the manufacture of high-performance chips at scale.

Could SpaceX Actually Hit $1T in Revenue by 2030?

During last week’s earnings call, Musk stated that SpaceX expects to hit $1 trillion in annual revenue by 2030, pulling the target forward a full year from previous projections of 2031. He also noted a non-zero chance of hitting the milestone as early as 2029.

Read more

“KLA wins no matter who wins. It doesn’t matter which hyperscaler wins, which semiconductor company wins, or which processor wins. AI doesn’t happen without process control.”

^{- KLA Corporation, 2026 Investor Day}

ASML makes the machine that prints the pattern. Applied Materials deposits the materials. Lam Research etches the structures. And KLA? KLA determines whether any of it actually worked - at the molecular level, across thousands of steps, on wafers worth up to $20,000 each. Still don’t get it? Let me dumb it down: KLAC makes machines that inspect chips for tiny defects, helping factories improve yields and produce advanced chips reliably. In this day and age a mistake can cost billions , KLAC is there to make sure theres no mistakes.

Subscribe now

That is why this is the highest-margin business in semiconductor equipment, and why its competitive position has only strengthened as chips become more complex. Q4 FY2026 results just confirmed it again - record revenue of $3.66B, EPS beat of $0.03, and full-year revenue of $13.58B.

KLA does not make the chip. It determines whether the chip can be manufactured profitably. The more expensive the wafer, the more valuable KLA’s answer becomes.

What KLA actually does?

Every modern chip requires hundreds to thousands of manufacturing steps. At each step, something can go wrong. A particle of dust. A misalignment of 0.001 microns. A chemical layer that’s 0.2nm too thin. Any of these defects - if undetected - will propagate through subsequent steps, eventually killing the chip entirely. And on an advanced AI accelerator wafer worth $15,000–$20,000, “kill rate” is everything.

Market Share

Before the numbers: your research cited 60–80%, which requires clarification. The overall process control category is ~56.5%. But individual sub-segments are where the 60–80% (and higher) figures are real:

KLA is roughly 6.5× larger than its nearest competitor in process control. That is not a normal competitive gap. That is a structural monopoly in an essential manufacturing step.

^{- KLA Corporation, 2026 Investor Day disclosure}

KLA vs. AMAT vs. LRCX: Why the Moat is different

Applied Materials and Lam Research are excellent companies. This is not a dismissal of them. But the nature of their competitive position is structurally different from KLA’s - and that difference shows up in gross margins, market share stability, and customer behavior.

The key insight: AMAT and LRCX sell tools that physically alter silicon. Fabs intentionally dual-source deposition and etch equipment - if Lam’s price is too high, TSMC can buy from Applied or Tokyo Electron instead. KLA sells measurement and diagnosis. There is no second option that fabs trust with $20,000 wafers. Swapping out a proven KLA recipe mid-process risks catastrophic yield loss worth hundreds of millions. That is a different class of moat.

Why AI makes KLA more important, not less

The market treats KLA as a secondary AI beneficiary. That framing is wrong in an important way.

The process control intensity trend - the most important long-term number

Process control as % of total wafer fab equipment spending:

Q4 FY2026 - the numbers released July 28, 2026

Breaking: Q4 FY2026 results (July 28, 2026) - EPS beat of $0.03 ($1.05 vs. $1.02 estimate). Revenue of $3.66B, slight miss vs. $3.67B estimate. Stock dropped ~9% in after-hours - a reaction to FY guidance framing and broader chip sector selloff, not fundamental deterioration. Q1 FY2027 guidance: revenue $4.0B ±$200M, EPS $1.16 ±$0.10 - both above consensus.

The 61.3% gross margin deserves its own moment of attention. Applied Materials is at ~49.9%. Lam Research is at ~47.5%. KLA runs 11–14 percentage points higher gross margin than its closest peers - not because it has lower costs, but because it charges more and customers don’t push back. That is the definition of pricing power.

Peer Financial Comparison

Risks - read before sizing

Valuation and price framework

⚠️ Disclaimer: For informational and educational purposes only. Not financial advice. KLAC is a cyclical semiconductor equipment company that can experience sharp drawdowns even with strong fundamentals. Market share data sourced from company presentations, Gartner, and TechInsights - subject to periodic revision. Always DYOR and consult a qualified financial advisor.

Technical Analysis & my Buy zone

Read more

A lot of people are bearish on August–October because of weak seasonality, midterm elections, rate hike fears, and geopolitical risk. I was bearish too, but I’m starting to change sides and take the contrarian view.

Join Discord for Faster alerts & Data

A lot of damage has already been done under the surface. High-beta and momentum stocks got crushed, leverage was flushed out, sentiment turned extremely negative, and many investors are now expecting another major leg lower.

At the same time, earnings are still holding up, AI capex remains strong, chip demand is healthy, inflation pressures are cooling, and rate hike expectations are starting to come down.

Today’s jobs data supports the Fed staying patient. Payrolls fell 23K vs. +80K expected, pushing September & 2026 hike odds below 50%. Bonds, stocks, and gold are rallying as short-term yields fall.

That said, the economy still looks solid, with unemployment falling and Atlanta Fed GDPNow tracking Q3 growth near 5.8%.

That doesn’t mean the market goes straight up from here. I still expect volatility and pullbacks, but I think the risk/reward is starting to improve.

When everyone starts expecting the same outcome, I usually start looking the other way.

^{Disclaimer: This is not financial advice. I am not a financial advisor. I’m simply sharing my personal opinions, research, portfolio decisions and investing strategy for educational and informational purposes. Always do your own research and make decisions based on your own financial situation and risk tolerance.}

Here are 6 reasons I’m becoming more bullish:

1. Huge margin calls already happened. A lot of leverage and speculation got flushed out, including Korea and highly leveraged traders.

2. Oil prices are falling, rate hike expectations are cooling, and inflation is starting to ease despite what many people think.

3. Momentum stocks just went through one of their worst sell-offs ever, even worse than parts of the Dot-Com era.

4. A lot of the new content creators have suddenly gone quiet. Fear and negativity are already extremely high. They are scared to sound overly bullish

5. AI capex spending is still strong, chip demand remains healthy, and shortages are still showing up across key parts of the supply chain.

6. Earnings remain strong, with roughly 90% of S&P 500 companies beating earnings expectations so far.
   

CASE STUDY OF THE AI BUBBLE

Right now, a lot of people are calling the top in memory and AI stocks, most notably names like Micron and Marvell. But history shows that major technology cycles rarely move straight up — and big corrections don’t automatically mean the trend is over.

Look at Cisco during the internet boom.

From roughly 1994–1997, CSCO exploded from around $0.80 to $9, then dropped nearly 40%. At the time, people called the top, said the internet was a bubble, and believed the run was finished.

Instead, Cisco V-shaped and rallied from roughly $5 to $18 between 1997–1998.

Then it crashed another ~40%, and once again investors said: “That’s it. This is the top. Sell everything.”

But the long-term technology cycle still wasn’t finished.

The lesson: even the biggest secular winners can suffer brutal 30–40% corrections multiple times before the real cycle ends. Similar to what we are seeing now.

My outlook for Aug- Sept remains bullish in the near term and believe the S&P 500 can reach 7,900–8,000 in August.

One major catalyst still ahead, in my view, is a decline in rate-hike fears and inflation concerns. If those fears fade, bond yields can move lower, financial conditions ease, and stock market valuations have room to expand.

Winston Churchill once wrote that war offices are often “preparing for the last war.” Financial markets tend to do the same thing: investors become so focused on the previous crisis that they use the wrong framework for what comes next.

We’ve seen this repeatedly:

• Post-Dot-Com: Investors spent years believing technology valuations were permanently broken. They weren’t.

• Post-GFC: Investors argued stocks would struggle for decades under a “new normal.” That proved far too pessimistic.

• Post-2022: Investors remain obsessed with the idea that inflation is about to come roaring back.

Today, that fear is showing up in expectations for another Fed rate hike.

Markets are still pricing a meaningful probability of a September hike, but I think those odds are too high. My base case is that the Fed does not hike in September.

The key tests are the upcoming economic data, particularly CPI.

If employment continues to come in softer than expected and core inflation continues cooling, September hike expectations should fall sharply.

That could push Treasury yields lower, support P/E expansion, and give equities another leg higher.

My August target remains S&P 500 7,900–8,000.

Inside My Portfolio: What I’m Buying, Selling & Positioning for Next

Read more

Its trailing-12-month free cash flow fell from positive $18.2 billion to negative $7.6 billion.

The stock responded by surging 15.3%.

That sounds contradictory until you look at one number: Amazon Web Services revenue accelerated 37% year over year, its fastest growth in 18 quarters.

The market’s message was clear.

Wall Street does not hate AI spending. It hates spending without measurable returns.

This may be the most important change in the AI trade since the infrastructure boom began.

SPONSOR POST

Disclosure: This post is sponsored by Tema ETFs. The information shared is for educational purposes only and should not be considered financial, investment, tax, or legal advice. Always do your own research and consult a qualified financial professional before making investment decisions.

THE BREAKDOWN

• Amazon reported Q2 2026 net sales of $200.6 billion, up 20% year over year.

• AWS revenue reached $42.2 billion, up 37% and above analysts’ expected growth rate of approximately 31.2%.

• AWS produced roughly 61% of Amazon’s operating income from only about 21% of company revenue.

• Amazon increased its planned 2026 capital spending by 10% to $220 billion.

• Trailing-12-month free cash flow fell to negative $7.6 billion.

• AMZN closed July 31 at $271.58, up 15.32%, on approximately 2.5 times its average volume.

At least 15 brokerages raised their Amazon price targets after the report.

What actually happened

The market entered Big Tech earnings worried that the AI infrastructure boom was becoming too expensive.

Alphabet intensified that concern on July 22.

Revenue increased 24% and Google Cloud revenue jumped 82%, but quarterly CapEx reached $44.9 billion, free cash flow was negative $5.9 billion, and full-year CapEx guidance increased to $195–$205 billion.

Alphabet’s shares subsequently fell about 7%.

Meta added to the anxiety on July 29.

Revenue grew 28%, but operating income declined 8%, quarterly free cash flow fell to just $784 million, and the company’s 2026 CapEx outlook became $130–$145 billion.

Meta also fell about 7% after investors questioned how quickly the spending could produce direct returns.

Microsoft then changed the mood.

Azure revenue grew 43%, quarterly free cash flow remained positive at $19.6 billion, and management forecast approximately 45% constant-currency Azure growth for the next quarter.

Microsoft gained more than 15% on July 30 and added roughly $450 billion in market value.

Amazon delivered the final answer on July 30.

AWS operating income rose 64% to $16.6 billion, Amazon’s total operating income increased 43% to $27.5 billion, and its cloud backlog reached $496 billion.

On July 31, AMZN gained 15.32% and closed at $271.58.

Reported figures and management guidance are confirmed in Amazon’s

One important warning: Amazon’s reported net income of $62.6 billion and diluted EPS of $5.75 included $53.4 billion of non-operating pre-tax income, primarily related to its Anthropic investment.

That gain should not be confused with Amazon’s normal operating performance.

What Wall Street may be missing

The most overlooked number was not Amazon’s EPS.

It was the $496 billion AWS backlog.

Backlog represents business that customers have contracted but Amazon has not yet recognized as revenue. It is not the same as guaranteed near-term sales because customers may use those commitments over several years.

However, Amazon’s backlog increased from $364 billion in the previous quarter—approximately 36% sequential growth.

Management also said the majority of available AWS capacity for 2027 and some 2028 capacity had already been reserved by customers.

That changes how investors can interpret the $220 billion spending plan.

Amazon is not simply constructing data centers and hoping demand appears. Based on management’s statements, a meaningful portion of the future capacity is already connected to customer commitments.

Here is the most shareable statistic from the entire report:

AWS represented only about 21% of Amazon’s Q2 revenue, yet it generated roughly 61% of total operating income.

That is why a 37% AWS growth rate matters far more than its revenue share suggests.

Leave a comment

What management said

CEO Andy Jassy said Amazon would still lack enough capacity to satisfy all expected 2026 demand—even after increasing its spending plan.

Management also reported that Amazon’s AI business and custom-chip business had each exceeded $25 billion annual revenue run rates, with both growing at triple-digit percentages.

Jassy explained that data centers can require approximately two years of spending before becoming operational. He also said AI servers may recover their cost in under three years and continue generating profit afterward.

These are management estimates, not guaranteed outcomes, but they explain the timing gap between cash spending and future revenue.

The cash-flow periods above are not identical, so they should not be treated as perfect like-for-like comparisons.

They are useful because they show what investors emphasized in each report

Market reaction

AMZN closed at $271.58 on July 31, up 15.32%.

Approximately 129.1 million shares traded—roughly 2.5 times the stock’s 65-day average volume.

That indicates the move involved unusually broad participation rather than thin trading.

At least 15 brokerages raised their price targets after the report, according to Reuters.

Reliable post-earnings data confirming specific unusual-options trades, changes in open interest, or a material short-interest shift could not be confirmed from a primary exchange source by publication time.

The broader market reaction was more complicated.

The S&P 500 gained 0.70% and the Nasdaq rose 1%, but declining S&P 500 stocks outnumbered gainers by approximately 1.3 to one.

Index strength was concentrated in large technology winners.

The Federal Reserve had also maintained its policy rate at 3.50%–3.75%, with three officials favoring an increase. Markets priced roughly a 65% probability of a September rate increase, raising the valuation hurdle for growth stocks.

Bull case

The bullish thesis depends on AWS remaining the clearest proof that Amazon’s AI investments are generating measurable returns.

• AWS growth accelerated to 37%.

• AWS operating margin reached approximately 39.4%.

• Backlog reached $496 billion.

• Future AWS capacity is reportedly heavily reserved.

• Amazon’s AI and custom-chip businesses each exceeded $25 billion annual revenue run rates.

The bull case strengthens if AWS growth and margins hold, backlog continues expanding, and AMZN confirms above $278.56.

After a 15% earnings gap, future upside requires continued execution—not simply more AI announcements.

Bear case

The bear case starts with cash flow.

Amazon’s trailing-12-month purchases of property and equipment increased 64% to $169 billion, pushing free cash flow to negative $7.6 billion.

If AWS growth slows before new data centers become productive, Amazon could face heavy spending, rising depreciation, and weaker returns.

Additional risks include:

• The $53.4 billion Anthropic-related gain inflated reported net income and could be volatile.

• Memory-chip inflation may make infrastructure more expensive.

• Backlog may take years to become recognized revenue.

• Higher interest rates can compress technology valuations.

• Competition from Microsoft Azure and Google Cloud remains intense.

A breakdown below $262.01 could turn the earnings gap into a short-term exhaustion move.

EagleTradesX view

My focus is not on chasing AMZN after a 15% earnings gap or trying to predict every move.

I am watching how buyers and sellers react around the key levels.

The immediate structure remains strong while buyers defend $262.01.

A break and hold above $273.23 would bring the $278.56 high into focus. A loss of $262.01 would increase the probability of a deeper gap retracement toward $239.82.

The bigger market lesson is more important than one stock.

AI is no longer a simple theme where every company rises because it announces higher spending.

Investors are now separating companies that can show direct revenue, margins, backlog, and cash generation from companies asking the market to wait for future returns.

This setup becomes more interesting only after confirmation.

A major slowdown in AWS growth or a confirmed break below the earnings gap would change the thesis.

Final verdict

Amazon reported a strong quarter, but the real story was not its headline EPS.

The company increased its 2026 capital-spending plan to $220 billion and allowed free cash flow to turn negative.

Investors accepted that trade-off because AWS growth accelerated to 37%, AWS operating profit climbed 64%, backlog reached $496 billion, and much of the company’s future capacity was already reserved.

Stay Ahead of the Next Major Market Move

Markets rarely warn everyone before the biggest opportunities appear.

Follow EagleTradesX on Instagram for:

📊 Daily market watchlists
📈 Technical chart analysis
🌍 Major macroeconomic updates
🦅 Clean setups with risk management first

Follow EagleTradesX — link in bio.

Subscribe now

Disclaimer: This article is provided for educational and informational purposes only. It is not financial advice or a recommendation to buy or sell, securities or any financial instrument. Markets involve substantial risk. Always conduct your own research and manage risk according to your personal financial circumstances.

Subscribe now

The real EPS was : ~$2.85 vs Est. $2.90 (miss) let me explain. Google’s $9.11 reported EPS is heavily inflated by a large paper gain on its investments. Google reports a $98B unrealized gain on its
$SPCX stake as of the end of Q2 & this is why Google isn't pumping 10% because they still have to wait end of the year to sell this stake.

Googles Cloud backlog just hit $467.6B, up 406% from last year. In simple terms, customers have already committed to spending hundreds of billions with Google Cloud in the future. That is a big deal because it gives Google much better visibility into future growth. Cloud revenue also grew 63% and passed $20B in a quarter for the first time. Backlog is not revenue today, but it shows demand for Google’s AI and cloud services is still accelerating.Management also said existing customers are consuming more than 50% above their original commitments while new customer acquisition has more than doubled YoY.

That demand is already translating into insane fundamentals with Cloud growing 82% while operating margins approach 36%.

Googles Capex continues to rise! $45B this quarter & $132B ytd up 100%. This is very Bullish for AI stocks & compute like $MU $NBIS $AMD $NVDA . Google is spending heavily because AI demand is growing faster than its available capacity. Management now expects $180B–$190B in 2026 CapEx, with spending increasing again in 2027. Demand is justifying one of the largest capex programs in corporate history with Cloud growing 82% and generating more operating income in a single quarter than it produced in total revenue less than three years ago.After hours we are already seeing chip stocks pump so this is definitely a positive.

Technical Analysis & buy zone

Google could go to test the trendline it’s been holding since April 2025 (liberation day). Our long-term buy zone remains 300-350 with a deep buy at 270 if Google breaks below 308. The long-term setup still looks attractive, but the stock needs to show more strength before the short-term chart improves.

Watch $341. Falling below that could push the stock toward $330, and a break there may open the door to $315.

Those lower levels could offer attractive buying opportunities. On the upside, a move back above $375 would be bullish and could lead to a test of $400. Charts still show a very nice uptrend for GOOGL, so I like owning and using any dips to buy. The area which lines up with the long-term trend lies near $315, but expect GOOGL to be higher into year-end and like buying weakness, if/when this happens after earnings.

We continue to love Google long-term & will continue to invest in this company. AI does not look fake, but parts of the market may still be overpriced. For AI stocks, this supports continued demand for chips, servers, networking, power and data centres. However, Google raised 2026 spending to $195–$205 billion, hurting free cash flow and showing how expensive the AI race has become. Google’s future looks strong if Cloud growth eventually justifies the massive spending.

After reaching an intraday record above $5,500 per ounce in January 2026, gold briefly dropped below $4,000 in late June. That represents a decline of approximately 25% from its peak.

To most investors, that looks like a broken trade.

But underneath the price decline, something completely different is happening.

China is importing gold at its fastest pace in more than two years. Central banks expect global gold reserves to keep rising. India is strengthening regulated, physically backed ways to own gold.

The price is falling—but the world’s biggest buyers are accumulating.

That raises one important question:

What if this is not the end of the gold bull market?

What if this correction is preparing gold for its next historic move?

China Is Buying the Dip

China’s latest customs data delivered a massive signal.

The country imported approximately 163 tonnes of gold in May—the highest monthly amount since March 2024. During the first five months of 2026, total imports reached roughly 692 tonnes, an increase of 76% compared with the same period last year.

This is not just government buying. It includes demand for physical bullion, gold bars and private accumulation plans inside China.

Subscribe now

The People’s Bank of China is buying too.

China’s central bank added approximately 10 tonnes to its official reserves in May, its largest reported monthly purchase since December 2024. That pushed its official gold holdings to around 2,332 tonnes—approximately 9% of the country’s total reserves. It was also China’s nineteenth consecutive month of reported gold accumulation.

Think about the timing.

Gold had already fallen sharply from its January high. Investor momentum was weakening and gold ETFs in China were experiencing outflows.

But instead of running away, China increased its physical exposure.

Retail traders often chase strength.

Strategic buyers accumulate weakness.

China May Be Preparing for a Different Monetary Future

Why would China want so much gold?

Because gold is more than a commodity. For governments, it is a reserve asset that does not depend on another country’s promise to pay.

China holds a substantial amount of U.S.-dollar assets. That creates exposure to American interest rates, fiscal policy, sanctions and the long-term purchasing power of the dollar.

Gold carries no direct counterparty risk. It cannot be printed by a central bank, frozen by a foreign government through the same mechanisms as financial assets, or diluted by deficit spending.

That makes it useful in a world becoming more divided economically and politically.

This does not mean the dollar is about to disappear. It remains the world’s dominant reserve and transaction currency. However, the dollar’s share of global reserves has gradually declined over the longer term as central banks diversify.

The important story is not the sudden death of the dollar.

It is the gradual search for alternatives—and gold remains the most established neutral reserve asset available.

The U.S. Is Trying to Balance Three Conflicting Goals

America is currently attempting to achieve three difficult objectives:

1. Rebuild domestic manufacturing.

2. Protect consumers from inflation.

3. Keep American exports globally competitive.

Tariffs and protectionist policies may encourage companies to manufacture inside the United States. But they can also raise input costs and consumer prices in the short term.

At the same time, a very strong dollar can make American products more expensive for international buyers, weakening export competitiveness.

This creates a complicated balancing act.

Some investors believe the U.S. may eventually tolerate a softer dollar to support manufacturing and exports. However, that is a market theory—not an announced policy. Treasury Secretary Scott Bessent has publicly stated that industrial strength and a “strong dollar policy” are both part of the administration’s strategy.

Still, currency values are determined by more than official statements.

Interest rates, deficits, inflation, global capital flows and market confidence all matter. If real interest rates decline or concerns surrounding U.S. debt rise, the dollar could weaken even without an official devaluation policy.

Historically, a weaker dollar and falling real yields have created a favourable environment for gold.

Central Banks Are Sending a Bigger Signal

China is not alone.

In the World Gold Council’s 2026 survey:

* 89% of reserve managers expected global central-bank gold holdings to increase over the following 12 months.

* A record 45% expected their own institution to increase its gold reserves.

* 74% expected the dollar to represent a moderately or significantly smaller share of global reserves within five years.

* 83% expected gold to account for a larger share of reserves five years from now.

Those are extraordinary numbers. Central Bank Gold Reserves Survey 2026⁠

Central banks are not trying to catch a quick trade.

They are positioning for the next decade.

India Is Quietly Changing How Gold Is Owned

Another major development is happening in India—one of the world’s largest gold-consuming nations.

In November 2025, India’s securities regulator, SEBI, warned investors about unregulated “digital gold” platforms. The regulator said these products could expose investors to counterparty and operational risks because they fall outside its investor-protection framework.

India is now expanding regulated Electronic Gold Receipts.

EGRs can be held in a demat account like securities, but each receipt represents ownership of underlying physical gold stored with an accredited vault manager. NSE began live EGR trading in May 2026.

This does not mean regulators are warning against every form of “paper gold.” Regulated gold ETFs and futures continue to play an important role.

But it does reveal something important:

Investors increasingly want transparency regarding whether their gold exposure is properly regulated and supported by identifiable assets.

Could Paper and Physical Gold Separate?

During normal markets, futures, ETFs and physical bullion remain closely connected through arbitrage.

But during periods of extreme demand or supply disruption, local premiums can expand. Coins and bars may trade above the international spot price because of refinery constraints, transportation problems, taxes or shortages.

That does not automatically mean the global gold market is failing.

It means physical availability can become more valuable when everyone wants delivery at the same time.

If central banks, institutions and retail investors all accelerate purchases together, physical premiums could rise even if the quoted spot price initially moves more slowly.

That is the scenario worth watching—not an automatic collapse of “paper gold,” but a possible rush toward directly backed and deliverable exposure.

Why This Correction Could Become an Opportunity

Gold’s drop from above $5,500 to around $4,000 has removed a large amount of speculative excitement from the market.

Read more

It only needs buyers to stop defending the same levels one more time.

Retail traders are still asking where to buy the dip. Meanwhile, corporate insiders—the executives and directors who understand their businesses better than anyone—are buying far less aggressively than normal.

At the same time, technology leadership is weakening, growth funds are losing money, oil is surging and some of the market’s biggest companies are about to report earnings.

That doesn’t guarantee a crash.

But it does tell me this is not the time to trade without a plan.

Friday’s Heat-map Was a Warning

Friday’s selloff wasn’t caused by one broken stock.

The weakness spread across technology, semiconductors, communication services, financials and higher-beta consumer names.

NVDA fell 2.21%. META dropped 2.79%. GOOGL lost 2.17%. TSLA fell 2.61%. HOOD dropped 5.72%, while NFLX got crushed by more than 7%.

That is broad de-risking—not an isolated earnings reaction.

Friday’s heatmap shows selling concentrated in technology, communication services and higher-beta names.

Share

The fund-flow data supports what we’re seeing on the screen.

Growth funds recorded approximately $7.18 billion in weekly outflows, while value funds attracted about $3 billion. The semiconductor index also lost roughly 8.5% for the week.

The same crowded technology stocks that carried the indexes higher are now being sold together.

That matters.

The Red Flag Nobody Wants to Discuss: Insider Selling

Corporate insiders don’t always sell because they expect their stock to fall.

They sell for taxes, diversification, estate planning, option exercises and scheduled Rule 10b5-1 trading plans.

So one executive selling shares means very little by itself.

But widespread selling combined with weak buying is different.

The overall U.S. market insider buy/sell ratio is currently approximately 0.23, compared with a long-term average of 0.39.

That puts the current ratio about 41% below its historical average, meaning insider buying remains unusually weak relative to selling

.Earlier this year, Reuters reported that the seller-to-buyer ratio reached 4.2—its highest level in 20 months.

During that period, there were 2,260 recorded instances of insiders selling versus only 543 instances of buying. The estimated value of shares sold was approximately $6.6 billion.

Those datasets use different methodologies, but they point in the same direction:

Insiders have been much more willing to sell than buy.

That is not a perfect market-timing signal. Insiders can sell for many personal reasons.

But insiders normally buy shares with their own money for one main reason: they believe the stock offers value.

When the market is near record highs, insiders are selling, growth funds are experiencing outflows and retail traders are still chasing—it deserves attention.

Most insider transactions must be publicly reported through SEC Form 4 filings, generally within two business days.

The insider activity I pay closest attention to is:

• Multiple executives selling around the same time

• Large discretionary open-market sales

• Selling that accelerates after a major rally

• Executives reducing a meaningful percentage of their ownership

• Very little insider buying across the same company or sector

• Sales not clearly connected to taxes, vesting or an existing 10b5-1 plan

Insider selling doesn’t tell us exactly when the market will drop.

But it can tell us that the people closest to these companies may not consider current prices attractive enough to buy.

My Base Case

I believe the market still has room to bleed more.

But I’m not blindly buying puts and predicting a crash.

I need confirmation from price.

SPY 740 and QQQ 686 are the two trapdoors.

If buyers defend them, we could see another squeeze higher.

If daily candles begin accepting below them, the next downside zones become real—not just dramatic numbers drawn on a chart.

SPY: 740 Is the Line in the Sand

SPY closed Friday at 743.29 after rejecting below the 760.40 all-time-high zone.

The chart is building lower highs beneath descending resistance, while 740 continues acting as the floor.

This is compression.

Compression normally leads to expansion. The question is which direction breaks first

A daily close below 740 would signal that buyers are losing control.

That opens the door toward 724, where horizontal demand meets the rising trendline.

If 724 fails, the entire structure weakens. From there, 702 and the high-690s near the rising 200-day moving average become the larger downside zone.

That doesn’t mean SPY will fall in a straight line.

We could see relief bounces, short squeezes and failed reclaim attempts along the way. Those failed reclaims are what I want to trade.

Bearish confirmation:

740 breaks and becomes resistance → 724 → 702 → high-690s

Bullish invalidation:

Reclaim 750–754 → retest 760.40

For the faster intraday plan, 744.01 is my main decision level.

Above 744.01:

745.33 → 746.60 → 747.36 → 748.72 → 750.24

A rejection brings:

742.57 → 740 → 738 → 736

The level itself isn’t the trade.

The reaction at the level is the trade.

QQQ: A Daily Close Below 686 Changes Everything

QQQ closed at 695.33 after dropping 1.50%.

Technology is already showing more stress than the broader market.

The Nasdaq-100 Volatility Index was 27.34 on July 16, compared with a VIX reading of only 16.73. Tech volatility is elevated, while the broader market still isn’t pricing complete panic.

That leaves room for broader volatility to catch up if support breaks.

The first QQQ pivot is 695/692.20.

The major structural support is 686.

QQQ has repeatedly found buyers around this area. But every time a support level gets tested, some of the demand sitting there gets used.

A daily close below 686 would turn old support into resistance.

If QQQ then attempts to reclaim 686 and fails, that would provide much stronger bearish confirmation.

Below 686, I’ll watch for reactions around 675–670.

The larger downside area is 650–642, where the rising 200-day moving average is approaching.

I’m not saying QQQ will immediately collapse to 642.

Markets rarely move in a straight line.

We could break 686, bounce toward 690–695, form a lower high and then continue lower.

That is the professional setup:

Break → failed reclaim → continuation

Bearish confirmation:

Daily close below 686 → failed reclaim → 675–670 → 650–642

Bullish invalidation:

Reclaim 700–702 → 715–720

For the faster plan, 692.20 is my pivot.

Above 692.20:

695.34 → 696.44 → 698.28 → 700.94

A rejection brings:

690 → 688 → 686 → 684

Earnings Could Decide Whether Support Holds

This is not a quiet earnings week.

Tuesday brings COF, NOC and SCHW.

Wednesday is the biggest day, with TSLA, GOOGL and IBM reporting.

Thursday brings INTC and NOK, followed by AXP on Friday

TSLA and GOOGL can reset sentiment across growth stocks.

INTC can influence positioning throughout the semiconductor sector.

COF and AXP will give us more information about consumer spending, credit conditions and potential loan stress.

The following week gets even heavier:

META, MSFT, AAPL, AMZN, ARM, QCOM, PYPL, LRCX and CVX are all on the calendar

That creates two major risks.

That creates two major risks.

First, companies can report good results and still sell off if expectations were already too high.

Second, traders may reduce exposure before the reports, putting additional pressure on technology stocks even before earnings are released.

I’m not interested in gambling through earnings with oversized positions.

I want to trade the market’s reaction after the numbers are released.

War Doesn’t Affect the Market Through Headlines Alone

The biggest macro risk is the U.S.–Iran conflict and its effect on global energy routes.

Brent crude finished Friday at $88.10, while WTI closed at $82.49. Both gained approximately 16% for the week as the conflict threatened oil flows through the Strait of Hormuz and surrounding shipping routes.

Before the conflict, approximately one-fifth of global oil supplies passed through the Strait of Hormuz.

But war doesn’t affect stocks through headlines alone.

The chain looks like this:

War → supply disruption → higher oil → higher transportation costs → higher inflation → fewer Fed cuts → higher yields → pressure on technology valuations

Higher energy prices raise fuel, freight and insurance expenses.

Companies must either absorb those costs, hurting profit margins, or pass them to consumers, keeping inflation elevated.

If inflation expectations rise, the market may reduce its expectations for interest-rate cuts.

Treasury yields can then move higher, putting pressure on expensive technology and growth stocks.

That is why war can hurt QQQ even when technology companies have no direct connection to the conflict.

Potential beneficiaries

Energy companies such as XOM, CVX and OXY can remain strong if crude stays elevated.

Defense companies may attract buyers as governments increase military spending.

Gold can benefit from safe-haven demand.

The U.S. dollar may also attract defensive flows.

Potential pressure areas

Airlines face higher fuel expenses.

Travel companies can weaken if consumers reduce discretionary spending.

Transportation businesses may face higher fuel, insurance and shipping costs.

Technology can struggle if yields rise and investors move away from expensive valuations.

Small-cap companies may experience additional pressure because they usually have less ability to absorb higher borrowing and operating costs.

The Fed Is the Next Major Risk

The next FOMC meeting is scheduled for July 28–29.

The Fed blackout period removes the usual stream of speeches and guidance before the meeting.

That leaves traders reacting more aggressively to earnings, oil, bond yields and economic data.

If oil continues rising, the Fed faces an uncomfortable situation.

Cutting rates could support economic growth—but it could also add to inflation.

Keeping rates elevated could control inflation—but it would pressure housing, consumers and corporate borrowing.

The worst combination for the market would be weaker growth while inflation remains elevated.

That is the stagflation risk traders cannot ignore.

Friday’s Economic Data Could Move the Market

Flash U.S. PMI is scheduled for 9:45 a.m. ET on Friday, July 24.

New-home sales are scheduled for 10:00 a.m. ET.

PMI will give us a faster reading on business activity, hiring, demand and price pressures.

New-home sales will show how consumers are handling mortgage rates and affordability.

Strong growth with controlled inflation would support the market.

Weak growth with lower inflation could increase rate-cut expectations.

But weak growth combined with sticky prices would be the ugliest result for risk assets.

Midterm History: Uncertainty First, Clarity Later

2026 is a U.S. midterm-election year.

Historically, these periods can bring weaker performance before the election and stronger returns after the results become clear.

A U.S. Bank study covering 31 midterm elections from 1900 through 2025 found:

• Average return during the 12 months before midterms: +2.9%

• Average return across all years: +8.9%

• Average return during the 12 months after midterms: +12.4%

But there is an important detail.

The study found the election effect was not statistically reliable enough to treat it as a guaranteed trading signal.
Economic growth, inflation, interest rates, earnings, war and financial stress had a larger influence on the weakest midterm years.

My interpretation is simple:

The election can add uncertainty and volatility, but it will not decide the market by itself.
The dangerous part of 2026 is the combination of midterm uncertainty, insider selling, war, higher oil, Fed pressure and weakening technology leadership happening at the same time.

History may rhyme. Price still has to confirm.

My Trading Plan & Watchlist

Follow EagleTradesX on Instagram

For daily watchlists, updated levels and new chart breakdowns throughout the week, follow EagleTradesX on Instagram:

@EagleTradesX

These weekend charts will be updated as price confirms or invalidates the original setup.

If this analysis helped, subscribe to the newsletter and share it with another trader who needs a plan before Monday’s open.

Subscribe now

Educational content only. This newsletter is not financial advice or a recommendation to buy or sell any security. Options involve substantial risk and can result in the loss of the entire investment.

Forgent Power Solutions IPO’d at $27 in February 2026. Five months later the stock trades near $47 and the backlog has grown to $2.4 billion. The company is fully booked for Q4 2026 and is operating at just 30% of its available production capacity. This is either the most asymmetric setup in AI infrastructure — or a warning about what happens when growth multiples meet a company that has never operated at scale.— SWP Equity Research, July 4, 2026

Most of you know Vertiv, most of you know Eaton. Almost nobody in retail investing is tracking Forgent Power Solutions (NYSE: FPS) — a custom electrical distribution equipment manufacturer that IPO’d in February 2026 and has since posted one of the most explosive growth trajectories of any industrial company to go public in the last decade. Here is the complete picture.

The business: the bottleneck nobody put in their AI infrastructure thesis

Vertiv makes power management and cooling. Eaton makes electrical distribution systems. Forgent makes the custom electrical switchgear, transformers, and power distribution units that sit between the utility grid and Vertiv’s equipment — the very first layer of the power chain inside every AI data center.

Without Forgent’s switchgear, the grid power cannot even enter the building. This is not a supporting role. It is the literal gateway through which all data center power must flow.

What Forgent actually makes

1. Data centers — 47% of backlog: Custom automatic transfer switches (ATS), gear eHouses, UPS eHouses, power distribution units, medium and low voltage switchgear, and substation transformers. Every AI data center that pulls more than 50 MW from the grid needs Forgent’s type of equipment to safely distribute that power at the rack level.

2. Power grid — utility and industrial Padmount transformers, switchboards, paralleling switchgear, and medium voltage VPI transformers for grid modernization, industrial facilities, EV charging infrastructure, and the broader energy transition. This is the grid modernization layer that complements the AI data center thesis.

3. The agile manufacturing edge: Forgent’s core competitive claim is industry-leading lead times on custom equipment. When hyperscalers need switchgear delivered in 12 weeks instead of 40, Forgent is the supplier they call. The book-to-bill of 2.3× in Q3 is direct evidence that this lead time advantage is being competitively recognized.

Every AI data center needs power delivered at scale from the utility grid. Before it reaches NVIDIA’s GPUs or Vertiv’s cooling systems, it passes through Forgent’s switchgear. This company is the first node in the AI power chain — and it is the least-followed name in the entire stack.

My Thesis & Why I bought

I want more exposure to the power grid theme behind AI.

The simple thesis is AI data centers need massive amounts of electricity before anything else. That means more demand for substations, transformers, switchgear, power systems and electrical equipment.

Forgent is benefiting from that bottleneck.

Last quarter earnings was strong:

Revenue: $379M, up 103% YoY
Bookings: $867M, up 308% YoY
Backlog: $1.98B, up 157% YoY
Book-to-bill: 2.3x

That means demand is coming in much faster than the company can ship.

What I like most is that Forgent is not just selling basic equipment. They are helping customers get power to data centers faster. In a market where delays can last 12–24 months, speed is a huge advantage.

Read more

Micron just reported gross margins higher than NVIDIA’s. Not close to NVIDIA’s — higher. From a company that makes memory chips, not AI accelerators. That single fact tells you everything about how tight the AI memory market has become.

Micron’s fiscal Q3 FY2026 print, released June 24, didn’t just beat estimates — it obliterated them. Revenue tripled ye…

Read more

“Total revenue increased 83% compared to 2024 to a record $456 million. We are targeting over $1 billion in revenue for 2026, driven by our 800G and 1.6T transceiver ramp.”— Dr. Chih-Hsiang (Thompson) Lin, CEO & Founder, Applied Optoelectronics

After years of watching semiconductor transitions play out, I can say this one is different. The shift from copper to fiber-optic interconnects inside AI data centers isn't gradual — it's physics-forced. Copper can't move data fast enough, far enough, or efficiently enough to connect tens of thousands of GPUs in a modern training cluster. Fiber isn't optional. Applied Optoelectronics (NASDAQ: AAOI) is one of the few companies in the world that makes the laser chips these transceivers require — in-house, in Texas, at scale.

Subscribe now

The stock has returned +441% year-to-date as of June 9, 2026. That is not a typo. From a 52-week low of $15.29 to a high of $233.67. And the fundamental story — if management executes — still has a meaningful runway ahead.

The business

Every time Microsoft, Amazon, or Google builds a new AI data center, they install racks of GPU servers. Those servers need to communicate — transferring enormous amounts of data between GPUs at speeds that make standard copper cables useless beyond a few meters. The answer is fiber-optic transceivers: small modules that convert electrical signals into beams of light, send them down optical fiber, and convert them back. AAOI makes those modules.

What makes AAOI structurally different from most transceiver companies is vertical integration. AAOI designs and manufactures its own Indium Phosphide (InP) laser chips — the single most expensive and difficult component inside a transceiver — at its Sugar Land, Texas facility. Competitors buy lasers externally. AAOI makes them. That is the cost advantage, supply chain control, and customization speed that has made AAOI the first-choice 800G supplier for Microsoft’s AI buildout.

The numbers: 5 consecutive record quarters

The quarterly progression is the story of a company arriving at an inflection:

• Q1 2025: $99.9M

• Q2 2025: $107.8M

• Q3 2025: $116.6M

• Q4 2025: $131.4M (+31.2% GAAP gross margin, record at the time)

• Q1 2026: $151.1M (+51.3% YoY) ✅ 5th consecutive record quarter

Q1 2026 gross margin dipped to 29.1% from 31.2% in Q4 2025. This is the number skeptics point to. The explanation is real: new product ramps always carry higher initial costs. Management guided 35% gross margin by Q4 2026 and 40%+ by 2027. If they deliver, the margin-compression narrative dissolves entirely.

The Q2 2026 guidance is $180–$198M (midpoint $189M) — another step-up of 19–31% sequential growth heading into the back half, where the real capacity ramp kicks in. Management is targeting 60–80% sequential growth in Q3 and Q4 as the new Pearland, Texas, facility and expanded Taiwan lines come online.

Read more

I trimmed 50% of my IREN position.

Main reason: the stock keeps struggling around that $55–$65 resistance zone while names like CIFR, WULF, NBIS, and CRWV are all showing much stronger breakouts. When the rest of the AI power/data centre trade is moving and IREN is lagging, I have to respect that.

BUT HERES THE RED FLAG

Read more

"Fiscal '26 marked another defining year for Credo. Revenue exceeded $1.3 billion, more than tripling year-over-year, while non-GAAP net income increased more than 5x to $662 million."
- Bill Brennan, President, CEO & Chairman, Credo Technology

After years watching semiconductors cycle through booms, busts, and everything in between — from the dot-com era to the mobile revolution to the cloud buildout — I can say without hesitation: what Credo Technology Group (NASDAQ: CRDO) is doing right now is among the most extraordinary financial performances I have ever seen from a hardware company. Revenue tripling in a single year. Net income is rising more than 5x. Gross margins above 68%. A ROIC of 70%+. These are software company metrics coming out of a semiconductor business.

Subscribe now

And then the stock dropped 12% after-hours on June 1st.

The Business

Credo makes the semiconductor chips embedded inside the purple cables connecting NVIDIA GPUs inside Amazon, Microsoft, Meta, and xAI’s AI data centers. These are called Active Electrical Cables (AECs) — copper cables with a smart signal-processing chip at each end that amplifies, cleans, and re-times the signal.

Think of a regular copper cable as a garden hose. Data leaks, weakens, and degrades over distance. Credo’s AEC is a smart pump system — it takes that degraded signal and restores it to perfection, allowing data to travel faster and farther with zero leakage.

Inside an AI training cluster housing thousands of NVIDIA H100 or GB200 GPUs, every chip needs to communicate with every other chip at blinding speed. AECs connect the GPUs to each other and to the network switches. The denser the cluster, the more AECs you need. And GPU clusters are getting denser every quarter.

This is Credo’s core business. And it is exploding.

FY2026: A Year That Rewrites the Semiconductor Playbook

Let the numbers speak:

The quarterly progression tells an even more dramatic story:

• Q1 FY2026: $223.1M (+274% YoY)

• Q2 FY2026: $268M (+272% YoY)

• Q3 FY2026: $407M (+200% YoY, +52% sequential — a blowout)

• Q4 FY2026: $437M (+157% YoY) ✅ Beat $433.3M consensus

Q4 non-GAAP EPS came in at $1.16 versus $1.03 consensus — a 12.6% beat. Non-GAAP gross margin hit 68.3%, above the guided 64–66% range. Non-GAAP net income for Q4 alone was $226.7 million.

Why the Stock Fell 12% After-Hours (And Why That’s Irrational)

The market was pricing in a Q1 FY2027 revenue guidance raise to $500M+. Management guided $465–475M (midpoint $470M) — exactly in line with analyst consensus of $470.4M. The stock dropped 12% for being... exactly right.

This is the “priced for perfection” tax on high-growth stocks. It is not a business deterioration. It is not a thesis crack. It is a classic post-earnings expectations reset on a beta-3.18 stock where investors had been bidding up the pre-earnings anticipation.

That is an exceptional result by any measure outside of the hyper-elevated expectations Credo itself has trained the market to expect. The stock fell not because anything went wrong, but because “extraordinary” wasn’t “extraordinary enough” for a single quarter of positioning.

The FY2027 Roadmap: From $1.34B to $2.4B+

Management’s FY2027 guidance is more than 80% revenue growth — implying $2.4B+ in revenue for the fiscal year ending April/May 2027. The key driver powering this acceleration is a dramatic pivot into optical.

The Optical Ramp: Credo’s Second Act

Here is what management committed to on the Q4 earnings call:

The cadence is deliberate: mid-single-digit sequential growth in the first half of FY2027, followed by a significantly stronger second-half ramp as the optical portfolio hits scale.

The DustPhotonics Acquisition: A Strategic Masterstroke

On May 28, 2026 — one week before earnings — Credo completed the acquisition of DustPhotonics for $750M in cash plus ~0.92 million shares. DustPhotonics is an Israeli company developing industry-leading Silicon Photonics PIC technology for optical transceivers & adding silicon photonics PIC technology and covering 800G, 1.6T, and 3.2T optical connectivity — so they now own the full stack: copper cables, DSPs, and silicon photonics.

What does this mean in practical terms?

Before DustPhotonics, Credo designed the DSP (the brain of an optical transceiver) but sourced the silicon photonics chip (the physical component that converts electricity to light) from third parties. DustPhotonics gives Credo that chip in-house. The result is complete vertical integration across the optical connectivity stack:

SerDes IP → Digital Signal Processor → Silicon Photonics PIC → System Integration

This is the same vertical integration playbook Broadcom and Marvell have used to build moats over decades — except Credo is doing it at warp speed in the most important segment of the AI infrastructure buildout.

The $750M price tag is significant — it reduces cash from $1.3B to approximately $550M post-close. But at $140M+ quarterly FCF generation, Credo replenishes that within two to three quarters. And the return on that investment, if the $600M+ optical revenue target for FY2027 is achieved, is genuinely compelling.

The Moat: Narrow and Rapidly Widening

After more than three decades of semiconductor analysis, I have a healthy skepticism about “moat” claims. Too many companies claim moats that evaporate at the first competitive challenge. Credo’s moat is real, and here is the evidence:

68.1% gross margins. Hardware companies with no pricing power do not sustain 68% gross margins. The premium Credo commands prove the IP advantage is real and that hyperscalers value the reliability differentiation enough to pay for it.

70%+ ROIC. For every dollar invested in the business, Credo generates 70 cents in returns. This is a software company metric coming from a semiconductor designer. It reflects the IP-driven, asset-light nature of the fabless model — design is everything, manufacturing scale comes from TSMC.

The N-1 Node Strategy. While every other AI semiconductor company is fighting for TSMC’s most advanced nodes (the same queue as Apple and NVIDIA), Credo deliberately designs on mature, proven nodes one generation back. This delivers lower wafer costs, better supply certainty, and frees Credo from the bottlenecks plaguing cutting-edge node customers. It is a supply chain moat hiding in plain sight.

Hyperscaler Integration Depth. Getting qualified for use inside Amazon’s AI clusters is not a week-long process. It takes quarters of engineering collaboration, reliability testing, and system-level integration work. Once qualified, the switching cost is enormous — hyperscalers do not swap out embedded connectivity chips mid-cluster. Credo is embedded in the infrastructure of at least five hyperscalers, with at least three each generating more than 10% of revenue.

The Purple Cable Brand. In AI data center circles, Credo’s purple AEC cable has achieved something rare in B2B semiconductors: brand recognition. That AWS CEO photo was not an accident. It was a signal of how deeply embedded Credo is in the world’s most important AI infrastructure buildout.

The Competitive Landscape

Credo is not without competition. Understanding who they compete with and where Credo wins is essential.

Broadcom (AVGO, ~$840B market cap): Broadcom competes across the full networking stack — optical transceivers, custom ASICs, networking chips. For AECs specifically, Credo’s dedicated focus and purpose-built technology for short-reach copper creates a speed and reliability advantage that Broadcom’s more generalist approach has not matched at hyperscale.

Marvell Technology (MRVL, ~$215B market cap): Marvell’s PAM4 DSPs and custom silicon solutions compete in the broader AI connectivity space. Their optical DSP products could pressure Credo’s optical segment, but Marvell’s diversified focus means AEC is not a priority product line for them.

Astera Labs (ALAB, ~$47B market cap): The most direct peer in market cap terms. Astera focuses on PCIe retimers and CXL connectivity — a different layer of the AI stack. The two companies largely serve complementary roles inside the same clusters.

The honest assessment: Credo has a 2+ year lead in AEC deployment at hyperscale, and the DustPhotonics acquisition is pulling them ahead in optical as well. The moat is building, not eroding.

Valuation: Fairly Valued at $210, Attractive at $150

Let me be direct: CRDO is not cheap on traditional metrics. It never has been, and it should not be for a company growing at these rates.

The most reliable valuation method for hyper-growth semiconductors is the PEG method. With FY2027 non-GAAP EPS estimated at $6.00–7.00 on 80%+ growth, a PEG of 0.8–1.0 implies a fair value of $240–350. At post-earnings prices of $200–215, CRDO is at or below the low end of fair value on a PEG basis.

Post-earnings analyst price target raises confirm this view:

The asymmetry at current prices: downside scenario (growth stalls, 10x FY27 P/S) implies ~$110–130. Upside scenario (80%+ growth delivers, multiple holds at 20x FY27 P/S) implies ~$340+. Risk/reward of approximately 2.5:1 favors the upside.

The Six Catalysts to Watch

1. Q1 FY2027 Earnings (August/September 2026): The first real test of whether the post-earnings dip was a buying opportunity or a warning. Watch for: Q2 FY2027 guidance above $490M, gross margin at 68%+, and DustPhotonics integration update. A guidance raise of this magnitude triggers a 15–20% re-rating.

2. DustPhotonics Optical Revenue Ramp (Q1–Q2 FY2027) Management committed to $600M+ in combined optical revenue for FY2027. Each quarter’s optical revenue disclosure will serve as a progress report on the $750M acquisition thesis. First hyperscaler design wins for the silicon photonics portfolio would be a material catalyst.

3. Customer Diversification (Ongoing FY2027) The single most important de-risking event for the stock’s multiple. A 5th or 6th hyperscaler crossing the 10% revenue threshold removes the concentration discount from the multiple. Watch quarterly filings closely.

4. 200G Per Lane AEC Ramp (FY2027–FY2028) Credo leads today in 100G per lane AECs. The next generation is 200G. First volume orders from hyperscalers for 200G products confirm Credo maintains its generational lead and extends the product cycle through 2028 and beyond.

5. Active LED Cable (ALC) Sampling (FY2027) The next product innovation: AECs that reach up to 30 meters using LEDs instead of lasers. This opens a new adjacent market for row-scale and rack-to-rack connections. Management is committed to customer samples in FY2027, with initial revenue in FY2028. Design win announcements here unlock a $2B+ TAM expansion.

6. Analyst Price Target Convergence With the new consensus target at $250.91 and the stock near $212 post-dip, there is an 18%+ implied return just from mean-reversion to consensus. Continued target raises as analysts model in the $600M optical guidance and DustPhotonics synergies would provide a near-term re-rating catalyst.

The Risks — And I Will Not Sugarcoat Them

Here is the unvarnished truth about CRDO:

Read more

Blue Origin’s New Glenn rocket tragically exploded Thursday night during a hot-fire test on its launch pad in Cape Canaveral, Florida. This event is dramatic for the whole space industry and reminds us of the risks and the complexity of going to space.

It is rumoured that the issue came from an engine that exploded.

Read more

Intuitive Machines (, IM) saw its stock crash intraday today, May 26, after NASA awarded the long-awaited Phase 1 of the Lunar Terrain Vehicle (LTV) contracts to competitors Astrolab and Lunar Outpost during its Moon Base plans press conference.

The dollar figures came in far smaller than expected, with $220M per LTV award and…

Read more

On one side, Trump wants lower interest rates to support growth, markets, and the economy heading into a crucial period. On the other, inflation pressures remain elevated, energy prices are volatile, and AI is creating an entirely new economic dynamic the Fed has never had to navigate before.

Warsh now has to balance politics, inflation, labor markets, asset prices, and investor expectations all at once — and every decision he makes could move stocks, bonds, housing, and the broader economy.

Here are the 5 biggest challenges Warsh faces as the new Fed Chair… and what I think happens next in the stockmarket.

NEW FED: 5 challenges facing Warsh — my Macro take

Read more

Today, SpaceX has released its long-awaited S-1 to go public on Nasdaq (and Nasdaq Texas!) under the symbol $SPCX.

Here are some information we can learn from the S-1 filing.

You’re buying three companies, not one

When reading “SpaceX IPO,” most investors think rockets and reusability, but what once distinguished SpaceX is now the past.

SpaceX is now three businesses welded together:

• Space: Falcon 9, Starship, Dragon Capsule

• Connectivity: Starlink, Starshield

• AI: xAI, Grok, and X (formerly Twitter)

The xAI/X merger closed in February 2026, with the combined entity valued at $1.25 trillion privately. The financial statements have been retroactively combined, but these were separate companies just three months ago.

SpaceX brought in $18.7B in revenue in 2025, up 33% year over year. But in terms of profitability , the company swung from $791M in net income in 2024 to a $4.9B net loss in 2025, a nearly $6B reversal in a single year!

The losses in 2025 appears attributed to management choosing to spend $9.5 billion on R&D (Starship + AI) when they could be reporting operating income.

The financials…

SpaceX made $18.7B in revenue in 2025 and $4.7B in Q1 2026. Most of that came from Starlink, which brought in $11.4B in 2025, grew 50% year over year, and generated $4.4B in operating income in 2025 and $1.2B in Q1 2026 alone.

The Space segment is still losing money, mostly because SpaceX spent $3B on Starship development in 2025 alone.

The AI segment is the newest and most expensive bet. It made $3.2B in revenue in 2025 but lost $6.4B at the operating level, and SpaceX spent $12.7B in capital expenditures to build out its AI compute infrastructure.

Overall, SpaceX runs at an operating loss, but still generated $6.6B in adjusted EBITDA in 2025. The losses are mostly coming from heavy investment in AI and the development of Starship.

The real cash machine is Starlink

Connectivity (...essentially Starlink) generated $11.4 billion in revenue in 2025, up 49.8%, with $4.4 billion in operating income.

Here are more numbres related to subscriber growth:

• 2.3M subscribers in 2023

• 4.4M in 2024

• 8.9M in 2025

• 10.3M as of Q1 2026

However, the average revenue per user (ARPU) appears to have decreased from $99/month in 2023 to $66/month in Q1 2026. That’s a 33% decline. The filing explicitly states ARPU will keep falling as Starlink pushes into emerging markets.

Why SpaceX is a durable business

What SpaceX has built is genuinely hard to replicate. From the filing:

• 99%+ Falcon mission success rate across ~650 launches

  • Falcon 9 is the most reliable rocket in the world.

• 24 flight-proven reusable rockets, each qualified for 40 flights

  • If reusing a Falcon 9 first-stage booster saves SpaceX an estimated $30 million per launch, this is more than one billion that can be saved after 40 flights per booster.

• Several Falcon 9 boosters have already been reflown up to 34 times

• SpaceX launches 80%+ of global mass to orbit every year

• 80% of Starship is manufactured in-house

• Approximately 75% of all active maneuverable satellites in orbit are Starlink

• Starlink performs over 1,000 automated collision avoidance maneuvers per day

According to NASA data cited in the filing, Falcon 9 reduced launch costs to ~$2,700 per kilogram, an 85% reduction from the historical industry average of $18,500/kg. Falcon Heavy got it to $1,400/kg. Starship targets a 99%+ reduction.

The customer base is way more diversified than I expected. “Customer A” (almost certainly the U.S. government in aggregate), was 25.2% of revenue in 2023, falling to 20.9% in 2025. No other customer exceeds 10%.

SpaceX also report a $27.6 billion backlog, with about 36% recognizable within a year, and roughly $23.7 billion in cash and short-term securities on hand pre-IPO.

Anthropic is paying SpaceX $15 billion/year

Per the S-1 filing, Anthropic has agreed to pay SpaceX $1.25 billion per month through May 2029 in exchange for access to cloud computing infrastructure.

SpaceX also noted that it “expect[s] to enter into additional similar services contracts” and “believe[s] [this] dual monetization strategy provides multiple pathways to generate returns on invested capital.”

SpaceX… an AI play?

The AI segment lost $6.4 billion in 2025 on $3.2 billion of revenue. Capex was $12.7 billion, and another $7.7 billion in just Q1 2026.

SpaceX brought COLOSSUS II’s first 210-megawatt cluster online in 91 days. The industry benchmark for a 100MW greenfield data center is roughly two years.

The Anthropic Cloud Services Agreement discussed above will generate $1.25 billion per month through May 2029 (roughly $45 billion in contracted revenue from a single customer (though terminable on 90 days’ notice)

SpaceX has also the right to acquire Cursor at an implied $60 billion valuation, with $10 billion in termination fees if they walk.

There’s also the Terafab framework with Tesla and Intel for chip manufacturing, but the filing explicitly notes nothing has been finalized.

Transactions with Tesla

Read more

Last month, you’ve likely heard or even seen live the historic achievement of four astronauts orbiting the Moon for the first time in 50 years. Artemis II marked the second mission from NASA’s ambitious program to have a permanent presence to the lunar surface. Collaborating with 67 other nations that signed the Artemis accord, the shared goal is to establish a permanent lunar base by 2030, a new priority set by Trump’s administration :

Unlike the Apollo program, NASA has decided to forge strategic partnerships with commercial entities to make this vision a reality. These commercial partners mainly consist of publicly traded companies, whose important role will be discussed in this article.

We might assume that because the Artemis program is primarily funded by taxpayers, it’s challenging to profit from it. But this is false! In fact, over half of the components and systems that enabled the four astronauts to launch, orbit the Moon, conduct scientific experiments, and safely return to Earth last month were designed and manufactured by publicly traded companies, not by the government! And as investor, this means you can benefit from this economy!

And you may not be aware yet, but several of the companies that you invest in your portfolio may have played an important role in making Artemis II a successful mission, even if they aren’t labeled as “space stock”, like GoPro ( )…

Let’s deep dive…

Space systems

When we think of a “space company,” we often think of rockets and satellites. However, it is much more than that. Components like solar panels, reaction wheels, and star trackers are what make any mission possible long before communication is even established.

Who would have thought a device to track stars could be so important?!

Mark my words, I strongly believe that this tiny device might be the pillar of the space economy. Satellites and spacecraft use star trackers to identify the positions of stars and calculate a highly precise 3D orientation. This data is essential for spacecraft to maintain their attitude (3D orientation) and ensure that solar panels remain pointed toward the Sun and communications dishes are aligned with Earth.

No star trackers, no space missions.

The Orion capsule for Artemis missions does not use star trackers from a publicly traded company but from a private entity named Jena Optronik, a subsidiary of Airbus Defense and Space ( EPA: AIR ) and leader in high-accuracy sensors.

…Or at least the Orion capsule does not use star trackers from a publicly-traded company just yet!

A particular aspect of space companies is that each build their own expertise and a publicly traded company would rather acquire companies specialized in a specific technology than consume cash on research & development to get to that level of expertise. Acquiring a company turns out less expensive down the road than developing new technologies. If a new acquisition can also help companies to sell more, this is even better!

A good example is Rocket Lab ( ) that is announcing the acquisition of new space companies now almost every quarter. As part of a $500 million contract for the Space Development Agency (SDA) to manufacture 18 satellites, they previously used Mynaric, a German company with expertise in optical inter-satellite links, as a supplier. One day, Mynaric found itself on the brink of bankruptcy. Rocket Lab could have lost their supplier, but instead they chose to acquire the company to not disrupt its supply chain and at the same time to enhance their vertically integrated suite of space systems. This also allowed Rocket Lab to have a presence in Europe.

Optical communication is not the only product in Rocket Lab’s space systems inventory. The company is also selling star trackers and reaction wheels, radios, space software, separation systems, solar panels and has recently added electric propulsion to their inventory.

I was mentioning that no missions can happen without star trackers, but in reality, what matters most is power, as star trackers obviously need power. That is especially true for crewed missions, where power is required for life support systems. While communication follows in the hierarchy of needs, it is not the most fundamental need. For example, when the Artemis II crew passed behind the Moon, they lost contact with Earth for 40 minutes, but they still had power and were able to conduct scientific experiments.

On deep space missions, power is primarily generated using radioisotope thermoelectric generators (commonly known as nuclear generators). L3harris ( ) just announced on May 15 that they have finalized the design of a next generation nuclear-based power source for future NASA deep space missions.

For Earth-to-Moon missions, no need for nuclear generators as the sun can still emit enough light for power to be generated through solar cells. On the Orion capsule, the four solar arrays were produced by Rocket Lab!

In January 2022, Rocket Lab acquired SolAero Holdings Inc. for $80 million to add solar solution to their space systems inventory. SolAero had been since 2001 one of the world’s leading manufacturers of highly-efficient radiation-hard solar cells and solar panels for space application. This acquisition has brought to Rocket Lab new expertise and a new base of customers.

Without this acquisition, Rocket Lab wouldn’t have solar solutions in their space systems inventory and therefore solar panels on the Orion capsule wouldn’t be those from Rocket Lab.

Having diverse revenue streams, as seen in companies like Rocket Lab, has become fundamental for a space company to survive. It’s easy for agencies to fund a rocket program or an ambitious project; they fund it through taxpayers, but for a company, when you do not have enough capital, the only solution you have is to first IPO, then raise capital and dilute shareholders. And raise and dilute again until you can’t anymore. This paragraph is literally the story of Astra Space (previously trading under the ticker $ASTR).

There is now unfortunately a proven track record of publicly traded space companies filing for bankruptcy only because they focused exclusively on building one thing e.g. an operational rocket ( $ASTR ) , a launch pad ( $MAXQ.NE ), etc. Without vertical integration, which leaves them reliant on external suppliers, and with no diverse products or services to sell than what their focus is, their only recourse when they need more cash is to dilute shareholders until facing either a delisting (when a reverse stock split is no longer an option) or ultimately a Chapter 11 filing.

A good example of this is Virgin Galactic ( SPCE 3.33%↑ ). They paused their commercial spaceflights in June 2024 to focus on the development of its next-generation Delta-class spaceships. With no other revenue streams (commercial spaceflights has been their only source of revenue) and the constant need for capital to develop their next spaceship, their only solution has been to dilute shareholders.

It is through strategic acquisitions to expand its space systems inventory that Rocket Lab could generate enough revenue to fund the Electron and Neutron rocket programs without having the need to dilute massively its shareholders. Research and development for a new launch vehicle is extremely cash-intensive because of how iterative the process is, so a publicly traded space company’s best solution to avoiding massive dilution and maintaining its reputation to shareholders is to sell more products and services. And to sell more, the first step is achieving vertical integration and the second step is to sell from its own supply chain. Rocket Lab understood this from the start by acquiring smaller space companies in specific domain like lasers, reaction wheels, solar solution, etc., to get that expertise but also to sell to a more diversified customer base.

Rocket Lab is no longer among the only few space companies that is vertically integrated and claims to be an end-to-end space solution. and are also two other publicly traded companies acquiring as many players as possible to reduce supplier reliance and diversify their revenue streams. While neither has to fund a highly expensive rocket or other ambitious programs, this diversification is particularly useful for mitigating risk. Much of the high-revenue portion of the space industry currently comes from government agencies and this is funding that can be cut or delayed at any time. By selling space systems components to companies it is still possible for these companies to generate revenue in the event that a program or funding is delayed or canceled.

During NASA’s Ignition event just a few days before the Artemis II mission, the agency announced to pause the Gateway program to focus instead on establishing a permanent lunar base. With the International Space Station set for decommissioning in 2030, the next space laboratory was originally intended to orbit the Moon aboard the Gateway; however, that laboratory will now be located on the lunar surface. This shift leaves billions of dollars in contracts paused indefinitely or repurposed.

MDA Space ( ) is contracted by the Canadian Space Agency for $1 billion to build the Canadarm3, the robotic arm for the Gateway. During their May 7 earnings call, the chief executive, Mike Greenlay, said to “continue discussions with the Canadian Space Agency on redefining the Canadarm3 robotic systems that will be required to support this new and exciting phase of moon exploration.”

One risk factor that I have learned the hard way as a space stock investor for the past three years is the extent to which a company’s backlog and revenue depend on government contracts versus commercial customers. You can find this data typically through quarterly reports. I often call this metric the commercial-to-government (CG) ratio. The higher the ratio, the more diversified the company's revenue streams. Let me explain…

Most of the time in the space sector, companies generally have less capital to deploy than agencies or other government entities. Because task orders from these government entities are generally substantial, a space company therefore needs a higher volume of commercial orders and from multiple companies to counterbalance.

When a company has a commercial-to-government ratio of 0.5 to 1, that implies a diversified base of commercial customers and is therefore less risky than a company with a CG ratio of 0 to 0.5, where one task order on pause could easily have serious impact on the business.

Back to space systems.

On the Orion capsule’s solar arrays made by Rocket Lab, there were GoPro ( ) cameras installed, yet another publicly-traded company! Four specialized, modified GoPro cameras were mounted on the Orion spacecraft’s solar array wings, where they captured high‑resolution views of the spacecraft, Earth, and the Moon. National Geographic has also equipped the Artemis II crew with GoPro cameras to document daily life on the mission from inside the spacecraft.

GoPros were not the only cameras aboard Orion. 11 internal and external inspection and navigation camera from Redwire ( ) to record 4k video, 12MP images and livestream of both the inside and outside of the vehicle were also installed.

More components like the barometric altimeter, the inertial measurement system, the GPS receiver and displays and controls inside the capsule and the flight software were produced by Honeywell ( )

On a larger scale, the Orion capsule itself was designed and assembled at the Kennedy Space Center by Lockheed Martin ( ), serving as prime contractor for NASA. Under the Orion Production and Operations Contract (OPOC) awarded to Lockheed Martin, NASA committed to ordering a minimum of six and a maximum of 12 Orion spacecraft.

The service module that provides power and propulsion was provided by the European Space Agency, with Airbus Defense and Space ( EPA: AIR ) acting as the prime contractor and with companies from 11 countries supplying spacecraft-specific parts made for the spacecraft.

What I just covered was just for the capsule and the service module, and as you can see, there are already tons of publicly traded companies that are involved. Now adding companies contracted by NASA like Amentum ( ) to assemble the Space Launch System (SLS) rocket and to transport it to the launch pad, Northrop Grumman ( ) that supplied the twin solid rocket boosters for the SLS rocket, and many more. As an investor, you can definitely take part in this growing economy.

If you cannot choose which stock to hold, there are now a few ETFs like and .

If these companies can build together the vehicle and its systems to safely bring a capsule around the Moon and back to Earth, these missions wouldn’t be possible without a lunar infrastructure. Let’s discuss that!

But before diving in, did you know you can follow my portfolio, trades and insights for free on Blossom?

The lunar infrastructure or the “239,000-mile marketplace”

The lunar infrastructure groups power generation, communication & navigation, surface mobility, in-situ resource utilization, landing sites & radiation-shielded habitats, and support sustainable human presence and activity around the Moon and at its surface.

Read more

Official Announcement: May 14th, 2026.

Zeta Global & Snowflake strike a partnership to develop an OSI with the purpose “to create a common, vendor-agnostic specification that defines semantic metadata.”

But, what the heck does that mean? And why the heck do we care?

That’s why I’m here!

Ironically, I made a post moments before this announcement was publish…

Read more

I’ve included each stock’s current performance, but if you want the full breakdown including:
• Buy averages
• Total capital invested
• Portfolio weightings
• My full investing strategy & research

…it’s all available inside the Discord.

This is where I share my real-time thinking, long-term theses, buy/sell zones, market updates, and deeper analysis with the community. 🚀

YTD performance : 11.66% vs SPY 8.66%
YTD moonshots : 29.23% vs IWM 13.6%

Subscribe to the substack to see my total holdings & why

Read more