STATE OF THE UNION

G.L.A INVE$T RE$EARCH

The Infinite Innovation Newsletter

STATE OF THE UNION — Market & Core Position Update

Issue Date: July 29, 2026

From The Desk Of Coach Kyle

Happy Wednesday to The Greatest Living American’s! — this is your State of the Union.

The calls, dm’s, and emails are pouring in. The panic is real. Growth stocks are getting hammered, Tesla is off almost 40% from its high, Palantir has been cut nearly in half, Spacex as well, MicroStrategy has been decimated, and Bitcoin is deep in a drawdown. Meanwhile the S&P 500 is barely off its high — which means the pain is concentrated exactly where we live: the innovation edge of the market.

I want to be very clear about something before we get into the tape:

Corrections don’t kill portfolios. Investor reactions kill portfolios.

We are long-term innovation investors seeking exponential 10X, 20X, 30X returns over the next 5–10 years. Volatility — and our tolerance for volatility — is the price of admission we agree to pay for that outperformance. If you can’t stomach a 30–40% drawdown in a name like TSLA or PLTR, you were never going to hold it long enough for a potential 5x. That is the deal. That has always been the deal.

Nothing in our core thesis has changed. In many cases, it has strengthened.

Let’s walk through the tape.

1. The Macro Backdrop — Why We’re Here

Three real, non-imaginary things are compressing multiples on long-duration growth right now:

• New Fed, new tone. Kevin Warsh is now running the Federal Reserve, and the market is still learning his reaction function. His panel at the ECB Sintra Forum this month, alongside Lagarde, Bailey and Macklem, was more hawkish-tinged than the doves wanted. The Fed Funds rate sits at 3.75%, headline CPI is still sticky at 3.5% even as core CPI has cooled to 2.6%, and the 10-year Treasury is at 4.63% — near the top of its 52-week range. Higher-for-longer rates are the single biggest headwind for long-duration cash-flow stories, which is what we own.

• Two active wars. Middle East fighting escalated again in mid-July with U.S. strikes on Iranian infrastructure, driving an oil spike and a defensive rotation. Ukraine remains a slow-burn drag. Geopolitical risk premium is elevated.

• AI capex anxiety. The market is starting to question whether the hyperscalers are over-spending. Asian markets sold off hard this week on AI-cost fears — the KOSPI fell 10%, SK Hynix and Samsung each dropped 12%, and Korea halted trading twice. This is the exact same misunderstanding that hit GOOG a year ago on TPU spend. Find the misunderstanding, make the money.

Meanwhile the actual economy is fine. GDP grew 2.7% last quarter, unemployment sits at 4.2%. This is not 2008. This is not a bubble bursting. This is a rate-driven, sentiment-driven, geopolitics-driven repricing of long-duration growth — the same movie we’ve seen every 18–24 months since we started.

2. The Bubble Narrative Is Dead Wrong

Every financial media outlet is running the same headline: “AI Bubble.” I want to address this head-on because it matters for how you behave over the next 90 days.

A bubble is a narrative detached from fundamentals. That is not what we have. What we have is:

• Hyperscaler cloud growth of 30–80% year-over-year with over $1 trillion in already-contracted backlog at Microsoft and Google combined.

• Palantir printing government + commercial revenue growth at levels most companies don’t see once in a decade.

• Bitcoin sitting inside a scheduled post-halving correction, exactly on the path every prior cycle has taken.

• Real, revenue-generating AI infrastructure being deployed — not vaporware, not a promise.

Dot-com had no earnings. This market has $1 trillion of already-contracted future revenue. Those are not the same thing. They will never be the same thing. Anyone telling you otherwise is selling you a book or a Substack subscription.

The bubble call is the consensus call. The consensus is almost always wrong at inflection points.

3. The Doctrine — Read This Twice

This is our Constitution. Print it, tape it to the monitor:

1. We are long-term innovation investors seeking 10X, 20X, 30X over the next 5–10 years.

2. Buy and hold. DCA on weakness. When a name breaks below its threshold (–20% for growth, –10% for broad ETFs, –30% for ultra-conservative), we accelerate, not retreat.

3. We do not let volatility shake our thesis. Research tells us what to do. Price tells us when.

4. We scrutinize our bull case daily. We are constantly mining the risk and the bear thesis. If the business changes, we sell. If only the price changes, we buy more.

5. Volatility tolerance is the price of admission for exponential returns. Full stop.

If you’re panicking right now, you didn’t size the position correctly. That’s a lesson, not a catastrophe.

4. Market State of Play

The index is fine. Under the hood is a growth-stock bear market. Every recovery of the last 15 years has started under exactly these conditions.

The 10 year treasury has been spiking due to inflation uncertainty, which has been caused by the war with Iran. This is the actual reason why our favorite innovation / long duration assets have been down. This to shall pass.

5. Core Position Update — The APEX 10

TSLA — $304.26 · –39.0% from 52W high of $498.83

Market cap $1.20T. This is the definition of a wonderful business on sale. Nothing meaningful about the business has changed — in fact, on the Q2 call Musk redirected investor focus toward autonomy and Optimus as the central long-term value drivers, and Tesla widened the Austin robotaxi geofence and opened access to a broader pool of public riders (Next Edition). The stock is punished because core auto sales slipped and the market is refusing to price autonomy. That is the entire opportunity. Thesis unchanged: Autonomy + AI + Energy. Range remains $498–$542 as we compound back through the cycle. 

PLTR — $123.47 · –40.5% from 52W high of $207.52

Market cap $283.5B. This is the most misunderstood position in the portfolio right now. PLTR just came off the strongest quarter in company history, FY26 guidance sits at $7.2B, and government + commercial adoption of AIP is accelerating. The multiple compressed from euphoric to merely expensive. The business is compounding faster than the multiple is compressing. This is exactly the setup that produces the next leg. Do not flinch. Add on strength through $130, add aggressively below $115.

MSTR / Bitcoin — MSTR $96.35 · –76.7% from $414.36 · BTC $64,220 · –49.1% from $126,198

Yes, MSTR is down almost 77% from its high. Yes, Bitcoin is roughly cut in half. This is a scheduled post-halving cycle correction, not a thesis break. Every prior cycle drew down 55–80% from the top before the next impulse. Michael Saylor’s $30B 2026 BTC accumulation plan is still executing. Our two-vehicle exposure — MSTR for leveraged BTC beta, ARKB for clean spot exposure — remains the right structure. 

GOOGL — $333.15 · –18.5% from 52W high of $408.61

Market cap $4.03T at a P/E of only 25.4. This is the cheapest hyperscaler on the board on a growth-adjusted basis. Google Cloud is still growing >70% YoY, TPUs took them off of 100% Nvidia dependency and are now a standalone product line, and cloud operating margin has expanded meaningfully. The market is punishing capex — the very thing that funds the next decade of moat.This is the tollbooth for AI, and it’s on sale. Overweight.

AMZN — $227.78 · –18.2% from 52W high of $278.56

Market cap $2.45T at P/E of 27.2. AWS is still the largest AI landlord on the planet, Trainium is scaling as vertical integration against Nvidia, and the retail flywheel is finally producing real operating income. AMZN has essentially been flat for a year while doubling revenue and quadrupling operating income over the last five years. This is the definition of a mispricing. Continue to accumulate; models point to 100%+ upside on a five-year horizon.

NVDA — $191.78

Still the picks-and-shovels king. The AI capex anxiety story is transient noise; the contracts say otherwise.

(HOOD, LMND, STRC, ARKB updates: no material change — continue standard DCA discipline. Full breakdowns in next week’s Weekly.)

6. Cybercab & Robotaxi — Slower Than We Wanted, Right When We Need It

Full transparency, family: the Cybercab rollout is behind our internal timeline. The unlimited public rollout many of us wanted by mid-2026 hasn’t happened. Instead we got a widened Austin geofence, broader public access, and Musk describing an “aggressive scale-up through 2H 2026” (Next Edition).

Here is the reframe: slower is correct. A single high-profile accident during a rushed rollout would set the whole autonomy category back 18–24 months and hand the narrative to lidar competitors. Tesla is doing this the way it must be done — geofence, expand, geofence, expand. This is what a durable, regulator-defensible autonomy business looks like. It’s also why the market is currently underpricing the option. Fine by us.

Zoom out. By the time this issue matters for the stock, we’ll already own it.

7. What Elon Has Been Saying

• Autonomy + Optimus are Tesla’s long-term value. Not vehicle unit sales. He said it plainly on the Q2 call. This is a category shift in the equity story.

• SpaceX has filed with the FCC to launch a third-generation 100,000-satellite constellation designed to power human connectivity and AI-fueled progress (YouTube — Elon news feed). Musk’s own words: “The artificial intelligence revolution promises a supersonic tsunami of progress.” Read that again. That framing directly underwrites our SpaceX thesis.

• Falcon 9 launched the 17th Transporter rideshare mission delivering 81 payloads to orbit — Starlink+Starshield cadence continues, cash-flow engine hums.

• The Best Long-Term Investment in History

  Ok G.L.A FAMILY!!! — I have to give SPCX its own section this week because the calls and the worries about its price won’t stop.

  Yes, SPCX is off its highs. Don’t let the stock price fool you. The stock is a wrapper. What we actually own is a piece of a Company that just came public and is still getting its footing.

Why This Is the Greatest Long-Term Investment in History

Not a phrase we throw around. Here is the case:

1. Starlink is the cash engine. A recurring-revenue, global-consumer, government-and-military internet utility with no serious competitor. It funds everything else.

2. Starship is the physics reset. Full reusability at scale is the single largest reduction in cost-to-orbit in human history. Every business built on top of that new cost curve — communications, defense, energy, manufacturing, planetary — is a SpaceX customer by default.

3. Starshield is the defense franchise. In a two-active-wars world, sovereign customers pay premium prices for hardened, distributed, low-earth-orbit infrastructure. That is a durable, high-margin annuity.

4. The 100,000-satellite third-gen constellation just filed with the FCC , is a bandwidth moat that no competitor can economically replicate within a decade.

5. Mars is optionality most investors refuse to price. That’s fine. We’ll price it for them.

SpaceX Is More Than Ten Years Ahead of Everyone Else

Blue Origin, Rocket Lab, Relativity, ULA, Arianespace, CNSA, Roscosmos — pick any of them. Look at cadence. Look at reusability. Look at payload economics. Look at active fleet size in orbit. The gap is not closing. It is widening. Every quarter SpaceX launches, that lead compounds. Every Starship iteration flattens the cost curve further and pushes the ceiling of what the next competitor even has to catch up to.

This is what a decade-plus moat looks like in real time.

The Musk Flywheel Is Working — Look Around the Ecosystem

SpaceX doesn’t operate in a vacuum. The other engines in Elon’s ecosystem are throwing off cash and compute that directly and indirectly benefit SPCX shareholders:

• x.com is now profitable. After the 2022 acquisition, the “X Corp is dying” narrative was consensus. It was consensus for three straight years. That was the trade. It’s now a functioning, cash-generative distribution and identity platform — and it is the real-time data pipe feeding Grok.

• Grok 4.5 is remarkable. Independent testing by Artificial Analysis ranks it #4 of 168 models on the Intelligence Index, scoring 54, at a fraction of the cost per task of frontier competitors — roughly $0.31 per task on the Intelligence Index and $2.59 per coding task versus around $11.80 for Claude Fable 5 (Codersera). It used 4.2× fewer output tokens than Opus 4.8 on SWE-Bench Pro. That is not “trying hard.” That is winning on the metrics that matter to enterprise customers: capability per dollar.

• xAI is dominating datacenters on Earth. Colossus 1 in Memphis was stood up in 122 days (CNBC). Colossus 2 construction is already underway. And here is the number that should stop you cold: Anthropic is renting compute capacity at Colossus 1 for up to $1.25 billion per month, and Google is renting for up to $920 million a month. Two of the three largest AI companies on the planet are cutting nine-figure monthly checks to Elon’s compute business. Digest that.

• Datacenters in orbit are next. The natural extension of Starlink + Starshield + Grok + Colossus is space-based compute. Solar-abundant, cold-cooled, latency-tuned, globally-served. SpaceX is the only company on Earth positioned to build the orbital compute layer, because they are the only company that owns the launch cost curve and the constellation and the AI stack. When that transition happens — and it’s a “when,” not an “if” — SpaceX is the pick, the shovel, and the land.

We are about to witness the golden era for IPO’s (IMHO)

8. Private Market

Anduril

• Capital raised to date: $11.96B (Forge Global)

• Latest evaluated price (7/28/26): $140.00/share → implied valuation $123.9B (Forge Global)

• Series H (May 2026): $5B at $68.95/share, $61B valuation

• Held broadly by BlackRock, Fidelity Contrafund, Franklin, StepStone

Anduril just doubled from its Series H valuation on the Forge tape. Defense tech is not slowing down — it is acceleratinginto the two-active-wars backdrop. Software-defined military infrastructure is the durable thesis of this decade.

Saronic

• Capital raised to date: $2.60B (Forge Global)

• Latest evaluated price (7/28/26): $29.62/share → implied valuation $9.45B (Forge Global)

• Series D (Mar 2026): $1.75B at $27.45, $9.25B valuation

Autonomous surface vessels for naval + maritime forces. This is the water complement to Anduril’s land + air stack. Saronic went from a $1B round to a $9.25B round in eight months. That is what asymmetric defense-tech compounding looks like.

9. AI’s Impact on the Economy and Human Life

This is bigger than the internet. I said it before, I’ll say it again. What we are witnessing is a cataclysmic global economic shift that will re-price labor, re-price capital, re-price time itself.

Three durable truths from this vantage point:

1. Diggers die, landlords compound. The chatbot layer will consolidate to 3–4 winners with brutally thin margins. The pick-and-shovel and land layers — hyperscalers, chip designers, energy providers, autonomy stacks — capture the profit pool.

2. Energy is the binding constraint. Distributed energy for AI is not optional. It is the substrate. This is why our energy-storage line remains in the doctrine.

3. The productivity dividend is going to arrive in waves, and each wave will re-rate a different tier of the market. We front-run the wave by owning the platforms, not the applications.

The human-life piece is harder and more personal. AI will displace tasks faster than jobs and will create abundance for those on the capital side of the ledger. It won’t be AI that takes your job, it will be a human using AI that will! Our job is to make sure our clients — you — are on the capital side. That is what this newsletter is for.

10. Zoom Out — The Big Picture

Every serious wealth event in modern history was built by people who did not sell during the drawdown. The Amazon investor who held through 2000–2002. The Apple investor who held through 2008. The Bitcoin holder who held through 2018 and 2022. The Palantir holders that held through 2022!

We do not know exactly when this correction ends. We do know:

• Fundamentals of our core positions are getting better, not worse.

• Multiples are getting cheaper.

• Backlogs are growing.

• The narrative is at maximum pessimism.

• The Fed will eventually turn. Wars end. Uncertainty resolves.

We are still in the early stages of the largest innovation boom in history, the companies that solve the largest problems will have the largest returns! you can’t make 10X by trading or by trying to time the market. The hall of fame for market timers is an empty room.

Action Items This Week

1. Rebalance mental accounting. 

2. Do not check the account daily. Weekly is enough. Daily is punishment.

3. Write down your thesis for each core name in your own words. If you can’t defend it in three sentences, you don’t own it — you’re renting it.

4. Reply with questions. The DMs, the calls, the “should I sell TSLA” texts — bring them. That is what this newsletter is for.

5. Dont look at things through a negative lens. I hear things like “ Im down xxxx from the highs” You did not buy at the high, and saying these things out loud will cause you to make a mistake. So zoom out and remember that we are investors not day traders, and corrections are normal and are literally part of the game. Watch the company not the stock!!!!

Volatility is the toll we pay on the road to exponential returns. We paid it in 2020. We paid it in 2022. We are paying it now. We will pay it again. And on the other side, we compound.

INVEST IN THE BEST, IGNORE THE REST, WITH A LITTLE HELP FROM G.L.A INVEST!

— G.L.A INVE$T RE$EARCH and Coach Kyle
The Infinite Innovation Newsletter