🤖 What OpenAI’s “GPT-6 Astra” Means for the Stock Market

Plus Dell soars 16%, Lululemon crashes to 8-year low, and the US strikes Iran again...

TOP STORY
🤖 What OpenAI’s “GPT-6 Astra” Means for the Stock Market

☀️ Happy Sunday everyone! I hope you're all having a lovely long weekend.

🛢️ It was another busy week in the markets, with Brent Crude rising 7% as the U.S. and Iran traded strikes for the first time since July, and the U.S. jobs report for August came back at nearly triple analyst expectations.

📊 Despite the strikes, the markets held green across the week:

  • The S&P 500 rose +0.09%

  • The Nasdaq 100 rose +0.38%

  • The TSX fell -0.11%

🤖 Macro aside, we have a lot to get into this week, starting with OpenAI’s Astra model sending shockwaves across the AI industry, Dell ($DELL) soaring 16% after unveiling a massive $95 billion AI backlog, and Lululemon ($LULU) crashing 17% to an 8-year low.

⚔️ But first, let’s take a quick look at the Iran war, because there were some big updates…

🛢️ Oil Prices Rise As U.S. and Iran Trade Strikes Again

Trump’s AI video of the U.S. military striking Kharg Island, posted on Truth Social

🚢 After a month of relative calm, the war in Iran came roaring back last Sunday, when U.S. forces struck Iranian rocket launchers in the Strait of Hormuz, saying Tehran was preparing to fire rockets carrying sea mines into the Strait.

😅 The same day, Trump posted an AI video on Truth Social of “Kharg Island being blown to smithereens,” although Kharg Island was not one of the targets of the strikes.

💥 On Tuesday, the U.S. launched a much larger wave of strikes targeting air defense sites, radar, maritime assets, mine-laying capabilities, and communications.

⚠️ Saturday morning things escalated with Iran launching ballistic missiles toward a U.S. aircraft carrier.

📈 Oil reacted exactly how you’d expect, with WTI jumping 10% to $91/barrel, and Brent Crude finishing up 7% to around $96/barrel.

⛽ U.S. diesel prices hit an ALL-TIME record of $5.85 a gallon, and Americans paid the highest Labor Day weekend gas prices in history. Diesel is a key price to watch because it fuels trucks, trains, and tractors, so it feeds into the price of basically everything (one economist called it a key reason government bond yields keep climbing around the world).

😬 As of the time of writing, no peace negotiation is in place after the June 17 memorandum of understanding expired unextended on August 17. Countries like Oman and Pakistan that were mediating have also now become idle.

💸 Interest Rate Hike Probability Increases to 58%

😰 Rising fuel costs and a strong jobs report make it even more likely the Fed will raise rates (which is bad for stocks), with the probability now at over 58%. But we won’t have to wait long, with the Fed’s decision coming in just 10 days on Sep 16.

🤖 Macroeconomics aside, let’s switch gears to talk about the biggest story this week: OpenAI’s GPT-6 Astra, its impact on the stock market, and who’s winning the AI model race…

✨ But first, a quick word from our sponsor this week: Harvest ETFs!

👋 P.S. For those new here, my name is Max and I'm the CEO of Blossom and the author of the Weekly Buzz (@maxstocks on Blossom), and every week I give you a breakdown of the top stories in the markets!

SPONSORED BY HARVEST ETFs
🌟 HHIC | 100% Canadian – High income – All-In-One ETF

📈 The Harvest Canadian High Income Shares ETF (TSX: HHIC) launched in August 2025. Since then, HHIC has climbed high alongside Canadians markets. Canada now stands at a key juncture after reporting GDP growth of 3.3% in the second quarter of 2026. That, along with strong earnings from Canada’s largest banks, may spark positive thinking among Canadians despite rising trade tensions.

🇨🇦 HHIC offers exposure to established and trending Canadian companies, included the country’s largest banks like Royal Bank and TD, energy giants like Suncor and Enbridge, and even exciting technology stories like Shopify.

💸 Access a multi-sector portfolio of Canada’s best companies, with monthly income from covered calls, with modest leverage to bolster income and growth potential.

✅ Benefits:

  • Diversified exposure to top Canadian companies

  • CAD stocks + High monthly income

  • Tax efficient

  • Lower purchase price to access underlying equity holdings

  • Modest leverage at ~25% to generate enhanced cashflow

*See Harvest ETF’s Disclaimer at the end of the newsletter

TOP STORY
🤖 OpenAI’s “GPT-6 Astra” Sends Shockwaves Across the AI Industry

🤖 This was a MASSIVE week in AI, with four of the world’s leading AI companies releasing their new frontier models:

  • 1️⃣ Tuesday: Anthropic released Claude Fable 5.1 and Mythos 5.1

  • 2️⃣ Wednesday: Meta launched Muse Spark 1.3, and Google launched Gemini 3.8 Flash

  • 3️⃣ Thursday: OpenAI launched GPT-6 Astra

🚀 Typically, these labs ship a new flagship every few months, so four frontier launches in three days is very rare.

🏆 But despite all the launches, there was one that stole the headlines: OpenAI’s Astra.

🤖 So let's dive into what's so special about Astra, the impact it's had on the stock market, and who's actually winning the AI race…

🌟 What's Astra?

🧠 GPT-6 Astra is OpenAI’s most powerful model yet, and the company is presenting it as a massive leap in what AI can do. OpenAI’s President even went so far as to call it AGI:

“I think it’s not unreasonable to feel that we are now in the AGI era.”

Greg Brockman, President, OpenAI

🖥️ The main capability is “computer use,” aka the model can control software, browsers, and desktop apps on its own to complete multi-step tasks like filling forms, updating records, building presentations, and reviewing financials (as seen in the company’s demo video here).

“Anything you can do on a computer, Astra can do for you. Fast.”

OpenAI

📈 And while OpenAI isn’t public (yet), SoftBank ($SFTB), one of OpenAI's biggest backers, surged more than 10% overnight in Tokyo, so the market is clearly quite excited about the new model.

📊 How Does Astra Stack Up

🥇 Using both OpenAI’s published numbers and independent testing, Astra leads on computer use, math, and cybersecurity (100% on ExploitBench), and also scored near-perfect (99.9%) on the reasoning benchmark that’s been used as the marker for “AGI-like” ability.

🥈 That said, Anthropic’s Fable 5.1 still leads on general reasoning, while coding is essentially a tie between Astra, Meta’s Muse Spark 1.3, Gemini 3.8 Flash, and Claude Opus 5.

⚡ Astra’s real commercial edge is efficiency, where, on Artificial Analysis’s Coding Agent Index, it matches Fable 5 on coding at less than half the cost thanks to big token-efficiency gains.

🧑‍🔬 All in all, analysts seemed impressed with the model, with Gary Marcus, one of the loudest critics of recent AI launches, saying that Astra “looks to be pretty impressive” and appears to be a “genuine advance,” but pushing back on the “this is AGI” statement from OpenAI’s president.

📈 What This Means for the Market

📊 Regardless of who "leads" the foundation model race, the rise of true AI agents (and ‘computer use’) will continue to drive increasing demand for compute, and the market responded accordingly:

💬 Goldman Sachs' head of trading Rich Privorotsky called Astra "precisely what the AI bull market has been waiting for," and explained why a smarter model matters more than a cheaper one:

"Price declines are one thing, but a true breakthrough in intelligence will reshape the demand curve once again… It forces all other laboratories to catch up, sustains the vitality of the spending cycle, and reactivates the belief that 'there are still better things worth building.'"

Rich Privorotsky, Head of Delta One Trading at Goldman Sachs

🏁 But notice what Privorotsky said: Astra forces every other lab to catch up. Which raises the trillion-dollar question… if the lead changes hands with every launch, who actually wins this race?

🤔 So Who’s Actually Winning the Race?

📈 The simple answer is that there’s no single “best” model anymore because each model leads in some domains and trails in others, and the leaderboard flips constantly.

💰 On the revenue side, Anthropic has overtaken OpenAI since April as they dominate enterprise sales, so OpenAI desperately needs Astra to help it retake its lead.

🩴 But this flip-flopping leads us to an interesting dilemma… if one update can change who’s in the lead, what is the moat?

💬 Tech analyst Benedict Evans has been making this argument since early last year:

“OpenAI and all the other foundation model labs have no moat or defensibility except access to capital.”

Benedict Evans, tech analyst

🤝 And Big Tech seems to agree, because nearly all of them have hedged their bets across labs rather than backing one winner (even the ones who have their own models):

  • ☁️ Microsoft ($MSFT) has revenue-sharing and licensing rights to OpenAI’s models, but OpenAI also represents a huge chunk of Azure’s backlog

  • 📦 Amazon ($AMZN) is Anthropic’s primary cloud and chip partner through AWS and Trainium, but now serves Astra via AWS too

  • 🔍 Google ($GOOGL) builds its own frontier models and owns its own distribution, yet still holds a stake in Anthropic

👀 This matters because Anthropic and OpenAI are spending billions more than they're making, with the rationale that this will be a winner-takes-all market (as was the case with Google for search).

🤔 But given how fast the lead flips, it's unclear whether these are future monopolies or brutally competitive commodities, and with OpenAI and Anthropic’s trillion-dollar IPOs coming up quickly (with Anthropic’s estimated for mid-October and OpenAI’s planned for 2027), the answer matters a lot.

😰 If it's the latter (commodities), those trillion-dollar price tags get much harder to justify.

💡 For retail investors, this is an important reminder that "current leader" and "long-term winner" are very different things.

👀 By the time these IPOs arrive, the leaderboard may have flipped again, and it's worth noting that even the smartest money in tech aren’t picking a winner, they’re hedging across the entire market.

🧘‍♂️ But enough about OpenAI, let’s switch gears to two very different companies: Lululemon’s 17% stock drop, Dell’s 16% surge, and more…

📊 But first, another quick word from our other sponsor this week, Global X Canada!

SPONSORED BY GLOBAL X
⚙️ The Chips Behind AI

⚡️Artificial Intelligence (AI) may be grabbing the headlines, but semiconductors is a key technology powering it.

From AI and data centres to gaming, autonomous vehicles and advanced computing, chips are at the heart of today’s digital economy.

Get Exposure to the Chips Powering AI

📊 For investors looking to gain access to the rather rapid growth of the digital economy, the Global X Artificial Intelligence Semiconductor Index ETF (CHPS) offers targeted exposure to one of the market’s most important industries.

🌐 CHPS provides direct exposure to some of the largest global companies that design, manufacture, and distribute semiconductors. It provides focused exposure to the semiconductor industry through a diversified portfolio of companies across the global chip ecosystem through a single ETF.

As of July 31, 2026, CHPS delivered a 44.44% YTD return and a 71.40% 1-year return (Global X Investments Canada Inc., As at July 31, 2026)

*See Global X’s disclaimer at the end of the newsletter

EARNINGS ROUNDUP
📊 Dell Soars 16% and Lululemon Crashes to 8-Year Low

🗓️ Aside from AI news, there were two high-profile companies that reported earnings this week that had very different reactions from the market… starting with Lululemon.

🧘 Lululemon Crashes to 8-Year Low

📉 First up, the biggest loser of the week: Lululemon ($LULU), whose stock fell 17% on Friday, bringing it to its lowest level in 8 years, with shares now down 80% from its all-time high in December 2023.

📊 On the surface, Earnings per Share of $2.92 crushed the $1.80 expected, but almost all of that beat came from a one-time $134.5 million tariff refund, worth $0.86 per share. Strip that out, and EPS was $2.06 while revenue actually fell 4% to $2.42 billion (vs. $2.46 billion expected).

🇨🇳 Mainland China, which was supposed to be the growth engine, was the single biggest source of the miss, with revenue up just 4% (and actually down 2% in constant currency, with comparable sales down 8%) after a wave of social media backlash.

🇺🇸 But this isn't just a China problem: revenue in the Americas, still by far Lululemon's biggest market, fell 8% with comparable sales down 9%.

🔮 Lululemon now expects Q3 revenue to drop 10-11%, cutting full-year revenue guidance to $10.35-10.5 billion (down 5-7% year-over-year).

👟 The company's new CEO Heidi O'Neill (a former top Nike executive) officially takes the reins next week, inheriting a brand losing share to cheaper competitors like Alo and Vuori. With analysts cutting price targets left and right, Heidi has a big challenge ahead to bring Lulu back to its glory days.

🌼 That said, not everyone is throwing in the towel. @bradleytalksmoney (a longtime Lulu bull on Blossom) published an honest thesis review this week, calling the quarter what it is, but arguing O'Neill is the silver lining, pointing out that management insists Lululemon's problem is marketing and brand, not product quality, and O’Neill’s 25+ years at Nike were spent running exactly that: product creativity, consumer strategy, and brand:

In the short term, it may continue to decline, and I'm sure I'll continue to be ridiculed… but I believe an inflection is coming, and patience will be rewarded."

Bradley (@bradleytalksmoney), on Blossom

🖥️ Dell Surges 16% After Reporting a $95 Billion AI Backlog

🔔 Next up was Dell ($DELL), which posted the most jaw-dropping numbers of the week on Tuesday, sending shares surging 16% and bringing the stock's year-to-date gain to more than 310%, making it one of the best-performing AI trades of 2026.

📊 Revenue hit a record $47.0 billion, up 58% year-over-year (vs. $44.9 billion expected), while adjusted EPS came in at $7.04, up 203% year-over-year (vs. ~$4.91 expected).

🚀 But the real headline was the insane demand: Dell booked a record $60.9 billion in AI server orders this quarter (more than an entire YEAR of AI orders in three months) and reported a $95 billion AI backlog, meaning the next several quarters are essentially already sold out, with management saying the pipeline beyond the backlog is "multiples" of that figure.

⚠️ It wasn't all perfect though. Management admitted part of the revenue growth comes from higher prices driven by the same memory shortage we've been covering for months (DRAM, NAND, CPUs, and rack components are all supply-constrained), and free cash flow actually fell 47%, showing just how expensive it is to build all this AI hardware before customers pay up.

💡 Still, with AI margins nearly doubling and a soaring backlog, Dell has cemented itself as one of the best picks-and-shovel plays in the market this year.

FROM THE BLOSSOM COMMUNITY
⭐️ Featured Posts of the Week

👇 Click to see the full post!

Harvest ETFs Disclaimer

Commissions, management fees and expenses all may be associated with investing in Harvest High Income Shares ETFs managed by Harvest Portfolios Group Inc. (the “Funds” or a “Fund”). Please read the relevant prospectus before investing. The Funds’ returns are not guaranteed, their values change frequently, and past performance may not be repeated. Tax investment and all other decisions should be made with guidance from a qualified professional.

Distributions are paid to you in cash unless you request, pursuant to your participation in a distribution reinvestment plan, that they be reinvested into available ETF Class A Units of the Fund. If a Fund earns less than the amounts distributed, the difference is a return of capital.

 The Fund is categorized as a liquid alternative ETF. This means it has the ability to use leverage and can invest more than 10% of its assets in a single issuer. The Fund employs modest leverage of approximately 25%, which can amplify both gains and losses.

Global X Disclaimer

The indicated rates of return are the historical annual compounded total returns, including changes in per unit value and reinvestment of all distributions and do not take into account sales, redemption, distribution, or optional charges or income taxes payable by any securityholder that would have reduced returns. The rates of return shown in the table are not intended to reflect future values of the Fund(s) or future returns on investment in the Fund(s). Only the returns for periods of one year or greater are annualized returns.