Iran talks stall and oil climbs… the real question inside Big Tech earnings… why the Mythos breach changes things… and what it means for your portfolioVIEW IN BROWSER As I write on Monday, oil is trading higher in the wake of this weekend’s failed peace negotiations. Over the weekend, President Donald Trump scrapped plans to send envoys Steve Witkoff and Jared Kushner to Islamabad for face-to-face talks, citing “tremendous infighting and confusion” within Tehran’s leadership. However, this morning, Reuters is reporting that Iran has proposed a new plan: end the war first, resolve the Strait of Hormuz dispute second, and only then address Iran’s nuclear program. That sequencing is almost certainly a non-starter in Washington. The Trump administration has been explicit that nuclear issues must be on the table from the outset. As Trump put it Sunday:
They cannot have a nuclear weapon; otherwise, there’s no reason to meet.
As I write, the two sides appear to remain far apart – and the energy market is pricing in the distance. Brent crude (the European benchmark) trades at nearly $109 a barrel, while West Texas Intermediate Crude is up to $97. Goldman Sachs has raised its Brent forecast to $90 by late 2026 – up from $80 previously – citing persistent Gulf disruptions and a supply drawdown it estimates is running at a record 11 to 12 million barrels per day here in April. Invesco puts $80 as a likely floor for Brent this year absent a full normalization of Hormuz flows. Even if the Strait eventually reopens, the inventory hole being dug right now will take months to refill. Bottom line: the longer oil prices stay at these elevated levels, the more pressure they put on the U.S. consumer, the greater the risk of inflation, and the harder it becomes for the Fed to justify the rate cut the market wants. We’ll keep tracking this as well as its economic consequences. Now, let’s zero in on one of the biggest potential market movers this week. This week’s Big Tech earnings aren’t really about Big TechThis week, four of the five most valuable companies in the world report quarterly results within 48 hours. Microsoft (MSFT), Alphabet (GOOGL), Meta Platforms (META) and Amazon.com (AMZN) all drop their numbers on Wednesday. Apple (AAPL) follows on Thursday. That’s roughly $16 trillion in market cap reporting inside 48 hours. Although earnings beats and misses will generate headlines, this week’s reports are less about any individual hyperscaler’s bottom line and more about the ongoing verdict on the entire AI trade. So, before Wednesday’s numbers hit, let’s frame what you should really be looking for and why… The billions of dollars flowing into AI infrastructure have been enormously profitable – for select chip companies like Nvidia (NVDA), data center operators, power companies, and for the whole ecosystem of picks-and-shovels plays that have driven much of the AI bull market. That spending is real. Those revenues are real. Those stock gains are real. But that spending by hyperscalers is not the same as consumer and enterprise demand for AI products. The hyperscalers have been building – aggressively, almost frantically – on the assumption that end users will eventually pay for AI at scale. Microsoft, Alphabet, Meta, and Amazon spent a combined $410 billion on capex last year, nearly triple what they spent in 2022. Wall Street expects as much as $674 billion from those four companies this year alone – the third consecutive year of combined growth exceeding 60%. 
Chart: Source Visible Alpha / WSJ That’s a staggering bet on future demand. And while the AI infrastructure trade has been pricing in the win, the revenue side will demonstrate if the win is real, which brings us back to this week’s earnings.
Recommended Link |
|
When DeepSeek hit the headlines, Nvidia dropped nearly $600 billion in a single session. Every investor felt it: What if China just won the AI race? But behind closed doors, America has already fired back — with a hidden project at the same Tennessee lab that built the atom bomb. 40,000 scientists. A device trillions of times more powerful than anything China has built. A $100 trillion AI reset on the way. Billion-dollar money manager Louis Navellier has identified the one stock that wins this arms race. Click here for the name and ticker — free. |
 |
|
The immediate diagnostic to watch this week: revenuesThree years into this arms race, the monetization question is becoming harder and harder to push off… Are consumer-facing AI products (whether to individuals or companies) making serious money? So, this week, we need to watch the revenue signals closely in hyperscalers’ earnings calls. Are cloud revenues accelerating in ways that can be credibly tied to AI workloads? Are enterprise customers moving from pilots into full deployments? Are any of these companies beginning to break out AI-specific revenue in ways they haven’t before? Strong signals here don’t prove the bull case permanently, nor do weak ones kill it. But we’re three years in now, and each earnings cycle without clear monetization progress shortens the runway for investors’ patience. The market has been willing to fund the build. But at some point, the build has to fund itself. This week is one more weight on the balance, and the balance is getting harder to ignore. But this week’s revenue picture is just the start… The longer-term vital sign: capexRevenues tell us what’s already happened. But capex guidance tells us what these companies believe is coming – and reveals whether conviction in AI’s ROI is holding or fading. Over the last few years, we’ve seen Wall Street panic about how much money the hyperscalers are spending to build out AI. But now, the market has fully bought the “you have to spend to win” argument. So, aggressive spending is the baseline. But I’d take it one step further… With many AI infrastructure stocks soaring thanks to the hyperscaler spending boom, aggressive spending is now the desired strategy. But this means that the far greater risk today isn’t overspending, but underspending. If a major hyperscaler quietly revises its capex outlook downward, softens its data center expansion timeline, or uses language suggesting a “reassessment” of its investment pace, Wall Street won’t just react to the reduced spending. It would start asking a more unsettling question… What does this company know that we don’t? Reduced spending wouldn’t read as fiscal responsibility, but as lost conviction. And that fear wouldn’t stay contained to one hyperscaler. It would ripple across the entire AI complex. To be clear, nobody is expecting that signal this week. So, we raise this issue not because Wednesday is the moment to worry about it, but because it’s the canary worth knowing about now – and tracking – well before it becomes obvious. Because when/if it does become obvious to the masses – and the end-user AI revenue side hasn’t picked up the slack by then – get ready for fireworks. We’ll be watching closely and will bring you the key takeaways as results roll in. For now, all eyes on revenues. The Mythos story just got more urgentTwo weeks ago, we covered a story that rattled the cybersecurity world… AI company Anthropic had created Claude Mythos, a model its own developers described as “currently far ahead of any other AI model in cyber capabilities.” Mythos is allegedly capable of finding and exploiting vulnerabilities in every major operating system and browser. In response, Treasury Secretary Bessent and Federal Reserve Chairman Jerome Powell summoned the CEOs of major U.S. banks to the Treasury Building to discuss the threat Mythos posed to financial infrastructure. Anthropic launched Project Glasswing the same day, giving eleven partners – among them Amazon, Apple, Google, Microsoft and Nvidia – early access to stress-test their own systems. Meanwhile, Wall Street’s verdict on what that meant for the cybersecurity industry was swift and brutal: Palo Alto Networks (PANW) fell roughly 10%, CrowdStrike (CRWD) dropped 11% and Zscaler (ZS) shed 14%. But the story isn’t over… Last week, a small group of unauthorized users from the online community Discord gained access to Mythos. They got in through a mix of insider contractor access, web-scouring bots, and educated guesses about the model’s online location – and they’ve been using it regularly since. Anthropic says it’s investigating and has found no evidence the breach was widespread. But the damage to confidence is already done. Here’s Fortune:
If a group of AI nerds could get into Mythos – allegedly without malicious intent – imagine the fallout if the next ones to slide through that door were actual criminals.
But the breach confirmed something biggerAI is now accelerating the discovery of online vulnerabilities faster than organizations can patch them. Here’s Fortune:
AI can now find flaws and exploit them so quickly that defenders may be the ones left truly exposed…
That’s the new reality Project Glasswing was designed to address. Whether it can move fast enough is a different question. We’ll keep tracking this as its outcome is critical not only for cybersecurity stocks, but for the safety of our sensitive data online. But there is a slight silver lining here… The same capability, pointed at marketsViewed through a “tech” lens, the Mythos story is about AI crossing a threshold in pattern recognition. As TradeSmith CEO Keith Kaplan put it this week:
Mythos finds patterns in computer programs that no human could see.
It reads millions of lines of code, identifies the specific combinations of conditions that point to a vulnerability. Then it acts on them.
Our friends at TradeSmith have spent years doing exactly this: building a system that applies this exact kind of AI-powered pattern recognition – but to financial markets. Back to Keith:
The stock market contains similar kinds of hidden structures. Buried inside decades of data for every stock are “signals” — specific combinations of conditions that have consistently preceded big moves.
For most of market history, they were invisible. The data existed… only nobody had the tools to read it.
But today, TradeSmith’s system has identified 30,000 of these signals across nearly 2,500 stocks, each with historical accuracy rates of 75% or better. Their flagship three-stock model portfolio – always three S&P 500 positions, rotated algorithmically when exit signals fire – produced a backtested compounded annual return of 54% from January 2020 through January 2026, versus roughly 15% for the S&P 500 over the same period. Backtests aren’t guarantees, but the underlying logic is sound: the same AI leap that makes Mythos alarming to cybersecurity professionals is the one making tools like this possible for regular investors. Last week, Keith walked nearly 9,000 people through the full system in his live AI Signals Trading Event. The replay is still available – trade examples, strategy details, and his current model portfolio picks included. The replay is available for a limited time right here. Have a good evening, Jeff Remsburg |
Post a Comment
Post a Comment