Listen to the audio version of this article (generated by artificial intelligence).
Iran Talks Falter and Oil Rise… The Real Question Inside Big Tech’s Earnings… Why the Mythos Breach Changes Things… and What It Means for Your Portfolio
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 canceled plans to send envoys Steve Witkoff and Jared Kushner to Islamabad for face-to-face talks, citing “infighting and massive confusion” within Tehran’s leadership.
However, this morning, Reuters Reports indicate that Iran has proposed a new plan: end the war first, resolve the dispute over the Strait of Hormuz second, and only then address Iran’s nuclear program.
This sequence is almost certainly unacceptable in Washington. The Trump administration has been clear that nuclear issues should be on the table from the beginning. As Trump said on Sunday:
They cannot have a nuclear weapon; Otherwise, there is no need to meet.
As I write this article, it appears that the two sides remain far apart – and the energy market is pricing in a distance. Brent crude (the European benchmark) is trading at around $109 per barrel, while the price of West Texas Intermediate crude oil is at $97.
Goldman Sachs raised its Brent crude forecast to $90 by late 2026 — up from $80 previously — citing ongoing Gulf unrest and a supply drawdown that it estimates will reach a record 11 to 12 million barrels per day here in April.
Invesco sets $80 as a possible floor for Brent crude this year in the absence of a full normalization of Hormuz flows. Even if the strait is eventually reopened, the stock pit being drilled now will take months to refill.
The bottom line: The longer oil prices remain at these high levels, the more pressure they will put on American consumers, the greater the risk of inflation, and the more difficult it will be for the Fed to justify the interest rate cuts the market wants. We will continue to track this as well as its economic consequences.
Now, let’s focus on one of the biggest market movers this week.
Big Tech earnings this week aren’t really about Big Tech
This week, four of the world’s five most valuable companies announced their quarterly results within 48 hours. Microsoft (MSFT), alphabet (Google), meta platforms (dead) and Amazon.com (Amzn) They all drop their numbers on Wednesday. apple (Apple) Followed on Thursday. This means approximately $16 trillion in market capitalization reported within 48 hours.
Although earnings beats and misses will make headlines, this week’s reports are less about the bottom line for any individual super-scaler and more about the ongoing judgment of the entire AI trade.
So, before Wednesday’s numbers hit, let’s define what you should really be looking for and why…
The billions of dollars pouring into AI infrastructure have been hugely profitable – for select chip companies like nvidia (NVDA)data center operators, energy companies, and for the entire ecosystem of pick-and-roll plays that have driven much of the bullish AI market. This spending is real. These revenues are real. These stock gains are real.
But this spending by hyperscale companies does not match consumer and enterprise demand for AI products.
Hyperscalers have been building — aggressively, almost feverishly — on assumption That end users will eventually pay for AI at scale.
Microsoft, Alphabet, Meta and Amazon spent a combined $410 billion on capital expenditures last year, nearly three times what they spent in 2022. Wall Street expects as much as $674 billion from those four companies this year alone — the third straight year of combined growth exceeding 60%.
Chart: Visual Source Alpha / Wall Street Journal
This is an amazing bet on future demand.
While AI Infrastructure Trading counts a win, the revenue side will show if the win is real, which brings us back to this week’s earnings.
The immediate diagnosis to watch this week: revenues
Three years into this arms race, monetization is getting harder to move forward.
Do direct-to-consumer AI products (whether for individuals or companies) make a lot of money?
So, this week we need to keep a close eye on revenue signals on super-fast corporate earnings calls.
Are cloud revenues accelerating in ways that can be reliably linked to AI workloads? Are enterprise customers moving from betas to full deployments? Are any of these companies starting to generate AI revenue in ways they haven’t before?
Strong signals here do not permanently stabilize the bull’s condition, nor do weak signals kill him. But three years on, and each earnings cycle without clear progress in generating income reduces investors’ patience.
The market was willing to finance construction. But at some point, the building has to finance itself. This week marks an additional weight on the scale, and the balance becomes difficult to ignore.
But this week’s revenue picture is just the beginning…
Long-term vital sign: capital expenditures
Revenue tells us what actually happened. But capex guidance tells us what these companies think is coming — and reveals whether conviction in AI ROI is holding up or fading.
Over the past few years, we’ve seen panic on Wall Street over how much money hyperscalers are spending to build AI. But now, the market has bought into the whole “you have to spend to win” argument. So, aggressive spending is the baseline.
But I’ll go one step further…
With many AI infrastructure stocks surging thanks to a superspending boom, aggressive spending is now the desired strategy. But this means that the biggest danger today is not overspending, but… Lack of spending.
If a large, aggressively expanding company quietly revised its capital forecasts, eased the timeline for its data center expansion, or used language suggesting a “reassessment” of its investment pace, Wall Street wouldn’t just react to lower spending. He will start asking a more disturbing question…
What does this company know that we don’t know?
Reducing spending will not be seen as a financial responsibility, but as a lost conviction.
This fear will not remain confined to one person. It will spread across the entire AI complex.
To be clear, no one is expecting this signal this week. So, we raise this issue not because Wednesday is the time to worry about it, but because it is Canary Day He deserves Know now – and track – before it becomes obvious.
Because when/if it becomes clear 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 bringing you the key takeaways as the results come in. For now, all eyes are on revenue.
The story of Mythos has become more compelling
Two weeks ago, we covered a story that shook the world of cybersecurity…
AI company Anthropic has created Cloud Mythos, a model its developers describe as “currently far ahead of any other AI model in cyber capabilities.” Mythos is claimed to be able to find and exploit vulnerabilities in every major operating system and browser.
In response, Treasury Secretary Bescent and Federal Reserve Chairman Jerome Powell summoned the CEOs of major U.S. banks to the Treasury Building to discuss the threat posed by Mithos to the financial infrastructure.
Anthropic launched the Glasswing project on the same day, giving eleven partners—among them Amazon, Apple, Google, Microsoft, and Nvidia—early access to stress testing their own systems.
Meanwhile, Wall Street’s judgment on what this means for the cybersecurity industry has been swift and brutal: Palo Alto Networks (Bano) fell about 10% Crowd Strike (Raw) Decreased 11% and ZeSkiller (ZS) 14% fall.
But the story is not over…
Last week, a small group of unauthorized users from the online Discord community gained access to Mythos. They gained entry through a combination of access to in-house contractors, web search bots, and educated guesses about the model’s online location — and they’ve been using it regularly ever since.
Anthropic says it is investigating and has found no evidence that the breach was widespread. But the damage to trust has already been done.
here luck:
If a group of AI nerds manage to get into Mythos – with no apparent malicious intent – imagine the repercussions if the next people to sneak through that door are actual criminals.
But the hack confirmed something bigger
AI is now accelerating the discovery of cyber vulnerabilities faster than organizations can patch them. here luck:
AI can now find and exploit flaws so quickly that defenders may be the ones truly at risk…
This is the new reality that the Glasswing project is designed to address. Whether it can move fast enough is a different question.
We will continue to track this because its results are important not only for us Cybersecurity stocksBut for the safety of our sensitive data online.
But there is a slight silver lining here…
The same ability is indicated in the markets
Through a “technical” lens, Mythos is about AI crossing the threshold of pattern recognition. As Keith Kaplan, CEO of TradeSmith, said this week:
Mythos finds patterns in computer programs that no human can see.
It reads millions of lines of code and identifies certain sets of conditions that indicate a vulnerability. Then he works on them.
Our friends at TradeSmith have spent years doing exactly this: building a system that applies this exact type of AI-powered pattern recognition – but to the financial markets.
Back to Keith:
The stock market has similar types of hidden structures. There are “signals” buried in decades of data for every stock — specific sets of circumstances that precede consistently big moves.
For most of the market’s history, it was invisible. The data was there… but no one had the tools to read it.
But today, the TradeSmith system has identified 30,000 of these signals across nearly 2,500 stocks, each with historical accuracy rates of 75% or better.
Their model portfolio of three stocks — always three positions on the S&P 500, rotated algorithmically when exit signals are triggered — produced a back-tested compound annual return of 54% from January 2020 through January 2026, versus roughly 15% for the S&P 500 over the same period.
Backtests are not guarantees, but the basic logic is sound: The same AI leap that makes Mythos so worrying to cybersecurity professionals is what makes tools like this possible for everyday investors.
In the past week, Keith has walked nearly 9,000 people through the complete system in his life Artificial intelligence signals trading event. Replay is still available – includes trade examples, strategy details, and his current selections from the model portfolio. The replay is available for a limited time here.
I wish you a good evening,
Jeff Remsburg
