News

Listen to the full episode now or follow the Saxo Market Call on your favorite podcast app.

(and yes, pardon the bad math in the podcast on how far the aggressive S&P 500 target discussed is from the current price!)

Today’s Links

• Michael McNair with a TLDR post on Sirius XM, the financially profitable, but stagnating broadcast media (from space) company, who might be an acquisition target from a deep-pocketed suitor, he proposes, due to its possession of a fat chunk of the radio broadcast spectrum. More below in today’s Chart of the Day.

• Don’t know to what degree Chinese manufacturers can make inroads into the big three RAM memory chip makers (Micron, Samsung and SK Hynix) but Tom’s Hardware has noted that Chinese chips are finally showing up in mainstream RAM memory modules for the first time. These chips are not suitable for the highest end modules, but could bring relief to PC suppliers or at least self-building enthusiasts if the prices move lower as well.

• “You know when you’re at a Kiwi BBQ when someone brings up house prices before the sausages are cooked”. New Zealand has been one of those small open economies plagued by extreme moves in house prices going back to the post-GFC extremely low interest rate era. This article from Bloomberg mostly discusses the pandemic era pump in prices from lowered rates and the subsequent bust, with policy makers struggling for the right policy to deal with the problem. Canada has seen a similar struggle and the US market is frozen, without yet seeing broad-based declines, but likely only a small recession away from being stuck in the same boat.

• Jeffrey Currie weighs in on the risks the hyperscalers are running by taking such a capital-intensive approach to trying to grow their businesses, comparing the approach of John Rockefeller and Standard Oil - Currie points out that you have to own the chokepoints (right Micron, Samsung and SK Hynix, oh and Western Digital and Seagate? - at least for now.)

• Who to believe on the Hormuz Strait? Jeffrey Currie the vaunted commodity strategist continues to bemoan the lack of opening and the risks, saying “sell the tweet [Trump social media posts], buy the molecule [oil]”, while James E. Thorne says that with every day that passes, Iran ‘s power is eroded as supply chains shift away from allowing the Strait to continue to so viciously constrict supply reaching market and boosting output elsewhere (especially US output).

• What will the Warsh Fed be about? Certainly not tightening policy, according to James E. Thorne, who suggests a battle will be afoot for the new Fed Chair relative to Wall Street expectations and others on the Fed board.

• AI Bears versus AI Bulls. Hedgie with a post on X discussing Uber’s experience with encouraging AI use, only to discover that heavy token usage with Claude Code is not seeing a productivity boost or more useful software shipping. Is it just too early days or is there simply too much compute required to drive meaningful productivity growth for larger scale organizations? And then we have James E. Thorne again waxing optimistic on the ability for an AI super-cycle to deliver massive continuing profit growth far superior to current expectations and possibly taking the S&P 500 to 14,000-16,000 by 2021.

• Cem Karsan puts in an appearance on the Thoughtful Money podcast, talking long term expected return outlook when we already sit atop high valuations and a strong market. Good long term perspective and he talks ways to position for coming risks (these I have not yet listened to).

Chart of the Day - Are you Sirius?

Not a company I have thought about in a long while, but the TLDR post from Michael McNair on the company as a potential target for those looking to buy it chiefly for its ownership of fat chunk of broadcasting spectrum far more than for its media business is an interesting one. The market may have gotten the memo recently as the stock price has picked up 50% or so from recent lows. Time will tell if this company becomes an acquisition target and, if so, at what price - it sells at 8-9 times earnings, yielding around 3.5% at current prices with no topline growth for years. The monster jump in late 2024 was on the culmination of Warren Buffett’s move to acquire a chunk of the company (revealed at 31% after he had been buying the company and the associated Liberty Media for quite some time) and the more recent surge might be on enthusiasm for all things space-related as we head toward SpaceX’s IPO soon. The company came from a near death experience in 2009, when its share traded hands for as little as 50 cents a share - down from as high as 89 dollars a share in 2004 (which was up from a low of under 4 dollars a share in 2003, down in turn from the all-time high in 2000 of…656 dollars a share). This is not an investment recommendation.

Questions and comments, please!

Key takeaways

• Nvidia delivered another very strong quarter, led by demand for artificial intelligence infrastructure.

• Guidance beat expectations, but China and high expectations kept the initial stock reaction restrained.

• The bigger investor question is now duration: how long can this spending cycle stay this strong?

Nvidia did not just report earnings on 20 May 2026. It handed investors another progress report on the artificial intelligence buildout.

The numbers were strong. Nvidia reported first-quarter revenue of 81.6 billion USD, up 85% from a year earlier. Data centre revenue, the core of the artificial intelligence story, reached 75.2 billion USD, up 92%. The company guided for second-quarter revenue of ahead of expectations, according to data compiled by Bloomberg.

The initial after-market reaction was mildly negative, suggesting investors liked the strength but were already looking beyond the headline beat. That may sound harsh. It is also what happens when a company becomes the unofficial scoreboard for one of the largest investment themes in markets. The earnings question is no longer whether demand is real. It is whether demand can stay this strong for long enough to justify the expectations already in the share price.

Strong numbers, stricter teachers

The main result was clear: demand remains powerful.

Revenue grew 20% from the previous quarter and 85% from a year earlier. Adjusted gross margin was 75%, which means Nvidia still keeps a very large share of each sales dollar after direct production costs. That matters because it suggests customers are still paying up for performance, supply and access to Nvidia’s platform.

Nvidia also announced an additional 80 billion USD share repurchase authorisation and raised its quarterly dividend from 0.01 USD to 0.25 USD per share. That will not be the main story for growth investors, but it does show how much cash the artificial intelligence boom is creating.

The market’s restrained initial reaction is therefore less about weak results and more about the bar. Nvidia has become so important that good is not enough. Investors want proof that the cycle extends into 2027 and beyond, that margins can stay high, and that customers are not simply ordering ahead before the next chip transition.

The data centre is now the centre of gravity

Data centres are the large facilities that store, process and move digital information. Artificial intelligence has turned them into industrial assets. They need chips, memory, networking, cooling, power and land. The cloud may sound weightless, but it increasingly looks like a very expensive electricity-hungry factory.

Nvidia’s data centre compute revenue reached 60.4 billion USD, up 77% from a year earlier. Networking revenue was 14.8 billion USD, up an eyepopping 199%. That second number matters because artificial intelligence systems do not rely only on one powerful chip. They rely on many chips working together quickly. Networking is the plumbing. Without it, the palace has no running water.

This is why Nvidia’s results matter beyond Nvidia. A strong data centre update supports the broader artificial intelligence supply chain: chip manufacturers, memory companies, networking specialists, data centre operators, power equipment makers and cloud platforms. It also helps explain why the theme has moved from software excitement to infrastructure reality.

For long-term investors, that shift is important. The artificial intelligence story is no longer just about chatbots and clever demos. It is about capital expenditure, supply chains and returns on invested capital. In plain English: companies are spending vast sums, and investors now need to ask who earns attractive profits from that spending.

China is still the quiet swing factor

Nvidia said its second-quarter outlook does not assume any data centre compute revenue from China. That is a key sentence. It means the guidance beat expectations even without including a contribution from one of the world’s largest technology markets.

That makes the result more impressive, but it also shows where uncertainty remains. Export controls, product restrictions and geopolitical decisions can change Nvidia’s addressable market. Addressable market simply means the revenue opportunity a company can realistically target. For a business this large, even small changes in market access can become very large numbers.

Competition is another factor. Advanced Micro Devices, Broadcom, custom chips from cloud companies and new internal designs all matter. They may not remove Nvidia’s leadership quickly, but they can reduce pricing power over time. The question is not whether Nvidia faces competition. The question is whether its full system, from chips to software and networking, remains hard enough to replace.

Risks to watch

The first risk is expectations. Nvidia can report excellent numbers and still see pressure if investors expected something close to perfect. That is not unfair. It is just what happens when a stock carries a very large share of market confidence.

The second risk is customer spending. The largest cloud companies are investing heavily in artificial intelligence infrastructure. If those companies slow spending, delay orders or struggle to turn artificial intelligence services into revenue, the whole supply chain could feel it.

The third risk is geopolitics. Watch China commentary, export licences and any change in the company’s assumption around China data centre revenue. In this story, policy can move faster than a product roadmap.

Investor playbook

• If shares move lower after strong results, compare the move with guidance, margins and China assumptions.
• If margins stay around current levels, it suggests Nvidia still has pricing power and customer urgency.
• If networking keeps growing faster than compute, it supports the idea that artificial intelligence systems are scaling.
• If cloud companies slow capital spending, stress-test the wider artificial intelligence supply chain, not just Nvidia.

The neat ending

Nvidia’s quarter shows that the artificial intelligence buildout remains very real, very large and very profitable for the companies sitting closest to the infrastructure layer. But the market’s first reaction also shows that investors have become harder to impress. That is healthy.

Great companies still need to clear great expectations, and Nvidia’s report is no longer just about one quarter of revenue. It is about the length, profitability and resilience of an entire investment cycle. The lesson for investors is not to guess tomorrow’s share price. It is to understand the machinery behind the story, because the machinery is now the story.

This material is marketing content and should not be regarded as investment advice. Trading financial instruments carries risks and historic performance is not a guarantee of future results.