Good morning investors,
Full edition this morning after Friday's abbreviated note. The charity golf event was a success, though we did not manage to defend our title from last year. My game is starting to feel good again, and I would love to get out on the course with some readers this summer. If you are in the Charlotte area and want to play a round, reach out.
Now, back to markets.
Futures are under pressure to start the week. Dow futures have dropped 373 points, or 0.8%. S&P 500 futures are down 0.4%. Nasdaq 100 futures have fallen 0.3%.
The 10-year Treasury yield remains firmly in focus after pushing past 4.5% on Friday. Global bond yields are rising in tandem. Japanese 10-year yields jumped over 9 basis points overnight to 2.79%. UK 30-year gilt yields touched levels not seen since the late 1990s.
Oil prices are climbing after President Trump warned over the weekend that Iran "better get moving" or "there won't be anything left of them." WTI has advanced to $106.65 per barrel. Brent is trading at $110.12.
The ceasefire remains fragile. The Strait remains largely closed. And negotiations appear deadlocked.
Friday's Pullback
The streak finally paused.
The S&P 500 fell 1.2% on Friday to finish the week up just 0.1%. The Nasdaq dropped 1.5% for a slight weekly loss of 0.08%. The Dow declined 1.1% for a weekly loss of 0.2%.
After seven consecutive weeks of gains, this counts as a breather. The S&P 500 briefly touched 7,550 during the week. The Dow reclaimed 50,000 before retreating. Records were set, then given back.
The proximate cause was rising bond yields. But the broader context matters more. Markets had run extraordinarily hard since late March. Semiconductors had gained 70% in roughly seven weeks. Profit-taking was overdue. I view this as healthy, and not alarming.
The Trump-Xi Aftermath
President Trump left Beijing with warm words but few concrete wins.
The trade truce holds. Xi accepted an invitation to visit Washington on September 24. Both sides agreed that the Strait of Hormuz must remain open. China reportedly committed to purchasing 200 Boeing aircraft, though that fell well short of the 500 initially expected.
On AI chips, the Commerce Department cleared 10 Chinese companies to purchase Nvidia's H200, including Alibaba, Tencent, ByteDance, and JD.com. But the U.S. Trade Representative confirmed that chip export controls were not a major topic of discussion at the bilateral meeting.
"We did not talk about chip export controls at the meeting," the Trade Representative told Bloomberg. That suggests any breakthrough on broader export restrictions remains distant.
The summit stabilized relations without transforming them. Markets had priced in more.
Nvidia Week
This is the most important earnings report of the quarter.
Nvidia reports Wednesday after the close. Analysts expect adjusted earnings of $1.78 per share on revenue of $79.2 billion. Any commentary from Jensen Huang about the China trip and potential H200 sales will dominate the discussion.
Despite the stock reaching new highs, some analysts sense muted enthusiasm. "We sense a marked apathy on this stock even among most big long-onlies," one noted. That setup, with expectations tempered despite elevated prices, could favor a positive reaction if results deliver.
Investors will also watch for commentary on competitive positioning. AMD and Broadcom remain the established alternatives. Cerebras just went public to significant fanfare. The competitive landscape is evolving.
When you directly compare Nvidia's valuation to some other chip names, Intel being one, there is a very reasonable argument that Nvidia is undervalued if the chip sector remains at these highs.
Target reports Wednesday alongside Nvidia. Walmart reports Thursday. Both retailers will provide insight into consumer behavior as inflation remains elevated and the K-shaped economy persists.
The Yield Problem
The 10-year Treasury yield crossing 4.5% matters as it gives us a yield problem.
This level has historically created gravitational pull on equity valuations. The math is straightforward: when risk-free returns reach 4.5%, the opportunity cost of holding equities increases. Capital on the margin flows toward bonds.
The driver is inflation. Core PPI for April came in at 1.0% month over month, well above the 0.3% consensus. Wholesale inflation running this hot typically filters through to consumer prices with a lag.
Fed funds futures now price a rate hike before year end as a likely possibility, at roughly 51% probability as of Friday afternoon.
"The financial markets expect interest rates to remain higher for longer, notwithstanding President Trump's demands that Kevin Warsh get rates down," one veteran strategist wrote. "But the macroeconomic backdrop no longer supports an easing bias, let alone a rate cut."
Warsh officially took over the Fed today. His first test arrives immediately.
The Commodity Supercycle Thesis
One of the most compelling arguments I am watching comes from the commodities side.
The thesis is straightforward and states that capital has chased the AI trade while ignoring the physical assets AI requires to run. Energy, metals, and compute capacity face bottlenecks that cannot be resolved quickly.
The Iran conflict has triggered the largest energy supply shock in history. The oil market has lost more than 13.7 million barrels per day of supply. Even after the war resolves, the playing field for the Persian Gulf has fundamentally changed.
The International Energy Agency cautioned last week that oil inventories globally are depleting at record pace. Inventories will near all-time lows of 7.6 billion barrels by end of May if demand remains constant.
One strategist called this "the most asymmetric trade in modern financial history." I would not go that far, but the supply-demand imbalance in physical commodities deserves more attention than it currently receives.
The hyperscalers are expected to spend over $700 billion on capital expenditure in 2026. That spending requires copper, electricity, concrete, and rare earths. The companies mining and producing those inputs have quietly become among the best performers of the decade.
Samsung Strike Risk
More than 41,000 Samsung Electronics workers are set to begin an 18-day walkout on Thursday.
If the strike proceeds, it could pull 3 to 4 percent of global DRAM supply offline. The Suwon District Court is expected to rule on Samsung's injunction request before May 20, which will determine whether the walkout proceeds on schedule.
Analysts have already cut Samsung's 2026 and 2027 operating profit estimates by 10% and 11% respectively on the strike risk.
This serves as a reminder that the memory supply chain carries concentrated risk. Micron, by contrast, faces none of the labor or geopolitical risk that hangs over South Korea's memory giants. Micron's HBM revenue share grew from 11% to 21% over the past year while Samsung's share fell from 35% to 22%.
For investors with heavy exposure to Korean equities or semiconductor ETFs, the strike outcome matters significantly.
The Two-Phase Framework
I want to share some thinking I have been developing on how to position for what comes next.
I believe we are in the early to middle innings of a multi-year AI infrastructure cycle that has a real chance of producing returns rivaling the late 1990s. The economic data, the earnings revisions, the capital expenditure curve, and the global breadth of the trade all support a constructive view over a three to five year horizon.
But there are two distinct phases to this story. A near-term phase where caution is warranted, and a 2027 to 2030 phase where the runway is wider than most investors appreciate.
The 1990s analogy is correct directionally but misleading on timing. The late 1990s tech run was not a clean monotonic move. It featured a 2.5x advance from 1995 to early 1998, followed by a 25% drawdown on LTCM and Russia, followed by the parabolic move from October 1998 to March 2000.
Investors who treated those years as one continuous upward move did fine. Investors who sized positions assuming volatility could not happen got carried out twice along the way.
The near-term risks are real. The 10-year yield approaching 4.5%. PPI accelerating. Oil above $100. The fiscal deficit running near 6% of GDP. And critically, 2026 is a midterm election year.
Midterm years have a historically distinct volatility signature. Since 1962, the S&P 500's average intra-year drawdown in midterm years runs roughly 17%, meaningfully larger than the 13% average for non-midterm years. The drawdowns tend to cluster in spring and summer.
The other side of that pattern is what makes it actionable. The 12 months following midterm elections have produced positive S&P 500 returns in roughly 85% of cycles since 1950, with an average gain in the high teens.
The right framing is not "the cycle is over." It is "the bumpy part is here, and the smoother part comes after."
How I Am Thinking About Positioning
Five points I would put in front of clients.
First, participate aggressively but size for volatility. The AI cycle has years to run. Going to cash would repeat the mistake investors have made for three years. But the multiples at the top of the cap-weighted indexes assume a benign rate environment that may not hold. Own the right exposures at sizes that survive a 15-20% drawdown without forcing emotional decisions.
Second, the midterm calendar is your friend if you plan for it. Clients with cash on the sidelines should be building shopping lists, not raising more cash. The drawdowns this calendar produces are statistical gifts to long-term capital.
Third, the picks and shovels with locked-in backlogs are where the asymmetry sits. Gas turbine manufacturers, grid equipment, select copper miners, the existing nuclear fleet bought on drawdowns, and the photonics complex as the next-leg bottleneck. These have earnings visibility through 2029-2030 that compounds regardless of the hyperscaler monetization timeline.
Fourth, inflation is the real near-term risk to multiples, not AI. A 10-year yield grinding toward 5% would re-rate every long-duration cash flow in the market. The hedges worth having protect against an inflation regime change, not an AI narrative break.
Fifth, the bigger opportunity is still ahead, not behind. The hyperscaler infrastructure build is the visible part of this cycle. The application layer, the companies using AI to compress costs and create new products, is where the next leg of returns will be created. Most of those companies are not yet public or are trading at sizes the market has not noticed.
Full piece here: https://open.substack.com/pub/dansheehan9/p/ai-has-years-to-run-but-the-setup?r=2n7yn9&utm_campaign=post-expanded-share&utm_medium=web
Final Thought
Markets are catching their breath after an extraordinary run. The S&P 500 has gained roughly 15% from the March lows. Semiconductors have nearly doubled. Records have been set across major indexes.
A pause here is not a reversal. It is the market processing gains before the next move.
Nvidia on Wednesday will likely set the tone for the weeks ahead. The retail earnings will reveal whether the consumer is bending under inflation pressure. The bond market will continue to exert its influence as yields test higher levels.
Stay invested. Size appropriately. Use volatility as opportunity rather than cause for panic.
The destination remains constructive. The path will not be a straight line.
Best regards,
Dan Sheehan [email protected]
Subscribe: https://substack.com/@dansheehan3
This newsletter is for informational purposes only and should not be considered investment advice. Market Pulse is an independent publication by Dan Sheehan and is not affiliated with, sponsored by, or associated with my employer. Please consult with your financial advisor about your specific situation.