A Market Priced for Perfection
The recent stretch on Wall Street looked like another winning week, capped by the best single trading day since April 8th. But that strength followed real volatility, and the cause of the preceding sell-off is instructive. The downward move began after Broadcom's results, which were genuinely positive on the numbers themselves — yet the market disliked the guidance. Specifically, investors balked at margin compression: as the business shifts from software toward semiconductors, margins fell from 77% to 74%. That reaction reveals something important about the current environment — the market is priced for perfection. Any deviation from a flawless outlook triggers a route. The recovery that followed was encouraging precisely because it signaled that the sell-off had been overblown.
There is broader optimism beneath the volatility. Earnings for the quarter have been impressive, running around 28% in the first quarter — well above historical norms. Commentary from Goldman Sachs leadership reinforces this: while the market may be elevated, it is not extremely elevated when measured against price-to-earnings ratios.
The Limits of Historical Valuation
A key argument here is that we habitually judge "elevated" against historical pasts — and that frame may be inadequate for what is unfolding. We do not yet know where this new technological frontier will take us. Monetization of new technology is always choppy and ugly in its early phases. But consider the precedents: the second industrial revolution, the arrival of rails and steam, the first wave of internet and e-commerce. Each rippled through every part of society. What is happening now is bigger than all of those. It is impossible to be too hyperbolic in saying this technology will change everything we do, because the very way everything is done is being changed. Because of that, we genuinely don't know where the appropriate valuation sits. Looking back five years from now, the likely conclusion is that we were merely at the infancy of monetizing this new frontier.
Two Live Concerns: Inflation and an AI Capex Pullback
Two specific worries temper the optimism. The first is a resurgence of inflation; the second is a potential pullback in AI capital expenditure.
On inflation, the data is troubling. CPI rose in May to its highest level in three years, and PPI posted its largest gain in three and a half years. Digging into PPI specifically: the core reading, which excludes energy, was still up 5.1% year-over-year, and up 8/10 of 1% month-over-month. That detail matters because it tells us inflation has moved beyond just the impact of crude and energy — it is showing up in the core, meaning it has spread into other parts of the economy. This raises real hope that the war can resolve itself as quickly as possible. But that resolution is unlikely to be quick, because the situation involves Iran and is therefore very complicated. Persistent inflation would be especially painful for the new Fed chief, who faces his first FOMC meeting next week. He may have expected to come in and be more accommodative, but the conditions simply are not there.
The New Financing Reality for AI
The capex concern centers on the hyperscalers and a striking shift in how AI investment is being financed. Last year, the hyperscalers wrote $121 billion of debt, against a previous average of roughly $30 billion. The realization sinking in is that the scale of required investment is so enormous that these companies must look beyond what they have normally done.
Google illustrates the new playbook. It frontloaded roughly $80 billion ahead of three AI-related IPOs that are coming, reasoning that an impending liquidity drain made it smart to be on the front end. The raise was oversubscribed — Google secured almost $85 billion. The takeaway is that AI capex will increasingly require new and different financing rounds beyond traditional structures.
The IPO Wave and the Liquidity Drain
This connects to a larger question: with so many offerings on deck — SpaceX, OpenAI, Anthropic, and Google's stock raise among them — is there enough capital to go around? Goldman Sachs President John Waldron has suggested this could be a record year for IPOs.
The marquee event is the largest IPO ever in history. If SpaceX comes in at a $1.77 trillion market cap, it would instantly become the seventh-largest public company in the world — right after IPOing. That scale makes it genuinely hard to earn ROI back. And it underscores a structural problem: you cannot have these massive IPOs without draining liquidity from current equities. Some of the selling already occurring in the market reflects rotation to make room for these new offerings.
Sticking With AI Winners — Because No One Knows the Winners Yet
Despite the capex concern, the recommended theme is to stick with AI winners. The reasoning is counterintuitive but deliberate: we actually don't know the winners and losers right now. The analogy is VHS versus Betamax — you couldn't know in advance which would prevail, so you had to buy both, accept that one would lose, and trust that the winner's gains would more than make up for the loser's decline. No one can pick the eventual victors at this stage.
Within that frame, a lot of ancillary companies are coming back into resurgence — the firms doing the compute, the infrastructure, the memory, and the actual technology to transfer data. Names like Micron and Dell, which are "really old," are benefiting. But this is also exactly where the capex concern bites: if those companies don't continue to receive AI data-center investment, they won't continue to deliver the results they are currently posting. The whole industry has been described as the fourth industrial revolution — the next chapter — and the overall mood is very optimistic.
Outlook: Volatility Now, Uptrend by Q4 — and a Possible Rate Hike
Will there be a dramatic pullback? The expectation is continued volatility through the quarter — absolutely — driven especially by the concentration of attention around the upcoming midterms and by the lack of a clear resolution on Iran, which will prolong the problem. The hope is for an Iran resolution past the midterms, with a positive uptrend arriving by Q4. But because the environment is volatile, investors should approach these positions with at least a five-year hold period, since it will take time to work through these valuations.
One notably contrarian call concerns interest rates. Given how hot the data has run — PPI as high as 6.5% year-over-year and CPI above 4%, a level not seen for a long time, since the Biden years — the market has swung. There is now only about a 30% chance of staying at current rates, and a 70% chance or greater of at least a 25-basis-point hike. This is a hike, not a cut, and it cuts against the more dovish expectations many held.
The Stock Picks
Reddit (RDDT) — AI licensing with a long runway. The attractive feature of Reddit, unlike Meta and Google, is that its AI-related average revenue per user is only $5.23. That low base implies a very long run rate, with the AI licensing business sitting on top of its core business as an additional layer. For that reason it offers a long runway and looks like a good buy.
Parker Hannifin — a non-pure-play space play. Offered with the full disclosure that this is non-scientific, the thesis rests on propulsion technology and thermodynamic technology — ancillary technology that would be used in many space launches. As SpaceX and other players ramp up their missions into space, these kinds of ancillary companies may get re-evaluated upward and have a long runway to become integral to the sector.
Procter & Gamble — a defensive value moat. This is a value play positioned around the technology boom rather than within it. P&G targets organic sales growth of about 3%. Its appeal is defensive: for consumer staples exposure, it offers 70 years of dividend raises and 136 years of paying dividends, with a current dividend yield of 2.9%. It is the kind of long-term company that pays yield and provides defensive value — a genuine moat at a time when everyone is talking about needing some kind of moat.
Expected Returns on Those Picks
When asked what percentage gains to expect, the answer differs by pick. For the first two — Reddit and Parker Hannifin — the hope is for re-evaluations and a real upward trend on the AI and data-center side. P&G is different: at 3% organic sales growth for a company hundreds of years old, the expectation is modest. It comes down to what kind of risk and what kind of return — what ROI — is being sought in each case. For P&G, the return expectation is "not so much."
Gold: Fundamentals Still Supportive
Gold has pulled back this week, down about 3%. But nothing has changed in the underlying fundamental global macro environment — in fact, it has gotten worse with the Iranian situation. The fundamentals remain supportive. Most analysts expect gold to end in the range of 5,300 to roughly 6,300, with only two bear cases keeping it in the 4,000s. The pullback is therefore viewed as a great dip to buy.