GOOD MORNING, The S&P 500 ended Friday just below a record, but the more revealing move was inside AI. Applied Materials delivered an upbeat forecast and still fell 5.1%, dragging other chip names lower as investors showed how high the bar has become for companies tied to the data-center buildout. Then came another reminder over the weekend: Anthropic's prospective IPO valuation is being built around a forecast of roughly $190 billion to $200 billion of revenue in 2028, more than four times its current run rate. AI growth remains extraordinary. So do the assumptions investors are already paying for. MARKETS | TLDR
EXPECTATION RESETApplied Materials Beat. The Stock Still Fell 5%.Applied Materials gave investors the kind of report that normally supports a semiconductor stock: demand tied to AI data-center construction remained strong and the company offered an upbeat quarterly forecast. Shares still fell 5.1% Friday. Broadcom dropped 5.9% and Intel lost 2% alongside it, pulling the chip complex lower even as the broader market remained close to record highs. The reaction matters because it shows how the AI trade is changing. When a stock has already doubled in a year and investors broadly agree that AI infrastructure demand is strong, a beat stops being new information. The market starts asking whether growth is accelerating faster than the expectations already built into the price. In that environment, "beat and raise" can become maintenance rather than upside. GLOBALT Investments' Thomas Martin described the setup directly: Applied Materials had a beat-and-raise quarter, but expectations were simply too high. That is a useful lens for the rest of the AI stack. Nvidia reports later this month with the same problem at much larger scale. The operating story can remain excellent while the stock reaction turns negative because the market is no longer comparing results with last year; it is comparing them with an increasingly ambitious future. FORWARD VALUATIONAnthropic's IPO Is Being Priced on 2028Anthropic is projecting roughly $190 billion to $200 billion of revenue in 2028 as bankers and investors prepare for a potential IPO, according to Reuters. That is more than four times the roughly $47 billion revenue run rate the company disclosed in May. Investors are reportedly using forward enterprise-value-to-revenue multiples based on those 2028 estimates rather than relying primarily on current earnings. That is unusual enough to be the story. High-growth software companies are often valued on revenue when profits are immature, but looking two years forward reflects how difficult it has become to value an AI company whose current economics are dominated by spending on GPUs, model training, inference and talent. The valuation case assumes revenue will grow much faster than those costs, allowing margins to expand as the business scales. Anthropic's trajectory gives investors some evidence for that bet: its revenue run rate rose from about $9 billion at the end of 2025 to more than $47 billion by May, and it expects its first quarterly operating profit in the second quarter of 2026. But the gap between current performance and 2028 expectations is still enormous. The broader signal is the same one Applied Materials sent Friday from the public market: AI valuations are increasingly about how much future growth has already been pulled into today's price. HEADLINES
UPCOMING
DEEP INSIGTHSAnthropic's IPO Valuation Hinges on $190-$200 Billion of 2028 RevenueRead this for a concrete look at how investors are valuing fast-growing AI companies when current earnings do not yet capture the economics they expect at scale. The use of 2028 revenue multiples shows how much future growth is being pulled forward into present valuations — and how sensitive those valuations could become if the growth curve changes. S&P 500 Ends Lower as Investors Weigh Data and Middle East TensionsThe useful detail here is not the index's 0.17% decline. It is the contrast between record-level equities, weaker retail and sentiment data, an oil shock and a 5.1% selloff in an AI supplier after a strong forecast. Together they show a market that still has earnings support but is becoming increasingly selective about what counts as genuinely positive news. |