Why AI Hardware Fell While Software Rallied?
The AI hardware trade suddenly hit the brakes
Coming into Monday, AI hardware had momentum. The $PHLX Semiconductor Index (.SOX.US)$ had surged 9.3% the previous week, with optical networking among the hottest parts of the AI infrastructure trade. Then the tape flipped.
By Monday's close, $iShares Semiconductor ETF (SOXX.US)$ was down 2.6%, $NVIDIA(NVDA)$
The underlying AI demand story did not suddenly disappear. What changed was how investors viewed the financing behind that demand.
Nvidia reopened the "who pays for all this?" debate
$NVIDIA (NVDA.US)$ announced partnerships with Apollo, $Blackrock (BLK.US)$ , $Blackstone (BX.US)$ , $Brookfield (BN.US)$ , $Goldman Sachs (GS.US)$ and KKR to create financing platforms designed to mobilize more than $500 billion for AI infrastructure. Importantly, the headline figure refers to third-party capital, so Nvidia itself is not writing a $500 billion check. The agreements are also still memorandums of understanding.
That distinction did little to calm the market.
The announcement reminded investors that AI infrastructure is becoming enormously capital intensive. If easier financing helps customers build data centers and buy more Nvidia systems, it can support demand. It also raises a harder question: how much AI demand can ultimately be funded by operating cash flow, and how much increasingly depends on debt, guarantees and private capital?
Credit markets had already been flashing this concern. Nvidia's five-year CDS surged. Monday's announcement landed in a market that was already highly sensitive to circular-financing risk.
Intel and the Middle East added fuel to the selloff
$Intel (INTC.US)$ delivered another uncomfortable financing headline, announcing a $15 billion common-stock offering, plus an option for underwriters to purchase another $2.25 billion. Intel said the proceeds could support capex and working capital. Its shares finished about 4% lower.
At the same time, hopes for a quick reopening of the Strait of Hormuz faded as negotiations between the U.S. and Iran became more complicated. WTI jumped about 5% to $82.13 and Brent rose 5% to $87.72. Higher oil prices revived inflation concerns and added another macro headwind for long-duration technology assets.
Put together, investors suddenly had three reasons to trim the most crowded AI hardware trades: financing risk, dilution risk and macro risk.
Money rushed back into software
The other side of Monday's trade was striking. The $iShares Expanded Tech-Software Sector ETF (IGV.US)$ gained 2.3%, even as $iShares Semiconductor ETF (SOXX.US)$ lost 2.6%. $Datadog (DDOG.US)$ rallied 11.4%.
Datadog is especially useful for understanding the move. Just days earlier, the stock had plunged about 19% after earnings, after management warned that revenue growth could slow in the third quarter. There was no similarly large new company-specific fundamental catalyst on Monday that suddenly erased those concerns.
That makes the rebound look heavily influenced by positioning and sector rotation. Investors were reducing exposure to capital-intensive AI infrastructure and moving money toward software, where business models generally require less physical capital.
One valuation nuance matters here. Software is still expensive in absolute terms, with IGV trading around 36x P/E as of August 7. But it would be misleading to describe the whole software sector as sitting at historical peak valuations or the entire hardware sector at historical lows. Forward valuation data vary significantly by metric and company. Parts of AI hardware, particularly names with rapidly rising earnings estimates, can look much cheaper on forward earnings than their recent price performance suggests.
Summary
Monday was less about a sudden change in AI fundamentals and more about a change in what investors were willing to pay for.
Last week, huge AI infrastructure spending looked like proof that hardware demand remained unstoppable. On Monday, the same spending raised questions about leverage, financing structures and future returns.
Software became the natural destination for that rotation.
So the simplest explanation for hardware down and software up is this: the market temporarily reduced exposure to the most capital-intensive part of the AI trade and rotated toward asset-light technology stocks that had recently lagged.
The AI story did not disappear. The market simply changed which part of the story it wanted to own.
@TigerStars @CaptainTiger @TigerWire @Daily_Discussion @Tiger_chat @Tiger_comments @MillionaireTiger
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
