Nvidia closed at $190.01 today on 145.2 million shares — nearly twice normal volume — suggesting something more than routine profit-taking. Meanwhile Alphabet closed at $336.71, up 0.90%, and its 16.75x TTM P/E now looks like the most compelling valuation in the entire Mag 7. The AI infrastructure trade and the AI monetization trade are beginning to diverge in ways that matter for positioning.
Let's start with the headline: Nvidia dropped 3.55% today, closing at $190.01 after opening at $195.85 and spending the entire session in retreat. Volume was 145.2 million shares — that's not a casual flush, that's institutional repositioning. The intraday range ($190.01 to $197.07) tells you sellers had the wheel all day, with the stock closing at its session low. YTD, NVDA is up just 0.74% and up 6.13% over the past 52 weeks — remarkably pedestrian for a company that just reported 78.3% annual revenue growth and a 47% trailing free cash flow margin. When price performance disconnects from fundamental momentum this sharply, the market is telling you something about expectations and valuation — not about the business.
On the fundamental side, NVDA's setup is genuinely complex. The 30.2x TTM P/E and 24.4x EV/EBITDA are not cheap in absolute terms, but relative to 78% revenue growth and nearly half of revenue converting to free cash flow, the multiple is defensible — arguably cheap on a PEG basis. What the market is wrestling with is duration risk: how long can Nvidia sustain hyperscaler capex at this level? The news that Nvidia is reportedly in talks to finance OpenAI's Ohio data center is a fascinating wrinkle — it signals Nvidia is willing to go beyond hardware supplier and become a capital partner in AI infrastructure, which either deepens the moat or adds balance sheet complexity depending on your lens. I'm watching this development closely. Revenue TTM of $253.5 billion at an 18.2x price-to-sales is not a value stock, but it's also not irrational for a company at the center of the most significant capital allocation cycle in enterprise tech history.
Now flip to Alphabet. GOOGL closed at $336.71 today, up 0.90% on 26.9 million shares — quiet, constructive, and frankly one of the better risk/reward setups I can identify in the Mag 7 right now. The 16.75x TTM P/E on $445.9 billion in revenue with a 54.77% profit margin and 34.03% operating margin is extraordinary. Let me say that again: Alphabet earns a profit margin north of 54% and trades at under 17x earnings. The 12.1x EV/EBITDA is cheaper than META, cheaper than AMZN, and dramatically cheaper than AAPL or NVDA. The 52-week return of +71.82% tells you the market has been repricing this — but the current valuation suggests the repricing isn't finished. For a company with dominant positions in Search, Cloud, and YouTube AND a credible frontier AI posture through Gemini, this multiple still prices in a lot of fear.
The narrative shift I'm tracking here is the bifurcation between AI infrastructure plays (NVDA, to some degree MSFT) and AI monetization plays (GOOGL, META). Nvidia's 3.55% drop today on massive volume while Alphabet gained 0.90% on a calm session is emblematic of this rotation. The market appears to be asking: who actually captures the economic surplus from AI deployment? Nvidia captures it at the hardware layer, but at 18.2x revenue, the market has been asking hard questions about competitive moats as custom silicon from Google TPUs, Amazon Trainium, and Microsoft's Maia mature. Alphabet, by contrast, is already monetizing AI through Search, Cloud AI revenue, and increasingly through Gemini-native products — and is doing it at a valuation that prices in meaningful competitive disruption risk that may be overstated.
My cross-Mag-7 read on today's action: the highest-conviction setup in the group remains GOOGL at 16.75x earnings with 54%+ profit margins and $445.9B in TTM revenue. NVDA needs to hold $190 technically and clarify the OpenAI financing story to rebuild momentum. AAPL at $338.19 and 41.1x P/E remains the most demanding multiple in the group. MSFT at 23.4x with its YTD decline of 17.06% is where I'd be building a position thesis if Azure reacceleration materializes. TSLA at 286.8x TTM P/E with 1.41% operating margins is a separate conversation entirely. The AI trade isn't over — it's just becoming more discriminating about who wins at each layer of the stack.