LearnBits

$METAB
One company is spending $145 billion this year to build AI infrastructure. The other made $120 billion in earnings off the exact same infrastructure boom. That single sentence captures why Meta (META) and Nvidia (NVDA) sit on opposite ends of the same AI trade, and why choosing between them right now is less about which business is better and more about which position in the food chain you want to own.
$NVDAB
Nvidia's numbers are almost absurd on paper. Fiscal year 2026 revenue hit $215.94 billion, up 65% from the prior year. Non-GAAP gross margins settled at 75%. The company authorized $80 billion in additional buybacks and raised its dividend. Wall Street's average price target sits near $298, implying roughly 49% upside from current levels around $200. The stock is up about 19% year to date, which sounds good until you realize the Philadelphia Semiconductor Index itself is up 94% over the same stretch, meaning Nvidia has actually underperformed its own sector. Michael Burry disclosed a short position against it last month. That's the uncomfortable footnote behind an otherwise dominant earnings story.
Meta's 2026 has gone the other way. Shares are down about 9% year to date despite a Q1 that showed 33% revenue growth, 3.56 billion daily active users across its apps, and ad impressions up 19%. The drag is almost entirely sentiment-driven, centered on whether $145 billion in annual capex can ever generate returns visible enough to satisfy investors who'd rather see buybacks than data centers. Reality Labs bled another $4 billion in operating losses, which doesn't help. But the core advertising engine that funds all of it keeps compounding quietly, and WhatsApp monetization plus AI glasses sales are starting to look like real incremental revenue rather than long-term roadmap promises.
The valuation gap between the two tells you where risk is currently priced. Nvidia carries a beta of 2.20, meaning it swings hard in both directions with the broader market. Meta's beta is 1.24, closer to a large-cap consumer staple than a hypergrowth tech stock, and it trades at what several analysts describe as a discount relative to its cash flow when heavy capex years are stripped out of the comparison. One is a high-conviction, high-volatility play on AI infrastructure spend continuing to accelerate. The other is a bet that the world's largest advertising business can absorb a historic capital expenditure cycle without breaking its own economics. Both theses are defensible. They just require very different kinds of patience.

Big Tech’s relentless spending on artificial intelligence infrastructure is starting to reshape the profit picture in ways that worry even the most bullish analysts at Goldman Sachs. Hyperscalers such as Microsoft, Amazon, Alphabet and Meta are pouring hundreds of billions into data centers, chips and related gear, with projections nearing $770 billion in capital expenditures across the sector for 2026 alone, roughly matching their entire operating cash flow.
$AMZN
This capital intensity marks a sharp departure from the asset-light model that powered these companies for years. Higher depreciation charges and declining sales-to-asset turnover are already pressuring return on equity at names like Microsoft and Amazon. Goldman points out that while current earnings look strong, the bill for all this buildout could weigh on future margins and shareholder returns if monetization lags behind the spending.
$MSFT
At the same time, investor expectations have raced ahead. AI-exposed stocks such as Nvidia, Broadcom and the major cloud providers now trade at premium valuations, and the S&P 500’s performance has become dangerously concentrated in this small group. Goldman sees signs of market exuberance but stops short of calling it a full bubble. Still, any slowdown in AI investment or disappointing revenue growth from these massive outlays could trigger a painful repricing.
$NVDA
The caution comes as the AI trade shifts from hype to heavy lifting. For companies like these hyperscalers and their suppliers, the real test will be whether the enormous upfront costs translate into sustainable profits before the momentum fades.

Investors pulled more than a billion dollars from Bitcoin exchange-traded funds in a single week recently, marking the largest weekly exit of 2026. At the same moment, capital flowed steadily into XRP products and surged into brand-new funds tracking Hyperliquid’s $HYPE token.
This isn’t a broad retreat from crypto. It’s a rotation. After years of Bitcoin dominating institutional portfolios, allocators appear hungry for assets that offer real activity and revenue. Hyperliquid, a decentralized perpetuals exchange, has captured attention with its high trading volumes and token buyback mechanics that funnel fees back to holders. Its ETFs have already drawn over $150 million in net inflows shortly after launch, outpacing early Bitcoin products on a size-adjusted basis.
$XRP funds, meanwhile, continue their steady climb. They added nearly $23 million in one recent week, extending a streak of positive flows even as the broader market wavered. Cumulative inflows for XRP ETFs now exceed $1.4 billion since their late 2025 debut, reflecting persistent belief in Ripple’s utility for cross-border payments.
The shift highlights a maturing market. Big players who once piled into Bitcoin as the safe entry point are now testing higher-upside names tied to actual usage. Whether this rotation sparks broader altcoin strength or remains a niche move will shape the next leg of crypto’s cycle. For now, the flows tell a clear story: diversification is back in fashion.