Mega-Cap AI Rally Hits Pause: SPY, QQQ Pull Back as Breadth Rotation Looms, NVDA Earnings Next

The mega-cap AI rally that powered the S&P 500 and Nasdaq to new highs for six to seven consecutive weeks came to an abrupt halt on May 15. What had been a classic 'melt-up' in heavily concentrated holdings (top 10 stocks now represent roughly 38% of SPY by some estimates) broke down sharply as bond yields spiked and inflationThe rate at which prices rise across an economy. fears re-emerged. NVDA, which had climbed 20% since May 5 alone, now faces the high bar of earnings season with the stock having added approximately $1 trillion in market cap in a matter of days. This level of anticipation creates headline risk: any sign that capex cycles are moderating, China demand is weaker, or supply constraints are easing would trigger a selloff.

Social mentions and market chatter underscored the melt-up nature of the prior week. Traders noted: 'Another life changing week for many. SPX and QQQ continue to print new highs the past 6-7 weeks. If the gap fills above for SPX QQQ another melt up rally incoming. NVDA to 250 in play after earnings next week.' That comment captures both the euphoria and the binary risk: if gaps fill (technical breakdown), the rally stalls; if they don't, 250 is theoretically possible. However, the pullback on May 15 (SPY fell, QQQ fell, NVDA down 2-3%, TSLA down 3.5%) suggests that the easy money has been made and profit-taking is setting in. The fact that even 'safe' mega caps like AAPL bled alongside risk assets indicates that the rally's foundation (valuation expansion, multiple re-rating) has eroded.

Breadth metrics are critical here. One trader observed: '10 green plays, 1 red all week. 10 tickers traded: ES SPY TSLA AMD AMZN GOOG OKLO NVDA AAPL MU. Market just closed Friday like a trap door opened. Everything red. No exceptions.' This points to a latent breadth problem: most of the gains were concentrated in a handful of mega-cap AI plays, while the broader market (Russell 2000, mid-caps, value) lagged significantly. A rotation into breadth names could dampen mega-cap outperformance and lower the S&P 500's year-to-date return despite new index highs. Ackman's Pershing Square increased MSFT by 5.65M shares and AMZN by 1.84M in Q1, signaling conviction in mega-cap infrastructure plays, but the timing of that filing (May 15) coincides with the selloff, suggesting his conviction may face near-term testing.

NVDA earnings on May 21 are the key catalyst. If the company guides to continued robust demand, China risk is overblown, and capex cycles remain on track, the stock could rebound sharply and reignite the rally. Conversely, if management signals caution on China, or if CapEx growth is moderating, the stock could break below the $100 level and trigger a broader tech unwind. The stakes are high because NVDA is the psychological leader of the mega-cap AI trade; a miss would force reassessment of valuations across the entire AI/semiconductor complex.