Nasdaq Breadth Deteriorates as AI Mega-Caps Dominate; S&P 500 Concentration Hits Record

The AI rally that has driven US equities to record highs is exhibiting classic signs of late-cycle concentration risk. The top 10 S&P 500 names, dominated by NVDA, MSFT, META, GOOGL, AAPL, AVGO, and a handful of AI infrastructure plays, now comprise a larger share of the index than at any time since the dot-com bubble. Active fund managers, despite having a brief window in early 2025 when it seemed they might finally outperform passive strategies, are once again confronted with the reality that a tiny cohort of mega-caps is driving all returns.

Data from multiple sources shows that Russell 2000 valuations are compressing relative to large-cap peers, while breadth (the percentage of stocks hitting 52-week highs) is rolling over despite the S&P 500 trading at all-time highs. This divergence is a classic warning signal. Retail flows, evidenced by mentions of Trump buying the dip on stocks like AMZN and HOOD, suggest retail investors are being drawn back in precisely when concentration risk is highest. One trader noted sardonically that 'just saw a guy try to convince his girl to full port into NVDA he literally took her phone and did it on the spot, which part of the market cycle is this?'

NVDA's valuation has become unmoored from traditional metrics; the stock has risen 20% in seven days, reaching a $5.5 trillion market cap. A $100,000 investment in NVDA in early 2023 would be worth roughly $1.5 million today, a 15x return that has few historical precedents outside of bubble conditions. Market technicians note that the stock is overbought on every major oscillator (RSIRelative Strength Index - momentum oscillator on a 0-100 scale. above 70, Bollinger Bands stretched), yet momentumThe empirical fact that winners keep winning over the medium term. remains so strong that traditional sell signals are being ignored.

The risk is that any catalyst, earnings disappointment, macro shock, or simply a profit-taking wave, could trigger a violent repricing. Taiwan stocks are being recommended by JPMorgan as 'the most pure-play exposure to global AI buildout,' suggesting that even consensus bullish strategists are hedging by recommending offshore proxies. The concentration narrative could unwind quickly if traders begin rotating out of mega-caps and into 'orphan' stocks that have been starved of capital.