Google Adds $1.5T in Market Cap in 6 Weeks; Mega-Caps Extend Run on AI and Earnings Beats

The numbers are staggering. Google added close to $1.5 trillion in market cap in six weeks. At $4.9 trillion, Google's valuation now exceeds that of all but three countries on Earth. This is not hyperbole; it is the scale of the asset concentration happening in real time. The S&P 500's top 10 stocks represent 38% of the index, a concentration level that has raised red flags among fund managers who worry about liquidity and downside tail risk if sentiment reverses. Yet the momentumThe empirical fact that winners keep winning over the medium term. persists because earnings are delivering.

Google's specific strength is rooted in a breakthrough that management framed on recent calls: the company has found a way to cut AI memory use by 6x, an efficiency gain that engineers describe as 'fitting a warehouse into a backpack.' If true, this solves one of the core constraints on AI scaling. Memory is now the binding variable in AI capex cycles (as CFOs of MSFT, AMZN, META, and AAPL all noted), so a 6x reduction in memory requirements per inference is transformative. For Google, this means either that the company can serve AI workloads at a margin Google hadn't anticipated, or it can depreciate its own internal AI infrastructure spend while maintaining performance, freeing capital for shareholder returns.

Microsoft is facing a different narrative challenge: the company's AI capex cycle is being misunderstood, and margins are being compressed by that capex spending upfront. The consensus is that Microsoft will eventually monetize AI demand through revenue growth, but the 'how quickly' is the real question. Most of the company's AI investments have yet to show up in earnings, a fact that could trigger a repricing if near-term margin pressure surprises to the downside.

The broader market dynamics are sustaining the rally: fund managers who underweighted mega-caps earlier in the year are now chasing the performance gap. Global asset allocators are recognizing that US tech leadership in AI is not priced into non-US equities, driving a rotation into dollar-denominated mega-cap tech. However, concentration risks are real. If any single mega-cap (MSFT, NVDA, GOOGL, AAPL) stumbles on earnings or AI monetization, the repricing could be swift. Tactical positioning suggests that the rally is vulnerable to profit-taking after the Trump-Xi summit concludes, but structural tailwinds from AI adoption remain intact.