Over the past several weeks, investors in artificial intelligence heavyweights Nvidia (NVDA) and Broadcom (AVGO) have experienced notable market turbulence. Both stocks have pulled back sharply from their recent highs, with Nvidia declining approximately 18.5% and Broadcom falling 24.4%. Throughout the AI investment wave, these two leading stocks have historically been viewed as attractive buying opportunities when correcting roughly 20% from their peaks.
The current market appears to be in a pessimistic cycle of waning AI enthusiasm, with investors growing increasingly concerned about hyperscalers’ massive capital expenditures poured into AI data centers. However, such anxiety may overlook a critical point: numerous AI giants have repeatedly emphasized that the risk of underinvesting in AI infrastructure far outweighs the risk of overinvesting. Spending is expected to maintain its growth trajectory through 2027. Broadcom has provided guidance projecting its AI semiconductor business will generate over $100 billion in revenue by 2027. Meanwhile, Wall Street expects Broadcom’s total revenue to reach $172 billion next year, up 62% from this year. Nvidia has also issued similarly optimistic forecasts, suggesting that AI hyperscaler capital expenditures could surpass $1 trillion next year. Given that many customers of these two companies are advancing according to multi-year construction plans, they possess relatively high visibility into development trends over the coming years.
Nvidia is undergoing its most dramatic valuation reset since the dawn of the AI boom, with its forward price-to-earnings ratio compressed to approximately 18 times, even below the average levels of the S&P 500 and Nasdaq 100. Yet this valuation contraction does not stem from deteriorating earnings prospects; analysts have actually been raising their earnings estimates for future quarters. The real implication of the sell-off lies in the shifting center of gravity of the AI trade: market capital is beginning to flow from Nvidia into the broader AI industry chain, such as memory chips. Nvidia’s periodic stock price weakness has instead become a “source” of funding for other AI investment themes in the market. Behind this dramatic rotation is the structural trend of AI applications diffusing from cloud-based training to edge-side inference.
Nvidia’s competitive landscape is becoming more complex, with major customers like Alphabet and Amazon accelerating the deployment of custom-designed chips, while AMD and Intel have also captured greater spending share in the CPU market. Despite this, Nvidia’s share of the server GPU market remains virtually unaffected, with its H100 and next-generation B200 GPUs still the dominant choice for AI data center workloads. Recently, rumors of product delays and geopolitical headwinds have exerted additional pressure on the stock, but the options market has flashed noteworthy signals: call option volume is significantly higher than put option volume, suggesting that some funds are betting on strong support near the $200 price level.
Given the rapid growth of both Broadcom and Nvidia, forward price-to-earnings ratios are the most appropriate valuation metric. Based on earnings expectations for fiscal 2027, both stocks appear attractively priced. Data shows that market expectations for Nvidia’s future profits have been revised upward by 13% over the past three months. Despite the stock price pressure, Wall Street’s bullish consensus remains strong, with multiple institutions reiterating “buy” ratings and arguing that, measured against Nvidia’s growth rate, the current valuation is highly compelling. Analysts generally believe that although the AI narrative has shifted from “sole” to “diverse,” Nvidia and Broadcom, by virtue of their core positions in AI infrastructure, remain the best investment partners for capturing the massive capital expenditure dividends from AI hyperscalers.
