英伟达财报前，花旗高喊：买入！AI需求远超供应，泡沫之说站不住脚

Core Viewpoint - Citigroup has maintained a "buy" rating on Nvidia and raised its target price from $210 to $220, indicating strong confidence in the company's upcoming earnings report expected on November 19, which is anticipated to exceed revenue expectations and provide an upward guidance [1][2]. Financial Performance Expectations - Citigroup predicts Nvidia's sales for the quarter ending in October will reach $57 billion, surpassing the market average expectation of approximately $55 billion [2]. - For the January quarter, Citigroup expects Nvidia's sales guidance to be $62 billion, also above the market expectation of around $61 billion [2]. - The optimistic forecast is supported by strong shipment momentum of Nvidia's Blackwell architecture GPUs, with 6 million GPUs shipped as reported during the GTC Washington conference [2]. AI Market Dynamics - Citigroup counters the narrative of an "AI bubble," asserting that the primary issue in the AI chip market is supply constraints rather than lack of demand [4]. - The report highlights that due to limitations in CoWoS capacity, AI supply will remain below demand until at least 2026, with a potential catch-up only around 2027 [4]. - Large cloud service providers are expected to see a turning point in cloud revenue starting in 2025, driven by the adoption of enterprise-level AI applications [4]. Target Price and Earnings Forecast - Based on strong growth prospects, Citigroup has raised its target price for Nvidia to $220, based on a projected EPS of $7.24 for the calendar year 2026 and applying a 30x price-to-earnings ratio [5]. - Citigroup has also increased its EPS forecasts for Nvidia for the fiscal years 2026, 2027, and 2028 by 2%, 7%, and 8% respectively, aligning with revised global AI capital expenditure models [5]. AI Semiconductor Market Outlook - Citigroup's bullish outlook extends beyond Nvidia to the entire AI semiconductor market, predicting that the global data center semiconductor market will reach $654 billion by 2028, a 16% increase from the previous estimate of $563 billion [7]. - The upward revision is attributed to higher-than-expected demand from key AI players like OpenAI [7].