Is Meta's $65 Billion Spending Spree a Good Idea? Here's What History Suggests.

Core Viewpoint - Businesses across various sectors are increasingly investing in artificial intelligence (AI), particularly in graphics processing units for training generative AI programs. Despite expectations of a peak in investments, significant spending continues, especially from major players like Meta Platforms [1][2]. Group 1: Meta's Capital Expenditure Plans - Meta Platforms announced a capital expenditure (capex) plan that could reach $65 billion in 2025, reflecting over 60% growth from the previous year, primarily focused on AI infrastructure investments [2][4]. - The spending will be allocated to three main categories: servers, data centers, and networking equipment, with a focus on building data centers equipped with large training clusters [5][4]. - Meta plans to enhance the adoption of its custom Meta Training and Inference Accelerator chips, developed in collaboration with Broadcom [5]. Group 2: Historical Context of Spending - The last significant spending spree by Meta was related to its investments in the metaverse, which began aggressively in 2022, leading to a substantial increase in capex [8][9]. - This previous investment resulted in a notable decline in profitability, with earnings per share falling 38% in 2022, as advertising revenue also decreased by 1% [12][9]. Group 3: Potential Differences This Time - The economic environment in 2022 was challenging for tech companies, with high inflation and rising interest rates impacting overall business spending [11]. - Unlike the metaverse investments, Meta's current infrastructure spending is designed to extend the useful life of its servers, potentially leading to better returns on investment [13]. - The effectiveness of the capex spending will be clearer this year, with a focus on monitoring earnings announcements and management commentary regarding the AI push [14].