The ongoing AI boom has set off a surge in investments from mega-cap tech companies, particularly in high-powered GPU chips, but the financial implications of these expenditures are starting to surface in ways that could dampen the enthusiasm. While the focus has largely been on the growth potential of AI, an overlooked factor—depreciation—threatens to erode profitability and valuation projections in the coming years.
Barclays analysts recently highlighted a significant yet underappreciated challenge: the depreciation of AI chip investments. As companies like Alphabet (GOOGL), Amazon (AMZN), and Meta Platforms (META) pour billions into these essential chips, the long-term impact of depreciation on earnings is becoming more evident. Depreciation, an accounting method that spreads the cost of capital investments over their useful life, is a critical but often understated expense. With AI chips experiencing rapid innovation cycles, their useful life is shorter than expected, leading to accelerated depreciation costs that could weigh heavily on earnings.
Barclays is already adjusting its earnings forecasts for major cloud hyperscalers, cutting estimates by up to 10% for 2025. This recalibration stems from the anticipated depreciation costs of GPU chips, which are expected to be significantly higher than Wall Street’s current projections. For instance, Alphabet’s depreciation expenses are expected to hit $28 billion in 2026, a stark increase from the consensus estimate of $22.6 billion. Meta Platforms faces an even larger discrepancy, with Barclays forecasting $30.8 billion in depreciation costs compared to the Street’s estimate of $21 billion.
These figures highlight a potential overvaluation of AI-driven tech stocks. “GOOGL, META, and AMZN shares may be 5% to 25% more expensive than perceived, due to miscalculations in depreciation modeling,” warned Ross Sandler, Barclays’ internet analyst. The issue is further exacerbated by Nvidia’s aggressive product launch schedule, which introduces new chips annually, shortening the lifespan and increasing the depreciation burden on companies that rely on these GPUs.
As a response, some companies have extended the useful life of their server assets from five to six years or more, an attempt to soften the impact of depreciation on their financials. However, this strategy has limitations. The rapid cycle of GPU innovation, driven by companies like Nvidia, means that further extensions in asset life are unlikely, forcing tech giants to absorb higher depreciation expenses in the coming years.
The implications of these financial adjustments are profound. Barclays’ recalibration of earnings estimates suggests that the AI-fueled stock rallies of companies like Alphabet, Amazon, and Meta may be overestimated. This could lead to downward pressure on their stock prices as the market adjusts to the reality of higher costs and lower-than-expected returns on invested AI capital.
Ted Mortonson, managing director and tech strategist at Baird, echoed these concerns, noting that the depreciation issue is significant enough to impact valuations and potentially push AI stocks lower. “This is a headwind that is big enough to drag down stock prices over the next year,” Mortonson said. He emphasized the importance of ROI on AI investments, which remains uncertain amid escalating costs and the complex accounting adjustments required to manage them.
The broader question for Wall Street now revolves around the return on these massive AI investments. With over $200 billion already spent and capital expenditures up by more than 50%, investors are increasingly asking where the payoff is. Mortonson suggested that a clear return on invested AI capital may not be realized until 2025 or 2026, leaving the market in a precarious position as it grapples with the current lack of transparency and the looming financial impact of depreciation.
In conclusion, while AI continues to drive innovation and growth in the tech sector, the financial realities of depreciation and the accelerating GPU cycle present a significant risk to the profitability and valuation of leading tech companies. As Wall Street begins to factor in these costs, the exuberance surrounding AI stocks may temper, leading to a more cautious and calculated approach to investing in this rapidly evolving space.
