In the rapidly evolving landscape of artificial intelligence, the current investment frenzy is reminiscent of past economic bubbles that ended in financial turmoil. As AI continues to integrate into daily workflows, the excitement around its potential has led to massive capital allocations. However, this surge in investment raises concerns about sustainability and profitability, echoing historical financial cycles.
Harris Kupperman, a seasoned investor and founder of Praetorian Capital Management, warns that the AI investment boom may end badly. Drawing parallels with past bubbles, Kupperman emphasizes the importance of cash flow and return on capital, which he believes are being overlooked in the current AI craze.
The Financial Dynamics of AI Investment
According to industry insiders, total datacenter spending for 2025 is projected to reach approximately $400 billion. This figure, however, could fluctuate due to construction delays or increased urgency to operationalize these centers. Datacenters comprise three main components: the building and land, power systems and infrastructure, and GPUs, each accounting for a significant portion of costs.
Depreciation of these assets is a critical factor, with AI datacenters expected to incur $40 billion in annual depreciation while generating only $15 to $20 billion in revenue. This stark contrast highlights a fundamental issue: the depreciation costs are double the revenue.
“The AI datacenters to be built in 2025 will suffer $40 billion of annual depreciation, while generating somewhere between $15 and $20 billion of revenue.”
Challenges in Achieving Profitability
The AI industry faces a daunting challenge in achieving profitability, as there is currently no gross margin in the sector. Companies are essentially giving away technology, hoping to drive adoption and eventually flip to positive returns. Kupperman speculates that margins might eventually reach 25%, but this remains uncertain.
To cover depreciation costs, the industry would need $160 billion in revenue at a 25% gross margin. However, current revenue levels are far from this target, necessitating a ten-fold increase. Achieving such growth is a formidable task, especially when compared to established companies like Netflix and Microsoft, which have struggled to expand beyond their market saturation points.
“You need revenue to grow roughly ten-fold, just to cover the depreciation.”
Historical Parallels and Future Implications
The AI investment boom draws parallels with past economic bubbles, such as the Dot Com bubble of the late 1990s and the shale oil boom of the 2010s. In both cases, companies overestimated the market’s willingness to pay for new technologies, leading to financial collapse.
During the Dot Com era, companies like Global Crossing invested heavily in fiber optics, only to face bankruptcy when demand fell short. Similarly, the shale boom saw companies pouring cash into oil production, only to encounter financial difficulties as returns dwindled.
These historical examples serve as cautionary tales for the current AI investment cycle. The risk of overbuilding without sufficient revenue is a real concern, and the potential for negative returns looms large.
“Fiber was the datacenter of that cycle, and Corning was the NVIDIA of its day (it lost 97% of its share price in the two years after it peaked).”
Looking Ahead: The Future of AI Investment
The future of AI investment remains uncertain. While the technology holds transformative potential, the financial dynamics pose significant challenges. Companies may be forced to choose between continuing the arms race of investment or writing off significant capital expenditures.
As Kupperman suggests, if the current trajectory continues, megacap tech stocks could face valuation challenges similar to those experienced by shale companies. The potential for shareholder backlash is a real threat, as investors may eventually demand accountability for capital destruction.
Ultimately, the AI investment boom may follow the pattern of previous bubbles, ending not with groundbreaking innovation but with a slow realization of negative returns. The lessons of history suggest that caution and financial discipline are crucial as the industry navigates this uncertain landscape.
“The datacenters will be built, the chips will hum, and some of the capacity will eventually prove mind-blowingly useful. But the investors footing the bill today will regret ever making the investment.”
As the AI investment cycle unfolds, stakeholders must remain vigilant, balancing enthusiasm for technological advancement with a realistic assessment of financial viability. The outcome of this cycle will have far-reaching implications for the tech industry and the broader economy.
Caveat Emptor…
Harris Kupperman is the Founder & Chief Investment Officer of Praetorian Capital Management, and author of Praetorian Capital’s public blog, Kuppy’s Korner, from which this article has been reproduced with permission. Information or statements provided here are opinions of the author and may not represent the opinions of Praetorian PR LLC or its affiliates. Furthermore, the information is for educational and entertainment purposes only and does not represent investment advice.
