Global tech behemoths are persisting in escalating their large-scale capital investments, propelling the AI industry to unprecedented levels of development.

Fueled by these substantial capital inflows, the stock prices of leading tech firms have continued their upward trajectory, frequently reaching all-time highs and demonstrating unprecedented market strength. Amidst this buoyant market sentiment, expectations for the AI infrastructure sector have soared, driving up industry valuations.

Yet, beneath the surface of market optimism, there are structural contradictions that warrant attention. While revenue at the AI application end is experiencing rapid growth, it is building on an extremely low base, and the overall revenue scale pales in comparison to the colossal capital investments in the AI infrastructure sector.

More critically, as tech giants persist in expanding their capital expenditures, corporate free cash flow remains under persistent pressure, and profit-generating capabilities are facing formidable challenges.

This scenario raises a pivotal market question: Should tech giants decelerate their capital expenditure pace, coupled with AI application-end revenue conversion lagging behind investment growth, does the asset bubble inflated by this round of AI market performance risk bursting?

From a historical vantage point of capital markets, all industry valuation bubbles ultimately necessitate digestion through the realization of tangible performance. Reflecting on the previous surge in the new energy sector, Tesla, the sector's frontrunner, witnessed its stock price plummet by over 70% at its peak drawdown, fully underscoring the correction risks associated with high valuations unsupported by performance.

In comparison to the prior new energy market performance, this round of AI sector involvement is more extensive, with deeper engagement from leading giants, corresponding to even larger-scale capital expenditures, and significantly greater industry leverage and investment pressure than before.

According to Goldman Sachs' latest research report, by 2027, capital expenditures by global hyperscale cloud computing companies are projected to reach between USD 1.1 trillion and USD 1.4 trillion.

Behind these substantial investments lurk continuously accumulating hidden liabilities: nearly USD 1 trillion in procurement commitments, over USD 800 billion in unexecuted lease contracts, and hundreds of billions of dollars in supplier financing arrangements, resulting in hyperscale cloud companies amassing approximately USD 1.8 trillion in off-balance-sheet risk exposures.

Morgan Stanley's cautionary data reveals that the leverage ratio of hyperscale cloud companies surged from 0.9x to 1.8x in just two quarters. Presently, industry capital expenditure growth continues to significantly outpace corporate revenue and free cash flow growth, while the profit pressure from large-scale asset depreciation has yet to be fully unleashed.

According to Morgan Stanley's AI Debt Financing Tracker report, as of the end of May 2026, global AI-related bond issuance reached USD 236 billion, marking a substantial increase of 357% compared to the same period in 2025.

As leading tech giants persist in ramping up capital expenditures, their subsequent financing needs will escalate further. Morgan Stanley predicts that total AI-related bond issuance for the full year of 2026 will surpass USD 570 billion, indicating that bond financing by tech companies will accelerate further in the second half of the year.

Notably, core companies such as OpenAI, Oracle, NVIDIA, Microsoft, AMD, and Amazon have formed a cyclical system intertwined with customer cooperation, equity investments, supplier financing, and share buybacks. A modest amount of capital circulates repeatedly among leading entities in a closed-loop operation, implying that once giant capital expenditure growth decelerates, the performance of upstream equipment and computing power suppliers—the 'shovel sellers'—will also weaken synchronously, creating chain pressure across the industry upstream and downstream.

Specifically, the five hyperscale cloud providers—Amazon, Meta, Google, Microsoft, and Oracle—currently account for 4% of the entire investment-grade bond index in terms of outstanding bond scale. From a free cash flow perspective, Morgan Stanley predicts that by 2026, Amazon and Meta's free cash flow will approach zero or even turn negative, with corporate incremental operating and investment funds relying almost entirely on new debt financing. This sustained debt-fueled spending model is unsustainable in the long run.

To maintain cash flow balance and sustain AI sector capital expenditures, large-scale layoffs have become a standard practice for global leading tech giants. Previously, Meta announced a global layoff of 10%, affecting approximately 8,000 employees. Simultaneously, Meta signaled plans to initiate multiple rounds of layoffs in the second half of 2026, with specific timing and scale yet to be finalized, as management may dynamically adjust and optimize personnel structure based on AI industry development progress.

In April 2026, Microsoft notified internal employees of a voluntary buyout plan targeting senior employees with a combined length of service and age totaling 70 years, expected to cover over 8,000 employees. In March of the same year, Oracle was reported to be planning a new round of large-scale layoffs, with the number of layoffs estimated to be in the range of 20,000 to 30,000.

As early as January 2026, Amazon announced the elimination of approximately 16,000 positions, marking the company's second round of large-scale staff reductions following October 2025.

Kanjian Finance believes that AI technology has indeed brought about revolutionary improvements in production efficiency, significantly optimizing corporate operational efficiency. Against this backdrop of transformation, the continuous compression and elimination of repetitive entry-level roles and redundant mid-level management positions in the industry are inevitable outcomes of sectoral iteration.

However, for leading tech companies, achieving sustained revenue growth and delivering on performance through AI remains the linchpin for long-term development. Once the AI industry encounters a phased bottleneck, with technology adoption and revenue conversion falling short of market expectations, the valuation bubble of this round of AI will ultimately backfire on all participating tech giants.