The AI Overspending Risk in Tech: Oracle's $15B Capex Shock and Market Reactions

In late 2025, Oracle's announcement of a $15 billion increase in capital expenditures for fiscal year 2026 sent shockwaves through the tech sector. The company, which now projects $50 billion in total capex for the year, cited surging demand for AI infrastructure and cloud services as the primary drivers. However, the market reacted swiftly and harshly, with Oracle's shares plummeting over 16% in after-hours trading as investors questioned the sustainability of debt-fueled expansion according to Bloomberg. This episode has reignited debates about whether the AI sector is entering a speculative bubble, with Oracle's aggressive spending serving as a cautionary case study.

Oracle's Strategic Rationale and Financial Health

Oracle executives defended the capex surge by emphasizing its alignment with revenue-generating infrastructure, particularly in data centers according to CIO Dive. The company highlighted a $523 billion backlog in customer commitments, including partnerships with OpenAI, Meta, and Nvidia, as evidence of long-term demand per investor reports. Yet, the financial risks are stark. Oracle reported negative free cash flow of $10 billion in Q2 2026 and a 25% rise in long-term debt to $99.9 billion. These figures, coupled with a 34% year-over-year increase in cloud revenue to $8 billion still below Wall Street expectations, underscore the tension between growth and profitability.

The market's skepticism is not unfounded. Oracle's stock price decline mirrored broader investor concerns about the efficiency of capital deployment in AI infrastructure. As Bloomberg noted, the company's strategy reflects a "debt-fueled expansion" that prioritizes market share over immediate returns. This raises a critical question: Can Oracle's AI investments translate into sustainable revenue, or are they merely inflating a sector-wide bubble?

Sector-Wide AI Capex Surge and Systemic Risks

Oracle's situation is emblematic of a broader trend. Major hyperscalers have increased AI-related capex by 67% since the release of ChatGPT in 2023. By 2025, AI-driven spending among these firms is projected to reach $350–$400 billion annually, with NVIDIA reaping significant benefits as a key GPU supplier according to Statista. This surge has fueled U.S. economic growth, with AI capex accounting for 1.2% of GDP in 2025 according to JPMorgan. However, the concentration of investment in a narrow set of firms has masked broader economic weaknesses, and productivity gains remain limited as Real Investment Advice notes.

The financial engineering underpinning this boom is equally concerning. Nearly half of the $2.9 trillion in AI investments from 2025 to 2028 is expected to be debt-funded. This leverage amplifies systemic risks, particularly as interest rates remain elevated and economic growth moderates. Oracle's debt spike is not an outlier; it reflects a sector-wide shift toward aggressive borrowing to fund speculative infrastructure projects according to Forbes.

Valuation Metrics and Bubble Indicators

Oracle's valuation multiples appear high but are not yet extreme. The company trades at a forward PE of 46.1x, above the software industry average of 32.0x but below the 71.2x average of high-growth AI peers according to Yahoo Finance. This suggests that while Oracle is more expensive than the sector average, it is not as overvalued as companies like Tesla, which trades at over 200x forward earnings according to P4SC4L.

However, Oracle's valuation is supported by strong fundamentals. The company has generated $19 billion in operating income over the past 12 months and maintains a 31.6% operating margin according to Oracle's earnings release. A discounted cash flow analysis estimates Oracle's intrinsic value at $252.33 per share, implying a 20.3% discount to its current price according to Yahoo Finance. This suggests the market may be underpricing Oracle's long-term cash flow potential, particularly given its expanding AI cloud backlog and strategic partnerships according to Aventis Advisors.

The broader AI sector, however, shows more troubling signs. Median valuation multiples for AI startups and public companies have reached 25–30x EV/Revenue according to Aventis Advisors, while venture capital inflows hit $120 billion in Q3 2025 according to Vestbee. These figures, combined with Oracle's stock plunge, indicate growing investor anxiety about overvaluation. As Robeco's analysis notes, the sector is "separating hard data from hype," with risks of a correction looming according to Robeco.

Conclusion: Balancing Growth and Risk

Oracle's $15 billion capex shock highlights the dual-edged nature of AI investment. While the company's strategic pivot to "chip neutrality" and cloud expansion demonstrates agility, its financial health and market reaction reveal vulnerabilities. The broader tech sector's AI spending surge, though economically significant, is increasingly reliant on debt and speculative growth narratives.

For investors, the key challenge lies in distinguishing between sustainable innovation and overleveraged speculation. Oracle's case suggests that while AI infrastructure is critical to the future, the path to profitability remains uncertain. As the market recalibrates, companies that balance aggressive capex with disciplined capital allocation-and avoid overreliance on debt-will likely emerge stronger. The AI boom is far from a bubble, but its risks are real and growing.