Introduction
Artificial intelligence (AI) has become the most compelling theme in global markets, with valuations soaring past $300 billion for firms like OpenAI and Anthropic. Yet, the critical question for investors is whether this flood of capital will translate into measurable productivity improvements or simply fuel a new financial bubble built on circular funding.- --
The AI Productivity Paradox
- MIT Study Findings: A recent MIT analysis shows that the majority of enterprise AI pilots fail to deliver quantifiable efficiency gains, and few progress to scalable production. The share of work reliably automatable by current models is lower than early hype suggested, creating a classic "J‑curve" where initial investment outpaces realized benefits.
- Sector‑wide Impact: While generative AI does accelerate research and routine tasks (e.g., data extraction, code generation, earnings‑call summarisation), the magnitude of these gains varies dramatically across use cases and depends heavily on proprietary data and governance structures.
- --
Circular Funding – A New Systemic Risk
Circular financing describes a situation where AI‑related capital circulates among a tight cluster of firms, inflating valuations without expanding the underlying demand.
| Indicator | Description | Why It Matters |
|---|---|---|
| Capex Circularity | Supplier‑customer revenue links, related‑party commitments | Highlights fragility when cash flows rely on continued funding rather than genuine market demand |
| Bubble Scores | AI stack‑layer valuation residuals | Detects localized overheating in specific technology layers |
| Debt‑Financed Rounds | Private‑equity leverage tied to AI asset appreciation | Increases exposure to a rapid de‑valuation shock |
These dynamics echo the dot‑com era, where concentrated capital and leveraged financing amplified a market correction when growth expectations fell short.
- --
Distinguishing Real Gains from Hype
Financial analysts must separate durable productivity improvements from momentum‑driven enthusiasm. Key approaches include:
1. Layered Valuation Analysis – Compare bubble scores across AI stack layers to spot downstream over‑valuation.
2. Network‑Panel Tests – Examine supplier‑customer revenue dependencies to gauge circularity risk.
3. Execution Risk Assessment – Evaluate a firm’s data assets, integration capability, and governance discipline.
"Differentiation increasingly depends on proprietary data, creative integration, and disciplined governance, rather than access to the same tools." – Sandback, CFA (Enterprising Investor, Dec 2025)
- --
Investment Strategies for a Mixed Outlook
- Focus on Companies with Proven Efficiency Gains: Look for firms where LLMs have demonstrably shortened research cycles or reduced routine‑task labor.
- Avoid Over‑Concentration in Circular Funding Loops: Diversify away from the small group of firms that dominate AI‑related capital flows.
- Prioritise Capital‑Intensive but Scalable Models: Companies investing in robust digital back‑bones and energy infrastructure are better positioned for sustainable productivity.
- --
Policy Implications and Macroprudential Safeguards
Policymakers are urged to:
- Promote skill development to broaden AI adoption across sectors.
- Strengthen competition policy to prevent excessive concentration of AI gains.
- Implement macro‑prudential tools that monitor circular financing and capex dependencies, similar to the IMF’s warnings on systemic risk.
- --
Lessons from the Dot‑Com Era
The 1990s boom showed that corporate profit spikes often lasted only a few years before a market correction driven by rising rates and waning sentiment. AI’s trajectory may follow a comparable pattern unless:
- Productivity gains become broad‑based and durable.
- Capital allocation shifts from speculative loops to real‑economy investment.
- --
Outlook for the Next Five Years
- Uneven Gains: Some firms will achieve notable efficiency improvements; many will fall short.
- Slower Pace: Overall productivity growth is likely to lag behind the most optimistic forecasts.
- Risk Monitoring: Circular financing indicators will be crucial barometers for investors and regulators alike.
- --
Conclusion
AI remains a transformative technology, but its true test lies in delivering tangible productivity gains while avoiding the pitfalls of circular funding that can amplify systemic risk. Investors who rigorously separate measurable efficiency from speculative hype will be best positioned to capture lasting value.
- --
References
- Sandback, A. W. (2025). The Two AI Stories: Measurable Gains and Hidden Balance‑Sheet Pressure. Enterprising Investor Blog. Link
- Gvozdeva, E. (2025). AI’s impact in investing: Value vs. hype. Russell Investments. Link
- DeepTech (2025). Caution and Opportunity: Understanding the Economic Risks of the AI Revolution. Link
- BIS (2026). Progress and Peril. Link
- Fidelity (2025). 5 signs of an AI bubble to watch for. Link
- Arxiv (2025). Boom, Bubble, or Buildout? A Multi‑Method Evaluation of Whether Artificial Intelligence Is in an Ongoing Financial Bubble. Link
- Commonfund (2025). AI and the Productivity Paradox: Are We Finally Seeing the Payoff? Link
- BNP Paribas (2025). Productivity, growth and employment in the AI era: a literature review. Link
