The Overlooked Dangers of Relying Solely on Valuation Metrics in Stock Selection

In the age of artificial intelligence, the stock market's reliance on traditional valuation metrics like P/E ratios and EV/EBITDA is becoming a double-edged sword. While these metrics remain foundational for many investors, their limitations in capturing AI-driven value creation are increasingly evident. The past year has exposed critical blind spots in these models, as companies adopting AI technologies often fail to see immediate financial returns—despite transformative operational gains. This disconnect poses significant risks for investors who prioritize short-term metrics over long-term strategic value.

The AI-Valuation Paradox: When EBITDA Falls Short

Consider the case of Wayfair, an online furniture retailer. By redefining its lost-sales KPI using AI, the company uncovered that 50-60% of "lost sales" were cross-category conversions, not dead-end losses. This insight allowed Wayfair to optimize product recommendations and align logistics with customer preferences, boosting customer loyalty and operational efficiency. Yet, these improvements—though measurable in customer satisfaction and long-term revenue—were not immediately reflected in EBITDA or P/E ratios. Traditional metrics, focused on short-term earnings, missed the broader operational and strategic value unlocked by AI.

Similarly, Tokopedia, Indonesia's largest e-commerce platform, leveraged AI to create a merchant-quality scoring system. By evaluating sellers on fulfillment efficiency and customer feedback, the platform enhanced trust and transaction volumes. While this boosted user engagement and long-term market dominance, the value was not captured in quarterly revenue figures or EV/EBITDA multiples.

The "gen AI paradox" further underscores this issue. A 2025 PwC report revealed that 78% of companies use generative AI in at least one function, yet over 80% report no material earnings impact. This gap arises because horizontal AI tools (e.g., chatbots, copilots) improve productivity but lack direct economic visibility, while vertical AI use cases—those embedded in core processes—remain underdeveloped due to technical and organizational barriers.

The Rise of AI Agents: A New Paradigm for Value Creation

The emergence of AI agents—autonomous systems that execute complex tasks—signals a shift from static AI tools to dynamic, goal-driven intelligence. These agents are redefining value creation by enabling proactive decision-making, operational agility, and new revenue streams. For example, a large bank reduced app modernization timelines by 50% using AI agents, while a retail bank improved credit-risk memo turnaround by 30%.

However, traditional valuation models struggle to quantify these benefits. AI agents create value through smart KPIs (e.g., customer retention rates, process automation efficiency) that are not part of standard financial reporting. This forces investors to rely on qualitative assessments or industry-specific benchmarks, which are often inconsistent.

The Case for a Holistic Investment Framework

To navigate AI-driven markets, investors must adopt a holistic framework that integrates traditional metrics with strategic, operational, and societal indicators. Key components include:

1. Strategic Alignment: Assess a company's AI roadmap, including investments in vertical use cases, data infrastructure, and talent. For example, Google's $32B acquisition of Wiz signals a strategic bet on AI security capabilities, a move that may not yet reflect in P/E ratios but aligns with long-term competitive positioning.

2. Operational Resilience: Evaluate AI-driven improvements in supply chain efficiency, customer personalization, and risk management. DBS Bank's cross-functional KPIs, which align customer and employee experiences with profitability, demonstrate how AI can enhance operational resilience.

3. Societal Impact: Consider metrics like public health outcomes (e.g., AI-driven cardiac risk prediction in healthcare) or environmental sustainability gains, which contribute to long-term value but are absent from traditional models.

4. Scenario Planning: Given AI's disruptive potential, investors should stress-test valuations against best- and worst-case AI adoption scenarios. For instance, a 2025 McKinsey analysis found that AI implementation risks (e.g., execution delays, cultural resistance) could reduce projected value by 30-50% in underprepared companies.

Investment Advice: Beyond the Numbers

For investors, the lesson is clear: do not let AI blind spots become blind investments. Prioritize companies that:
- Balance horizontal and vertical AI use cases (e.g., Pernod Ricard's interconnected KPIs).
- Demonstrate AI maturity through scalable infrastructure and governance (e.g., Sanofi's AI-powered Plai app for real-time KPI insights).
- Align AI goals with long-term value (e.g., DBS Bank's customer journey metrics).

Avoid over-reliance on EV/EBITDA multiples for AI-driven firms, particularly in sectors where intangible assets dominate. Instead, incorporate AI-specific metrics into your due diligence. For example, track AI ROI ratios (value created per dollar invested in AI) or agent-based productivity gains.

Conclusion

The AI revolution is reshaping value creation in ways traditional metrics cannot fully capture. While P/E and EV/EBITDA remain useful, they are insufficient in an era where strategic agility and operational innovation drive long-term returns. By adopting a holistic framework that accounts for AI's transformative potential, investors can avoid the pitfalls of short-termism and position themselves for sustained growth in the agentic age.