Anton Nicholas at Nasdaq: The AI Trade Now Demands Measurable ROI, Not Hype

Anton Nicholas at Nasdaq: investors now want proof, not AI branding

At Nasdaq MarketSite earlier this month, Anton Nicholas said investor focus has shifted from whether companies are exploring AI to how effectively they're deploying it, with businesses expected to show measurable AI strategies. In markets, that changes the conversation quickly: AI narratives are being tested against execution.

The ROI clock is shorter

The pressure is not abstract. 95% of enterprise generative AI projects have failed to show measurable financial returns within six months, and 53% of investors expect positive ROI in six months or less. The market is increasingly focused on evidence, not roadmaps.

CEO accountability is rising

Nearly three quarters of CEOs say they are the main decision maker on AI, double the share from last year, and half of CEOs believe their job is on the line if AI does not pay off. That helps explain why capital allocation around AI is becoming more disciplined.

Spending is still climbing, but patience is not

Companies plan to double AI spending from 0.8% to about 1.7% of revenue, while hyperscalers have committed upward of $500 billion to AI infrastructure. The setup is clear: massive supply-side investment is colliding with tougher demand-side proof requirements.

Measurable AI is becoming the investor screen

At Nasdaq, Nicholas also stressed that organizations should connect AI efforts to concrete business outcomes, with emphasis on clear performance metrics before deploying AI initiatives. That fits a broader market signal: 61% of senior leaders say they feel more pressure to prove AI ROI than a year ago.

Only a minority of large companies are citing AI benefits

Only 21% of S&P 500 companies now cite AI benefits, even though companies delivering measurable results are seeing cash-flow margin expansion at roughly 2x the global average. The implication is straightforward: investors are rewarding demonstrated operating impact more than AI branding.

A practical three-step screen

1) Look for pre-defined metrics

Seek companies that set the scorecard before rollout rather than describing impact in general terms. Evidence-backed signals include:

• Clear performance metrics defined before deployment.
• AI linked to broader organizational goals and evaluated through established impact frameworks.
• Named KPIs tied to revenue, cost, quality, or cycle time, with baselines or measurement windows where possible.

2) Look for transparent reporting

If results cannot be traced, the market is less likely to reward them. Watch for:

• Quantified outcomes in earnings calls, investor decks, or operational reporting.
• Specific use cases and scale markers, not just pilot announcements.
• Explicit links between results and the company's original framework for evaluating impact.

3) Look for financial follow-through

This is the decisive screen. The evidence base is still selective:

• Only 21% of S&P 500 companies cite AI benefits.
• But adopters delivering measurable results are seeing cash-flow margin expansion at roughly 2x the global average.

What to scan for:

• Productivity or quality gains that show up in margins or working capital, not just employee-tooling headlines.
• Results large enough to matter to the valuation model, especially against a backdrop of record spending where ROI is still proving harder to surface.

Bulls are right that AI can be a structural value driver when results are measurable. Bears are right that much corporate AI commentary is still too vague. The practical edge is to favor companies that combine clear metrics, transparent reporting, and demonstrable outcomes.

Positioning for the next phase of AI value creation

The buildout is far from over, with nearly $3 trillion of AI-related infrastructure investment projected through 2028. That leaves room for market leadership, but the question has changed. It is no longer just "Is AI important?" It is "Who captures value while that spend is deployed?"

Infrastructure: spend certainty, but still a monetization test

Chips, cloud, networking, power, and data-center enablement remain the clearest spend-capture lanes while the buildout continues. Watch for:

• Revenue tied to compute, storage, optical, power delivery, cooling, or site development.
• Capacity bookings, lead times, and utilization commentary that confirm demand.
• Supply constraints that can support pricing and mix.

Break signal:

• Spending slows before utilization or margins improve.
• Management replaces shipped-volume commentary with vague pipeline language.

Application adopters: higher variance, bigger multiple upside if proof lands

This is the less certain bucket, but also the one where multiples can expand fastest if companies can show real operating impact. The market wants companies that defined clear performance metrics before deploying AI initiatives and can now point to trackable, verifiable results.

Watch for:

• Revenue protection, win-rate improvement, lower handling costs, fewer defects, or faster cycle times.
• Production deployments, customer references, and repeatable industry templates beyond pilots.
• Margin impact that holds beyond a single quarter.

Break signal:

• Commentary slips back to "experimentation" or broad transformation language.
• Productivity claims appear, but financial outcomes still do not show up.

Sponsored and adjacent themes: narrative only matters when it supports real adoption

This is the most perception-sensitive bucket. It warrants attention only when enthusiasm is feeding real orders, financing, or downstream deployment.

Confirmation signal:

• Investor, policy, or customer enthusiasm is translating into real commercial activity.

Break signal:

• Hype is carrying the story while actual adoption remains thin.

The positioning lesson is straightforward: stay constructive on AI, but rotate toward proof. Infrastructure offers more spend certainty; application adopters offer more multiple-expansion upside if they can demonstrate results; adjacent themes deserve attention mainly when perception is backed by real economics.