The AI Capital Flow Divergence: Where Money Is Moving and Why

The capital expenditure wave for AI infrastructure is accelerating at a staggering pace. The four largest U.S. tech giants are expected to collectively invest about $650 billion this year, a sharp increase from $410 billion in 2025. This spending is creating significant downside risk, as companies have curbed share buybacks more aggressively to fund the surge. The scale of investment is so large that it is now a major driver of U.S. economic growth, with Bridgewater estimating it could provide around 100 basis points of support this year.

Analyst estimates for this spending have consistently been too low, with actual hyperscaler capex exceeding 50% growth in both 2024 and 2025. The consensus for 2026 capex is now $527 billion, but history suggests this figure will likely be revised higher. This divergence between expectations and reality is already showing up in stock performance, as investors rotate away from infrastructure companies where earnings growth is under pressure and capex is debt-funded.

The bottom line is that the AI buildout is a massive, accelerating flow of capital. While it fuels near-term growth and stock gains, the sheer scale and reliance on external financing introduce new vulnerabilities. The market is beginning to price in the risk that not all of this spending will translate into proportional returns.

The ROI Reality Check: Where Financial Returns Are Failing

The massive capital outlay for AI is hitting a wall of disappointing returns. Enterprise AI initiatives achieved a mere 5.9% ROI in 2023, a figure that fell short of the 10% capital investment required to fund them. This stark disconnect between spending and payoff is forcing a strategic reckoning. The experimentation phase is over, and the pressure to demonstrate value is intense.

That pressure is now acute, with 61% of senior business leaders feeling more pressure to prove AI ROI now than a year ago. Boards have stopped counting pilots and started counting dollars. This urgency is creating a bifurcation in corporate strategy, as companies must choose between using AI primarily for cost-cutting through workforce reduction or investing to augment human capability and create new value.

The failure to scale is evident in the data. While many organizations are using AI, most are still in early stages. Only 39% report EBIT impact at the enterprise level, despite widespread adoption. This siloed value creation-where benefits are seen in specific use cases but not yet across the entire business-highlights the gap between technological capability and strategic execution. The path to proving ROI requires moving beyond pilots to deeply embed AI into core workflows.

The Market's Pivot: Capital Flow to Productivity and Platforms

The market has issued a clear verdict on the AI capex surge. Investors have rotated away from AI infrastructure companies where growth in operating earnings is under pressure and capex spending is debt-funded. This shift is the key signal of the trade's evolution, as seen in the divergence of stock performance. The average stock in Goldman Sachs' infrastructure basket returned 44% year-to-date, while the consensus two-year forward earnings estimate for the group grew just 9%. The disconnect is stark.

The next phase of the AI trade is expected to favor AI platform stocks and productivity beneficiaries. This is already showing in the data, as the average stock price correlation across large public AI hyperscalers has collapsed from 80% to just 20% since June. Investors are being more selective, rewarding companies that demonstrate a clear link between capex and revenues. This includes some of the world's biggest cloud platform operators, where the path to bottom-line impact appears faster.

While boardrooms are debating CEO-led prioritization, the financial flow is already moving. Capital is chasing a faster path to earnings, not just a bigger buildout. The market is betting that the real winners will be the software and services companies that can monetize AI adoption sooner, and the productivity beneficiaries that can show measurable operational gains. The era of indiscriminate infrastructure spending is giving way to one of targeted, value-driven investment.