The Infrastructure Trap
Capital is impatient. It demands immediate returns. In the first five months of this year, those returns lived almost exclusively in silicon. While the broader technology sector surged on the back of hardware demand, the software and internet services layer remained stagnant. This is the divergence that currently defines the equity markets. Investors have poured billions into the physical architecture of the next economy while ignoring the companies that must actually use it. The gap has reached a breaking point.
The data does not lie. As of late May, the PHLX Semiconductor Index has outperformed the Nasdaq Internet Index by a factor of seven. This is not a market in balance. It is a market obsessed with the shovel makers while the miners are left in the cold. Per recent Bloomberg market data, the valuation spread between hardware providers and internet platforms has reached its widest level since the early 2000s. The narrative suggests that internet companies are lagging because they are spending too much on capital expenditures and not seeing enough immediate top-line growth. This cynical view misses the cyclical nature of tech adoption.
The Callahan Thesis
Peter Callahan, the specialist at Goldman Sachs, sees a shift coming. His argument is technical. It focuses on the transition from the build phase to the deployment phase. For the past eighteen months, the market has rewarded the companies building data centers. Now, the focus must shift to the companies that will monetize the compute power within them. Goldman Sachs suggests that the internet sector is currently a coiled spring. The underperformance is not a sign of weakness but a period of digestion.
Margins are the key metric. While the market focused on the high costs of AI integration, it ignored the efficiency gains starting to appear in quarterly filings. According to Reuters financial reporting, several mid-cap internet firms have quietly reduced operational overhead by 15 percent through automated customer service and code generation. These are real savings. They are not yet fully priced into the stock valuations. The market is still looking at the expense line rather than the margin expansion.
YTD Performance Comparison: Semiconductors vs. Internet Index
Valuation and Mean Reversion
The math of mean reversion is brutal. When one sector trades at 45 times forward earnings while its primary customers trade at 18 times, something has to give. Either the hardware providers will see a massive multiple contraction, or the internet companies will see a rapid re-rating. Callahan bets on the latter. He points to the historical lag between infrastructure investment and application revenue. We are entering that window now.
Institutional positioning supports this view. Hedge fund exposure to internet stocks is at a three-year low. This is a contrarian signal. When the largest players are underweight a sector that is still growing its user base and revenue, the risk is to the upside. The following table illustrates the current valuation gap between the leaders of the hardware surge and the laggards of the internet space as of May 2026.
| Sector Leader | Forward P/E Ratio | YTD Return (%) | Revenue Growth (YoY) |
|---|---|---|---|
| Hardware/Chips | 44.2 | 28.4 | 38.5% |
| Cloud Infrastructure | 31.5 | 15.2 | 22.1% |
| Internet/SaaS | 18.8 | 4.2 | 14.8% |
| Social Media/AdTech | 16.5 | 3.8 | 12.2% |
The discount is palpable. You are paying less than half the multiple for an internet company compared to a chip designer. Yet, the chip designer relies on the internet company to keep buying their products. This circularity is the fundamental flaw in the current market narrative. If the internet companies fail to rally, they will stop buying chips. If they continue to buy chips, their earnings will eventually reflect the utility of that investment. Either way, the current spread is unsustainable.
The Regulatory Shadow
Fear is the primary driver of this discount. Investors are terrified of the regulatory environment. The Department of Justice and the FTC have been relentless in their pursuit of big tech. This has created a permanent risk premium on internet stocks. However, the legal process is slow. Markets are fast. The discount for regulatory risk has likely been over-applied. We are seeing a classic case of price discovery being hampered by political noise.
Smart money is looking past the headlines. They are looking at the cash flow. Per SEC filings from the first quarter, the top ten internet companies have increased their share buyback programs by a combined 40 percent. They are buying their own stock because they know it is cheap. They are the ultimate insiders. When the companies themselves are the most aggressive buyers in the market, retail investors should take notice.
The next major catalyst is the June 15 FOMC meeting. If the Federal Reserve signals a pause or a pivot in response to the cooling labor data seen this week, the internet sector will be the primary beneficiary. These companies are long-duration assets. They are sensitive to interest rates. A stable rate environment is the fuel needed for the Callahan rally to ignite. Watch the 10-year Treasury yield. If it breaks below 4.1 percent, the rotation out of hardware and into internet software will accelerate. The divergence is ending. The convergence will be fast and expensive for those who missed the signal.
