The correlation between the performance of U.S. semiconductor stocks and software stocks has fallen below zero for the first time on record, underscoring the deepening divergence within the technology sector driven by artificial intelligence.

According to data, a metric tracking the 40-day correlation between the iShares Semiconductor ETF and the iShares Expanded Tech-Software Sector ETF closed in negative territory for the first time last Friday and further declined to -0.05 by Tuesday. At the level of S&P 500 constituents, the correlation between chip and software stocks remains positive but has dropped to its lowest point since 1994.

The divergence is particularly pronounced at the individual stock level. Memory chip leader Micron Technology has surged over 150% year-to-date, while enterprise software company ServiceNow has fallen more than 33%, resulting in a performance gap exceeding 180 percentage points.

In recent months, investors have been buying semiconductor stocks while selling software shares. This strategy has delivered substantial returns as spending on AI computing infrastructure surges and profit growth accelerates for chipmakers like NVIDIA and Micron Technology.

In contrast, software stocks have faced pressure, primarily due to concerns that competition from AI companies such as OpenAI and Anthropic could erode their business and hinder long-term revenue growth.

Brad Conger, Chief Investment Officer at Hirtle Callaghan, noted that investors have been "betting on AI capital expenditures and shorting stocks vulnerable to AI disruption." He added that momentum-based trading strategies appear to have been drawn into this hedge trade as well.