Minutes from the Federal Reserve's April policy meeting show several officials recognized that recent price gains in the software segment have pushed up core inflation readings, while also cautioning that such moves "may not be good predictors of future overall inflation." Those internal discussions underscore a tension for policymakers weighing whether the recent inflation uptick is broad-based or concentrated in narrowly defined technology inputs.
The Fed minutes stated: "Some participants noted that recent price increases in the information technology sector had contributed to higher inflation. A few of these participants remarked that, while price increases in the software category were contributing meaningfully to the increase in core inflation, price increases in that category may not be good predictors of future overall inflation."
Citi economists this week highlighted a significant gap between the latest core consumer price index, a widely watched monthly inflation statistic, and the Federal Reserve's preferred gauge, the core personal consumption expenditures price index. Core CPI rose 2.8% year-over-year in April, while core PCE was reported at 3.2% year-over-year. The Fed targets 2% inflation.
According to Citi, a large portion of the April divergence between core CPI and core PCE is attributable to surging memory prices that show up in a small sub-index of core CPI labelled "Computer Software and Accessories." Historically this specific index has trended downward over the past 25 years but is currently up 13.9% year-over-year, the analysts noted.
The technology industry has been operating under a supply squeeze for memory chips as demand has accelerated since the broad adoption of artificial intelligence technologies began in late 2022. That rapid demand expansion has complicated capacity planning for chip manufacturers and major technology companies investing in large-scale data center capacity.
At the same time, the so-called AI trade has driven an intense equity rally this year, helping U.S. stock indexes reach record levels. The Philadelphia Semiconductor Index, often used as a proxy for semiconductor equities, recorded an 18-day winning streak from late March through the end of April, the longest run in the index's history.
That equity advance occurred despite geopolitical tensions between Washington and Iran and a global sell-off in bonds tied to higher expectations for rate increases intended to counter rising inflation from elevated oil prices.
Citi's team also pointed to a transmission channel from equity prices into core PCE via portfolio management fees. In other words, rising equity valuations associated with AI-related investment activity are feeding into a component of the Fed's preferred inflation measure.
The analysts argued this pattern - where higher prices reflect a tighter supply for a narrowly defined business input, such as memory used in AI infrastructure - creates what they described as a "dovish opening" for the central bank's new leadership. The minutes' caution that AI-related price spikes could distort inflation metrics upward aligns with the view that these moves may not translate into sustained consumer inflation.
Kevin Warsh, a former Federal Reserve governor, was sworn in as the central bank's new chair on Friday in a ceremony overseen by President Donald Trump, who had selected Warsh for the position. Warsh assumes leadership at a complex moment: the president has pushed publicly for rate cuts, while rising oil prices linked to the Middle East conflict have made near-term easing less likely.
Citi emphasized that the memory price increase is not the same as generalized consumer inflation. Instead, it represents a higher relative price for a narrowly specified good experiencing short supply due to strong business demand. That distinction is central to the argument that measures lifted by AI-driven price moves may be poor indicators of future broad inflation trends.
For market participants interested in exchange-traded products tied to artificial intelligence and related technologies, the analysts pointed to a few AI-linked ETFs: Global X Artificial Intelligence & Technology ETF, iShares Future Exponential Technologies ETF, and Global X Robotics & Artificial Intelligence ETF.
Summary
The Fed's April minutes flagged software price increases as a contributor to higher core inflation, but noted those changes may not predict wider inflation. Citi analysts attribute much of the CPI-PCE divergence in April to a sharp rise in a "Computer Software and Accessories" sub-index driven by memory price increases linked to AI demand. Equity gains tied to AI also feed into core PCE through portfolio management fees. Citi says this concentrated, business-driven price pressure could provide the new Fed chair a "dovish opening" in communications.