Tax refunds and AI boom have offset some US economic pain from Iran war and high gas prices, so far

U.S. inflation accelerated to 3.5% year over year in March, the fastest pace in nearly three years, as gasoline prices jumped 21% month over month following the Iran-related oil supply shock. GDP still grew at a 2.0% annual rate and first-quarter business investment rose 10.4% on the AI boom, while consumer spending increased just 1.6% and tax refunds temporarily cushioned households. The Fed and other central banks are holding rates steady for now as they weigh higher inflation against growing downside risks to growth.

Analysis

The market is getting a temporary cushion from two sources that are both non-recurring in the same way: refund season and capex tied to AI. That matters because the inflation shock is hitting the same household balance sheet twice — higher fuel outlays plus still-sticky non-energy prices — so the eventual slowdown should show up first in discretionary categories rather than in headline employment. The key second-order effect is that the consumer can look “fine” in aggregate while lower-income and ex-urban cohorts quietly retrench, which is usually when retail dispersion widens.

The more important macro implication is that this is a stagflationary bridge, not a clean growth scare. If energy remains elevated into the next payroll and CPI prints, the Fed is boxed into prioritizing inflation credibility over growth support, which keeps real rates higher for longer even if growth expectations fall. That is typically a bad setup for long-duration assets broadly, but it can be selectively constructive for balance-sheet-resilient energy and software/hardware capex beneficiaries with pricing power.

The AI capex story is also vulnerable to an over-earnings-risk premium reset. If businesses keep spending on compute while consumer demand softens, the market will start to distinguish between true monetizers and “capacity buyers,” pressuring the latter if management teams can’t show near-term ROI. Meanwhile, lower bond yields won’t rescue cyclicals quickly because the inflation impulse is coming from an exogenous supply shock, so the usual recession playbook may lag by one to two quarters.

Consensus is probably underestimating how quickly the tax-refund offset rolls off. Once that temporary support fades, the hit to unit volumes should emerge with a delay in May/June retail data, then flow into Q2 guidance cuts. The setup favors a short window where earnings revisions matter more than macro headlines, especially for consumer-facing names with weak margin buffers and for AI-linked hardware where expectations are already stretched.

Key Decisions for Investors

• Short XLY vs long XLE over the next 4-8 weeks: consumer demand should decelerate as refund support fades, while energy retains direct pricing power; target 8-12% relative underperformance if gasoline stays above prior highs.
• Buy put spreads on discretionary retail names with high lower-income exposure (e.g., M, NKE, HD) into the next earnings cycle: use 60-90 DTE structures to capture margin compression and volume softness; risk is a fast reversal in fuel prices.
• Maintain a tactical long in quality mega-cap AI infrastructure beneficiaries (e.g., NVDA, AVGO, ANET) but hedge with short baskets of high-multiple AI software/semicap names: capex spend should keep flowing, but valuation dispersion will widen if monetization lags.
• Add a small long-energy hedge via XLE or CVX against any broad equity exposure for the next 1-2 months: this is a cleaner inflation hedge than duration, with downside limited if oil retraces on policy de-escalation.
• Avoid chasing duration-heavy growth until the next inflation and retail print confirms disinflation; if 2Q data show spending rollover, use that as the entry point for long-duration assets rather than buying the dip now.