War, inflation and Trump's tariffs have shaken the US. Why does the stock market keep going up?

US equities have recovered sharply despite the Iran war, high oil prices, and renewed inflation, with the Nasdaq up 11% year to date and the Dow and S&P 500 near record highs. The article argues that investor confidence is being sustained by expectations of Federal Reserve and government backstops, a K-shaped economy, and massive AI capital spending from seven mega-cap stocks that now represent 30% of the S&P 500. It also highlights rising inflation at 3.8% in April and escalating concerns that an AI bubble could eventually trigger broad market selling.

Analysis

The market is no longer trading the macro headline; it is trading the policy backstop plus index concentration. When a narrow set of mega-cap balance sheets becomes the marginal buyer of risk assets through AI capex, breadth deteriorates while indices still levitate — a classic late-cycle tell that can persist for quarters before it breaks. That makes passive exposure look safer than it is: the benchmark is being carried by a handful of names whose fundamental outcomes are increasingly correlated through the same spending cycle.

The more interesting second-order effect is that AI has become a quasi-fiscal stimulus for the private sector. Data-center and compute buildout supports semis, cloud, networking, and power infrastructure, but it also creates a crowded capital-allocation trade where every hyperscaler is racing to avoid underinvestment. That raises the probability of mispriced duration: if expected monetization slips even modestly, the market could re-rate capex-heavy winners first, then de-risk the broader growth complex through index-weighted flows.

The political overlay matters mainly as a volatility catalyst, not a trend driver. Investors are assuming intervention will cap downside in crises, but that also means the asymmetry is shifting toward abrupt repricing if policy credibility breaks — especially around tariffs, energy, or Fed independence. In the near term, the cleanest upside remains in the companies selling picks-and-shovels to the AI buildout; the cleanest downside is in crowded, high-multiple names whose valuation depends on flawless execution and stable discount rates.

Consensus is probably underestimating how much of the current market resilience is mechanical rather than cyclical. If the top of the market continues to outperform, that can coexist with weakening consumer demand and soft breadth for a while; the true risk is not an immediate crash but a sharp dispersion event when capex returns disappoint or rates stay higher for longer. That argues for staying long the infrastructure winners while selectively hedging the most crowded megacap exposure.

Key Decisions for Investors

• Long NVDA vs. short equal-weight S&P 500 for 1-2 quarters: captures continued AI capex momentum while reducing dependence on broad market breadth; risk is a sudden multiple reset if AI spend guidance rolls over.
• Pair long MSFT/GOOGL with short TSLA over the next 3-6 months: favors recurring AI monetization and cloud infrastructure over the most sentiment-driven megacap with weaker earnings quality; stop if TSLA reaccelerates on margin expansion.
• Initiate a basket long in AI infrastructure beneficiaries not named in the article (semis, networking, data-center power) for 6-12 months: best risk/reward if capex stays elevated even as headline AI enthusiasm cools.
• Buy put spreads on a concentrated mega-cap growth proxy into the next earnings cycle: cheap hedge against a breadth unwind and capex disappointment; target payoff is a 10-15% index drawdown with limited premium at risk.
• Trim low-quality AI-adjacent winners after strong prints and rotate into cash-flowing hyperscalers: the market is rewarding spend today, but the first sign of slower capex growth will likely hit the most expensive names hardest.