Everything Shaken, Little Really Stirred

Michael Grant

The first half of 2026 has delivered the same cocktail as 2025—everything is shaken but little is really stirred. In both years, a negative shock was absorbed quickly and for the same reason: While the Trump Presidency is inclined to disrupt the status quo, there is acute aversion to prolonged financial pain. The curtailed shock has become one of the signatures of Trump 2.0. Through all of this, the defining features of the investment landscape remain in place—dominated by the technological and financial supremacy of America, now amplified to the point of caricature by the AI revolution.

After a brief struggle, the curtailment of the Iran conflict has allowed the AI investment mania to rule supreme once more. Market leadership across technology has broadened, but only partially. The economic background is unchanged: a two-tier, K-shaped world of corporate prosperity and consumer discomfort in which genuine price stability has been lost but recession risk remains low. Our interpretation of peak reflation in Q1 appears validated—the expectations of Central Bank assistance with which the year began are gone. Bond markets have been remarkably unperturbed: their message is that interest rates remain in the higher-for-longer world, with only short interruptions.

Three features of the first half appear relevant for the second: the AI mid-cycle transition, the Warsh Fed, and the approaching political season. On AI, the fireworks are over—what follows is digestion rather than the unwind of a bubble. It is a pullback driven by secular rotation out of the Mag 7 and into the enablers and beneficiaries of the capex cycle. We have not yet reached peak investment, but free cash flow at the hyperscalers is deteriorating at an accelerating pace, feeding anxiety about funding requirements. The shift in leadership from the spenders to the suppliers is not fortuitous—it is the market signaling that the next phase of value creation must soon emerge.

The core tension is price compression at the model layer versus inflation at the physical layer. AI may feel like software to the user, but economically it begins as a real-world industry. Only an estimated fifth of a roughly $8 trillion buildout has been spent, and the bottlenecks are already visible and spreading—HBM, racks, liquid cooling, copper, substations, gas turbines and so forth. The profitability surge is unsustainable at its present pace. It is not yet clear that the AI boom has reached the phase at which it can be judged unreasonable—but the margin for error is narrowing with every leg higher.

The arrival of Kevin Warsh at the Federal Reserve has shaken consensus thinking about the monetary outlook. Warsh has long been uncomfortable with the hyper-interventionist practices that became the norm after the GFC of 2008. He has announced a comprehensive reassessment—Fed communications, balance sheet policy, data reliance, productivity and inflation frameworks. The weak-dollar narrative that has been so seductive within the financial community has begun to crumble: The spot USD Index has risen above 100 to its highest level since May last year.

Bond markets through the first half of 2026 have signaled that we remain in a "higher for longer" world that has not become "higher and higher." Supply constraints will not produce a wage-price spiral. Demand in most major economies remains constrained. Labor productivity promises to remain healthy. The 10-year US Treasury has failed to breach 4.65%, remaining within its zone of fluctuation since 2022. Bonds should not threaten the equity bull market unless that threshold gives way—this would be the signal for investors to return to fixed income.

The inflation picture reinforces this view. The consensus has anchored its expectations to hot Q1 prints, but residual seasonality in the data is well documented and payback comes in H2. Our forecast is for the inflation data to surprise on the downside in the coming months. The labor market is not tight as evidenced by unemployment at 4%–4.5%, wage growth below 4%, and weak job turnover. We believe core PCE can decelerate toward 2.9% by year-end, meaningfully below consensus. If correct, markets will begin repricing Fed cuts in H2. In other words, a Warsh Fed is more likely to reduce rates than hike them in 2026.

Looking into Q3, US equities typically consolidate or correct through the summer of midterm years. The AI theme may produce—and has so far in 2026—a reverse of market tempo this year. There are no obvious warning flags from bonds or credit. The AI profitability surge shows little immediate sign of fading. Political considerations will probably not cause disruption before the midterms. The influence of AI has broadened, extending into industrials, financials, and utilities—but it remains dominant. Given the absence of macro risk, our primary concern for the AI thematic is simply positioning: The risk is rotational in nature.

The argument for portfolio diversification is compelling given the concentration of returns, but it remains dangerous to over-anticipate an authentic change of market leadership. During an investment mania, there is no significant move away from leadership assets until the climax. Climax requires a catalyst—and three kinds have the power to produce discontinuity: policy, politics, and investor paroxysm. The November midterms, focused on control of the Senate, represent the next major checkpoint. Between July and October, we expect election uncertainty to weigh on sentiment—this is the window in which we will re-emphasize defense.

By the standards of investment manias, recent years will be judged as unprecedented. In this setting, many expect the S&P 500 to reach 7,900–8,000 by the midterms. The catalyst that might reverse the trajectory of the past quarter is not yet visible, yet the degree of volatility across and inside the equity world is itself a message: Positioning risk is rising because everyone is already on board. And when everything is shaken, it is not a time for complacency. Our forecast is for a topping process to unfold into autumn—with increasing challenges into 2027.