Jun 2, 2026
The main worry in markets today is not the direction of travel but the level of conviction. Investors are not wrong to lean towards the optimistic case – the de-escalation in the Gulf, a US economy that keeps growing, and an AI cycle that could eventually lift productivity more broadly. Our concern is more precise, however: while those possibilities remain contingent probabilities, markets are pricing them as if they were close to certain. A great deal of positive news has already been pulled forward into this year’s growth, this year’s earnings, and a long-term future being discounted back to the present at a generous rate. When good news is front-loaded, the bar for disappointment falls. We are not bearish on the world. We are sceptical of certainty.
The quality of US growth is lower than the quantity suggests
The headline US data remains respectable, and second-quarter GDP estimates have been revised higher. But the composition is somewhat concerning. The “growth” is increasingly being driven by consumers running down savings and companies building inventories, as they pull forward spending in anticipation of likely disruptions and higher input costs — none of which is the same as durable, income-led demand.
The household side is the first warning light. Consumers are still spending, but increasingly on credit rather than on rising real incomes. The personal savings rate fell to 2.6% in April from 4.3% in January, even as disposable income slipped 0.1% on the month and spending rose 0.5%. Real disposable income was down 1.1% year on year, the weakest since late 2022, with PCE inflation at 3.8% and core at 3.3%. Consumption can remain resilient for a while after household finances start to deteriorate; the difficulty is that such resilience is borrowed from the future. Credit data tell the same story: the New York Fed reports that 4.8% of household debt is in some stage of delinquency, with credit-card early delinquency transitions running at 8.6% annualised. These are not crisis readings, but they are evidence of a consumer less robust beneath the surface than the aggregate spending line implies.
The corporate side may be flashing amber, too. Inventory accumulation is supporting GDP — the Atlanta Fed’s GDPNow inventory component is currently adding around a percentage point to second-quarter growth — but some of that build looks like fear rather than confidence, with purchasing managers stocking ahead of likely input-cost increases. Inventory build is not final demand; it lifts growth today and can become tomorrow’s drag if demand fails to clear it. The US is still growing, but more of that growth than usual looks pulled forward rather than created.
Chart 1: Atlanta FedNow GDP growth Indicator – flattered by Inventory Build
Index
Source: Bloomberg
Oil: the peace dividend may arrive before the barrels
The oil price dropped 10% over the past week on hopes of some negotiation between the US and Iran, and Brent has shed much of its war premium. We think the market is right to brush aside some of the geopolitical panic, but it is possibly too quick to fade the underlying logistics problem. This is no longer a fear story; it is more an inventory-and-shipping story now. The US Energy Information Administration estimates that Middle East crude shut-ins reached 10.5 million barrels per day in April and expects global inventory drawdown to be an extraordinary 8.5 million barrels per day in the second quarter. Even its base case that Hormuz traffic resumes gradually from June, does not see most shut-in production restored until January 2027.
The industry is more cautious still: The Abu Dhabi National Oil Company’s Sultan Al Jaber has warned that full Hormuz flows may not return before the first half of 2027, while Chevron CEO Mike Wirth notes that trade cannot normalise until shipowners and insurers are willing to go back to work. Exxon’s Neil Chapman supplies the sting in the tail: once inventory buffers are exhausted, prices do not rise gently, they jump. A ceasefire can be announced over a weekend; however, shipping confidence, insurance capacity, and commercial inventories take far longer to rebuild. The peace dividend is real — the barrels are slower than the headlines.
Iran: the most likely outcome, done the President’s way
A deal remains, in our view, the most probable path. Negotiations appear to have edged towards a 60-day ceasefire extension and a route back to nuclear talks. But the obstacle has shifted from drafting to presentation. President Trump has reportedly asked for changes, particularly on how Iran’s nuclear material will be handled. The decisive question is no longer whether an agreement is technically achievable but whether it can be made to look like the US achieved it through strength rather than compromise.
That requires several things to line up: Iran giving visible ground on nuclear material, Israel avoiding actions in Lebanon that would undermine – or collapse – the process, and the White House being able to claim that pressure produced the result. It is possible, but it is not a clean signature, and the most likely route there runs through a harder public posturing first. The calendar adds an incentive: with the 250th-anniversary celebrations approaching and the programme already thinned by performer withdrawals, a foreign-policy win offers a stronger narrative than ceremony alone. If the deal can be sold as “Iran blinked,” it can probably be done; if it is perceived as a reciprocal compromise, the odds of delay rise. Markets are pricing the outcome. They are underpricing the path.
Chart 2: Brent Oil Discounts some Good News on Negotiations
Source: Bloomberg
AI: the cycle’s shock absorber and its stress point
The AI debate has moved on from whether AI is a bubble to how far AI capital expenditure has become the shock absorber for a more fragile US economy. Combined 2026 capex at Alphabet, Amazon, Meta, and Microsoft is estimated at roughly $635bn to more than $700bn, with much of it tied to AI infrastructure. One recent estimate suggests that if we strip out AI infrastructure and energy, forward S&P 500 earnings growth falls close to zero. Put simply, the market is not buying the US economy; it is buying the AI build-out and everything that supports it.
That makes the index more concentrated and the risk more layered than the level implies. AI capex is now simultaneously an earnings bridge, an electricity-demand shock, a construction cycle, a chip cycle, and a potential political flashpoint. The market is pricing the upside associated with AI deployment but not yet the constraints at hand: power availability, grid-connection delays, water, rare earths, local resistance to data centres—and the possibility that households face a fresh cost-of-living squeeze just as white-collar displacement accelerates. The lagged cost pressure is already visible upstream where chip prices have risen substantially: sulphur has reportedly moved from $650 to $1,060 a tonne and sulphuric acid is up around 158%, with the effect likely to surface in aspects of normal life- fertiliser, crops, and grocery prices over the next couple of harvest and retail cycles. AI may be supporting the cycle at precisely the moment it begins to destabilise parts of it.
Chart 3: Ai’s Bull Run
Source: Bloomberg
China: a strong headline built on a fragile consumer
China is the photographic negative of the American consumer, and at first glance the better story. First-quarter GDP grew 5.0% year on year, beating the consensus’ roughly 4.8% projection and accelerating from 4.5% in the final quarter of 2025, with quarter-on-quarter growth of 1.3%. But the engine pulling the growth was external and industrial: Q1 foreign trade rose around 15%, auto exports jumped close to 58%, and “new quality productive forces” such as high-tech manufacturing and robotics, where China became a net robot exporter for the first time, did the heavy lifting. The domestic front stayed soft. Fixed-asset investment rose just 1.7%, property investment is still contracting (if less steeply), and retail sales managed only a modest recovery.
The inflation data flatter more than they reassure. April CPI of 1.2% was heavily inflated by a surge in gold prices; strip that impact out and the core is closer to 1%, while factory-gate prices have now been in deflation for over three years. The Chinese household is hoarding savings that it will not deploy — the mirror image of the American household spending savings that it does not have. Interestingly, both reflect fragile confidence. The read-through is the same as our central theme: the good news in Chinese assets (the 5% print, the export and AI-hardware boom, the promise of stimulus) is again being front-loaded ahead of the domestic demand that would make it durable, and Beijing’s policy support is priced with more conviction than the rebalancing has so far earned.
Japan: reflation priced as a destination, not a journey
The Bank of Japan voted 6-3 to hold rates at 0.75% on 28 April, the highest in three decades, but the split decision exposed pressure to move higher, with dissenters calling for 1%. More telling was the shift in its forecast: the BoJ raised its FY2026 core inflation projection to about 2.8% while cutting growth to 0.5%. That is the problem. Inflation is rising partly for the wrong reason — an oil and terms-of-trade shock from the Middle East that is squeezing real incomes — rather than solely through a healthy wage-price cycle, even as this year’s Shunto round again targets pay rises near 5% and real wages have only just stopped falling.
The fiscal arithmetic is tightening, too. The 10-year JGB yield hit 2.68% this week, the highest since 1997, against debt near 230% of GDP. A softer yen, still just below the roughly 162 “line in the sand,” is flattering exporter earnings in a way that could reverse abruptly on a single hawkish surprise. Corporate-governance reform is real, and we like the structural direction, but the market is pricing the best-case version of a transition that is still only half complete and now complicated by stagflation-lite arithmetic: 0.5% growth and near-3% inflation.
The week ahead
The coming week tests the front-loading of growth thesis directly. The key release is the US May employment report on Friday 5 June — payrolls, the unemployment rate and average hourly earnings will show whether the labour market is still underwriting consumption or starting to confirm the income squeeze. Around it:
• ISM manufacturing and services surveys, JOLTS job openings, and the ADP private payrolls print — the clearest read on whether demand is cooling beneath the headline.
• Euro-area flash inflation, and China’s private-sector Caixin PMIs, as a cross-check on the official data.
• Oil prices and any US–Iran framework headlines, which remain the swing factor for risk sentiment.
The ECB decision falls on the following Thursday, 11 June; a hold is the likely outcome, but its updated projections will matter given the current energy backdrop.
What it means for portfolios
Our posture is not to step out of risk but to stop paying for certainty — and to be specific about the direction.
Equities. We would keep equity positioning simple. In the US, active managers may still struggle given index concentration and rapid theme rotation. Japan, with clearer momentum and better structural support, is easier to like. China remains the clearest market for active management, where selectivity still matters. Overall, we would reduce US concentration, stay constructive on Japan, and remain selective and hedged in China.
Fixed income. Move up in quality: we prefer investment grade to high yield, where spreads do not compensate for rising consumer-credit stress, and cut CCC, leveraged loans, and consumer ABS. In rates, we prefer the intermediate part of the US curve over the long end, given fiscal risk and sticky inflation. We advise holding inflation-linked bonds against lagged cost-push pressure from energy and fertiliser, and we stay underweight long-dated JGBs with yields at multi-decade highs and the BoJ still normalising.
Real assets and optionality. Keep a tactical allocation to energy and oil equities as cheap insurance against a renewed Hormuz logistics premium, with gold as a portfolio hedge. We also want more portfolio optionality than usual in the form of modest, inexpensive downside protection, because the market is treating probabilities as certainties. The base case is constructive; the mispricing is in certainty.
