Carmignac – H2 2026 Outlook

Economic perspectives:

• Middle East developments still pose key stagflationary risks for the global economy. The energy price shock so far has been limited by a massive drawdown in inventories, which cannot continue at the same pace into July, should no deal emerge by then to reopen the Strait of Hormuz. Inflation pressures compel central banks in Asia to hike rates and restrict capital outflows, which tightens external constraint on the US AI capex cycle.
• US growth is one big bet on AI through the infrastructure buildout and the related equity wealth effects. The capex boom adds to the inflation pressure in the goods sector. Consumers suffer from the real income shock but should benefit from easier access to credit (deregulation) and a stabilised labour market (with a contracting labour force). The Federal Reserve (Fed) is behind the curve, and Fed Chair, Warsh will have to choose between pleasing Trump or the bond vigilantes.
• China’s policy mix projects stability but its economy is turning Japanese. Xi erroneously believes China’s status as a superpower makes its predatory export model sustainable. The AI boom is too small to offset the deflationary impact from the housing market. The Peoples Bank of China faces a liquidity trap. Fiscal policy will have to come to the rescue to safeguard the GDP growth target.
• Europe is the collateral victim of the tech war, the energy war and the trade war. The European Central Bank (ECB) is playing the credibility card well, but executive powers appear to fail to stand up to the structural challenges.

Investment strategy:

• In short: own growth, collect income, and stay short the part of the market we view as most vulnerable to repricing – long-term government bonds.
• We want exposure to nominal growth, but selectively. The US economy is still running at a strong nominal pace, and earnings expectations remain robust. But this is not a broad-based equity story – growth is increasingly concentrated in areas tied to AI investment.
• AI remains the centre of gravity. The first leg of the trade rewarded hardware. The next phase should broaden toward hyperscalers, data centres, infrastructure and companies sitting at the bottlenecks of the AI value chain. Compute remains scarce, demand is accelerating, and economics are improving.
• Europe offers optionality if geopolitics calm down. The European catch-up trade has been interrupted, not necessarily broken. A de-escalation in the Middle East could revive appetite for European cyclicals, with aeronautics and banks among the clearest beneficiaries in our view.
• The weakest asset is duration. Resilient growth, sticky inflation and deteriorating fiscal dynamics all point to higher long-term sovereign yields. The issue is both fiscal and monetary credibility.
• Credit still deserves a place, but not blind trust. Spreads are tight, but carry remains valuable. The key is to own credits backed by hard assets and resilient cash flows, while avoiding sectors exposed to AI-driven disruption.

Oil shock curbs the acceleration in global growth
After a promising start to the year, global growth was hit with the shock of a new major conflict in the Middle East. The closure of the Strait of Hormuz and the damage to energy infrastructure in the Gulf region are a massive negative supply shock affecting a long list of industrial sectors, given the widespread use of hydrocarbons as energy, but also as feedstock for industrial chemistry and metallurgy.

Shortages of certain varieties of crude, of LPG, LNG and specific refined products (e.g. naphtha) are disrupting entire industries in some emerging markets (e.g. fisheries in the Philippines, restaurants in India), and creating cascading disruptions in global supply chains for fertilisers (urea), plastics (PET), synthetic fibres (propylene), metal refining (sulphur), and semiconductors (helium, photoresist solvents). The inflation shock is forcing central banks to pause their planned easing moves which were a key plank of our optimistic outlook at the end of last year. Overall, we assess a 0.5% reduction in global GDP growth this year, from our earlier 3% growth scenario.

Two months into a fragile ceasefire, Iran and the US remain at loggerheads on all the topics that led them to war: Iran’s nuclear capabilities (stockpile of enriched uranium, enrichment capabilities), ballistic missiles, and regional proxies (in Lebanon, Iraq and Yemen). The fact that Trump did not escalate the conflict after judging Iran’s negotiating stance “totally unacceptable” on May 10, or that he cancelled “Operation Freedom” after only 24 hours have convinced the Iranians that they possess escalation dominance, and can therefore play the clock as the damage caused by a closed Strait spreads through the global system. Trump faces a Cornelian choice between military escalation or humiliating concessions.

Our central case is that, as the summer driving season picks up, falling approval ratings will push Trump to offer financial relief to Iran and negotiate a weaker version of the 2015 deal (JCPOA) in exchange for reopening the Strait.

In that scenario, the need to clear all mines from the waterway, gradual reopening of insurance channels and remaining sporadic flareups could mean traffic only normalises by Q4. Supply chain disruptions would linger well into 2027 and keep global inflation under upward pressure. Moreover, the way the global economy adjusted to the oil supply loss so far has relied disproportionately on inventory drawdowns and not on price-induced demand destruction, notably because governments and many market participants anticipated the shock would be short-lived. As a result, should the reopening of Hormuz be delayed further towards September, there would no longer be much leeway to absorb the shock through inventories, and thus the price response could be non-linear.

US: a giant AI play
Among major economies, the US will be the least affected by the Hormuz standoff, both as an oil exporter and as the frontrunner in the AI race. AI capex has surpassed consumer spending as the main driver of the economy, accounting for 87% of GDP expansion over the past 6 months¹. The AI engine is further boosted by the wealth effects generated by the surge in AI-related equities. We believe that equity wealth effects can be estimated to bolster consumption by 0.8%, albeit part of that is offset by the contraction in housing wealth.

Apart from these virtual equity gains, households are still grappling with the price shock of tariffs and the incoming Hormuz-related supply disruptions. Thanks to rising equity wealth and easier consumer lending standards (driven by bank deregulation), households have so far limited cuts in spending by slashing their savings rate to 2.7%², the floor reached during both the subprime bubble and the post-lockdown euphoria. Given the recent signs of an upturn in hiring, there is no reason to fear further weakening of consumer spending until Q4. However, with inflation reaccelerating and a lenient Warsh at the Fed, the term premium at the long end of the US Treasury curve is likely to rise further, threatening the long-duration assets that are the major AI equity plays. AI has cushioned US growth from the oil shock, but makes it more vulnerable to an interest-rate aftershock.

China: the hubris of a global hegemon
Although the largest importer of Middle East hydrocarbons, China has to date managed to cushion its economy from the Hormuz shock by drawing down on its massive inventories. In addition, like the US, it benefits from the domestic boom in AI investment. However, the AI capex wave is unable to transmit to the wider economy as effectively as it does in the US. Despite China’s comparative advantages in human capital, faster adoption in manufacturing (thanks to open-source models) and cheap electricity, its AI boom is limited by the import dependence of its infrastructure build-out, at a time when export controls on critical US chips are tightening. The related wealth effects are also capped by the technology sector’s relatively low market capitalisation. We estimate that direct AI investment in China will reach around 0.8% of GDP in 2026, compared with 2.7% in the US.

In addition, the geographical concentration of the AI boom in Tier 1 cities exacerbates the negative wealth effects stemming from the housing sector. Most excess housing capacity is located in lower-tier cities suffering from demographic decline and economic marginalisation. Lower land-transfer revenues mean local governments are forced into counter-cyclical fiscal tightening, with no offsetting response from the central government. All in all, Beijing appears to assume that the export boom and AI build-out will be sufficient to keep GDP growth at 4.5%, despite the global protectionist backlash and tightening US export controls. We expect disappointing activity data to force an abrupt policy pivot towards public infrastructure spending in Q3, which could prove a challenging moment for the Chinese leadership.

Euro area: collateral damage
Europe is once again being shaken by external shocks. After the tech war and the trade war in previous years, the Iran conflict adds a clearly stagflationary twist to the outlook, already visible in business and consumer surveys. We have halved our euro area growth forecast to 0.6% year-on-year, as a renewed squeeze on purchasing power and heightened uncertainty weigh heavily on household consumption. Investment should prove less cyclical than usual, but for reasons that say more about public crutches than private dynamism. German special funds will be the only meaningful support to German growth in 2026, while the final year of NextGenerationEU disbursements will still irrigate Italy and Spain.

On the inflation front, stronger energy prices lift our forecast to 2.9% year-on-year. Together with mounting indirect effects and higher consumer inflation expectations, this pushes the ECB into credibility-defending mode. We expect two hikes by the end of the summer, though not a 2022-like cycle: a softer labour market makes wage dynamics less threatening than in the early post-Covid recovery, limiting the risk of second-round effects.

Europe is fighting for credibility at the institutional level too. Tentative moves towards a capital markets union and tougher instruments against China show that the EU has finally understood both the scale of its structural challenges and the nature of the China threat. But recognition is not leadership. The lack of political unity and executive resolve throw into question the effectiveness of this fight.

Equities powered by earnings growth expectations
With US equities expected to deliver annualised mid-teens earnings growth over the next three years, earnings momentum continues to outweigh the headwind from higher bond yields. In an economy growing at roughly 6% in nominal terms, largely powered by AI-related capital expenditure, technology remains at the heart of our equity allocation.

While the AI story may still be in its early chapters, adoption and inference are accelerating at a remarkable pace. Anthropic's annual recurring revenue surged from $10 billion at the end of last year to $47 billion by the end of May 2026³. Its revenues are set to be on par with those of Salesforce and well ahead of many established software leaders.

The race for compute is a defining feature of the AI investment cycle, with demand continuing to outstrip supply. Investors have largely viewed hardware suppliers as the primary beneficiaries of the AI boom. We believe the opportunity is broadening. Data centre economics are becoming increasingly attractive, suggesting that hyperscalers could emerge as some of the next major winners as returns on massive AI infrastructure investments materialise faster than expected.

Within technology, we favour high-quality hardware and infrastructure assets, particularly where bottlenecks remain acute and supply is constrained. TSMC, the critical backbone of the AI value chain, is expected to expand capacity well below the pace of AI-driven demand growth.

The rising nominal growth rate is not lifting all boats
And outside of technology? Consumer stocks have been underperforming the rest of the market for several years and are among the most underweight sectors globally, but that alone is not enough to make them an attractive opportunity. Indeed, consumer stocks are facing deeper challenges than high energy prices.

Luxury remains supported by a resilient US consumer and the wealth effect generated by record equity markets, yet this strength rests on increasingly narrow foundations. In the meantime, European luxury continues to suffer from the absence of a meaningful recovery in Chinese, Russian and Middle Eastern demand. As Richemont's Johann Rupert put it, the sector can feel like "a black-tie dinner on top of a volcano."

Consumer staples look equally challenged. High leverage, weak demographic tailwinds, the rise of GLP-1 drugs and slowing volume growth all raise questions about the sustainability of their attractive dividend yields.

The catch-up of European equities that began in late 2024 was fully reversed in the wake of the Iran war, yet valuations have already absorbed much of the expected rise in bond yields. A normalisation of the situation in the Middle East could reignite optimism around European growth and support a renewed rotation into the region. Our preferred exposures remain aeronautics and banks.

European banks in particular offer optionality to a “peace play”. Any easing in geopolitical tensions could unlock further upside for a sector that still combines attractive valuations, strong capital returns and earnings that tend to benefit from higher rates. We see further re-rating potential, especially for institutions combining traditional banking with wealth management, where returns are less dependent on the credit cycle and more resilient across market environments.

Growth, inflation, deficits: the three horsemen of higher sovereign yields
Long-term bond yields have moved sharply higher since their 2020 historical lows, but we believe this trend is not over. The ingredients behind higher yields remain firmly in place: resilient growth, sticky inflation and deteriorating fiscal dynamics. This is particularly true in the US.

First, the economy continues to defy expectations, with US growth even showing signs of reacceleration, keeping real rates elevated. Second, inflation has proven more persistent than anticipated. Beyond the short-term impact of geopolitical tensions on commodities prices, strong labour markets, robust consumption, and rising trade frictions are challenging the disinflation narrative. The growing debate around the independence and credibility of future monetary policy is also beginning to feed into long-term inflation expectations. Finally, investors are increasingly questioning the fiscal path ahead, especially with elections coming up. Governments are issuing record amounts of debt at a time when natural buyers are becoming scarcer. Deficits do not lower yields, and the Fed cannot print term premia away.

Credit markets: too expensive to love, but investors are paid to stay
Credit markets have somewhat defied the traditional playbook. Historically, rising rates and volatility led to wider spreads. Since 2022, the opposite has happened. The return of income, positive real yields and strong demand, combined with the cushion embedded in a decent carry, has allowed long-term investors to be less preoccupied by mark-to-market moves and has hence kept credit markets remarkably resilient. Against a backdrop of expensive equities, many investors continue to view credit as the most attractive source of income.

The result is a market where spreads are close to historical tights, where carry remains appealing, but where valuations leave little room to digest bad news.

The main risk is complacency. Credit is inherently asymmetric: upside is capped at the coupon, while downside can be substantial if fundamentals hint at the potential for an existential crisis. In such an environment, avoiding losers becomes more important than finding winners.

Our preference is therefore to avoid sectors facing technological disruption, where history shows innovation waves tend to create their own credit casualties. Cable operators suffered during the telecom revolution, energy companies during the shale boom, and retailers and office landlords during the rise of e-commerce and remote work. The next fault line could be software, where AI is beginning to challenge business models, pricing power and competitive moats.

Conversely, we favour issuers backed by hard assets, strong cash flows and tangible recovery value, particularly in financials, energy, industrials and selected consumer sectors. This allows us to build diversified portfolios with investment-grade characteristics while still generating yields around 5%. The love affair has matured, but the income remains.

What could go wrong?
There are always reasons to be cautious. The recent parabolic moves in parts of the market, from Korean equities to Intel and pockets of non-profitable technology, bear some of the hallmarks of FOMO-driven behaviour. Valuations are becoming increasingly demanding and the surge in both equity and bond issuance raises questions about how much new supply investors can absorb. Yet history teaches that late-cycle dynamics can persist far longer than many expect, as the experience of the late 1990s reminds us.

The key variable remains long-term bond yields. Higher yields sit at the nexus of most market risks today: they challenge the valuations of long-duration assets, weaken the wealth effect that has supported consumption and increase financing costs for both companies and governments. Rising yields remain the most likely catalyst for a repricing.