At first glance, 2026 looks like a good year for equities. Wall Street reached new records during the summer, Europe extended the momentum that began in 2025, and several Asian markets benefited from accelerating technology investment. Yet looking only at headline indices gives a deceptively uniform picture of what has actually happened since January. Beneath the averages lies a far more selective market, in which a handful of companies, sectors and, above all, major investment cycles have generated a disproportionate share of the gains.

The question, then, is not simply which stock markets have risen. It is who made them rise — and who prevented them from rising further.

In the United States, that distinction is particularly important. By late August, the S&P 500 was up roughly 12% since the beginning of the year and had recently reached new records. The move rested on a reality that has become increasingly difficult to ignore: US corporate earnings have grown much faster than expected. According to LSEG data compiled after the second-quarter reporting season, adjusted earnings for S&P 500 companies were on course to rise 33.5% year on year, their fastest pace since 2021.

But that aggregate strength conceals a more specific engine. A crucial part of the American advance remains tied to the enormous investment cycle surrounding artificial intelligence, data centres, semiconductors, power grids and the electricity required to sustain them. Nvidia has become its most visible expression. The company no longer represents merely a successful technology stock: its size has become large enough to influence the trajectory of global indices directly. By late August, its growth outlook was still moving Asian suppliers, component manufacturers and companies associated with AI infrastructure.

It would nevertheless be reductive to describe Wall Street as another repetition of the “Magnificent Seven” phenomenon. What increasingly distinguishes 2026 is precisely the broadening of the movement. Technology investment remains central, but its effects are spreading into electrical infrastructure, industrial equipment, power generation and companies capable of addressing the physical constraints of the digital economy. Behind artificial intelligence models now stand transformers, cables, turbines, grids and power plants. Software continues to fascinate; the market is rediscovering that even the cloud eventually ends up inside a building connected to a socket.

This expansion helps explain why the S&P 500 has resisted a succession of shocks that might otherwise have interrupted its advance. The war between the United States and Iran, disruptions around the Strait of Hormuz, trade tensions and rising bond yields have shaken markets without, so far, destroying their underlying trend. Earnings have provided a powerful cushion.

Europe tells a different story.

By late August, the STOXX 600 was heading towards an annual gain of more than 13% if levels anticipated for December were reached. More importantly, European corporate earnings had increased by 24.1% in the second quarter, their strongest growth since 2022 and, excluding the post-pandemic rebound, one of the most vigorous expansions in more than a decade. The familiar narrative of a Europe structurally incapable of generating equity growth has therefore collided this year with a more complicated reality.

Europe's engines, however, are not Wall Street's.

Defence represents one of the most visible changes. The sustained increase in European military budgets has transformed companies once regarded as specialised industrial stocks into strategic assets. Rheinmetall provides perhaps the most spectacular example. Despite occasional setbacks — including the cancellation of Germany's F126 frigate programme, which led the company to reduce its 2026 forecasts — Rheinmetall recorded nearly 70% revenue growth in the second quarter and continues to build its expansion around enormous military order books.

Alongside defence, a second engine is emerging: infrastructure investment. Germany has opened a new fiscal cycle through its €500 billion special infrastructure fund and the easing of constraints on military expenditure. Markets have begun to incorporate the potential consequences of this spending for industry, construction, electricity and capital equipment.

Banks have also regained a central position. Persistently higher interest rates have improved some sources of banking income, while the resilience of the European economy has so far limited the prospect of a severe deterioration in credit quality. But what supports a sector can quickly become its problem. When concerns over French public finances intensified in late August, BNP Paribas, Société Générale and Crédit Agricole lost between 4% and 5% in a single session. Europe's 2026 market rewards banks when they benefit from higher rates; it punishes them as soon as those same rates become a thermometer for sovereign risk.

That contradiction is particularly visible in Paris. The CAC 40 operates in an environment where several of its largest companies remain exposed to forces France controls only marginally: Chinese demand, American consumption, European interest rates, currencies and global trade. Luxury and some consumer stocks must contend with a less predictable China, while French fiscal tensions increase the risk premium demanded by investors. Pernod Ricard's weakness following disappointing US and Chinese sales in August illustrates that vulnerability. In Paris, perhaps more than elsewhere in Europe, the international quality of its corporations is not always enough to neutralise domestic risk.

London offers almost the opposite picture. The FTSE 100 benefits from its unusual composition: banks, energy, commodities, pharmaceuticals and multinational companies generating much of their revenue outside the United Kingdom. When metals rise, mining groups can pull the index higher; when oil rises, the energy majors can take over. A structure long considered unattractive compared with American technology becomes an advantage when global markets rediscover the strategic value of physical resources.

Asia adds a third story — or, more accurately, several of them.

Japan remains supported by changes in corporate governance, technology investment and the gradual normalisation of its economy, but a stronger yen and rising Japanese bond yields introduce a new constraint for major exporters. For an automaker or industrial company generating a substantial share of its sales abroad, currency appreciation can quickly turn an improvement in Japan's macroeconomic position into bad news for its share price.

South Korea occupies almost the opposite position. Its equity market has become one of the most direct gauges of the global semiconductor cycle. Samsung Electronics, SK Hynix and the industrial ecosystem surrounding them are exposed to booming demand for advanced memory and equipment required by artificial intelligence. In late August, another favourable Nvidia outlook immediately pushed the KOSPI 1.8% higher in a single session. An announcement in California can now move Seoul before Wall Street has even opened.

China and Hong Kong present another case entirely. Here, artificial intelligence is intertwined with technological sovereignty. US semiconductor restrictions have accelerated Beijing's efforts to construct a domestic supply chain capable of competing with Western suppliers. Huawei and a new generation of Chinese AI processor manufacturers are progressively occupying that space. In 2026, several young chip producers have recorded spectacular growth, while Nvidia's dominance of the Chinese market has eroded.

This gives Chinese technology stocks a different meaning from their American counterparts. In the United States, markets are financing a race for computing power. In China, they are financing both a race for computing power and a strategy of industrial independence. The same GPUs tell two different political stories.

A fairly clear geography of the 2026 winners is therefore emerging. America is rewarding artificial intelligence and the infrastructure required to sustain it. Europe is rediscovering defence, banks, industry and public investment. London is benefiting from its exposure to physical resources. South Korea is turning the semiconductor cycle into a national market engine. China is attempting to convert technological constraints into domestic industry.

The losers display a similar coherence.

Companies heavily dependent on hesitant Chinese consumption have struggled more. Parts of the luxury, spirits and discretionary consumer sectors have had to contend with less predictable demand. Rate-sensitive sectors have suffered whenever bond yields have risen, with real estate and indebted utilities repeatedly exposed. Exporters become vulnerable when their domestic currency strengthens. And companies whose valuations already incorporated near-perfect growth scenarios are discovering that good results are no longer necessarily enough: sometimes they must exceed expectations that have themselves ceased to be reasonable.

This may be where the year's most important change lies.

For much of the previous decade, markets learned to value scarce growth in a world of cheap money. In 2026, they are gradually moving towards a different configuration. Global growth has proved more resilient than expected, investment related to AI and defence remains enormous, public spending remains elevated and corporate profits are rising strongly. Yet that same strength simultaneously sustains inflation and keeps interest rates high.

In September, that contradiction stopped being theoretical.

Brent crude's return above $100 a barrel, driven by worsening tensions in the Middle East and disruptions to export routes, abruptly revived inflation expectations. The European Central Bank raised its policy rate from 2.25% to 2.50%. In the United States, the ten-year Treasury yield moved towards 5%, while markets sharply increased the probability of further tightening by the Federal Reserve. On 11 September, US indices rebounded by more than 1%, but were still heading towards a negative week.

The paradox is striking. An economy strong enough to sustain corporate profits can also become strong enough to prevent central banks from easing monetary policy. Investment that generates growth can increase demand for electricity, copper and capital to the point of reinforcing inflationary pressures. And higher oil prices can enrich energy producers while reducing the disposable income of almost everyone who buys from them.

Markets are therefore beginning to ask a different question.

At the beginning of 2026, the challenge was primarily to identify the companies capable of producing growth. As autumn approaches, investors must also determine which companies can afford its price.

That is what makes the current hierarchy more revealing than the ranking of the indices themselves. Nvidia, semiconductor manufacturers, defence groups, selected banks, energy producers, mining companies and infrastructure suppliers appear to belong to very different worlds. Yet they have something in common: they sit close to the expenditures that governments and corporations increasingly regard as difficult to postpone.

Computing power, energy, defence, networks, metals, infrastructure.

By contrast, companies primarily dependent on cheap capital, abundant discretionary consumption or geopolitical stability are discovering a less accommodating world.

The indices may continue rising together for some time. That no longer means their constituents are operating in the same market.

Since January, 2026 has rewarded not so much a particular region as a particular position within the global economy: being located where increasingly necessary investment flows. American technology, European defence, Asian semiconductors and energy infrastructure are not four independent phenomena. They are different expressions of the same migration of capital towards physical and strategic capacity.

The question now is how much those capacities are worth when the money financing them is becoming more expensive too.

Main Sources

Reuters, 26 August 2026 — strategist survey on the S&P 500 outlook; LSEG I/B/E/S data on US corporate earnings.

Reuters, 26 August 2026 — STOXX 600 outlook and LSEG I/B/E/S data on European corporate earnings.

Reuters, 27 August 2026 — Nvidia results and outlook; reaction among Asian semiconductor stocks.

Reuters, 6 August 2026 — Rheinmetall results, outlook and order backlog.

Reuters, 27 August 2026 — European markets, French banks, interest rates and French fiscal risk.

Reuters, 4 August 2026 — FTSE 100, mining groups and developments in the UK equity market.

Reuters, 7 September 2026 — China's artificial-intelligence processor industry and the evolving competitive position of Nvidia.

Reuters, 9–11 September 2026 — oil, global equity markets, inflation and rising bond yields.

Reuters, 10 September 2026 — European Central Bank decision and reaction across European bond markets.