AI Capex Drives Divergence as Korea Tech Stays Up 80%

Key Takeaways

• Goldman Sachs estimates US artificial-intelligence capital expenditure at around $755 billion this year, versus around $100 billion in China; Morgan Stanley estimates European spending at approximately 5% of the US level.
• China’s second-quarter GDP growth was 3.6% at a seasonally adjusted annual rate, constrained by property-sector and consumer-demand weakness.
• Emerging-market technology hardware and semiconductor stocks were up nearly 120% from January at their peak this year; Korea exceeded 250% before the late-June correction.
• After that correction, the same group remained up nearly 80% in aggregate as of 6 August 2026, while the non-technology portion of the MSCI All Country World Index was up 10% year to date.

US AI Investment Gap Keeps Global Equity Leadership Uneven

BNP Paribas Asset Management’s latest global equity outlook, published with returns measured as of 6 August 2026, argues that the anticipated broadening of global growth has not materialised as expected. The United States continues to outpace other major economies, supported by heavy investment in artificial intelligence and solid consumer demand.

The scale of the investment divide is central to the thesis. Goldman Sachs estimates US AI capital expenditure at around $755 billion this year, compared with around $100 billion in China. Morgan Stanley estimates that Europe is spending approximately 5% of the US total. That disparity helps explain why US growth has retained an edge even as investors had expected more balanced international expansion in 2026.

BNP Paribas AM also points to the latest US GDP figures, where business investment made a significant ongoing contribution to growth. Semiconductor purchases from abroad created a net-export drag, but this was partly offset by resilient domestic demand. For traders studying whether the equity leadership remains narrow or broadens further, the market institutional positioning data can provide a complementary lens on how professional participants respond when technology leadership and macro growth diverge.

The report’s underlying message is not that every market outside the US is weak. Instead, it describes a two-speed environment: a US economy backed by AI investment and demand, versus regions facing different constraints and recovery timelines.

Korea Semiconductor Correction Leaves AI Hardware Returns Elevated

The primary beneficiary of the AI capital-expenditure boom has been emerging-market technology hardware and semiconductor shares, with Korea singled out as the most pronounced example. At their high point this year, an index of these stocks had risen nearly 120% from January. For Korea, the advance exceeded 250%.

The subsequent correction from late June was substantial. Yet the retracement did not erase the broader advance: the hardware and semiconductor group remained up nearly 80% in aggregate as of 6 August 2026. US hardware stocks also fell 6% in sympathy, according to BNP Paribas AM, while retaining good year-to-date gains.

This sequence matters for active traders because a large correction after an outsized rally can create elevated volatility without necessarily settling the broader trend. our research does not provide intraday levels, technical thresholds, or current market prices, so trade planning should focus on confirmed session behaviour, liquidity conditions and firm-specific execution rules rather than assumed support or resistance.

For traders seeking to interpret whether the correction reflects a shift in broader participation or a concentrated repricing in AI-linked hardware, order flow analysis around prop-firms events can help frame the distinction between a temporary risk reduction and a more persistent change in market appetite.

Software and Hyperscalers Rebound as AI Revenue Takes Focus

BNP Paribas AM describes software and hyperscaler shares as, at times, the mirror image of the hardware trade. Software business models have faced perceived threats from AI, while the capital expenditure of hyperscalers-defined in the report as Meta, Alphabet, Microsoft and Amazon-has weighed on near-term earnings prospects.

More recently, investor attention has turned to revenue generated by AI models, supporting a rebound in these shares. That rotation is important because it broadens the AI debate beyond semiconductor and hardware demand. Investors are increasingly weighing the immediate cost of spending against the potential revenue outcome from AI-related products and services.

For prop traders, this creates a market structure in which technology subsectors may not move together. Hardware weakness, software resilience and hyperscaler recovery can coexist, increasing the risk of treating a broad technology index as a single uniform exposure. Traders carrying correlated index and single-stock positions should review their AI-correction trading restriction comparison before adding exposure, particularly where rules apply to open equity, concentrated positions or daily loss thresholds.

our research does not claim that the rebound will persist. A constructive scenario would require continued confidence in AI-related revenue generation. A bearish scenario would emerge if investors again prioritise the cost and near-term earnings impact of elevated capital expenditure over monetisation prospects.

Europe’s Energy Shock and China’s 3.6% Growth Complicate the Rotation

The contrast with Europe is stark. BNP Paribas AM says the region has been more affected by the Iran war energy shock, with services moving into contractionary territory in many countries. The European Union’s strategic-autonomy initiative is taking effect more slowly than expected.

Still, the report does not frame Europe as uniformly deteriorating. Growth has remained resilient, and improving economic surprises could support investor sentiment. For European equity and currency traders, this introduces a conditional setup: sentiment could improve if the anticipated surprises continue, but energy-related uncertainty and services weakness remain important counterweights.

China, meanwhile, is benefiting from AI-linked exports but remains constrained by a weak property sector and consumer demand. Its economy grew 3.6% in the second quarter on a seasonally adjusted annual-rate basis. The most recent purchasing managers’ indices point to slower domestic momentum, although BNP Paribas AM cautions that high temperatures, typhoons and flooding may have distorted the readings. The firm expects those weather effects to reverse in coming months and therefore anticipates only incremental government efforts to boost growth.

For global equity traders, these regional differences support a selective rather than uniform approach. Challenge rules for AI-driven equity volatility are especially relevant when a seemingly broad risk-on or risk-off move masks materially different conditions across US technology, Korean semiconductors, Europe and China.

Market Impact Snapshot

The directional classifications above reflect our research’s stated performance and macro backdrop rather than new price forecasts. BNP Paribas AM says oil-price and bond-yield swings, together with renewed Middle East conflict and a dovish, uncommunicative Federal Reserve, have restrained gains. Even so, most markets continued to advance modestly, and the non-technology component of the MSCI All Country World Index was up 10% year to date.

Oil, Bond Yields and Fed Communication Remain the Cross-Market Risks

The report highlights several forces that could interrupt the equity advance outside the technology hardware sell-off: renewed Middle East conflict, fluctuations in oil prices, swings in bond yields, and a US Federal Reserve described as dovish and uncommunicative.

These factors may increase cross-asset volatility and reduce the reliability of one-directional equity trades. A rally in non-technology equities can coexist with rapid reversals in technology hardware, while changes in energy and rates expectations can affect regional indices differently. Session selection matters: traders may prefer to wait for liquid market windows and confirmed follow-through rather than initiating positions during abrupt headline-driven moves.

The practical question is whether a trader’s evaluation structure can withstand a reversal after entry. Before trading regional equity or technology volatility, use a technology-volatility drawdown rule comparison to identify how firms treat daily losses, overall limits and related restrictions. Traders preparing position exposure can also use funded account maths tools to plan size relative to the maximum loss permitted by their programme.

Prop Trader Playbook for a Two-Speed Equity Market

The BNP Paribas AM outlook supports a disciplined, selective approach rather than a blanket long or short view on equities. Technology hardware has delivered extraordinary gains, but the late-June correction shows that strong year-to-date performance does not remove downside volatility. At the same time, non-technology global equities have continued to advance modestly, suggesting the broader market picture is not defined solely by the AI trade.

For intraday traders, consider concentrating on the session in which the relevant underlying market is most liquid, while avoiding the assumption that US technology, Korean semiconductor shares, European indices and China-linked assets will react identically. For swing traders, energy developments, bond-yield movements, Federal Reserve communication, China’s domestic data and evidence of AI revenue generation are the major forward-looking catalysts cited or implied by our research.

Traders starting a new evaluation should consider prop firm options suited for prop-firms market conditions and match their approach to the likely volatility profile. A trader who specialises in fast technology rotations may need a different rule set from someone trading slower regional-index themes; a trading style matcher for AI-led market dispersion can help identify a better fit.

Avoid chasing an already extended move solely because the year-to-date return remains positive. Instead, define invalidation before entry, account for correlated exposure across technology products, and preserve loss capacity for high-conviction setups. Traders who are close to a withdrawal threshold should also account for processing times across top prop firms when planning whether to reduce exposure after a volatile session.

Frequently Asked Questions

Why are US equities outperforming other major markets

BNP Paribas Asset Management attributes the US advantage largely to massive AI capital expenditure and solid consumer demand. It says US business investment has continued to contribute significantly to growth, although semiconductor purchases from abroad have created a net-export drag.

What happened to Korean semiconductor stocks in 2026

At their peak this year, the relevant emerging-market technology hardware and semiconductor index had risen nearly 120% from January, while Korea had gained more than 250%. Following a large correction from late June, the broader group was still up nearly 80% in aggregate as of 6 August 2026.

Is Europe’s equity outlook weakening because of the energy shock

BNP Paribas AM says Europe has suffered more from the Iran war energy shock and that services entered contractionary territory in many countries. However, it also says growth has remained resilient and that improving economic surprises could lift investor sentiment.

What does China’s 3.6% second-quarter GDP growth mean for markets

China’s GDP grew 3.6% in the second quarter at a seasonally adjusted annual rate, with property-sector weakness and soft consumer demand remaining constraints. The firm says recent PMI data indicate slower domestic momentum, but weather distortions may reverse in coming months and policy support is expected to be incremental.