BCA Macro Outlook: How Much Longer Can U.S. Stocks Rise? The AI Investment Cycle May Only Have Completed Two-Thirds

TL;DR · BCA believes that the recent three rounds of selling in U.S. Treasuries were primarily triggered by uncertainties in U.S. fiscal, trade, and foreign policies, rather than a sudden deterioration in economic fundamentals or Federal Reserve policies. · The U.S. economy remains in a strong expansion phase, with corporate profits rather than valuation expansion being the main support for the current rise in U.S. stocks, making it premature to turn fully bearish. · AI capital expenditures may still have 3 to 5 years of expansion space; even with intensified competition among leading large models, declining inference costs may actually increase demand for computing power. · The Strait of Hormuz has not been completely disrupted, and oil transportation is recovering, but low refined oil inventories and high crack spreads indicate that energy pressures have not completely dissipated. · The Russia-Ukraine conflict is currently a more concerning geopolitical risk, as Ukraine's drone operations are breaking the previous balance, and Russia may escalate its hybrid warfare or energy countermeasures. · The relative growth advantage of the U.S. over the past few years has largely stemmed from fiscal expansion; as the political environment shifts towards fiscal constraints, the U.S.'s growth advantage relative to other economies may narrow. · BCA is optimistic about the long-term trade of "buying markets outside the U.S.,” but this does not equate to a wholesale sell-off of U.S. assets, nor does it mean that the dollar will quickly lose its status as a reserve currency. · The 2020s are still in a capital expenditure cycle dominated by commodities and physical assets; inflation may only turn downward again after global capacity gradually becomes excessive in the 2030s.

Bonds Trade on "Sentiment," Economic Fundamentals Have Not Deteriorated Significantly

BCA's core judgment of the current market is that U.S. stocks reflect strong growth, while Treasuries are trading on policy sentiment from the White House.

Over the past two years, the U.S. bond market has experienced three distinct rounds of selling. The first round occurred around the 2024 U.S. presidential election, as investors worried that Trump’s potential return to power would further expand the fiscal deficit; the second round followed the announcement of the "Liberation Day" tariff policy, with concerns that excessively high tariffs would harm both growth and fiscal revenue; the third round was related to uncertainties in U.S. policy towards Iran.

According to BCA, the common catalysts for these three adjustments were political or geopolitical conflicts, rather than economic recession, runaway inflation, or a sudden hawkish turn by the Federal Reserve. The overall scale of U.S. Treasuries held by foreign investors remains stable, with no signs of large-scale selling from China and Japan, indicating that there is currently no clear global capital "flight from U.S. Treasuries."

Increased bond supply has indeed brought some pressure, especially as AI companies expand financing and corporate bond issuance grows rapidly, but this alone cannot fully explain the rise in yields. More importantly, the nominal and real economic growth rates in the U.S. remain above the 10-year Treasury yield, household leverage is low, and corporate financing activities have not significantly frozen. Current interest rates are not yet high enough to end economic expansion.

BCA therefore expects that if U.S. policy towards Iran gradually comes under the leadership of officials like Treasury Secretary Yellen, who prioritize market stability, the risk premium brought by geopolitical factors may decrease, limiting the further significant rise in Treasury yields.

AI Capital Expenditures Have Not Reached Their End

The U.S. economy can continue to maintain resilience, with AI capital expenditures being one of the important supports. In the first quarter of 2026, software and hardware investments contributed 0.8 percentage points to the quarterly growth of U.S. real GDP, the highest level since the beginning of this century. However, compared to the information technology investment cycle of the 1990s, the current intensity of capital expenditures is only just approaching that level.

Based on this, BCA judges that the AI capital expenditure cycle may still have 3 to 5 years left, rather than being close to its end.

This judgment does not rest on the assumption that "all large models will ultimately achieve high profits." The report presents a seemingly contradictory view: even if the business models of leading large models come under pressure, underlying computing power investments may continue to expand.

Open-source models and price competition will lower the cost of AI usage, harming the profit margins of some model companies, but will also enable more businesses to deploy AI. In other words, falling model prices may lead to greater computing demand and data center needs through demand elasticity. Whether large model companies are profitable is not entirely the same issue as whether investments in chips, electricity, and data centers can continue to grow.

Currently, the proportion of enterprises adopting AI is still rising, and data centers are beginning to demonstrate their commercial viability. Meanwhile, electricity demand has resumed growth after years of stagnation, indicating that AI investments have begun to have observable impacts on the real economy.

BCA acknowledges that every round of technological investment ultimately experiences overbuilding, as seen in the railways, internet, and telecommunications infrastructure. However, if this cycle is compared to the 1990s, AI capital expenditures may have only completed about two-thirds of their journey. Even if the market has entered the latter half of the cycle, leaving too early may still miss the concentrated gains in the final phase.

The real turning point to watch may be the concentrated IPOs of large tech companies. Historically, large IPOs often signify a rapid increase in market stock supply and have frequently approached phase tops. If global central banks simultaneously tighten liquidity at that time, a wave of large IPOs may become a clearer risk signal.

U.S. Stocks Still Have Fundamental Support, but Risks Come from Abroad

BCA continues to maintain a tactically optimistic view on stocks. Currently, U.S. economic growth is strong, global liquidity remains ample, and private sector leverage is not high. The current rise in U.S. stocks is primarily driven by corporate profit growth, rather than solely relying on valuation multiple expansion, making it different from the bubble phase at the end of the 1990s.

Inflation also does not yet pose a major threat. As long as energy prices do not consistently break through existing ranges, inflationary pressures are likely to have peaked. The U.S. labor market model is strengthening, which may push wages higher in the future, but BCA believes this is more of a risk for 2027 rather than an immediate variable that needs to be traded.

At the same time, U.S. consumer willingness to spend remains strong. The market has long expected the savings rate to rebound, but household consumption concepts may have undergone structural changes, meaning that the savings rate may not recover according to traditional models. This suggests that consumption can still support growth, but it also means that household buffer space is shrinking.

The report also views China's fiscal expansion as a potential "positive black swan." The pace of local government bond issuance is insufficient, while investment growth has significantly slowed, which may force the central government to increase support around the October Politburo meeting. If the policy strength exceeds market expectations, it will benefit Chinese assets, global manufacturing, and commodity demand simultaneously.

Decreasing Risks in the Strait of Hormuz, but Refined Oil Pressures Persist

BCA believes that the market tends to focus on the absolute level of geopolitical risks while overlooking the direction of risk changes. When conflicts remain severe but the pace of deterioration begins to slow, risk assets often rebound from the bottom.

The Strait of Hormuz exemplifies this change. According to shipping information obtained by BCA, vessels can still pass through the strait, but transportation costs have risen from about $1 under normal circumstances to $12 to $15. Trade flows, U.S. commercial crude oil inventories, and Chinese import data also indicate that oil is still traversing the strait, and the degree of supply disruption has eased.

This has gradually formed a new "dynamic equilibrium" in the Persian Gulf: localized military actions will still recur, but all parties are constrained by oil prices, domestic politics, and global energy demand, unwilling to truly sever strait transportation.

BCA even believes that at this stage, oil prices are not just a result of the conflict but also constrain the conflict in reverse. When oil prices fall, the U.S. and Iran have greater military action space; when oil prices rise to levels that could impact the global economy, all parties will instead restrain themselves. Brent crude may thus form a new volatility range around $85 to $100 per barrel.

However, the recovery of oil transportation does not mean that the energy shock has ended. Due to the combined effects of the Hormuz crisis and the Russia-Ukraine conflict, U.S. refined oil inventories remain low, and refining crack spreads remain high. Gasoline and diesel prices may continue to exert a tax-like drag on household purchasing power, further impacting the U.S. midterm elections.

Greater Geopolitical Risks May Come from Russia

Compared to Iran, BCA is more concerned about the potential escalation of the Russia-Ukraine conflict. Ukraine's expansion of drone operations is breaking the battlefield balance established over the past three to four years and directly affecting Russia's energy exports and domestic political stability.

The report argues that the pressure the Russia-Ukraine conflict has placed on the Russian economy and society is already significantly higher than the relative burden of the Vietnam War on the U.S. If oil prices remain high, Russia will have more war financing resources on one hand, while also judging that the West will find it more difficult to impose severe sanctions on its energy exports, which may increase the willingness to escalate further.

In the short term, Russia may prioritize actions that have "deniability," including drone strikes, cyberattacks, destruction of energy facilities, and other hybrid warfare methods; however, if domestic pressures continue to increase, actions may shift from covert to overt. Therefore, European energy costs and Russian export facilities will become key indicators to observe in the coming months.

The U.S. Advantage is Weakening, Funds Will Flow Back to the Global Market

The most important long-term judgment of the report is to remain bullish on markets outside the U.S.

BCA emphasizes that this is not a simple "sell the U.S." trade. The U.S. will still be a major economic and financial power globally, and the dollar will not suddenly lose its status as a reserve currency. However, the weight of U.S. assets in global investment portfolios has become too high, and as growth gaps and fiscal policies change, funds need to rebalance.

2025 is an important signal: although the U.S. remains the center of AI investment, U.S. asset performance has lagged behind other markets, with the dollar declining about 10% over the year. BCA believes this is not coincidental but the beginning of a long-term trend.

The U.S. post-pandemic growth advantage is often explained as productivity improvement, but the report argues that fiscal expansion is the more critical variable. The U.S. invested far more fiscal resources during the pandemic than other major economies, which boosted economic output, corporate profits, and labor productivity per hour, also supporting the performance of U.S. stocks relative to global markets.

Now, this logic is reversing. The bond market has already issued warnings about fiscal expansion, and U.S. voters' concerns about deficits and debt are nearing levels seen during the "Tea Party" movement. The fiscal expansion efforts during Trump's second term are actually under significant constraints, with government spending consistently falling short of expectations and fiscal impulses tending to flatten.

If the U.S. no longer relies on large-scale fiscal spending to maintain its growth lead, its growth advantage relative to Europe, China, and other economies may narrow. Exchange rates and cross-border capital flows typically follow changes in relative growth, which will weaken the foundation for the dollar and U.S. assets to outperform global markets in the long run.

The 2020s Still Belong to Commodities and Physical Assets

BCA believes that the world has entered a long-term capital expenditure cycle driven by multipolarization, supply chain restructuring, and national security spending.

Countries are redistributing supply chains from a single center, reducing dependence on China, and expanding investments in defense, energy, manufacturing, and infrastructure. China is unlikely to be completely removed from the global supply chain, but its central position may decline. Establishing new production networks requires a large number of factories, equipment, electricity, transportation, and raw materials, making this process inherently commodity-intensive.

Reindustrialization is not only happening in the U.S. Europe has room to further expand fiscal spending, and China has low financing costs, while pension funds and private capital in various countries can also be directed towards domestic investment. Economies outside the U.S. are fully capable of expanding investment and consumption.

Therefore, BCA recommends a long-term allocation to commodities and other physical assets for the remainder of this decade, while reducing excessive concentration in expensive U.S. financial assets. The report summarizes the 2020s as a decade of "building atoms rather than just producing bytes": while AI is undoubtedly important, data centers, power grids, energy, factories, and supply chain restructuring are the broader investment themes.

However, the capital expenditure cycle will ultimately also lead to overcapacity. When sufficient new capacity is formed globally in the 2030s, inflation may turn downward again. At that time, the currently high bond yields may provide attractive allocation opportunities for long-term investors.

Overall, BCA's combination strategy can be summarized as: continue to hold risk assets and gain bond yields in the short term, increase allocation to markets outside the U.S., commodities, and physical assets in the medium to long term, while remaining vigilant about large tech IPOs, tightening global liquidity, and the cyclical turning points brought about by the escalation of the Russia-Ukraine conflict.