UBS Raises AI Capital Expenditure Forecast: Approaching $1 Trillion by 2026, 90% of Growth Driven by Memory Price Increases

BlockBeats News, September 19 – According to UBS's latest estimates, global AI capital expenditure will reach $998 billion by 2026, nearly doubling from $506 billion in 2025, and further rising to $1.447 trillion in 2027. This significant upward revision is primarily attributed to the rapid surge in memory prices rather than an overall expansion in infrastructure investment. Memory spending is projected to jump from $71 billion in 2025 to $367 billion in 2026, and further expand to $923 billion in 2027. During the same period, non-memory AI capital expenditure is expected to reach approximately $631 billion in 2026 but decline to $525 billion in 2027. The proportion of memory in AI capital expenditure will increase from about 14% in 2025 to 37% in 2026, and further to 64% in 2027. UBS estimates that around 60% of the year-on-year increase in AI capital expenditure in 2026 will come from rising memory costs. By 2027, the increase in memory spending will even exceed the net overall increase in AI capital expenditure, as spending on other components declines during the same period. Of the nearly $1 trillion increase in AI capital expenditure between 2025 and 2027, approximately 90% will be driven by rising memory costs. UBS specifically noted that if the incremental expenditure is primarily driven by price increases, the impact on U.S. real GDP will be relatively limited, instead reflecting a shift of income and profits to Asian memory manufacturers, thereby contributing more directly to the GDP of related economies. Since the world's major memory manufacturers are concentrated in Asia, this suggests that as the AI investment cycle deepens, the benefits to the Asian memory supply chain may exceed previous market expectations. This structural shift indicates that the economic driving force of the AI investment cycle is transitioning from mere infrastructure expansion to the rising costs of critical components needed to support more powerful computing systems. Investors will need to adjust their frameworks for tracking AI capital expenditure accordingly. [Original Link]