UBS Equities-Global Equity Strategy _US exceptionalism exceptional or no...-118348843

ab20 October 2025Global ResearchGlobal Equity StrategyUS exceptionalism: exceptional or not…Stock market exceptionalism: We are benchmark the US (within a 100% dedicated equity portfolio). There are 4 supports: i) The US economy and stock market are particularly well suited to AI adoption: US tech+ accounts for c82% of global tech, AI added c65bp to GDP in 1H post Q2 GDP revisions (see here), and the US is likely to adopt AI quicker than elsewhere (owing to both scale and looser labour regulations).ii) The US outperforms around c80% of the time if tech outperforms, and right now the P/E relative of the Mag 6 is close to its post-2018 norm (40% premium cf to a norm of 3%) but both tech+ earnings revisions remain superior to the market, and earnings growth is also forecast by consensus better than non-tech (until Q1 27 on consensus numbers). We think there is upside risk to government spend on datacentres (forecast to be c10% of total datacentre spend) as governments globally will likely aim for datacentre independence. In aggregate, hyperscalers' capex could rise c40% before they have to fund out of debt. TMT investment as a percent of GDP is half of 1999 levels.iii) The US economy particularly benefits from a bubble that we are entering (we have potentially all 7 preconditions we are not fully in yet); see here. This is because US households have nearly $61trn in equities (hence 10% on equities=0.6% to 1% on GDP). iv) Historically, when global growth slows to be sub-3.5% (and UBS predicts GDP growth globally of 2.9% in 2026 cf to 3.2% in 2025), the US outperforms because: it has the lowest operational leverage; it cuts labour costs more quickly; and the Fed, with a dual mandate, tends to become the most dovish major central bank into slowdowns, as is the case now (with the Fed restarting a cutting cycle despite revising nominal GDP up by 60bp). We see 6 factors that currently stop us from recommending an overweight (within a 100% dedicated portfolio):i) A lot is in the price. The sector-adjusted P/E is on a 40% premium to global markets (compared to a norm of 3%). While valuation arguments have not worked, both we and Gerry Fowler note that the total yield in the US (dividend and buyback) is now nearly 42% below the total yield of Europe having up to 2022 been on a par. ii) There are reasons to be more selective on tech, namely: a) the clear-cut undervaluation of the Mag 6 in mid-April has gone; b) the capex to sales of hyperscalers is now 26% - with an unclear revenue model of Gen AI; c) UBS forecasts that hyperscalers' capex slows to 16% in 2026 from 60% in 2025; and d) there may be some inherent circularity in eg OpenAI deals (that brings back memories of vendors financing the telecom operators during the TMT period). iii) The net buyback yield is now 0.3% below Japan and on par with the Stoxx 600, having since 2010 on average been 1.4% and 1.5% higher - this impacts both valuation but also corporate sector net buying and th