Global Macro Research
เศรษฐศาสตร์มหภาคGlobal Economic Scenarios Q4 2026
Our baseline scenario — “resilience despite the risks” — combines constrained Gulf oil supply with a powerful AI capex cycle. We forecast global growth to accelerate from 3.1% this year to 3.6% in 2027, but inflation to remain too high, so central banks will do a few more insurance hikes. The key risks, both to the upside as well as the downside, are from geopolitics — “Peacemaker-in-chief” and “New World Disorder” — and AI — “AI equity & capex collapse”, “AI labour market disruption” and “productivity boom”. We also consider an “investment boom”, a “bond market rout” and a US-China “technological decoupling”.
Author
Global Macro Research
ระยะเวลา: 1 นาที
วันที่: 21 ก.ย. 2569
Key Takeaways
- Our baseline scenario, “resilience despite the risks”, has a 55% probability. Geopolitically, it is conditioned on six months of “no war, no peace” between the US and Iran, followed by a deal that allows Middle East oil supply to normalise. During the impasse, around half of normal crude volumes, but no LNG, still escape the Middle East. We are therefore conditioning on Brent crude oil prices around $90 per barrel at the end of 2026, easing towards $75 by the end of 2027. But the spread of oil price outcomes across scenarios is unusually large.
- In the baseline, the lagged effects of the fuel cost increases that have already happened keep headline inflation elevated. But the underlying disinflationary trend is intact, and second-round effects are limited enough to keep inflation expectations anchored.
- Global economic activity is more resilient than the scale of the supply shock would imply. AI-related capex is a powerful tailwind in the US and, via exports, for China and Asia. World trade has hit a fresh high as a share of global GDP, fiscal policy is supportive in Germany and Japan, and China is insulated by its strategic oil reserves. So, in our baseline, we forecast global growth to accelerate from 3.1% this year to 3.6% in 2027, while inflation moderates from 4.5% to 3.8%.
- The adjustment therefore falls on interest rates rather than on activity. The Federal Reserve (Fed) follows September’s hike with another in December, the Bank of England (BoE) hikes in November and in February next year, the European Central Bank (ECB) hikes again in December and then March, and the Bank of Japan (BoJ) hikes again in December and April. We think these are insurance hikes rather than a sustained tightening cycle, in keeping with the latest Fed “dots”, but there is scope for a larger rates cycle, in both directions, across the alternative scenarios.
- Geopolitics is the first axis of uncertainty around the baseline, and the alternatives scenarios are both to the upside as well as the downside.
- In “Peacemaker-in-chief” (20% probability), Trump refocuses on diplomacy sooner rather than later. A return to the memorandum of understanding (MoU) with Iran normalises traffic through Hormuz, the US China relationship is put on a more stable footing, and diplomatic progress towards a Russia-Ukraine ceasefire could also be made. Energy chokepoint constraints that have weighed on activity and pushed up prices since 2022 ease near-simultaneously, oil falls below $70, and central banks’ next moves are cuts rather than hikes.
- The mirror image is “New World Disorder” (25% probability), in which conflict risks escalate and converge. A shortage of air defence munitions and a diminished US deterrent embolden multiple state actors at once, Gulf crude exports fall below a quarter of normal with both Hormuz and Red Sea workarounds disrupted, so oil prices breach $120 per barrel and stay elevated for much longer. The global economy experiences a large stagflationary shock alongside a scramble to rearm. Central banks undertake a sustained hiking cycle, while asset prices fall.
- AI is the second main risk axis, and it too cuts both ways. In “AI equity & capex collapse” (25% probability), weaker monetisation, competition between closed- and open weight models, a deliberate slowdown in frontier model development, or funding and permitting constraints, cause capital investment and investor optimism to fall off a cliff. Mega-cap equities correct violently and private credit defaults rise, but the recession resembles the 2001 dotcom bust rather than a 2008-style systemic crisis, because the losses sit outside the core of the banking system. Nevertheless, it is a key risk given current market pricing.
- On the other hand, in “productivity boom” (also at 25% probability), AI gains arrive faster than expected and are reinforced by deregulation, vindicating Chair Warsh’s argument that they are ultimately disinflationary. Growth is higher while inflation and interest rates are lower. There is a significant further equity rally, while the dollar strengthens given the US is at the technological frontier, and inflation expectations fall.
- Sitting in between these two outcomes is “AI labour market disruption” (5% probability), in which agentic AI delivers on its productivity promise, but with very disruptive labour market impacts. These occur much faster than policy can respond, so the economy and therefore financial markets are ultimately weaker than they would otherwise be. We think this outcome would require significant policy mistakes to be sustained over a long period of time, hence the relatively low probability.
- Three scenarios shape the wider distribution. In “investment boom” (15%), hyperscaler capex approaches $1trn and is reinforced by public spending, lifting growth, inflation and rates together, but in an orderly way. In “bond market rout” (15%), a faster Fed balance sheet run-down and heavy issuance create a “savings dearth” that pushes yields sharply higher, with bond market dysfunction spilling into risk assets. And in “technological decoupling” (10%), a US-China “Sputnik moment” balkanises the global economy along competing technological lines.
- Common to many of the scenarios is a world of more frequent supply-side shocks, structurally higher defence spending and shorter supply chains. This keeps inflation volatility elevated, term premia rebuilding, and bond-equity correlations more frequently positive. Note that our scenario probabilities sum to more than 100%, because we don’t think of these scenarios as necessarily being mutually exclusive. Nor are they a spanning set of scenarios. These are the most important macro risks we see over a one-to-three-year cyclical time horizon.
Read the full article.
